Difficulty balancing: a conundrum

[Doom-X-Machina]

I have no set method for balance other than my experience as a Doom player and what "feels right".

I do tend to populate enemies for the easiest difficulty first and what I think would be appropriate for a beginner. After that, I add the next difficulty monsters, then the next, up to Ultra-Violence where ALL the monsters are usually set to appear. Sometimes I may change a monster type instead of changing the amount of monsters - like swapping out a Hell Knight in a corridor for a Baron Of Hell in higher difficulties. It really comes down to the feel of the map and what sort of combat situation would play well in that area.

I also tend to use lower tier monsters on early maps, slowing introducing the higher tier monsters gradually as you play. I've never been keen on Barons or Archviles in an opening map, even on higher difficulties.

As far as populating items, health and ammo - I load out the same amounts of everything across all difficulties. There is MORE than enough for easier difficulties but as you go higher up to Ultra-Violence, the higher monster counts start to put a crunch on ammo and health levels, making ammo conservation and life preservation more important.

Sometimes I may also swap out the odd weapon for example, introducing the super-shotgun earlier in a lower difficulty where it might be just a regular shotgun on a higher skill level.

It's all about what feels right really. There's no real "right" or "wrong".

[Xaser]

Difficulty balance is one of things that's more of an art than a science, but I've gathered a handful of (subjective!) tips/principles/whatnot I tend to stick to. Some of these are more on the topic of "how to make a map challenging in general" rather than the OP's "how to balance across difficulty settings", but there's a good bit of overlap.

 

Emphasis on "subjective" above -- for anything that reads like a "do" or "don't" here, there's someone who's done the polar opposite and made something amazing out of it. Perhaps that next someone is you. ;)

 

Incoming SFALL1 of Text:

 

It's completely reasonable to expect UV players to be familiar with the map, or be prepared to learn it.

 

By "learn", of course, I mean "die and try again". :P

 

A big part of the enjoyment of any map, IMO, is the challenge of discovery -- figuring out how the map works, how fights play out, finding where to go, unearthing secrets, and all those good bits. On UV, it's totally reasonable to expect players to engage with the mechanic of discovery to a higher degree than HMP and below; it's just as valid as requiring a UV player to have good knowledge of monster movement patterns, finely-tuned dodging skills, and all that other skilled-player magic.

 

Relatedly, I typically thingplace for Ultra-Violence by "default", not 'cause I'm super-skilled (I definitely ain't), but because I already know where all the traps, secrets, and side routes are. ;) -- UV is basically the "I know what I'm doing" difficulty, HMP takes the edge off the "sneaky" evil bits, and HNTR and below are the comfy fun zone.

 

NOTE: There are occasionally people who make the counter-claim (intentionally or not) that every map must be completable on UV blind saveless FDA or The Map is Wrong -- ignore their ass. This is a vocal minority whose existence should not influence your designs in any way. :P

 

DOUBLE NOTE: Just to be clear, the above doesn't mean it's automatically a good idea to add deathtraps and unwinnable situations carte blanche and go "just retry or savescum!" to anyone who voices frustration. You should definitely ask yourself whether or not that instakill pit or surprise archvile mob trigger is actually fun -- BUT sometimes, amazingly, it actually is. ;) There's a fine line between trolling the player and pulling off a fantastic "gotcha!" moment, and straddling the line is an art of its own. There are entire challenge-based mapsets dedicated to the "try, fail, and try again" concept, often with hilariously unfair failure conditions, and they work. :P

 

Ammo starvation is a legitimate tactic...

 

...on UV, at least. ;)

 

Sure, some players don't care for this, but you can't please everyone and there's a very unique flavor to ammo-management-heavy combat that might complement your map well. Got a fight toward the end that's trivialized by the BFG or plasma gun? Maybe the problem isn't the monster placement, but there's too many cells in the map and players are always overstocked when they get there. Players trivializing your cramped maze bit with the SSG? Limit the number of shells the player has. And so forth.

 

...But don't flood easier difficulties with too much ammo.

 

The flip-side of the above. You usually want to provide easier difficulties with more of an "ammo cushion", but don't overdo it. Although it's a bit counter-intuitive, sometimes you may want to _remove_ a few ammo pickups on lower difficulties if there's significantly less HP the player has to wade through.

