Doom Roguelike Class Analysis: Universal Traits
Traits are core to Doom Roguelike's levelup system. Unlike a lot of games with leveling mechanics, leveling has no innate benefits: all you get is a single skill point that you then invest into leveling a Trait to acquire or improve its benefits. A Doomguy at Level 10 with no skill points invested is no better off than he was at Level 1!
Traits themselves come in three categories: Basic Traits, which are available immediately and have multiple ranks, Advanced Traits, which normally require investing 1-2 points into a specific Basic Trait to unlock a given Advanced Trait (But each class specializes in an Advanced Trait by removing this requirement) and don't necessarily have multiple ranks, and Masteries, which are class-specific, have complex and stringent requirements I'll be talking about in other posts, and always have exactly 1 rank and are mutually exclusive with each other.
Basic Traits in particular also have two different caps: the default cap, and the 'extended ranks' cap. In this post, I list the extended rank cap in parentheses after the default cap. The default cap is the initial cap, and if you're playing on a lower difficulty and/or not particularly thorough about killing enemies it can easily be the only cap of relevance to you. The extended ranks cap comes into play if you reach Level 12: at that point, you're allowed to raise Traits past the default cap, up to the extended rank cap. This is pretty important, as Doom Roguelike's Traits are often designed so they have accelerating returns for additional ranks: if you could just hit the extended ranks immediately, there'd be a pretty strong risk that optimal play would be doing stuff like simply maxing Son of a Gun before bothering to invest in anything else.
Note that you actually start a run with a single skill point, which the game prompts you to invest before even starting the run.
Also note that it is actually possible to hold onto a skill point, exiting the Trait screen, and that in turn you can spend it at will anytime afterward by reentering the Trait screen. This is generally a dubious idea, but is good to know for runs that do get to extended ranks: if you hit Level 11 and don't really have any good options for investing your skill point, holding onto it with intent to spend it once you're Level 12 and have extended ranks can be worth considering.
Now on to Basic Traits!
Ironman
Also note that Ironman is slightly less valuable for a Marine than the other classes, between easy access to Badass (Which is, for many purposes, a larger increase in effective maximum HP per rank while also providing protection against knockback) and already having 10 more HP than base. (For example, the first rank of Ironman is a 20% increase in durability for Scouts and Technicians and only a 16.67~% increase in durability for a Marine)
Ironman
Max rank: 3 (5)
Effect: Each rank raises 100% max HP by 10 points. Current HP is immediately adjusted so the percentage is consistent. Also increases Bullet, Shrapnel, and Melee resistance by 10% per rank.
HP in Doom Roguelike is a bit awkwardly handled, in that the framing looks exactly like classic Doom's, but the actual mechanics noticeably diverge in ways that mean a learning player's intuition is going to be fairly wrong, but subtly enough a player could easily go a long time without realizing their initial assumption is wrong.
In classic Doom, each percentage point is a hit point, with the percentage framing elegantly communicating that going over 100 HP has special rules attached to it. (Specifically: only certain healing sources can add more HP when at or above 100%)
In Doom Roguelike, the percentage is in place of telling you your true HP.
True base HP is actually 50; thus, at true base HP your displayed HP will only ever increment in units of 2%, because 2% of 50 is 1. So for one thing even if you have excellent defenses such that weak enemies can only do 1 damage per hit, you'll die twice as fast as you might intuit. This isn't too bad once you know it...
... but the obscured information aspect becomes a lot more of a nuisance once you have an HP value that doesn't fit cleanly into a whole-numbers-only percent system. Such as by taking a rank in Ironman; now each hit point is 1.67% of your HP, meaning a single point of damage will variably remove 1% or 2% of displayed HP. This won't fully clear up unless you get Ironman to rank 5, where 1% does in fact represent 1 HP.
Oh, and Marines -the class designed to be easiest to jump into for a new player- have a base HP of 60 in the first place, so a new player is liable to have this learning curve issue exacerbated further, since they inherently have the issue of 1 HP being 1.67% of max HP. Oops.
Also clunky, though much less so, is that 'overhealing' works a bit differently from classic Doom; in addition to HP over 100% not being accessible via your 'default' healing tools (Medpacks, mostly), HP over 100% 'rots', with you losing 1 HP every 5 game seconds until you're at or below 100%. There's another Trait that can do away with this behavior, but by default you should only be overhealing either in preparation for a fight you'll enter shortly, or when you're ready to leave a floor and the floor's overhealing pickups will be just left behind if unused. Otherwise you're wasting hit points.
As for Ironman itself...
Well, for starters Ironman isn't a prerequisite for any other Traits, and thus you'll only ever take points in it for its own sake. This is pretty killer to its potential as an early pick -generally, a run should have a Mastery decided on from the beginning and beeline directly for it, since Masteries have a high impact on your effectiveness. On Ultraviolence difficulty, if you thoroughly clear every floor (Including Special Levels) while building directly to your Mastery, you'll normally have your Mastery online shortly before the Phobos Anomaly, which is to say in time for the game's first mandatory boss level.
Once you have your Mastery online there's more room to pick Traits as the situation calls for it...
... but Ironman's peak period is early in a run, when you definitely don't have strong enough offenses to trash entire groups of enemies before they get in damage and you also definitely don't have strong enough damage reduction to shrug off most individual enemy attacks. In those conditions, 10 more HP can easily be the difference between skating through a rough fight and healing up afterward vs dying outright. By the time you're past the level your Mastery comes online, you may well have your offense and/or defense situation substantially solved such that 10 more HP is unlikely to be decisive.
In the extreme long haul of an Angel of 100 or Archangel of 666 run you're liable to grab ranks in it simply because you maxed out your highest priorities, then maxed out your secondary priorities, and now are just casting about for something worth putting points into.
But for other runs, it can easily be questionable to take even a single rank of Ironman.
On the plus side, it's worth pointing out that most sources of healing in Doom Roguelike are percentile, and so increasing maximum HP does actually increase how much HP you gain from healing, especially in the later stages of a run; the percentile forms of healing become overall more common as a run progresses.
It should also be noted that Ironman is unusual for never being blocked by any Mastery. Among other points, this makes it a relatively safe choice to take ranks in if you want to keep your options open while leveling; in most runs this isn't worth caring about, but in an Angel of 100 or Archangel of 666 run the power spike provided by a Mastery is less essential to get online early, and it can be nice to not commit to a build too quickly. Maybe you find a Unique or one of the less common Exotics and keeping your options open means you get to actually build around it instead of being frustrated at finding a Laser Chaingun after you've already blocked off your rapid-fire weapon Mastery.
Now, all the above was written predominately off my experiences in version 0.9.9.7: version 0.9.9.8 improved Ironman by having each rank increase Melee, Bullet, and Shrapnel resistances by 10%, and while this doesn't erase many of the problems with Ironman I just covered, it does have several notable implications that make Ironman less 'a filler pick when you're running out of better options'.
First of all, the resistance boosting makes Ironman inherently more appealing for melee combat builds, which inevitably will get hit in melee a lot throughout the game and so the resistance boosting is evergreen. It's probably better per rank than Tough as Nails for such builds, as +10% Melee resistance will often work out similarly or better than +1 Protection, even aside that Ironman gives HP.
Second, some Masteries are sufficiently generous in their requirements that there's an 'empty' Level where you've met all the requirements but aren't allowed to take the Mastery yet because all Masteries require you reach Level 6 before you can take them. For such Masteries, an early rank in Ironman can be very valuable: the early game is dominated by Bullet, Shrapnel, and Melee damage threats, and even 10% resistance will reliably shave off at least 1 point of damage from such if no other source of such resistances is in play. (Having a rank in Ironman makes it more straightforwardly good to swap out a Green Armor for a Blue Armor, for example)
Third, if a run successfully gets its energy resistance situation solved (Such as by Assembling Cerberus Armor), quite often this means that it's now the Bullet/Shrapnel/Melee threats that have become the most dangerous ones. It's entirely possible to find yourself laughing off Mancubi, Archviles, and Barons galore, only to lose half your HP abruptly from some Former Captains getting a turn of fire in: ranks in Ironman to cover that kind of situation become very appealing at that point.
Fourth, for a run that isn't really a full and proper melee build but that wants to clear the Unholy Cathedral (Maybe you're wanting to pull off a 'full 100% run'), dipping a rank in Ironman for the Melee resistance is worth considering if you can slip it in.
A more minor detail is how this improvement to Ironman intersects with another 0.9.9.8 change: that Marines now have innate 20% Fire/Plasma/Acid resistance. This means that Marines can more readily reach a point of fearing Melee/Bullet Shrapnel more than the energy damage types, hitting scenario 3 pretty readily where a rank in Ironman is disproportionately valuable.
Finesse
Hellrunner
Tough as Nails
Son of a Bitch
This is basically a specialty in rapid-fire weapons, and conversely is mildly anti-specialized in Shotguns: if you're trying to improve your Shotgun damage output, you're generally better off pursuing Reloader, Finesse, or Shottyman. Among other points, Shotguns lowering damage at longer ranges absolutely applies to Son of a Bitch-derived damage, and so a given shot may get literally no value from a given Son of a Bitch rank. Say you fire a basic Shotgun at range 7 with 2 ranks in Son of a Bitch, and it rolls 10 damage before the range drop-off and ranks in Son of a Bitch are accounted for; in that case, the second rank of Son of a Bitch didn't actually affect the final damage at all, because 11 damage at range 7 results in 5.61 damage that rounds up to 6 damage, while 12 damage results in 6.12 damage that rounds down to also be 6 damage.
Son of a Gun
Notice that this is Son of a Bitch rolled together with a superior version of Finesse. If you're going to all-in on Pistols, there's no reason to bother with either of those Traits unless you're unable to gain more ranks in Son of a Gun... and Finesse in particular is a really bad investment in runs that expect to hit extended ranks, because Son of a Gun 5 knocks your Pistol firing speed down to the absolute fastest the game allows of 0.1 second all by itself. (If it weren't for this enforced minimum, rank 5 Son of a Gun would in fact make firing a Pistol take literally no time at all) Finesse does literally nothing to your Pistol fire rate at that point.
Reloader
This is most heavily a specialization in Shotguns, since by default they alternate firing and reloading while taking at least as long to reload as to fire: thus, for several Shotguns Reloader is a more significant boost to overall fire rate than Finesse provides. This is particularly emphasized by the fact that two ranks of Reloader unlocks Shottyman, an explicitly Shotgun-supporting skill. (Albeit one that noticeably reduces Reloader's value, which is a bit awkward)
Eagle Eye
The extended ranks of Eagle Eye are almost never worth pursuing: for a weapon with no inherent modifier to accuracy, 3 points of Eagle Eye already puts you at maximum accuracy on targets within three tiles... and very, very few weapons have less than +2 to Accuracy, so actually you're generally at max accuracy even at line of sight. The only penalties to Accuracy the player can suffer are the distance penalties (-1 every 3 tiles) and a -2 penalty when Running: there aren't enemies that lower your Accuracy against them or anything like that. As such, Eagle Eye's extended ranks are only particularly worth considering if you're using low-Accuracy weapons like Chainguns and regularly fire while Running and/or have Cateye and/or regularly fire into the dark -and in the latter case, if you have two Sniper Mod Packs on your weapon of choice to do away with the in-the-dark penalty you also ignore range penalties, neutering the value of additional Eagle Eye ranks.
Brute
Berserker
Before 0.9.9.8, this was actually anti-synergistic with Ironman, as getting past 60 HP would mean that Archviles switched from 'always triggers Berserk with their ranged attack' to 'never triggers Berserk with their ranged attack'. Berserker Marines thus couldn't take even one rank in Ironman without being crippled!
Dualgunner
If you're going the Pistol specialization route, unless you're specifically going for the Technician's Sharpshooter Mastery you're going to eventually take Dualgunner; it's just too much of compression in time taken on killing things, with the only real disadvantage to using it being that you'll waste a bullet here and there. (Because sometimes the first shot kills your target but Doomguy still fires both shots) Pistols are sufficiently ammo-efficient that this ammo wastage is only particularly plausible to be a concern to a Marine that went Bullet Dance, though.
