Lessons from Doom (2010)

I'm glad the article mentioned movement. Games today make so many decisions when it comes to movement where the goal is to make it feel 'realistic', with sprinting, ducking and the like, but they end up feeling like they don't have any depth. An average player will have fairly similar movement to a fairly good one, with incremental differences on small things, and a player with amazing movement won't hugely stand out. In Quake 3? A player with amazing movement could do way more to control the map. Quake 3 really mastered that depth, especially with how much there was to learn between the two physics rule-sets (with CPM allowing for a bit more than VQ3). And the reason I say that is because the classic arena shooters (Doom, Quake, Unreal Tournament, etc) had their movement systems as a key part of the game. It's why Defrag became such a (relatively) big thing. And it's why Overwatch just doesn't feel as fun as it could, to me at least. For a game to look like that, but to leave out bunny hopping? Disappointing.

Can I just say how unbelievably good the new Doom is. That game has single-handedly restored my faith in the first person shooter genre.

"This creates a feeling that’s quite rare in modern FPS: that you are powerful because you are agile, not because you’re a tank."

Thanks. I lost interest in FPS and after reading that chapter now I know why.

The article mentions that the level design was "empowered", but it doesn't mention specific level design decisions. Here's a recent article about that: http://kotaku.com/surprise-doom-is-still-an-incredible-game-... The author has lots more discussion of level design and game mechanics at https://www.patreon.com/DocGames

This coming out in 2010 the new Doom 2016 runs like an itemized list of his explanations of what made the original great.

Being agile instead of being a tank. Doging over cover. Low barrier of entry to modding (which the new one didn't totally do but the snapmap system is at least an attempt at that.) Fighting priority of large bosses while kiting around smaller enemies.

Great read!

I disagree on the point of no other games having the same level of user generated content. Specifically, Starcraft and then Warcraft III modding birthed multiple centi-million dollar industries (Tower Defense, MOBAs). Minecraft didn't exist at the time this article was written but obviously it has similar depth of user modding.

I read other comments here, and would like to point something.

Wolfenstein 3D, Doom, and other early FPS, got the name "FPS", because people saw them as a first person version of shooters, and shooters were those games where you control some vehicle from the top view and would dodge a hail of bullets and shoot back.

Currently, most poeple think the important thing about "first person shooter" and "third person shooter" is the camera, and completely forget why "first person shooter" had that name, the "first person" thing was just an adjective added to the base genre, "shooter".

Most modern games aren't shooters... I am not sure how I would name them, but the focus is not on shooting or avoiding get shot, the focus is something else.

The descriptions I am seeing of Doom 4 single-player and why people are loving it (I am waiting for Vulkan version to try it) are the descriptions of a game that nailed the "shooter" genre correctly, Doom 4 can proudly call itself a FPS, because it is a true FPS.

So some people might wonder, what I am calling here an FPS, why people feel it is so fun and whatnot?

1. Base mechanic, is shooting.

2. Movement is very important, both to aim, and dodge.

3. Stealth should be discouraged.

4. Cover should only be part of the game if you dodge in and out of it, it would be acceptable in a FPS with parkour mechanics for example.

5. No Halo-style health regen, this requires you hide, stop and wait, and when you do that you are not shooting.

6. Weapon variety is important, even one of the shooters known to be "minimalist", Ikaruga, has 4 weapons for 4 different uses (dark bullet, light bullet, dark laser, light laser), the point of shooting games is the shooting, even if the only variety comes from using powerups (1942 for example) it makes shooting in different ways possible and interesting.

7. If game has few weapons, it should nail it REALLY well, and the few weapons should solve any situation, games with "realistic" limit of weapons that result in you picking up a shotgun and a pistol and then having to shoot something 3km away is not a shooter. (maybe it is some other genre, war simulator or something, but it is not a shooter).

8. For the thinking part, the important is tactics, not strategy, the player should only be thinking of immediate decisions while shooting, and not stopping to think what path (in the map, narrative or upgrade) he will have to take several hours on from now (unless he is replaying the game for completionism, then it is fine, but usually he will preplan before starting the playthrough anyway).

Glad to see this article making the rounds again.

Here's a little background to the article that might be interesting to some of you, concerning the evolution of shooters.

I'm the Nathan McKenzie the blog author thanks at the end of the article. I worked at Raven Software, off and on, from the spring of 1997 through the summer of 2004. I think JP, who wrote the blog, worked at Human Head in Madison, WI for a time. So we were both kind of in the id / fps diaspora, to some degree. I think this conversation took place on JP's old anti-factory blog, way back in the day.

Anyway, enough real-life background. The more interesting, possibly, background, is that I carried a lot of the responsibility for the single-player game balance and feel of Raven's and Activision's 2000 FPS, Soldier of Fortune. I basically was responsible for tuning player weapon damage, player input responses, enemy response times, enemy damage, high level enemy logic, enemy style differentiation, overall game progression curves, user settable difficulty levels, etc. I worked with other super talented people, of course, but at the end of the day, I was the main person in Raven's office at 2 am tweaking variables and subtle enemy aiming algorithms and properties to try to make the game match what appealed to me aesthetically. I really didn't work on SoF2 at all, so in so far as the games feel quite different, I would say that at least partially comes down to SoF being much closer to my own tastes.