If you don't keep an eye out for this one, the player may end up sleepwalking through the map while holding mouse1. ;P

 

DISCLAIMER: These two issues are often a bitch in the ass to deal with if you're targeting both pistol-start and continuous play (i.e. basically every megawad/episode ;) -- there's really no easy solution there, aside from Test Both and just do your best. :P

 

DOUBLE DISCLAIMER: Backpacks also make this one considerably harder to deal with. There are even some folks around that explicitly recommend never giving the player a backpack ever -- "never" is a strong word, but it may be worth considering whether your map really needs that backpack; how does it play without it? Might work better. ;)

 

RELATED NOTE: ITYTD already has double ammo, so if "survival horror ammo starvation" is an explicit Design Goal of your map, the game at least provides a built-in outlet for players who want it. It's not a catch-all, but it _is_ always there. Unless you're doing fancy GZDoom stuff and rip it out I guess. :P

 

Don't remove the most interesting parts of a fight.

 

e.g. if you've got a well-crafted scene where a few archviles are unleashed in a sea of fresh corpses, and it's a bit too spicy on lower difficulties, it's probably better to subtract a vile or to, or give the player more resources, than to demote the archies to revenants or somesuch, since it completely removes the entire Thing that makes the fight interesting.

 

There are tons of different variants of this one -- it's a broad principle, but a good one to keep in mind.

 

Invulnerability spheres are trump cards for lower difficulties.

 

Can't figure out how to make your awesome monster mosh pit not a massive pain in the dick on ITYTD/HNTR? Try tossing in an invuln. 30 seconds of OP power fantasy is often more fun than a nerfed battle.

 

Megaspheres work in a similar fashion -- giving players a free 4x effective-HP boost is extremely powerful. I've gotten a lot of mileage out of the "Invuln on Easy, Megasphere on Medium, Medkits/Berserk on Hard" pattern; you might not even need to shuffle any monsters around if you're lucky ;)

 

"Meat Reduction" is only as effective as your "combat arena" allows.

 

This one takes a bit of explaining. :P

 

There are probably better terms for these, but by "meat reduction" I mean reducing the total HP pool of a group of same-type monsters by removing a few (e.g. reducing a crowd of 8 revenants to just 2), and "combat arena" is a catch-all for any part of a map a particular fight takes place in, whether it be a literal arena or a series of rooms/hallways or a mindbending abstract fuckscape.

 

When it comes to monster crowds, there's sometimes a "critical mass" where adding or subtracting Meat to the pool doesn't really make any difference in difficulty -- this is heavily influenced by your Arena shape. If you've got a very large circular room with some revenants in the center, it doesn't actually make a ton of difference difficulty-wise whether you've got 10 revs or 100, since in either case the fight is basically "circle strafe and don't stop moving or you Just Die". Got more of a maze-like area, though? The threat is now getting boxed into a corner, and adding more enemies absolutely increases the pressure on the player.

 

The above examples are a bit contrived, but there are all sorts of scenarios where adjusting the monster count in an area only affects the time it takes to complete the fight, not how hard the fight actually is. If done wrong, this could lead to UV simply being a "more tedious" version of HMP, or HMP accidentally becoming "whoops it's actually still as hard as UV" if the reduced meat is still just as lethal. ;P

 

A lot of MtPain's "monster window dressing" callouts fall into this category (i.e. a lot of time is spent cleaning up enemies on ledges which will never hit you if you're moving in a circle; the threat is the same whether there's 5 monsters or 50), and there's a phenomenon that often happens in high-monstercount maps where players lament "I've already beaten the fight but it takes 10 minutes to finish cleaning up." Less is more, sometimes literally. ;)

 

This is another general guideline with no hard and fast rules, but you can easily fall into this trap by leaning too heavily on formulas like "3 monsters on UV, 2 on HMP, 1 on HNTR". Guidelines, not rules -- they don't always work.

 

EXCEPTION OF SORTS: Monsters with built in "time pressure" mechanics (e.g. archviles and pain elementals) buck this trend a bit, since leaving them alone for long periods of time can have disastrous results for the player. You'll get much more mileage out of "meat-reducing" crowds of these types than others.

 

Archviles in particular need careful consideration -- they have a very steep"meat to difficulty" (lol) curve, i.e. the difference between 0 archviles, 1 archvile, and a trio of archviles is pretty massive. Without a rocket launcher or BFG, this goes straight from "no threat" to "seek cover or take heavy damage" to "seek cover or Become Die." ;)

 

 

 

That's all for now. I may jot down some more ramblings later, but this has been churning in my brain for long enough. So here, have a post.