Dodgemaster
Intuition
The Scout getting to skip Eagle Eye is huge, not only for the obvious reason of Eagle Eye not being particularly great but also because Intuition 2 goes great with Shotguns being excellent at corner-shooting -and Shotguns get no benefit whatsoever from Eagle Eye ranks. Furthermore, the beginning of the game is when the Lever-related benefits are most relevant; at the beginning of a run, a Lever filling the room with Acid or Lava can be essentially an instant run-wipe, whereas late in a run the player can easily be outright immune to floor fluids, and even if they aren't they probably at least have some protection and enough spare Medpacks to be able to power through and still be ready for fighting actual enemies. There's a very solid argument for the Scout to prioritize Intuition immediately, even if it isn't progress toward their intended Mastery. (Which, unfortunately, only Cateye uses Intuition as a requirement)
Whizkid
Badass
Great on the Marine, harder to justify on the other classes -the Scout in particular blocks Badass with every Mastery, and there's no particular synergy between Badass and their innate advantages, certainly not enough synergy to make up for giving up having a Mastery. (Scout Masteries only include one 'dud' Mastery, for one, and it's the 'generalist' Mastery, so no matter what weapon you focus on, you've got a viable Mastery)
Shottyman
This is particularly helpful with Combat Shotguns, as the move-reload effect skips the need to 'pump' the Combat Shotgun before you can reload or fire.
Triggerhappy
Altogether, this means Ironman is in an okay place as of 0.9.9.8: still usually a low-priority pick, but with some clear niches. It's also helped by another change I'll be covering later in this post.
It would be nice if it wasn't so heavily sidelined in the early game by Mastery design, but I'll talk about that whole thing when we're talking about Masteries.
Finesse
Max rank: 2 (3)
Effect: Attacks take 15% less time per rank.
So to get into what this really means, we need to talk a bit about Doom Roguelike's handling of time, particularly of how it handles modifiers to time taken.
The 'default' turn or action length in Doom Roguelike is one 'second'. Doom Roguelike then tracks time to within a tenth of a 'second', where the lowest amount of time that can pass at once is 0.1 seconds. One might wonder how -15% works out in that context, since of course 15% of 1 second is 0.15 seconds.
The thing is, Doom Roguelike invisibly manages turns through an 'energy' system. This mostly isn't terribly important to the player, but basically each 0.1s every unit (Including the player) generates an amount of Energy defined by their base Speed, and gets a turn when their Energy meter fills up. This Energy meter uses much larger numbers and thus is more precise; a turn is achieved when at or above 5000 energy, the player's base Speed and thus Energy-per-tick is defined as 100, and when a unit takes an action it subtracts Energy in proportion to the 'true' time passage, with 100 Energy being equivalent to 0.1 seconds of time passage. Thus, rank 1 Finesse causing Doomguy to take 0.85 seconds to fire is implemented as 850 Energy being expended when he fires; this will result in Doomguy going from 5000 energy to 4150 Energy, and then 0.9 seconds later he'll have 5050 Energy and get a turn and spend 850 again, taking him to 4200 Energy, and then 0.8 seconds later he'll be at 5000 Energy and get a turn. The end result is that you'll never take 0.85 seconds on a base-1-second attack, but you will in fact average a firing speed of 0.85 seconds by virtue of alternating 0.8 and 0.9 second delays.
This is a really impressive system that evades a lot of common jank of turn-based games that try to have turn mechanics more granular and largely results in intuitive behavior that doesn't require a bunch of number-crunching to figure out what will actually happen with your next action. It's great! My only complaint is how non-obvious this whole thing is to the average player who doesn't go digging around for this info on the wiki or the like.
Anyway, wrapping back closer to Finesse itself, speed modifiers have two implementations, depending on whether the game considers multiple modifiers to be from the same 'source' or from different sources. If they're the same source -which multiple ranks in a Trait are considered to be- you just add them all together, meaning that rank 2 Finesse takes you from 85% fire time to 70% fire time, and the extended rank then takes you to 55%; its per-rank impact gets higher the more ranks you have!
If modifiers are considered to be separate sources, they instead get multiplied together; for example, Son of a Gun reduces Pistol firing time by 20%, so a rank in Finesse and a rank in Son of a Gun brings Pistol fire time to 0.85s*0.8=0.68s, rather than bringing it down to 0.65s. This mechanic has less overall application than you might expect, as each class of time expenditure has a pretty limited array of things that can modify it to thus be considered different sources; reload speed only has one general improvement (The Reloader Trait), and one situational improvement. (Bulk mod Packs, which improve reload time on weapons that have one or two ammo capacity; not many weapons qualify)
Also, we'll talk about this in another post, but Mod Packs have multiple different rules for time modification, specific to Modpacks, making the above 'different sources' rule even narrower than it seems.
Anyway, Finesse itself is generally useful, applicable to every form of attacking (including melee attacks), but overall lower in value than you might expect due to what other Traits exist and how weapons are designed.
First of all, Pistol builds don't want ranks in Finesse: Son of a Gun blows it out of the water and will ultimately hit the game's cap on speed modification, rendering Finesse ranks literally worthless in the long haul. (Awkwardly, Marines and Scouts that want Whizkid will have to get Finesse 2 regardless, leaving them with 'dead' levels when doing Pistol runs in Angel of 100/Archangel of 666)
Second, Shotgun builds do want Finesse ranks, but not early: with a basic Shotgun, it's much more useful to get ranks in Reloader and then grab Shottyman to improve time efficiency, as Reloader has superior time efficiency to Finesse and Shottyman is an incredible all-around improvement. Unless a run gets an unexpectedly early Combat Shotgun (Preferably also the Mod Packs to turn it into a Tactical Shotgun early) or Assault Shotgun, Finesse just isn't as good as Reloader-into-Shottyman.
Third, melee builds do want Finesse ranks, but once again, not early: they desperately need Brute ranks and to get their Mastery online and to improve their survivability. Hitting things in a shorter amount of time is nice, but a lower priority than all that. The Scout especially places it as a low priority since their Mastery makes lethal strikes not spend time at all!
Fourth, rapid-fire builds do want Finesse ranks, but later: Son of a Bitch and Eagle Eye improving their reliability at killing things is better for things like 'preventing enemies from firing back', and they also improve ammo efficiency, which is very important if you're not an Ammo Chain Marine. Finesse thus should wait until later.
That's all the 'major builds' the game provides clear, official support for. As such, Finesse is only really a clearly high-priority choice in the event you get one of the really non-standard Uniques that doesn't count as one of these weapon categories and lean into it, as it's one of the few Traits that improves stuff like the Railgun. This itself has the problem that such Uniques are unable to spawn early, all having a minimum floor of at least 12: by the time one shows up, you've probably already committed pretty heavily to a specific Mastery and very possibly can't swerve effectively to the Unique. So Finesse tends to be just plain a late pick: nice once you're getting around to it, yes, but not prioritized early unless your Mastery specifically calls for it or your actual motive is to get Whizkid online because you want a specific Advanced or Master Assembly. (Or just want to be able to mod your Basic Assemblies)
Hellrunner
Max rank: 2 (3)
Effect: Movement time is reduced by 15%, and 15 points are added to Dodge chance, per rank.
Hellrunner is a lot more powerful than you might intuitively expect.
Hellrunner is a lot more powerful than you might intuitively expect.
So first of all, we need to discuss the Dodge mechanic. The basic idea of Dodge is straightforward, but the actual mechanical execution is fiddly and a bit complicated to accurately describe. The clear concept is that Doom Roguelike is attempting to recapture the aspect of classic Doom where a major part of gameplay is avoiding being hit by ranged attacks by virtue of keeping on the move, where enemies fire where you currently are and miss you because you're not actually there by the time the projectile arrives. However, projectiles in Doom Roguelike never have travel time; if an Imp's fireball travelled, say, 1 tile every 0.5 seconds, then this aspect of Doom would be recreated organically, but projectiles in Doom Roguelike instantly cover infinite distance until they impact something, so Doom Roguelike has to get creative to approximate this aspect of classic Doom combat.
Doom Roguelike's solution is that anytime an enemy chooses to fire, there is a percent chance they will randomly target the tile Doomguy was at on their last turn rather than the tile Doomguy is currently standing in. This percent chance is variable, defined by the weapon the enemy is using (Which matters with the handful of enemies that randomize between weapon types when generated), with a base chance for a Dodge to occur that is increased for each tile of distance between Doomguy and the enemy attacking him. (This distance bonus is also custom for each weapon) I'll be listing these chances on the individual enemies when we get to them, but if you've played classic Doom it's mostly pretty intuitive, aside Archviles. (Doom Roguelike's attempt to represent the Archvile's unique channeled attack is... strange, but that's for later)
Hellrunner then adds 15 points to the final chance per rank (eg if the pre-Hellrunner Dodge chance is 40% and you have 2 ranks of Hellrunner then your final Dodge chance is 70%), capping at 95%. (Exceptions: Mancubi shots never pass the Dodge check, and Dodgemaster's forced Dodge also bypasses this cap) Note that 10% is the minimum Dodge chance, and because enemies refuse to use ranged attacks at point-blank, in practice that's an absolute minimum of 13%; with 2 ranks in Hellrunner, that means a 43% Dodge chance at minimum, and with 3 ranks it rises to 58%! Also, the enemies with that minimal chance are primarily early-game enemies that largely vanish; the Arachnotron's minimum of 24% is often the real lowest-chance-to-Dodge past the early game, which is thus 54% Dodge at minimum with 2 ranks of Hellrunner and 69% at minimum with 3 ranks. Ranks in Hellrunner make moving pretty strongly protective!
Also, the Running state adds another 20 points to Dodge, so Running with 3 ranks in Hellrunner is already a bare minimum Dodge chance of 78%. Note that this bonus is tied very specifically to the Running state, and so Berserk prevents access to it, and healing will put a stop to it.
Big caveat to all this: Former Sergeants and Elite Former Sergeants use Shotguns, and completely ignore Dodge as a mechanic. Since Shotguns also can't miss, the only way for them shooting to not hit Doomguy is in terrain-based edge cases where the AI thinks it has a line of fire but actually doesn't. Similarly, I've already implied this, but I should explicitly state that melee attacks completely ignore the Dodge mechanic.
Anyway, to get into nitty-gritty a bit more, I bolded the bit about enemies targeting where you were during their last turn because this detail is not actually entirely obvious in playing the game and has implications that can catch a player off guard if they don't properly understand this detail. For example, a player might arrive at the reasonable-but-not-quite-correct theory that Dodge triggers based off whether the player's last action was movement. As a concrete example, if Doomguy shuffles left and then shuffles right, arriving at the tile he started our example in, normally what will happen if, say, an Imp is in sight and attacks every time its turn comes about, is that it'll have a chance to throw a fireball at Doomguy's starting position followed by a chance of throwing a fireball at his second position...
... but if Doomguy's movement speed is very high such that he can walk twice in between the Imp's own turns, this can result in the Imp being guaranteed to target Doomguy's current position because Doomguy is standing in Position A during the Imp's first turn, and then sidesteps out and back into Position A before the Imp's second turn, and so whether a Dodge triggers or not the Imp ends up targeting Position A.
As such, when Doomguy's movement speed is quite high, it's generally better to keep walking in more or less a straight line if the goal is to be protected by Dodges, not dance back and forth. Notably, a zig-zagging advance on a target is a good idea at base speed (eg when advancing with intent to melee a target), but if Doomguy is very fast this can result in Dodges still having projectiles coming right for Doomguy!
Also, a wrinkle I should point out explicitly is that Dodge is -aside Archviles and their variants- normally functionally irrelevant if Doomguy's movement is directly toward or away from an attacker, because even if the Dodge check technically results in a Cacodemon targeting the tile behind Doomguy the projectile will still end up on the same course and make an accuracy check against Doomguy. (Unless the projectile ends up hitting something before reaching Doomguy, but this scenario has nothing to do with Dodge as a mechanic)
Also, something that should be pointed out about Dodge checks is that the game's AI clearly performs the check before checking if there is a clear line of fire to Doomguy. This mostly isn't very important, as enemies don't care about friendly fire in the first place (That is, they're perfectly happy to fire at Doomguy even if fellow monsters are in the way and liable to be hit instead) and completely breaking line of sight will (Aside Archviles, Mancubi, and Nightmare Arachnotrons being weirdos) always result in an enemy refusing to attack, but it's very visible when moving to a position that breaks line of fire but does not break line of sight; if the Dodge triggers, the enemy will fire at Doomguy's last position, but if it does not trigger the enemy will instead move in an attempt to get a line of fire. This is most readily seen when ducking around corners, such as into a doorway.