I actually don't remember the conversation in great detail that JP and I had that led to this blog post, but I do know that a lot of what I was saying at the time about Doom was related to my experience of trying to do my part to balance Soldier of Fortune and make it fun and fair, and the challenges I faced in doing that.

SoF is a fine game, but I think it occupies an interesting point, historically, for shooters. I LOVED Doom. I just loved it. I played so much of it. It was hugely influential on me. If it had been up to me, I would have just been making more Doom-esque games.

But once you're making a realistic, bullet weapon based game with human enemies in relatively recognizable environments, much of what Doom got right kind of HAS to go out the window in the name of realism. Player movement speed. Enemies very strongly differentiated by attacks, properties, and novel behaviors. Enemies very cleanly parsable with strongly different silhouettes to signify that they are of different kinds. Enemies with radically different amounts of health or interesting weaknesses or weak points. Weird, interesting, abstract traps and interactive gadgets that can be combined with enemies to create novel situations. Game pacing controlled by the rate of introduction and variation of these things. Fairly abstracted environments that interconnect in weird ways, custom built to make for interesting combat experiences. Ratios of player size to monster size to projectile size to environment size that maximize player ability to move gracefully. Player interaction styles that maximize the aesthetic appeal of moving through a hostile environment that looks nothing like how real people interact with being shot at (specifically, using cover). There's an entire grammar that Doom and similar shooters from that era rely on that kind of has to be tossed (or at the least VERY dampened down) once you want some realistic verisimilitude.

And it just gets worse when you're trying to tell a linear story in the middle of the game and so require hyper linear level flow to make sure all the plot points are hit, and so you have to largely toss exploration as an aesthetic pleasure. And it gets worse still when you have dumb AI buddies that have to stay alive and get in the player's way in the middle of combat, and end up causing players to lose because the AI decided to do dumb things and die.

We were in the middle of all these broader genre transitions while working on SoF, and I don't think we entirely understood the consequences of how the individual changes were adding up, nor did the broader industry. It was like we had a recipe for a perfectly good dish, and then said, "Yeah, we're going to totally make that exact recipe again! Except we're not going to use salt, chicken or flour, and we will be adding some grapes and some rhubarb. And no use of an oven!" You might be able to hill climb your way back to a new recipe that works, but there's no guarantee, and once you do, it will be something totally different.

The biggest thing that plagued me while trying to balance SoF, and what caused me to look so much more closely at Doom later, was the issue of fairness and legitimacy of challenge. In Doom, because of a variety of very carefully considered factors (many of which JP talks about), when players take damage, it's nearly always because they did something dumb that was avoidable. When you get hit in Doom, it is punishment. Because of this, it's very difficult to get yourself into an unrecoverable state in Doom. No matter how low on health you get, you can almost always finish a level as long as you are sufficiently savvy. And because of this style of design, Doom levels can really ratchet the difficult up while still presenting a fair challenge. And, in turn, the distance between casual play and really skillful play can be pretty immense.

I really love and admire that quality in a game. But it really does take a lot of very precisely balanced, often game-y choices to make that possible.

I think JP's mention of the player-as-damage-mop versus player-as-one-who-dodges gets right at the heart of this. Obviously most shooters with regenerating health in the last decade or so are fair in some important sense of the word. If you play reasonably, you will probably win. But they're often just... sloppy. The connection between player input and resulting world states is often looser, more chaotic, more often collapsing into a kind of spectacle. Obviously you can ship a game someone can play through with those design choices. But I do think games have lost something in the transition. Given the success of the Dark Souls series, I think there's some contingent of gamers out there how agree.

In the end, I think SoF was pretty fun. I was certainly proud of my work on it. It's hard not to look back on it as a kind of intermediatary step, digging its heels in as a traditional run-and-gun shooter as the future of modern military shooters came barreling down. I know there are plenty of people who like the direction that more modern shooters have gone, but for my part, I always lament that I never got to work on a shooter more closely aligned with the original Doom's design philosophy.

There's a series on YouTube called "Game Maker's Toolkit" which does an excellent analysis of why Doom worked so well: "What We Can Learn From Doom", https://www.youtube.com/watch?v=yuOObGjCA7Q

He hits on many of the same points, but it goes into more depth.

Previous discussions: https://hn.algolia.com/?query=Lessons%20from%20Doom%20vector....

Bloodborne seems to meet all of these pretty well. Especially "speed over tanking", levels not needing to make sense if they're functional, weapon and enemy variety.

Cool to read it all laid out why we like certain games so much more than others.

From a previous HN link on Doom, I remember discovering this : https://github.com/jmtd/wadc

I Love that the author mentioned M.U.L.E.

If you have never heard of it, get a C64 ROM emulator and grab a version of the game. It's wonderfully addicting.