[Fonze]

There's a lot of good thoughts in here already, so I'll try to keep my reply short-ish. For starters, I'd like to bring up that it's good to think about the capabilities of players playing at lower difficulties. That is to say, maybe UV can require people to have a good 6th sense wrt sound effects and dodging turreted enemies behind them while facing off against a tough encounter in the other direction, but maybe for HMP/etc that turreted enemy group could be changed to something easier to kill, or less punishing to fail to dodge, as an example. Of course it's a bit time consuming to attempt to tackle the underlying issues that lead to different difficulty needs and it'll never be perfect, so to that end it can also sometimes work out to make the difficulty levels match you as the designer on keyboard-only or on mobile controls. I suppose that's the wonderful thing about art: we can do it however we like and every way is as valid as the others. Just have fun with it 😃 That said, I tend to design my maps with:

UV is for me and the 0.5 people that replay my maps. Don't play the hardest a map has to offer then complain it's too hard. I've never not implemented interesting difficulty settings... use them! Lol

HMP is much the same, but with more supplies and perhaps some fewer key monsters. Sometimes reworked encounters altogether. I think making maps play out differently on each difficulty is a fun way to combat the fomo that leads to the UV-or-bust mindset and a good mapping exercise to see how different mobs and encounters play out in the space one has designed, while not feeling like you're totally divorcing yourself from the encounter you've planned to that point.

HNTR/ITYTD is either morer supplies and morest toned down resistance (less need for advanced tactics too) or it's a meme where I fill the map with rockets and pain elementals 😄🚀💥🫠

Alternatively, make HNTR/ITYTD the secret hard difficulty so people can satisfy that UV-or-bust mindset under the protection of half dmg and double ammo.

Edited March 13 by Fonze

[blueyosh43]

How often do people just not take into consideration other difficulties? When I start mapping I'm most likely just going to flag everything to appear on every difficulty. I'd rather just make my intended vision and have it at that, I'm not interested in compromising or adding stuff. What's the consensus on this? 

[Stupid Bunny]

Doing that obviously narrows the accessibility of the map but there’s nothing intrinsically wrong with that. I think it’s ok as long as you’re up front about it and expect that some people will be frustrated that the map is outside their skill range if it’s balanced to be quite difficult. Some people even do the thing where they fill the map with archviles or something on lower difficulties (PROTIP: don’t do this)

[Jayextee]

7 hours ago, Xaser said:

Ammo starvation is a legitimate tactic...

 

...on UV, at least. ;)

This. So hard. I'm probably a pain in the ass for resource-starvation on UV. Intentionally so. You find yourself without enough to continue? Coulda played smarter -- believe me when I say I test my maps on all difficulties. I don't run out of ammo, and if I did at some point during mapping/testing I'd add a cache somewhere to balance things out.

I give excess resources on skill 1/2, ample on skill 3, 'enough' on skill 4/5.

[baja blast rd.]

 

 

 

Guess why and you'll learn something.

 

edit: adding a hint 

 

@DoomGappy 

Spoiler

This is not in a player-accessible part of the map.

 

[DoomGappy]

18 minutes ago, baja blast rd. said:

 

 

 

Guess why and you'll learn something.

Gray trees become Revenants when they die?
 

Spoiler

I actually think it's because revvies have a stationery attack where they stop, so you substitute that for a path blocking object that the player will also have to strafe around?

@baja blast rd. now with the hint I'm supposing it's just because revenants sniping you from afar is annoying. Though that looks like a monster closet and you may use the trees as a spacer or marker of sorts for better organization. 

 

Edited March 13 by DoomGappy

[thelamp]

11 minutes ago, baja blast rd. said:

Guess why and you'll learn something.

 

The monster closet has more than one teleporter and the gray tree makes one of them unavailable? This controls the difficulty by changing where monsters will enter the play area

[Aurelius]

38 minutes ago, baja blast rd. said:

Guess why and you'll learn something.

Not sure if the same intent, but I personally use this to keep closets tight. Early on I noticed that if you just remove closeted monsters on lower difficulties, the closets become "loose" and monsters spend more time wandering in the closet than perhaps originally intended (especially true with big closets), which slows down throughput. That's not always a bad thing for lower difficulties, but if you specifically want quickly dispersing closets, this is a good way to keep things consistent across difficulties.

[Li'l devil]

First, I balance UV. I should be able to beat my map in one go with no saves. If it's not realiably doable, then I lower the difficulty.

Then, for HMP, first of all, I balance it such that I should be able to beat the map in the same way, but this time without using any help from the secrets. And then of course I also reduce the monster count by about 20%, put a bit more resources, maybe replace harder monsters with easier ones.

For HNTR I just reduce monster count by about 40% from UV and give even more resources. I don't think I ever test on HNTR, I just always assume if HMP is more than doable, then HNTR should be a cakewalk.

 

However, one rule I follow is that when I reduce the enemy count, I always make sure that the core of each encounter stays the same. That means if a map has a pain elemental encounter by design, I won't take out this PE on easier difficulties, no matter how much pain it causes.

 

But there were some maps I just didn't know how to balance properly on easier skills without taking away the fun, so they either had virtually no changes between skill levels, or, as is the case with one map, I simply changed the damage the player receives in MAPINFO.

[baja blast rd.]