(This is easy to misunderstand, mind, since the AI can think it has a line of fire and be wrong, and the result can potentially look the same as this oddness with Dodge triggering, depending on exact positions)
Also note that, contrary to what you might intuitively expect, the enemy doesn't have to have seen Doomguy at his previous position, and enemies can in fact end up targeting a tile that's on the other side of a solid wall if that's where Doomguy was on that enemy's previous turn!
Anyway, all this fiddlyness with Dodge mechanics is made particularly relevant by the fact that Hellrunner of course significantly increases Doomguy's movement speed. Rank 2 Hellrunner already has Doomguy take only 0.7 seconds to walk a tile, and rank 3 means he only takes 0.55 seconds to walk a tile; with just a little bit more help to his speed (eg Tactical Boots) he can easily end up moving more than twice as fast as base movement!
The movement speed part of Hellrunner is itself quite valuable. First of all, there's a surprisingly large number of enemies that are just a little bit faster than Doomguy; a Hell Knight, for example, has a base speed of 110%, meaning by default roughly every 10 seconds the Hell Knight will slip in an extra turn. If Doomguy is at base speed, this means that sometimes Doomguy will walk forward one step, gain sight on a Hell Knight, and be attacked twice with no chance to respond! Taking a rank in Hellrunner takes away this kind of possibility... before taking into account the fact that Armor tends to slow down Doomguy, mind, so it's a little more complicated than I'm making it sound.
More generally, ranks in Hellrunner -and otherwise increasing Doomguy's speed- reduce how often enemies get the jump on Doomguy. If Doomguy is at base speed, then even base speed enemies have a roughly 50% chance of getting to attack without warning when they first enter line of sight. (The coin flip being 'did Doomguy enter their line of sight with his move, letting them immediately attack, or did Doomguy move and then the enemy walked into his line of sight') If Doomguy has two ranks in Hellrunner (and no other speed modifiers), then walking a tile takes 0.7s, meaning the odds of a base speed enemy getting the jump on Doomguy have been slashed significantly; this is very noticeable just playing the game, where walking into enemy sight gets punished much less consistently if Doomguy's speed has been bumped up.
Also worth noting is Pinkie Demons, who are melee-only enemies with a base speed of 130% and so at default speed even Running is not enough to actually increase the gap between them and Doomguy. Early ranks in Hellrunner can be a lifesaver if Doomguy encounters a blob of Pinkie Demons and for whatever reason can't simply blast them before they get into melee, letting him start Running, create some space, and resume fire. (Especially if wielding a Shotgun, where the accuracy penalty from Running is irrelevant)
It's also worth pointing out that certain events are running relative to the game's overall clock, where covering more ground in less time can be huge when these events come up. For example, one Special Level and a rare Level Feeling involve the floor being flooded by deadly fluids: the time pressure created by such is much lower if Doomguy spends less time per move. This effect is pretty easy to underestimate: in playing the game, the player's time tends to be eaten up pretty heavily by combat, with pure movement finishing up really quickly, but Doomguy's time is often eaten up much more heavily by covering ground: say you walk into the darkness for five seconds, and then spend 30 seconds working through a fight. This can easily be that Doomguy spent 20-30 seconds on walking and only 15 seconds on fighting if you've got no modifiers: ranks in Hellrunner thus tend to save Doomguy's time by a lot more than other time-savers, even when those other timesavers are much more impactful on the player's time spent on the game.
And of course Hellrunner is hugely helpful to melee builds, letting Doomguy get into reach quicker to start killing enemies so they stop hitting him.
In the very early game improvements to Doomguy's offenses tend to be more useful -especially up on higher difficulties where getting swarmed is more common at every stage of the game- but Hellrunner is overall probably the most reliably great Basic Trait, which you basically always want maxed at some point. (Unless your chosen Mastery blocks it, anyway)
One final aside about Dodge: for some reason, enemies can Dodge! This is easy to overlook (Your own Shotguns ignore it, just like enemy Shotguns, for one), and if you know about it is easy to make essentially untrue: the game only makes performs a Dodge check on Doomguy's shooting action if the shooting action was actually targeted on a specific enemy. (Right-clicking an enemy's tile count) For the most part, there's no need to do that: aim in front of behind your target, and Dodge is irrelevant...
... but there's some wonky exceptions to this generalization. I'll point them out as we get to them.
It's also worth pointing out that Dodge can have really bizarre results when teleportation gets involved, where attacks get aimed in completely the opposite direction the intended target is at because it was last dozens of tiles behind the attacker. This looks really ridiculous when multi-hit attacks have it trigger...
Tough as Nails
Max rank: 2 (3)
Effect: 'Internal' Protection stat is increased by 1 point per rank.
'Internal' Protection is unusual in that it applies to both conventional attacks and damage taken from Acid or Lava on the floor. If you're grabbing Tough as Nails ranks early, this can make it more practical to wade through Acid, potentially without even bothering to Run if you get Protective Boots early or are willing to slap a Power Mod on Steel Boots.
That said, that's not a primary draw of Tough as Nails, and you shouldn't be planning as if it will be an important piece in letting you wade through the Lava in Mount Erebus/the Lava Pits, or anything like that. The primary point is to help protect you from enemy attacks -emphasis on help, as Tough as Nails is not, by itself, a particularly significant boost to your survivability against much of anything. Stacked atop Protection provided by Armor and/or combined with relevant resistances, though, it can be knocking damage down to 1 point instead of 3.
On that note, it should be pointed out that in Doom Roguelike damage cannot normally be lowered below 1 point. In conjunction with how low max HP actually is, how Doom Roguelike is very fond of escalating enemy count as part of rising difficulty/danger, and the fact that some enemies outright fire multiple projectiles a shot, you can lose a lot of HP fairly rapidly if you try to rely primarily on passive defenses.
This damage minimum is part of why Tough As Nails won't let you effortlessly wade through Acid or Lava: even if your feet have more Protection than the floor has damage, Doomguy will still be losing 1 HP every single action, and so by default will die in at most 50 actions.
Before 0.9.9.8 added resistances to Ironman, Tough As Nails was pretty clearly the better choice for improving Doomguy's survivability. Now, it's probably the weakest, or at least second-weakest, of the Basic Traits. Among other points, the biggest rapid-fire damage threats in the game all do Plasma damage, and so substantially ignore Tough As Nails...
Son of a Bitch
Max rank: 3 (5)
Effect: Each individual hit on all attacks gains +1 damage per rank.
Son of a Bitch is also secondarily oriented toward Pistols, but due to Son of a Gun being flatly better at that this takes a bit to be relevant; you shouldn't be taking ranks in Son of a Bitch on the idea of boosting your Pistol performance unless either you can't take any more ranks in Son of a Gun at that point, or the idea is that you're using Pistols right now but don't expect to stick with them. (eg you're intending to go for your Rapid Fire Mastery but haven't found a Chaingun yet)
A similar-ish point is that if you're intending to go melee, Son of a Bitch is really inferior to Brute, and only maybe makes sense to take ranks in if you're currently capped on Brute ranks.
I should also point out that Son of a Bitch's damage is simply added directly to final damage, not made part an attack's dice-rolling. This makes weapons with wide damage ranges more consistent (A roll of 1-6 damage backed by two ranks of Son of a Bitch will become 3-8 damage, where the lowest possible result is 38~% of the highest possible result, rather than 16~% percent of the highest possible result), and is important for correctly assessing the actual impact of Son of a Bitch. (eg 1-6 damage averages 3.5 damage, and 2 ranks of Son of a Bitch raises the average damage to 5.5: that's almost a 60% increase in average damage, even though +2 damage is only a 33% increase to max damage)
Son of a Bitch also, it should be noted, tends to drop off in impact as a run progresses and Doomguy starts getting a hold of stronger weapons. Better weapons are more prone to raising damage per-shot than they are raising number of shots fired, and so Son of a Bitch's flat damage boost becomes a smaller proportion of damage added when compared against, say, Finesse.
On the other hand, Son of a Bitch is also an improvement to ammo efficiency. Ranks in Finesse kill enemies faster, but aside occasionally saving ammo by virtue of killing an enemy before it can use a Medpack to heal itself, you end up spending the same ammo on killing a target, whereas Son of a Bitch can actually reduce how much ammo you burn on killing enemies. How important this is depends on things like which Mastery you took, but unless you Assemble a Nanomanufacture weapon only melee weapons don't have ammo efficiency as a relevant concern.
Something wonkier to keep in mind is knockback thresholds. Bumping your damage up can create problems if it's high enough to keep knocking enemies back, causing chaos that messes up your plans, and in Angel of Max Carnage in particular can result in your damage going down. (Because your Chaingun reliably knocks enemies out of sight with its first shot and then probably misses at least one of its next three shots due to the 'firing into the darkness' penalty) The tuning on numbers is such that this is largely not a concern (A Chaingun needs +4 damage to have a chance of doing knockback; this can't come up without one or both of extended skill ranks and/or Power Mod application), but it's worth keeping in mind, and, again, if you're doing Angel of Max Carnage it's really important to maintain awareness of.
Son of a Gun
Max rank: 3 (5)
Effect: Pistols take 20% less time to fire, and do one more point of damage per shot, per rank.
Notice that this is Son of a Bitch rolled together with a superior version of Finesse. If you're going to all-in on Pistols, there's no reason to bother with either of those Traits unless you're unable to gain more ranks in Son of a Gun... and Finesse in particular is a really bad investment in runs that expect to hit extended ranks, because Son of a Gun 5 knocks your Pistol firing speed down to the absolute fastest the game allows of 0.1 second all by itself. (If it weren't for this enforced minimum, rank 5 Son of a Gun would in fact make firing a Pistol take literally no time at all) Finesse does literally nothing to your Pistol fire rate at that point.
Son of a Gun is, itself, pretty straightforward of a Trait as far as strategic planning goes. Either you're intending to commit to Pistols and should always prioritize Son of a Gun over other Traits, or you're not intending to commit to Pistols and should just ignore it. Son of a Gun is a prerequisite to an Advanced Trait, but said Trait is also a hard-specialty in Pistols so that just reinforces the 'are you specializing in Pistols or not?' dynamic; there are Basic Traits it can make sense to invest in even though they're personally worthless to your build because you want the Advanced Trait they unlock, but Son of a Gun is not one of them.
Which means I don't actually have a lot to say about Son of a Gun itself, in terms of player-perspective stuff.
From more of a design perspective, I quite appreciate Son of a Gun: by default in Doom Roguelike, Pistols are designed just as the Pistol is in classic Doom, which is to say they're designed to be that weapon you start with that is largely inferior to the alternatives so you wish to displace it as soon as possible. Giving the player a realistic ability to specialize in Pistols thus really requires fairly enormous support, and Son of a Gun delivers: rank 3 of Son of a Gun means 250% as much shooting while increasing expected per-shot damage by around 60% (For a total DPS equal to 400% of base), while rank 5 means 1,000% as much shooting for roughly double expected damage, ie 2,000% DPS relative to base.
I've seen other games that tried to let the player specialize in The Basic And Bad Weapon Type, and it's frustratingly normal for the math behind such an attempt to be very obviously inadequate: if The Basic And Bad Weapon has been successfully designed to be truly so terrible, this automatically implies that options for specializing in it have to be outrageously powerful to compensate. Games do typically recognize that a deliberately weak weapon needs stronger support to compete with good weapons, but for whatever reason it seems to be normal to not recognize the 'opportunity cost' aspect.
That is, if I take 3 ranks in Son of a Gun, and then compare my damage output to a Chaingun, the correct comparison point is not 'a basic, unsupported Chaingun', but rather is 'a Chaingun backed by 3 ranks of relevant Traits'. Since a Chaingun has a better foundation than a Pistol (4-24 damage per volley vs 2-8 damage per Pistol shot), it can stay ahead off of weaker boosts: say Doom Roguelike had made the Pistol specialty add 20% of base DPS per rank, while some more generic trait added 10% of base DPS to the Chaingun, with both these hypothetical Traits having 5 ranks. At rank 5, Pistols would rise to 4-16 DPS off their +100% damage, while Chainguns would rise to 6-36 DPS off their +50% damage: in this imaginary version of Doom Roguelike, Pistol specialization wouldn't make any sense to actually pursue! And this kind of math is pretty common when games attempt this sort of concept.
So I very much appreciate Doom Roguelike being willing to make Son of a Gun such a drastic performance boost. It's also impressive how it stays true to classic Doom's design as the starting point while using the level-up system to open up the option of deviating and making the humble Pistol your best weapon; in general, a big part of what I like about Doom Roguelike is that it successfully balances 'meaningfully parallels classic Doom in its dynamics' with 'works within a roguelike framework'. I always like this kind of thing when I see a game be adapted into a radically different format, but generally speaking I tend to feel most such adaptations either compromise their own gameplay by too heavily prioritizing faithfulness, or are too quick to discard faithfulness such that the final product being ostensibly an adaptation feels basically dishonest. Doom Roguelike managing the balancing act as well as it does is a rare thing to see.
Reloader
Max rank: 2 (3)
Effect: Reload time of all weapons reduced by 30% per rank.
This is most heavily a specialization in Shotguns, since by default they alternate firing and reloading while taking at least as long to reload as to fire: thus, for several Shotguns Reloader is a more significant boost to overall fire rate than Finesse provides. This is particularly emphasized by the fact that two ranks of Reloader unlocks Shottyman, an explicitly Shotgun-supporting skill. (Albeit one that noticeably reduces Reloader's value, which is a bit awkward)
That said, only melee weapons get zero benefit, and Pistols in particular tend to be pained by their reload time once you've got multiple ranks in Son of a Gun. For example, if you have rank 3 in Son of a Gun, a basic Pistol will take 2.4 seconds to unload its clip and then 1.2 seconds to reload, and if you have rank 5 in Son of a Gun the Pistol will take 0.6 seconds to unload its clip and then 1.2 seconds to reload; a rank in Reloader knocking that reload time from 1.2 seconds to 0.84 seconds is a pretty appreciable improvement when reloading accounts for 1/3rd or 2/3rds of your firing cycle!
It's also worth pointing out that rapid fire weapons tend to have long individual reload times: their proportion of time spent reloading is generally low (A basic Chaingun spends 10 seconds firing and then 2.5 seconds reloading by default), but the long individual reload time can be quite deadly, where you think you're safe, start a reload, and then one or more enemies enters sight and attacks once or twice before you have a chance to respond. Reloader reducing the odds of that kind of outcome can be a pretty big help, especially if your run is just not managing to get a strong defensive setup built.
Eagle Eye
Max rank: 3 (5)
Effect: +2 Accuracy with all attacks per rank.
But how does Accuracy actually work in Doom Roguelike? The game itself never explains the system, unfortunately...
... but the wiki gives the following description:
To determine if the attack hits, the game rolls 3d6. If the accuracy is greater than or equal to the roll then the attack hits, otherwise it misses. There is one exception to this rule: if the roll is 3 or 4 then the attack always hits, and if the roll is 17 or 18 then the attack always misses no matter how high the accuracy is.
It further states the default target number is 10 and so base accuracy is 50%... which means something about its explanation is incorrect, because 'greater than or equal to' would produce 50% accuracy if the target number was 11. 10 as the target number would produce 50% accuracy if 'greater than' was strictly held to. I'm unsure where the incorrectness of the explanation lies -if it's that the 'equal to' part is wrong, then everything is fine aside that this handful of words shouldn't exist. If instead the description of the formula is fully accurate, then instead the conclusion is wrong and all accuracy in-game is greater than what the game claims by 1 step, where base accuracy is actually 62.5% and +4 accuracy is already maxing out hit chance rather than requiring +5 accuracy, and so on.
You can use this site to readily see the numbers yourself. It's also worth noting that the wiki and in-game numbers truncate the stated accuracy and so are consistently not quite accurate to the true odds. (For example, the wiki will inform you that minimum hit chance is 2%, but it's actually 1.85%) Given it's always less than a percent of difference this isn't a big thing, but it's worth pointing out regardless.
Anyway, leaving aside the question of what the wiki has wrong here, have a table followed by some implications! (Note that the following table is assuming the 'just greater than' scenario; if 'equal to' is in fact how the game works, then treat all accuracy scores as 1 step above what they're stated to be)
-6 4/216 1.85%
-5 10/216 4.62%
-4 20/216 9.25%
-3 35/216 16.2%
-2 56/216 25.92%
-1 81/216 37.5%
0 108/216 50%
+1 135/216 62.5%
+2 160/216 74.07%
+3 181/216 83.79%
+4 196/216 90.74%
+5 206/216 95.37%
+6 212/216 98.14%
First, notice that Accuracy increases have significant diminishing returns in Doom Roguelike. Going up 2 steps of Accuracy from the baseline value raises your chance to hit from 50% to 74.07% -slightly over a 48% increase in hit chance. Going from +4 Accuracy to +6 Accuracy is raising your chance to hit from 90.74% to 98.14%. That's an improvement, but only about an 8% increase in how many shots hit, even though in both cases you added 2 points of Accuracy.
Second, notice that +6 Accuracy is the cap on benefits, yet Eagle Eye can give up to +10 with extended Trait ranks. Its extended ranks are bad! It's worth emphasizing here that no weapon in the game has negative base Accuracy: there are Assemblies that can lower Accuracy, but most of them only lower it by 2 points, and the one exception is exclusive to a melee weapon, when Eagle Eye is worthless to melee builds because Brute supersedes it. As such, the fifth rank in Eagle Eye is really difficult to get actual use out of -almost impossible if you're unwilling to fire blindly into the darkness- and even the fourth rank is difficult to leverage to any real extent, since the maximum aim penalty while firing with line of sight is only -2, and almost all (non-melee) weapons have at least +1 base accuracy, with most of them having at least +2. The Chaingun and Plasma Rifle are the least accurate common-tier weapons at +2! And most Exotics and Uniques go up in Accuracy; the Minigun dropping to a mere +1 is anomalous, and pretty erratic in its relevancy to a run to boot.
I really wish the extended ranks of Eagle Eye did something else entirely. If the Sniper Mod Pack benefits were provided by Eagle Eye 4 and 5, that would open up interesting options for runs hitting extended ranks, and make those ranks of Eagle Eye a lot less of a trap for ignorant newbies who haven't looked up spoilers and collated all the Accuracy numbers and realized its extended ranks rarely have a chance to do anything at all.
Brute
Max rank: 3 (5)
Effect: Melee attacks get +3 damage and +2 Accuracy per rank.
The hard melee specialty. Notice that it's worth 1 rank of Eagle Eye and 3 ranks of Son of a Bitch; there's no reason to bother with either when focusing on melee unless Brute can't get any more ranks... and Eagle Eye never makes sense to take, since Brute 3 is already maximum accuracy essentially all the time.
Brute itself is very straightforward to talk about. Generally, either your run is a melee run and it's a priority, or your run is not a melee run and you should ignore it. The main qualifier here is the Unholy Cathedral: if you're doing a non-melee run but still want to take on the Unholy Cathedral (Maybe you want to do a run that's truly 100%), you'll probably want ranks in Brute in the mid-late game in preparation for that, if only the two ranks necessary to unlock Berserker.
I should more explicitly point out that all melee weapons have a natural base Accuracy of +0 and therefore have only a 50% chance to land a hit. This is one reason why melee weapons tend to be dubious to break out without ranks in Brute: you need either Brute or Eagle Eye ranks for melee to be reasonably reliable. Preferably, you have 3 ranks in Brute, so you're at the Accuracy cap: 1 rank of Brute still has you missing about 26% of the time (Which is way too unreliable), and even 2 ranks has you missing over 9% of the time, which is tolerable but still means you'll miss regularly in extended fights.
This does in turn mean that melee weapons become noticeably more appealing as a backup weapon if you're taking ranks in Eagle Eye: it's better to have Brute's damage bonus and especially to have gotten to Berserker, but melee weapons mostly have pretty good base damage in the first place. Using the Chainsaw to rip through Former Humans without expending ammo can be a worthwhile use-case to keep in mind, especially if you've gone for an ammo-hungry build.
Advanced Traits
None of these have extended ranks. Only four of them have more than one rank at all, in fact!
Juggler
None of these have extended ranks. Only four of them have more than one rank at all, in fact!
Juggler
Requirements: Finesse 1
Max rank: 1
Effect: Switching weapons in any way no longer takes any time. Attacking a target in melee while a melee weapon is in the Prepared slot automatically attacks with the Prepared weapon as if it was in the Weapon slot (Instead of using the ranged weapon in the Weapon slot), at no additional time cost.
Before 0.9.9.8, I really didn't like Juggler. It had three effects: it made swapping between your Weapon and Prepared slots take 0.1 seconds, it had the Prepared melee weapon effect it currently has, and it made Quickkey swaps take 0.1 seconds. This latter one... well.
Okay, Quickkeys were a cute callback to weapon hotkeys in Doom, and... a pretty terrible idea to attach mechanical significance to given this is a turn-based game and all. In short: prior to 0.9.9.8 the number keys were assigned to various weapons where pressing the appropriate key would, if you had a copy of that weapon in your inventory, swap to that weapon without needing to hassle with going into your inventory. Where manually swapping to a weapon takes 1 second by default, a Quickkey swap took 0.8 seconds by default, so it was mechanically optimal to only ever switch to a weapon via Quickkey input. Juggler exaggerated this still further by knocking the swap time to 0.1 seconds, ie to the closest to instant the game is normally willing to get.
I really didn't like this since it was basically just a stealthy punishment for a learning player, where you'd start out not knowing this is a thing and magically pick up a bit of mechanical time efficiency in-game not because you made a good build decision or anything but just because you became aware that a convenience feature exists and/or became aware it's actually mechanically significant rather than just a convenience feature. It was especially frustrating since one of the nice things about Doom Roguelike is that in most regards it's equally supportive of almost-pure mouse play and no-mouse play; it's a smoother experience to use both, but for the most part it's not important to use both, making it an unusually accessible experience... and then Quickkeys notably undermined this, where you were punished for not making heavy use of the keyboard, especially if you ever took a rank in Juggler.
Thankfully, 0.9.9.8 did away with this system -you can now manually assign hotkeys to specific weapons, instead- and in the process made Juggler simply upgrade all weapons swaps super-fast, and indeed upgraded these 'instant' swaps to actually be instant; you can in fact sit in front of thirty Barons, entertaining them with your circus skills forever in total safety. (By which I mean you can swap weapons dozens or hundreds of times if you feel like) This is very much appreciated, especially if you like to make use of a decent array of weapons, making it more practical to use one weapon for certain situations, as you can easily swap it in and out safely as necessary.
It's also worth pointing out that Juggler can make it more optimal to swap weapons, rather than reloading a weapon, once a weapon runs out. This is very worth being aware of very early in a run if you get Juggler so early, as there's a decent chunk of the early game where you have plenty of empty inventory slots: instead of firing your Shotgun, having your target live, reloading and probably being attacked, and then finishing them, you just switch to a loaded Shotgun instantly and blow away your target before it can respond. Even later in a run it can be worth keeping in mind, especially if your run is having odd luck with drops in any number of ways: not enough Medpacks generating to clog up your inventory, not getting non-basic weapons generating for a weirdly long time, having two Assault Shotguns generate close together... all kinds of things can result in it being dramatically more efficient to swap instead of reload, especially if you're not taking ranks in Reloader.
Oh, and there's the attack-with-prepared-melee-weapon thing. I never rated this as high-value effect, but with all swaps being truly instant, it's now just a minor convenience feature, since you could get the exact same outcome by manually swapping in your melee weapon. You get it for free, so whatever.
Berserker
Requirements: Brute 2
Max rank: 1
Effect: You can now spontaneously Berserk in two ways: the first way is by attacking in melee, where every 4 attacks landed will trigger 20 actions worth of Berserk time, up to a maximum of 100 actions of Berserk stored. The second is by being hit by an attack whose base damage is at least 10 and which would remove at least a third of your maximum health; this is before accounting for resistances and damage reduction, and the Berserk it triggers applies its resistances before you actually take the damage. In this case, you get 20 actions of Berserk by default, up to a maximum of 100 actions of Berserk, with diminishing returns the more actions of Berserk you already have stockpiled.
As of 0.9.9.8, it now runs off 'nominal HP', which... I'm not clear what that means, but supposedly Ironman and the Marine's innate HP advantage no longer affect the Berserker threshold for damage, removing that newbie trap. (ie it should now trigger when the original damage roll is 17 damage, regardless of your actual max HP) Which is a huge relief. (Unless it was miscoded in some problematic way I haven't noticed yet, anyway) This particularly irons out some odd jank regarding specific enemy/class threat profile intersections: in 0.9.9.7, a Berserker Scout or Technician was usually in less danger from Barons of Hell than a Berserker Marine, because a Marine could get Berserker triggering off a Baron's ranged attack, but only if it rolled literally max damage per se, whereas a Scout or Technician would Berserk off as little as 17 damage, which thanks to the Baron ranged attack using 4 dice was vastly more probable than them rolling 20 damage. (20 damage per se is a 0.16% chance to happen. 17 or more is a 5.6% chance: 35 times more likely!)
Anyway, as for the Berserk status' actual effects, they are threefold.
Firstly, Doomguy's melee attacks double their final damage before the target's Protection and Resistances reduce it. (That is, roll damage, add flat modifiers like Brute, then double it via Berserk, then start on damage reduction from the target's defenses)
Secondly, it adds +60% to all of Doomguy's resistances. This is, to be clear, simply added to other resistances, not a further multiplier; if Doomguy has no other resistances, Berserk effectively extends Doomguy's HP by 250% because he's only taking 40% of what he'd otherwise take. If instead Doomguy already has 20% resistance to an incoming attack (Such as because he's a Marine and it's Fire, Plasma, or Acid), then Berserk is raising that resistance to 80%, and so Berserk triggering is effectively quadrupling Doomguy's durability. (Because instead of taking 80% damage he's taking 20% damage, which is a quarter of 80%) Note that resistances being added together is not allowed to produce a higher result than 95%; Berserk is in fact the main context this comes up in. Regardless, note that Berserker's effectiveness is functionally raised by having other resistances unless the non-Berserk resistances are already very close to that cap of 95%. (eg having 90% resistance would result in Berserk only doubling Doomguy's durability, because it would knock damage taken from 10% to 5%)
Thirdly, it causes Doomguy to generate 50% more Energy per game tick. This functionally means Doomguy gets turns 50% faster when Berserking than when not, and it effectively chain-multiplies with other action speed modifiers since it's not an additive modifier of their effects. (That is, if you have Finesse 2 so you attack in 70% of normal time, and you're Berserking, what happens is you spend 700 Energy on attacks and generate 150 Energy every 0.1 seconds, and so get a turn after 0.5 seconds by virtue of having generated 750 Energy, instead of getting a turn in 0.7 seconds by virtue of generating 100 Energy every 0.1 seconds) Since it is a modification to Energy generation, it affects all actions, and if you're really fast it can partially bypass the time minimum of 0.1 second in non-obvious ways.
Say you have Son of a Gun 5, so you fire your Pistol in 0.1 second. If you're Berserking, what happens is you spend 100 Energy, then 0.1 seconds pass and you gain 150 Energy and get another turn. Say you repeat this 4 times: now you're 200 Energy over what's necessary to actually get a turn. If you perform an action that's supposed to occur in 0.3 seconds, you'll get your next turn 0.1 seconds later, because you already had 0.2 seconds of Energy in storage!
Anyway, Berserker as a Trait is most obviously a melee specialty, what with being able to trigger Berserk specifically off of melee attacks and the Berserk status itself being hugely advantageous to melee in particular, but is actually more generally beneficial than you might expect. If it weren't for the demand that you invest 2 points in Brute to unlock Berserker, Berserker would be very straightforwardly a great unlock for every run that didn't block it off via Mastery choice. Even with the Brute requirement being a modest toll for non-melee builds, Berserker is still very worth considering going out of your way to get, especially in Angel of 100/Archangel of 666 runs where you expect to end close or at the level cap and so have more skill points to spare on 'dead weight' Brute ranks.
This is because quite a few late-game enemies can trigger Berserker at range, with Archviles in particular standing out because their damage is reliable and so they always trigger Berserker with their ranged attack. Doing everything 50% faster while taking far less damage is great no matter what you're doing, and crucially Berserker is something of a shield against bad luck: if a late-game enemy rolls high on its damage roll, instead of this potentially killing you when you thought you had enough HP to survive, you actually take less damage than if they'd rolled closer to their average (A Revenant's average damage is 12.5. Their maximum damage is 25. If they roll max damage when you have no Protection or innate Fire resistance, Berserker activating will reduce the hit to 10 damage!) and will in fact be supercharged and able to take on all comers more readily for the near future.
This gets especially dramatic and reliable as you climb difficulties and swarms of enemies are more common: having Berserker might mean the difference between descending and almost unavoidably dying to a horde of enemies that are right on top of you vs shredding them all with minimal HP lost!
Berserker is also important if you care about a specific secret thing, but I'll talk about it more explicitly when we get to that secret thing.
Dualgunner
Requirements: Son of a Gun 2
Max rank: 1
Effect: If your Weapon slot and Prepared slot both contain Pistols, you will automatically fire both at the same time, taking 20% longer than firing either one. (Different firing times will be averaged together and then raised by 20%) Dual-firing will always take at least 0.12 seconds, even with Son of a Gun 5. Additionally, if both your equipped weapons are Pistols, then switching between them occurs instantly.
Note that most Pistols have their Alternate Reload be Dual Reload, but not all of them do: in such cases you ideally have a Dual Reload Pistol sitting in your Weapon slot, as Dual Reload will work even if the Prepared slot doesn't have Dual Reload, so long as they're both Pistols and neither of them recharges ammo for free.
Like Son of a Gun itself, I don't really have a lot to say about Dualgunner: it's pretty binarily either that you're a Pistol run and want it or not a Pistol run and don't want it, aside the odd case of the Technician's Pistol Mastery giving you reason to consider ignoring it.
The one thing I do have to say about it is that 0.9.9.8 made an unmentioned quality-of-life improvement: in 0.9.9.7, the 'instant' swap Dualgunner provided was actually an 0.1-second-long swap, not its current truly instant swap. This is especially nice if you prefer to minimize keyboard usage and rely nearly solely on the mouse; it's now mechanically fine to just reload, swap, and reload in place of using the Dual Reload function. Indeed, there's an argument to be made that now that is the optimal thing to do, since it lets you respond to stuff happening partway through the reload! (Except when under an action-timer effect, mind: you should still use Dual Reload when Berserking, for example)
Dodgemaster
Requirements: Hellrunner 2
Max rank: 1
Effect: Your first Dodge attempt per action automatically succeeds.
This sounds amazing, but is actually severely hamstrung by the very fact that it requires Hellrunner 2. Let's say you want to Dodge a Baron of Hell's acid ball: since you need at least one tile between you and the Baron for them to use a projectile at all, you have a 53% chance to Dodge before other modifiers. Hellrunner 2 then adds 30 points, putting us at 83% Dodge if you're basically right in the Baron's face. More realistic is that you've got 5 or more tiles of distance between you and the Baron, giving you an at least 92% chance to dodge. At that point, Dodgemaster has only rarely done anything at all -and it gets worse if you're running Tactical Boots/Armor (+10 to Dodge from either), Run in combat regularly (+20 Dodge), or get into extended ranks and take Hellrunner 3. (+15 more Dodge)
Now, the fireball-lobbers have the most generous Dodge formulas, starting from a base of 50% for you to Dodge, while other enemies start from lower chances... but Archviles have more generous distance-based modifiers so if you have Hellrunner 2 you'll generally be hitting the cap of 95% Dodge chance anyway, and various other enemies are either non-threats that are phased out in short order (eg basic Former Humans) or multi-shot attackers where Dodge is unreliable in general and Dodgemaster in particular can never protect from an individual turn of attacking. (eg Former Commandos, Arachnotrons) Or are enemies where Dodge bonuses are irrelevant for one reason or another. (Dedicated melee, Mancubi, Pain Elementals)
In practice, Dodgemaster is only reliably significant against Revenants, Cyberdemons, and Elite Commandos equipped with Rocket Launcher variants: these are all attacks with low base Dodge chances that fire a single shot... oh, but they're also all splash damage attacks, and in the Revenant's case the projectile stops on exactly the targeted tile, where a Dodge isn't adequate to avoid the attack even in the open unless you're fast enough to move twice between Revenant attacks. Which generally means combining Hellrunner ranks with Tactical Assemblies, returning to the issue that those reduce the utility of Dodgemaster.
This isn't even getting into the issue that Dodgemaster's utility goes down the more enemies have a line of fire on you, and for the most part Doom Roguelike defaults to trying to seriously threaten you by throwing groups at you. 100% reliably Dodging a single attack can let you solo a lone Baron. It barely matters if you descend the stairs and find yourself in sight of a pack of 9 enemies.
This is all particularly killer to Dodgemaster in Angel of 100 and Archangel of 666 runs. In standard runs, and most challenge runs, you're at least guaranteed to get a comparatively early encounter against a solo Cyberdemon, where it's the only threat on the floor, alternates firing and reloading so you can reliably move and then shoot between its shots, and is very dangerous: Dodgemaster can be a run-saver in that particular fight, and thus potentially well worth spending a level on. In Angel of 100 and Archangel of 666, there's nothing like that to guarantee its utility, and indeed the mechanics of floor generation mean it's increasingly normal to be getting swarmed in such runs, increasingly depressing Dodgemaster's value.
But even in a standard run, Dodgemaster often ends up being something of a trap choice, sounding much better than it actually is.
To be fair, this is also difficulty-dependent. On lower difficulties, you're more likely to have stuff happen like encountering a lone early Hellknight, or a lone Baron of Hell in the midgame, with the floor being empty enough you can plausibly dance around to keep triggering Dodgemaster. But up on Ultraviolence, where I largely prefer to play at, the swarm issue pretty heavily drags it down.
I'm not sure what a good solution would be, but Dodgemaster could really use a mechanics overhaul of some kind.
Intuition
Requirements: Eagle Eye 2. No requirement for Scout class.
Max rank: 2
Effect: First rank causes Powerups to be visible at all times, from anywhere on the floor, and provides general descriptions of Lever effects. ('Beneficial', 'Neutral', or 'Dangerous') Second rank changes Lever descriptions to be precise, and reveals the position (But not the type) of enemies that are nearby but not currently visible, in addition to Powerups remaining revealed. The enemy-position-revealing effect is defined by sight radius, plus 2 tiles beyond that, except that it ignores line of sight obstructions.
For the Marine and to a lesser extent the Technician Intuition is unfortunately much harder to justify putting points into. The need to tech through Eagle Eye -which is itself a low-value Trait- delaying getting Intuition online means you'll often be past the portion of the game in which reading Levers can be what saves a run (Because you're willing to use a Lever and so get healing, for example) or prevents its doom. (Because you don't pull a Lever that would kill you with lava, among other possible dire fates to avert) The Marine especially has a hard time justifying Intuition, since only their Survivalist Mastery doesn't block Intuition, directly or by blocking Eagle Eye. (And Survivalist is probably their weakest Mastery within a standard run)
For a Technician it's a bit better, in that while two of their Masteries block Eagle Eye (And thus block Intuition), Sharpshooter actually requires Eagle Eye and Scavenger outright requires Intuition 1. Squeezing in Intuition 2 isn't a big leap from either of those, and with Scavenger it can in fact be done before the Mastery is online without delaying the Mastery any. Entrenchment is still unlikely to be able to fit Intuition in within a standard run, though, not unless making the strange decision to skip Whizkid (What kind of Technician skips Whizkid, though? It's most of their class identity!), but that's still a couple Masteries it makes sense for.
Of course, both the Marine and Technician can ignore Masteries entirely, and have decent arguments for doing so given how awkward the design of several of their Masteries is. If you're skipping Masteries, then only the awkwardness of needing Eagle Eye remains... and Eagle Eye 2 is actually good for rapid-fire weapons, so a Mastery-less Technician going for rapid-fire weapons can fit in Intuition 2 relatively easily.
Anyway, do note that by default the game's convenience behavior of automatically stopping your movement upon spotting enemies does not apply to enemies detected via Intuition. If you intend to use Intuition to snipe enemies in the dark a lot, you'll need to either mess with the config file to change this (... I'm actually not sure the config file has such a setting. I've yet to get such a thing working myself, even though I've seen people say things that imply it's possible) or get used to moving only a tile or two at a time if there's any possibility of enemies nearby so you can actually get the jump on them. If you're not willing to do either of those things, the second rank of Intuition is a lot less useful.
Also, Intuition 2 is probably a bit less amazing than I'm making it sound, as 0.9.9.8 reduced its effectiveness: it previously detected enemies up to 3 tiles past vision radius, not 2 tiles. This writeup was written primarily off my 0.9.9.7 experiences.
On the other hand, at least in the graphical version, 0.9.9.8 fixed a frustrating bit of jank: in 0.9.9.7, the effects that imposed a visual filter on the screen (ie Berserk, Invulnerable, and Envirosuit Packs) would make Intuition markers not display at all when they were placed in unmapped tiles (I assume they were displaying, but being rendered the same color as the black tiles), and were still really easy to overlook even in mapped tiles. In 0.9.9.8, Intuition markers always display correctly, and are easier to see through the filters in general.
Actually, in general 0.9.9.8 made the visual filters much better, a fact that the patch notes don't communicate at all. In 0.9.9.7, the Nightmare enemy set had their own visual filter overrule such filters, which was sort of convenient but clearly not intentional, the filter effects even interfered with UI visual clarity, and the nature of their color modification made it easy to overlook items and enemies of all types. (Aside Nightmare enemies) The new approach to the visual filters does less 'washing out' of colors, with notable changes like Berserk flipping green and blue colors instead of just rendering them all in shades of red; this can be a bit unfortunate as far as spotting a Green Armor and mistakenly assuming it's a Blue Armor because you're Berserk, but among other points it includes Doomguy's helmet visor being rendered in bright green so it's a lot harder to lose track of your own position. This is by far one of my most appreciated improvements brought on by 0.9.9.8, and it annoys me that reading the patch notes will lead a new player to have no idea that this piece functioned poorly in 0.9.9.7.
Whizkid
Requirements: Finesse 2. No requirement for Technician class, and for Technician gets an extended rank, capping at 3 starting from Level 12.
Max rank: 2
Effect: Each rank allows 2 more Mod Packs to be added to ranged weapons, and 1 more Mod Pack to be added to Armor, Boots, and melee weapons. The first rank also allows you to make Advanced Assemblies, while the second rank lets you construct Master Assemblies. The second rank also allows you to add 1 Mod Pack to Assemblies. The third rank lets you add a second Mod Pack to Assemblies.
Whizkid's nature makes its actual implications difficult to intuit, because it's purely tied up in the mechanics of Mod Packs and Assemblies, which is a big system to master all by itself. Mod Packs and Assemblies will in fact be several posts unto themselves!
Still, a broad overview to start, to give at least some context.
Mod Packs are, firstly, not immediately relevant; a Technician can apply literally their first level to Whizkid, and it's understandable to suspect this might be a desirable thing to do, but no. Mod Packs literally can't spawn in the earliest floors, and Whizkid has zero relevance until you have at least two Mod Packs. Even with the Technician starting with a Technical Mod Pack, that still means Whizkid doesn't matter until finding your first Mod Pack -and even then it'll only matter if you're wanting to apply both of those Mod Packs to a single piece of gear, which is not remotely guaranteed. (And if you're making a Basic Assembly, Whizkid is still not relevant at that point!) Whizkid should generally not be charged towards immediately.
Second, Mod Packs come in two tiers: there's the four common Mod Packs you expect to pick up several copies of each type over the course of a standard run, and there's the Exotic Mod Packs where you can easily end up with exactly one copy of one type across the entire standard run. When looking over Assemblies or otherwise making plans around Whizkid, you can plan as if you'll have enough of any given common Mod Pack -because you almost always will- but should treat Exotic Mod Packs as a bonus. (Unless you already have one on hand, of course)
Third, most of Whizkid's power is tied up in Assemblies in particular. Doom Roguelike is designed so that a decent number of Assemblies are overall superior to simply piling Mod Packs onto a piece of gear. (There's a depressingly large number that miss this mark, unfortunately, but that's for later posts) Among other points, Mod Pack stacking is actually fairly harshly constrained; if you could put 3 Power Mod Packs onto an Armor, that would be incredibly powerful and worth doing! But you can't; you'll need to turn to Assemblies if you want to raise your Armor's Protection by more than 2 points.
Fourth, Assemblies are pretty good! (Well, enough of them are pretty good) By which I specifically mean that I've played other games with a comparable crafting system where crafting was extraordinarily niche (eg "You have 50 recipes, only 2 of which have any use, and they're niche uses only relevant to a couple specific playstyles") such that a learning player was honestly better off simply ignoring the crafting system. If you're going to immediately loot a better thing than what you just crafted, why bother at all?
Whereas in Doom Roguelike there are Assemblies that are arguably better than even the best possible loot to find randomly, and a reasonably large number of Assemblies that can potentially be displaced by loot have it be a low-odds event within a given run's framework, or at least have it likely to occur far enough along to have gotten significant use out of the Assembly prior to it being replaced while having used only low-value Mod Packs so it's not a huge waste to do. (eg a Tactical Shotgun will, in a standard run that diligently explores most Special Levels, be inevitably displaced by an Assault Shotgun, but normally you'll Assemble a Tactical Shotgun many, many floors before you get an Assault Shotgun)
That said, Assemblies locked behind Whizkid have some qualifiers. The big one is that more advanced Assemblies automatically require more Mod Packs; you create Basic Assemblies by trying to apply 2 Mod Packs to a piece of gear, Advanced Assemblies by trying to apply 3 Mod packs to a piece of gear, and you create Master Assemblies by trying to apply 4 Mod Packs to a piece of gear. (Which is a bit weird with weapons, since Whizkid 2 lets you apply 5 Mod Packs to a weapon)
This means Advanced and Master Assemblies are not only locked behind Whizkid ranks but also behind significant material needs. You literally can't leverage Whizkid 1 without a minimum of 3 Mod Packs, and making your entire kit out of Master Assemblies will require a total of 12 Mod Packs. (15 if you want to also slot a Mod Pack into each of the completed Assemblies) That's a big spike that means a standard run has to be a lot more careful about how it uses its Mod Packs; with no ranks in Whizkid, you can pretty carelessly make Basic Assemblies that don't use Exotic Mod Packs and it's unlikely to result in you being unable to make some other Basic Assembly later. If you're shooting for Master Assemblies, though, any Assembly you make that isn't part of your final kit is risking destroying your ability to make your final kit!
Master Assemblies labor under the additional difficulty that most of them require at least one Exotic Mod Pack; only Cerberus Armor, Cerberus Boots, and the Ripper can be made purely with common Mod Packs. As such, planning around building a given Master Assembly has a very strong risk of simply not looting the necessary Exotic Mod Pack(s) before the run completes. If it's not one of the prior exceptions, of course, but while Cerberus Boots are worth considering planning around in a standard run shooting for the best possible victory, as they make navigating Mount Erebus/The Lava Pits reliably possible at no resource expenditure, Cerberus Armor is often unnecessary, and the Ripper has flaws I'll get to when I get to it.
In spite of all these qualifiers, Whizkid unfortunately is pretty deleterious to Mastery balance, and to a lesser extent general Trait/build balance; a standard run benefits hugely from Whizkid, more so as you go up in difficulty and/or take on more Special Levels, and is basically 100% mandatory for Angel of 100 or Archangel of 666 runs; in turn, this means Masteries that block Whizkid have inherently depressed viability, while Masteries that have easy access to Whizkid have inherently inflated viability, especially in Angel of 100/Archangel of 666. I kind of wish Whizkid wasn't a Trait at all and your ability to improve at Mod Pack usage was simply tied to Doomguy's core Level, or at least that Whizkid was a Basic Trait for every class that was never blocked by Masteries.
Especially since Whizkid's benefits scale funkily. Whizkid 1 in particular can easily end up being a bit of a dud if you either don't really want any of the Advanced Assemblies in your run or do want one but fail to get the required elements to make that Assembly. Whizkid 2 is more reliably good since it adds the ability to straightforwardly upgrade your Basic Assemblies: those Tactical Boots you Assembled fairly early on can now be bolstered with an Agility Mod Pack for even more running speed, for example. But Whizkid 2 is pretty reliant on this feature to justify itself since only 3 Master Assemblies don't require at least one Exotic Mod Pack, and you can easily go multiple runs in a row without seeing even one Exotic Mod Pack. But even before that point, the combination of Whizkid ranks requiring increasingly large numbers of Mod Packs to be benefitting from them and Mod Packs being unreliable about giving you what you want (Common Mod Packs are common, but if your plans call for 6+ Agility Mod Packs, you're going to be frustrated when you get 6+ Bulk Mod Packs that don't really help your build at all) makes Whizkid ranks dependent on outside RNG factors to scale as intended, which is a flaw unique to this Trait; disconnecting Whizkid's benefits from the Trait system per se could smooth out a lot of jank.
On the plus side, 0.9.9.8 fixed a really frustrating bit of wonkiness in Whizkid's design, in that previously the ability to slot in a Mod Pack into an Assembly at Whizkid 2 required the Assembly was created while already having Whizkid 2. As such, there was janky pressure to consider holding off on finishing an Assembly until you got to Whizkid 2, especially if you were already pretty close, and similarly it was depressingly optimal to Assemble Tactical Boots basically immediately, knowing you'd need 3 more Agility Mod Packs and another Steel Boots at some later point so you could Assemble another Tactical Boots that you were allowed to add an Agility Mod Pack to. I very much appreciate this fix having been made.
I'm also a fan of the new extended skill rank for Technicians: it's not a huge thing overall, but being able to add 2 Mod Packs to an Assembly is a notable niche, and part of what was wonky about the Technician's specialty in Mod Packs/Assemblies was that they were merely faster to achieve competency at such, not actually better than the other classes once everybody had Whizkid 2, outside the edge case point that a few Uniques will accept Mod Packs to some extent from Technicians in specific... and before 0.9.9.8, there were only 3 such Uniques, making it pretty rare to have any chance of being relevant.
I do hope future updates iron some of this out more without destroying the strengths of the current system.
Badass
Requirements: Tough as Nails 2. No requirement for Marine class.
Max rank: 2
Effect: The first rank allows perfect retention of HP up to 150%, and the second rank negates the HP 'rot' mechanic entirely. Additionally, each rank lowers knockback's effectiveness against Doomguy by 1 tile.
Part of the reason why Badass is less notable on the other classes is that Badass has a lot of early value from letting you use 'feast' periods to get through 'famine' periods -that is, if you're a Marine with the first rank of Badass, and you fully clear a floor with full HP and three Small Health Globes, you hoover those up and that HP lasts until something actually hurts you, where without Badass that 30 HP would potentially completely vanish before anything got around to hurting you and so have provided zero benefit. Early in a run, when your gear is poor and you probably have few or no Medpacks, this can allow you to comfortably survive a floor that would've killed you if you'd arrived at only 100% HP, or mean you don't feel the need to burn precious Medpacks too quickly and thus get the opportunity to instead heal up entirely off Powerups, saving Medpacks for combat healing or a really dire low point.
Later in a run... well, for one thing Supercharges, Megaspheres, and Invulnerability Globes all get noticeably more common. Supercharges literally can't spawn at all until your fourth regular floor, and when they first enter rotation it's surprising to see one at all. When you're deep in Hell floors, it's not particularly surprising to find two Supercharges on the same floor. So getting Badass maxed early can turn a Supercharge into something that benefits you for several floors, where later on a non-Badass may well grab a Supercharge, run into the thick of combat before it can decay much, and then grab a second Supercharge before descending to repeat this on the next floor.
For another, a run that's far along is generally pretty secure in terms of having a decent supply of Medpacks and good gear for efficiently taking on even endgame enemies, and in fact may well be regularly leaving Large Medpacks lying on the ground because their inventory is filled to the gills with essentials. In such a case, it doesn't necessarily meaningfully cost anything to use even a Large Medpack: if you've already spotted and left behind three, using a couple in clearing the floor isn't leaving you less prepared for a later floor. This makes the benefits of pushing your HP over its standard maximum less significant -the fact that the Technician uses items super-fast particularly exaggerates this, since they can get away with hanging at slightly-low HP values on the idea they'll instantly use a Large Medpack if they drop to a truly dangerous level of HP.
Put another way, Badass has significant strategic benefits that become less usefully relevant the deeper a run gets; a Technician who rushes for their Mastery and only then turns to getting ranks in Badass is liable to find they're already hitting the point these strategic benefits don't really apply anymore.
As for the knockback benefits, well... we need to talk about how knockback works:
Melee attacks never cause knockback.
Most ranged weapons require 10 points of damage per tile of knockback. (Note: the wiki historically claimed 12 points. I'm unsure if 0.9.9.8 silently changed this or the wiki was always wrong, but I can say with certainty it's 10 points in 0.9.9.8)
Shotgun weapons and most explosions (as well as the direct impact of an explosive missile) require 7 points of damage per tile of knockback.
BFG explosions require 14 points of damage per tile of knockback.
Rocket jump explosions require 2 points of damage per tile of knockback.
If Doomguy is wielding the Butcher's Cleaver, Longinus Spear, or Dragonslayer, any knockback he takes is halved.
If Doomguy has ranks in Badass, the number of tiles he is knocked back is reduced by one per rank.
Armor can affect knockback as a percentage, which then rounds to the nearest tile for knockback purposes.
Note that all of this is the rolled damage number: resistances and Protection lowering incoming damage doesn't affect knockback taken.
In the early game, especially on lower difficulties, Doomguy is generally only going to be knocked back if he's standing near a barrel when it detonates, if he fires a Rocket Launcher from up close, or is shot at by a Former Sergeant, and Doomguy will only rarely be knocked back more than 1 tile.
In the late game, most ranged enemies are actually pretty likely to knock Doomguy around a tile, and several of them can knock him two tiles, with some enemies (eg Archviles) being outright guaranteed to knock him around if they hit him at all. (But few enemies can apply 3 tiles of knockback: Badass 2 is close to complete immunity to knockback, as even the enemies that can apply 3 tiles of knockback won't do so very often... unless you're doing an Angel of Max Carnage run, anyway)
How valuable immunity to knockback is... is kind of difficult to judge.
On the one hand, knockback can be really dangerous. It can result in Doomguy being shoved into Lava, it can result in him taking damage from barrels he was a safe distance from during his turn, it can result in him being attacked by enemies he didn't even know existed, it can prevent him from reaching exit stairs or a door or some other goal...
... but immunity to knockback can actually get Doomguy killed, such as if a bunch of enemies are in view but the first one to act knocks him out of reach of the rest before they get their turns. Notably, Mancubi by default attack a location twice in a row with three shots apiece: with full knockback, Doomguy is often pushed out of reach of the second volley by the first volley, and indeed doesn't even necessarily take full damage from the first volley, whereas with Badass maxed Doomguy can end up taking all six hits right to the face!
On average, Badass protecting from knockback is usually more protective than it is problematic, but... I kind of wish it had a different secondary benefit, or knockback's mechanics were somehow more reliably Bad To Have Happen To You, or both. The current state isn't horrible, but it is janky. The one context in which it's reliably 100% desirable is as backing to melee builds, since knockback on a charging melee Doomguy almost always means he's being knocked further away from his enemies, delaying his ability to kill them and so stop them from killing him. For ranged builds, though? Janky.
Shottyman
Requirements: Reloader 2.
Max rank: 1
Effect: If Doomguy moves while equipped with a Shotgun-type weapon or a Rocket Launcher-type weapon, a reload action will be performed at no time cost.
Also note that this reload is a free action aside the requirement of movement to trigger it. Shottyman reloads are thus not actually affected by Reloader nor by a Bulk Mod improving the weapon's reload time. If you're trying to improve your Shottyman reload speed, you should be trying to improve your movement speed, such as through ranks in Hellrunner, applying Agility Mods to Boots/Assembling Tactical Boots/Assembling Antigrav Boots, equipping naturally-fast Boots, and so on.
Mind, you're forced to take two ranks in Reloader to get Shottyman, but eg if you reach extended ranks you shouldn't grab the third Reloader rank if you're specifically trying to improve Shottyman reloads.
Shottyman itself makes the Dodge mechanics considerably more relevant than normal, since you can alternate firing and walking to trigger a reload while also fishing for a Dodge attempt: non-melee, non-Shotgun styles have much stronger incentives to hold still and keep outputting damage. This is undermined a bit by corner-shooting existing and Shotguns being great at corner-shooting, but if nothing else it matters anytime conditions force a fight to occur with no opportunity for corner-shooting. (Such as because you descended stairs and found yourself in an open area with a horde of enemies in range)
Shottyman also opens up a lot of potential to kite enemies, where you blast an enemy and walk away from them to reload for free while walking out of their vision range. In conjunction with Shotguns having high knockback, it's often possible to prevent an enemy from attacking for several turns in a row, especially if you've gotten hold of one of the Shotguns that performs better at longer ranges.
And of course Shottyman benefitting Rocket Launchers makes it fairly natural to pair the weapons together in your run. Which is good, given Rocket Launchers are very impractical to try to use as a primary weapon: the game would have to do something pretty wild with Traits to truly make a pure Rocket Launcher specialization actually viable.
Do note that Rocket Launchers and most Shotguns only reload one unit of ammo per reload action, and Shottyman specifically performs a regular reload. A Combat Shotgun's maximum of 5 ammo, for example, requires 5 tiles of movement to fully reload from empty. This is no big deal when just travelling about, as Doomguy will often fully reload before reaching the next enemy and in fact Shottyman makes time-pressure situations a little easier to deal with since you can just keep running away from the Lava or whatever and still be ready to shoot as more enemies come on-screen, but if you're expecting to be able to walk one step and blast 5 times... that'll only work out for you if it's something like a Plasma Shotgun, where it reloads fully in one go.
In terms of design, I wish Shottyman was more representative of Advanced Traits: it's a weapon specialization that dramatically changes how you use the weapon you're specializing in, creating new strategies and meaning there's a much starker playstyle difference between 'non-Shotgun build that uses a Tactical Shotgun for specific situations' and 'Shotgun build that uses a Tactical Shotgun as its core weapon for much of the game' than with the weapon-specialization Advanced Traits. Dualgunner, for example, does have implications worth keeping in mind, like 'leveraging Dualgunner means not leveraging Ammo Chain quick-reloads and also means not having a completely different weapon ready to swap in', but the biggest difference in how Dualgunner Pistol usage plays is the constant need to use Dual Reload instead of a regular reload: in most respects, it's just 'regular Pistol usage, but stronger'.
Triggerhappy
Requirements: Son of a Bitch 2.
Max rank: 2
Effect: Weapons that fire 3 or more projectiles at a time will fire an additional projectile per attack per rank. (If a weapon fires exactly 2 shots, but is not a Shotgun or Pistol, Triggerhappy will also apply. Bullet Dance on the Marine overrules this logic in regard to Pistols) These additional shots still cost ammo.
Triggerhappy is fantastic for the Ammo Chain Marine Mastery (Which causes the added bullets to be completely free), is required for the Marine's okay Pistol Mastery (Which is centered on Triggerhappy), and is difficult to justify otherwise.
First of all is the issue of overkill: if your Chaingun kills the target with the first bullet, the other bullets are just wasted shots if there's nothing else in the firing line. Same for if the second, third, fourth, or fifth bullets kill: Triggerhappy thus adds a very noticeable amount of ammo wastage, when rapid-fire weapons already are prone to running out of ammo! (Unless supported by the Ammo Chain Mastery)
Second is the issue that Son of a Bitch, Finesse, Eagle Eye, and to a lesser extent Reloader provide the same basic benefit of improved DPS, but without inherently increasing ammo wastage. (In fact, Son of a Bitch and Eagle Eye reduce ammo wastage)
Third, weapon numbers are against it. A regular Chaingun is 1d6*4; a rank in Triggerhappy raises your damage by 25%, while a rank in Son of a Bitch is on average an increase of almost 30%, because on average a D6 will roll 3.5 damage. If it helps to understand, imagine that exactly half the dice land on 3 and half the dice land on 4; that results in a total damage of 14, which Son of a Bitch then adds 4 to, resulting in 18; a 30% increase of 14 results in exactly 18.2. So like I said; almost a 30% increase.
Other weapons don't exactly help. A Minigun is 1d6*8, meaning Son of a Bitch is still almost a 30% increase in damage whereas Triggerhappy drops to a 12.5% increase. Plasma Rifles are 1d7*6, making Son of a Bitch exactly a 25% increase on average while Triggerhappy is a 16.66~% increase. A Laser Rifle's 1d7*5 is less bad, but it's still a 25% increase on Son of a Bitch vs a 20% increase on Triggerhappy.
Only the Unique rapid fire weapons are actually favorable to Triggerhappy as a damage increase; the BFG 10K's 6d4*5 places Son of a Bitch as on average an increase of less than 7% (+1 damage to an average of 15 damage per shot) vs Triggerhappy being 20%. The Megabuster's fire modes all fire 3 shots -and thus Triggerhappy is a 33% increase- where their average damage per shot consistently hovers around 4 damage. (ie Son of a Bitch is on average something like a 25% increase) And honestly with the Megabuster the numbers are close enough Son of a Bitch's ammo efficiency advantage is far more important than Triggerhappy averaging a slightly higher damage increase; it's only really the BFG 10K where Triggerhappy is genuinely a much stronger choice for improving damage.
There is at least the Jackhammer, which is a Unique Shotgun that benefits from Triggerhappy: the Jackhammer's 8d3 damage averages 8 damage, with range drop-off lowering the final damage, meaning +1 damage from Son of a Bitch will generally be less than a 16% increase in damage vs the first rank in Triggerhappy increasing shots by 33%, and Shotgun ammo is less prone to being torn through... but this all comes with the big qualifier that the Jackhammer very reliably knocks targets back over its volley and so additional shots will often be suffering harsher range penalties, making Triggerhappy not truly a 33% increase.
Finesse as the comparison point is less stacked against Triggerhappy, but... it's still not great. 2 ranks in Triggerhappy is a 50% increase in Chaingun damage, and its absolute best-case scenario is that it's boosting a weapon that fires exactly 3 times (The minimum for Triggerhappy to apply) where it rises to a 66% increase in damage. 2 ranks of Finesse reduces firing times to 70%, which works out to roughly a 42% increase in DPS: only a little behind. But then Finesse can ultimately get to a third rank: at rank 3 Finesse, you're spending 55% of base firing time, which is to say it increases DPS by around 85%!
Eagle Eye is a bit of a nuisance to properly describe given how you can affect accuracy with Mod Packs and Assemblies, by getting closer or farther, by Running... but let's just provide the simple examples of an unmodified Chaingun and unmodified Plasma Rifle fired at line of sight. They both have +2 Accuracy, so at line of sight this decays to +0 Accuracy: looking at our earlier table on what Accuracy amounts mean, we find that they expect to land exactly 50% of shots fired on a target. Taking a rank of Eagle Eye bumps them up to landing 74.07% of their shots: this is very slightly over a 48% increase in average damage, which you'll notice is barely worse than taking both ranks of Triggerhappy for improving the Chaingun's damage, and is cleanly superior for boosting the Plasma Rifle's damage. (Since 2 ranks in Triggerhappy only increases its shot count by 33%) The second rank of Eagle Eye is less dramatic, bringing the hit rate to 90.74%, which is only about 22% more damage than the previous rank, and the third rank of Eagle Eye is even less notable, bringing the hit rate to 98.14%, which isn't even a 10% increase in damage over the previous rank... on the other hand, if I frame it as '2 ranks of Eagle Eye vs 2 ranks in Triggerhappy', we find that Eagle Eye increased damage by 80%, vs Triggerhappy 2 being a 50% increase for a Chaingun.
A further issue is that Triggerhappy's damage advantage gets reduced in longer fights by it increasing downtime spent on reloading; a standard Chaingun by default spends 10 seconds shooting before needing to reload for 2.5 seconds. 2 ranks in Triggerhappy results in that same Chaingun spending 7 seconds on shooting before running out of ammo and spending 2.5 seconds to reload; that's switching from 20% of your time being spent not shooting to over 26% of your time being spent not shooting.
It's also worth pointing out the consideration of volley-to-ammo-storage ratios; a Chaingun normally spends its ammo 'cleanly', where all 10 volleys are a full 4 shots. 2 ranks in Triggerhappy results in firing 6 volleys of 6 shots, then the 7th volley is actually only 4 shots; you effectively don't benefit from Triggerhappy at all on that volley!
To be fair, this latter point isn't a universal problem; 2 ranks of Triggerhappy puts Plasma Rifles at a clean 5 volleys with no partial volleys, and the Minigun ends up at a clean 10 volleys with no partial volleys. Bulk mods can also mess with this... but they can do so both ways, so that's a bit of a wash.
The Trait comparisons especially matter for the sheer number of relevant Traits that can be grabbed before bothering with Triggerhappy. Imagine I modified the game to throw in a comparably mediocre Trait for boosting melee damage: once you're at Brute 3, Finesse 2, and Berserker (6 points), you're done boosting melee damage until extended ranks, which starts from Level 12, so a melee build would have actual reason to consider grabbing some mediocre Trait for boosting melee damage. And extended ranks just adds 3 points to spend for reaching Brute 5 and Finesse 3: again, it would potentially be legitimately tempting to dip points into a mediocre Trait to increase melee damage, especially if it did so in a way that made it easier push through swarms. (Maybe it's Finesse, But Only For Melee, With A Weaker Modifier, And Considered A Separate Source From Finesse)
Whereas rapid-fire weapons have 8 points they can dump in Eagle Eye/Finesse/Son of a Bitch before extended ranks and even considering that Eagle Eye 4 and 5 are basically worthless that's still 3 more points in extended ranks, for a total of 11 points you can burn on Traits that are better at boosting Triggerhappy before bothering with it. Add in that you'll normally want a couple ranks in Whizkid and ranks in defensive Traits and/or Hellrunner... a Mastery-less run focusing on Rapid Fire weapons could easily win before it makes any sense to dump points into Triggerhappy.
Overall, Triggerhappy is a really flawed Trait that could use an overhaul. Just making the extra shots magically not spend ammo would help a lot, but personally I wish it got replaced entirely with some other Advanced Trait for specializing in rapid-fire weapons. Its current state is not great, and its problems are such that fiddling with numbers isn't liable to help: making it add 2 shots per volley would help with the damage output comparisons, but worsen the ammo efficiency problems. Right now, you should normally ignore it unless your Mastery requires it, and of the 4 Masteries that do require it, Ammochain and Bullet Dance are the only ones not dragged down by having Triggerhappy as a requirement... and in Bullet Dance's case, this is only true in the sense that its effect runs off Triggerhappy and therefore just removing Triggerhappy from the rank requirement list wouldn't help Bullet Dance any. Bullet Dance is still very much harmed by requiring Triggerhappy.
---------------------------------
An interesting aspect of Doom Roguelike is that its leveling system is, overall, fairly prone to accelerating returns.
In most games with a leveling system, leveling has diminishing returns, at least past a certain point. Plenty of systems use linear mathematical progressions (eg 'each rank of The Toughness Skill adds so and so amount of Toughness Points'), which usually works out to being a system of diminishing returns; adding +10 HP when you only have 10 HP is doubling your durability! Whereas adding +10 HP when you have 200 HP is an increase of 5% to your durability, ie barely anything. Most mathematical systems that games like to make use of will pretty reliably produce such a progression: +1 of Useful Stat is just less valuable the more of said Useful Stat you already have.
Even when games don't work in this particular way, diminishing returns often ends up being the default model anyway, just via slightly less obvious mechanisms. Games with a system akin to Doom Roguelike's Trait system often end up with diminishing returns when talking what I'll call 'realistic good play' by virtue of a player who knows what they're doing reliably grabbing the highest-value skill first, then the second-highest, then the third-highest, and so on, where the diminishing returns happen because they of course grab the best stuff first. ("I'll grab the skill that doubles my damage first, then the one that increases my damage by 50%, then the one that increases my damage by 35%...")
Games with skill trees, where you have to work through some skills to get through other skills, often produce the 'at least past a certain point' qualifier, because eg you spend levels 3, 4, and 5 on junk skills because points in them is necessary progress to The Best Skill In The Tree and so level 6 is a much bigger boost to power than those prior three levels, but typically once the tree is fairly filled out or the key Best Skills have already been grabbed, you're back to the diminishing returns point, and games with such systems typically don't correctly identify the highest-value skills; a player who knows the game will typically unlock all the best skills long before the endgame of such a game, because even if the devs intended for the best skills to only unlock right as the player is reaching the end of the game, the actual best skills generally only have partial overlap with the dev intentions, sometimes no overlap, and so a player has gotten all the biggest power spikes when they're 50% of the way through the game, not starting on those power spikes when they're 96% of the way through the game the way the devs intended.
And of course games with leveling systems often just make early levels come faster in the first place, where your first hour of gameplay nets you 10 levels and your second hour only nets you 5 more levels.
I could go on for a while, but the point is this is a pretty widespread final trend due to the underlying approaches common to how devs tend to handle such progression systems; the player experiences massive power growth in the early game, and then it slows down and generally keeps slowing down, regardless of the exact details for why.
So it's very interesting that this is not true of Doom Roguelike.
Part of why is just its approach to the math of the system is prone to implicit accelerating returns. Since time modifiers in Doom Roguelike are subtractive, they get more impactful as they get stacked on; Son of a Gun, as noted earlier, is a 25% increase in firing speed for gaining the first rank, whereas the fourth and fifth ranks each double the firing speed over their respective previous ranks. (And the fifth rank would be a literally infinite improvement to fire speed if Doom Roguelike didn't force actions to take a minimum of 0.1 seconds; without this, the fifth rank would reduce firing time to zero!)
Another part is that the game is prone to designing skills so that if they fairly directly interact with another skill, that interaction tends to be what I'll call 'positive'; that is, Badass and Ironman both increase Doomguy's effective HP, and they interact so that points in one increases the impact of points in the other. (eg taking 5 ranks in Ironman to double your base HP in turn doubles the effective increase in max HP you get out of a rank in Badass) A lot of games with such interactions will make an effort to blunt the synergies that result, at times all the way to arbitrarily ruling that A and B effects simply refuse to interact, precisely because the devs recognize that logically they should be an extremely powerful combination.
This is all particularly worth noting due to how it interacts with an aspect of Doom Roguelike's design that's actually pretty standard for games: that the threat profile of enemies grows in a non-linear, accelerating returns aspect! After all, as you go deeper into a run, the Danger value of floors goes up (resulting in more enemies per floor) and new, more deadly types of enemies enter rotation: if you compare two floors and find one has twice the Danger of the other, the higher-Danger floor is almost always going to be more than twice as threatening.
Plenty of games using this type of progression run into problems with getting the difficulty curve to work out as a gentle upward curve the way they want, precisely due to the aforementioned tendency for the player's growth to be designed as diminishing returns. Without lots of careful, custom-designed interventions (eg power spikes for the player at specific levels), the default result of these two points intersecting is that enemies grow faster in threat than the player. Up to a certain point, this is actually a good thing: if you start the player off kind of unreasonably strong, then a new player fumbling their way through the early game will hopefully still get through okay because their character (or army, or whatever) is so powerful that even pretty bad play is adequate to win, and by the time enemies are catching up to their character hopefully their own skills are reaching the point that they can cope with such a challenge, which is good!
But plenty of games end up travelling well past that point and accidentally (or sometimes intentionally) produce a dynamic where the enemies rocket past 'reasonable challenge for an on-level player character' and so the late game is only doable for players who can wring out as much value as possible from the game systems, at times requiring more value be extracted than the devs really intended to be possible. There's absolutely players who enjoy that kind of design, of course, but... it's pretty clear it's usually not what the devs were trying to produce.
Whereas Doom Roguelike doesn't have this problem: the accelerating returns on Traits and parts of Mod Packs/Assemblies means that a well-built Doomguy who is staying on top of his kill rate so he's gaining good experience can keep up with -or even stay ahead of- the enemy threat rising!
Though it also helps that Doom Roguelike has an unusually low 'ceiling' on individual enemy power.
But that's for other posts.
----------------------------------
Next time, we start on classes and their Mastery Traits, starting with the Marine.
See you then.
I feel like you understate Finesse a little bit; one rank may not be great, but finesse + juggler is a solid competition for reloader in time saved, as long as you’re willing to carry around spare weapons instead of reloading.
ReplyDeleteAnd it’s more versatile than reloader + shottyman; shottyman is only useful for a shotgun build, but juggler works with most builds.
I quite like it on melee builds, especially, where having a shotgun ready without compromising your ability to melee opponents is quite handy.
You also, I think, understate the value of melee weapons without Brute. The inaccuracy is annoying, but a lot of enemies are much less threatening in melee than at range, and even with 50% accuracy, the chainsaw can do more damage than a chain gun, which is taking armor off every hit.
(Brute is still essential for melee builds, of course. But even without it, the chainsaw is useful)
Finesse into Juggler usually doesn't fit smoothly into progressing directly to your Mastery, and by the time I have my Mastery I don't have the inventory space for spare Shotguns/I reliably already have a Combat Shotgun in a standard run so such swapping is much less relevant. (Especially since I probably also have a Shell Box: in conjunction with Reloader 2, reloading occurs in 0.1 seconds just like a Juggler-swap) Juggler-swapping to evade reloading in combat is also primarily *relevant* to builds that want Shottyman -not that there's no utility to swapping a Chaingun for another Chaingun, but the utility is much more erratic in its relevancy compared to doing so with a Shotgun or Rocket Launcher. I'd much rather have Shottyman ready for Hell's Arena than Finesse and Juggler.
DeleteVarious enemies being less dangerous in melee than at range isn't really an argument to use *melee weapons* supplementally except specifically to Shotgun builds. Your Dualgunner should just keep shooting. Your Ammochain Marine should just keep shooting. (Son of a Bitch 3 Chainsaw: 50% chance to do 7-27 damage. Son of a Bitch 3+Triggerhappy 2 Chaingun fired point-blank: 74% accuracy while doing 6 shots of 4-9 damage apiece, or 24-54 damage before accuracy. If I say you're firing into a Baron of Hell with its Protection of 2, you drop to 2-7 damage per shot, or 12-42 damage. A Baron's 2 Protection is very rarely exceeded: this is not a contest)
Note that if I instead assume you're Masteryless and ignoring Triggerhappy as bad (Which is fair on a Technician especially), the Chainsaw still compares poorly. Son of a Bitch 3 Chainsaw's 7-27 (50% hit rate) is now competing against a Chaingun hitting 4 times for 4-9 damage apiece at 74% hit rate, or 16-36. Against a Baron, it drops to 8-28. While passing the Accuracy check almost 50% more often.
Your Shotgun guy absolutely should consider hitting things in melee -especially Arachnotrons, where knocking them away makes them outrageously more dangerous- but other ranged builds only want to do so if ammo starvation is a concern, and committing to carrying a Chainsaw about is increasing one's risk of ammo starvation in an attempt to stave it off.
That said, I should emphasize that these posts are primarily written from the perspective of standard runs. In Angel of 100 and Archangel of 666, I'm much more of an advocate for getting Brute 2 and Berserker at some point if your Mastery doesn't block it. Ammo starvation is a much more constant specter in such runs if you don't luck into a Nano Mod, and you WILL get to Level 20+, so you can easily spare the Trait points at some point, and even just punching with Brute 2 and Berserker is surprisingly lethal. (7-9, doubled by Berserk into 14-18 is pretty good for a free attack, and against an Arachnotron it doubles again to 28-36 damage, meaning you kill them in two punches so long as you don't miss)
DeleteI agree that once most builds really start coming on line, they should be shooting instead of using a melee weapon whenever possible.
DeleteThat said, SoB3 is doing a lot of work in your example, and it's a significant investment - even the ammochain marine won't have it until after getting their mastery, and it hits a sweet spot for lots of damage without risk of knockback.
And cases where the chainsaw performs better than a chaingun with SoB2 or less aren't *that* uncommon; item-using enemies who've grabbed a green armor are something you will have to deal with pretty much every run.
Not... really? Like, looking at the SoB 3 no Triggerhappy numbers against a Baron, you might notice they're 1 point of damage range over the Chainsaw while hitting way more consistently. Just SoB 2 means a Chaingun against a Baron is 4 hits of 1-6 damage apiece (ie 4-24) vs a Chainsaw is 6-26 with SoB 2... with the Chaingun hitting 74% of the time in melee vs the Chainsaw hitting 50% of the time. So the Chainsaw is slightly harder hitting *if* if you ignore Accuracy, but way worse if you account for Accuracy. An Ammochain Marine who decides to prioritize SoB over Reloader hits this state at level 2, well before they reach the Chained Court.
DeleteEnemies picking up Armor is somewhat relevant, but has two problems. Firstly, only Commandos, Hell Knights, and Barons of Hell are at all relevant to these early game numbers while having innate Protection: if a Former Human/Sergeant/Captain picks up a Green or Blue Armor, then my Baron of Hell numbers are the result you get, ie Chaingun wins at point-blank still. Second, as far as I'm aware there's no way to tell whether an enemy is equipped with Armor outside seeing the Armor vanish yourself or doing weirdly little damage. If you do find yourself in melee with a Baron equipped with Green Armor, the Chainsaw gains a lot of appeal... but you often have no idea this is so until it's way too late for it to make sense to switch. (Unless you have Juggler, of course)
Just a heads up that in current release (0.9.9.8a), Juggler and Dualgunner's swapping benefits are truly instant and do not progress turn/time.
ReplyDelete... huh. I'll test that as soon as I can, but that would be a much-appreciated change if so.
DeleteHave now confirmed Juggler is indeed truly instant. Will be updating post shortly. (I've yet to per se confirm Dualgunner, but while I will test it manually I'll be very surprised if it turns out to not be truly instant as well)
Delete