Aurelius got it. That's a teleport closet setup and the blocking object keeps the closet from being unnecessarily spacious when many monsters were removed. 

 

More generally, it's a good idea to pay attention to how (removed) monsters interact with your geometry. Like, if you have those SWTW-style "dormant monster" closet setups with a fake wall tightly surrounding a group of enemies, unflag enemies from the back (or middle) rather than the front (or borders) so the invisible wall now isn't suddenly strange. 

[OliveTree]

ive said before that i playtest my maps on HNTR, that "if its not fun on HNTR its not fun at all", and i stand by that. the overall grammar of the map should be dramatic and surprising and enjoyable even without constant friction. That being said, i tend to give myself certain handicaps. Restricting myself to keyboard only controls is a common shortcut for me to simulate a lower-skilled player.

[Li'l devil]

Forgot to mention, but I think it's actually a good idea to make your map obscenely hard on UV only so that you can annoy the UV-only players!

[Dheta]

1 hour ago, Li'l devil said:

Forgot to mention, but I think it's actually a good idea to make your map obscenely hard on UV only so that you can annoy the UV-only players!

as a UV-only player that's what i want from maps so mission failed.

[Budoka]

15 minutes ago, Dheta said:

as a UV-only player that's what i want from maps so mission failed.

So do I, but you'd be surprised how many people refuse to play anything but UV and then whine that maps are unfair at the first sign of resistance. At any rate, I'll echo the general approval of the "balancing through power-ups" approach, it's a pretty great idea.

[rita remton]

1 hour ago, Li'l devil said:

Forgot to mention, but I think it's actually a good idea to make your map obscenely hard on UV only so that you can annoy the UV-only players!

i want the opposite, but have yet to succeed in finding that sweet spot of my maps being hard and fun at the same time. players could just opt for lower difficulties if they find a level too grindy, but i suppose uv is a fomo and ego thing for some, including myself :P
 

13 hours ago, Xaser said:

Difficulty balance ...

thanks! learnt some precious tips especially on the overflooding of ammo and "critical mass" thingy. will try to implement what i've learnt in my future maps :)

Edited March 13 by rita remton

[bioshockfan90]

Just from a player standpoint, some mapsets nail it right on the head, others can be tricky but still fun; I liken it to something like a Guitar Hero chart; different charters are gonna have different standards for what they would consider hard, and obviously this applies to mapping as well. A lot of the older megaWADs do it pretty well like Hell Revealed or Alien Vendetta - I think back then difficulty balancing was just a given, and then as time went on there was a brief stretch of time where people kinda just did whatever the hell they wanted; not to mention the rise of speedmapping events where it's hard to include difficulty settings under a time limit

 

It's definitely gotten a lot better over the past 10 or so years, a couple notable recent examples of good HNTR balancing I can think of are Machete by @A2Rob and Nostalgia by @myolden. They don't diminish the challenge per-se, just more powerups and maybe an invul or two to allow for a Cyber shot or two, just leeway for errors. Then you have sets like Tetanus or Daylight District, really all the Squonker works that kinda make it more of a run-and-gun affair on lower difficulties as opposed to really having to put some effort in. 

 

With the slaughtery stuff ala Ribbiks or what have you, I'm willing to up the scale a bit when it comes to difficulty scaling because it's a genre of map that's inherently harder, there do exist sets that forego the slaughter on HNTR for more incidental shit but it's pretty rare, I think most Ribbiks sets give you the same option of like "if you play on a lower difficulty, it'll still be hard just less exacting" and cr2.wad that came out this year did it pretty damn well I think. 

 

Sorry I didn't really mention anything from a mapper's perspective, just kinda putting out there that I know people put work into this kind of stuff and it's usually done at least somewhat well, sometimes exceedingly well and I appreciate all mappers that add lower difficulty support for those that play doom for fun first and foremost and don't feel like a massive grind or for it to be hell. 

[NecrumWarrior]

I usually make UV such that I can beat it while having fun. Lower difficulties depend on the kind of map. For more traditional levels I usually tweak number and types of enemies without changing with the resources. (On easy those Revenants can be Hell Knights or Imps.) But when I make Slaughter maps I usually don't mess too much with monster composition and instead provide more health and armor for lower difficulties. I almost never change the amount of ammo, though sometimes I will if a map is meant to starve the player on UV.

 

In truth I find implementation of difficulties very tedious and unfun, but necessary. I find myself more and more trying to do some of the work while I make a level so that I can shorten the tedium. To not do so is to gate off players who might otherwise enjoy your maps. 

[Stabbey]

 

3 hours ago, thelamp said:

 

The monster closet has more than one teleporter and the gray tree makes one of them unavailable? This controls the difficulty by changing where monsters will enter the play area

 

Even if that was the case, I wouldn't do it like that. Teleport destinations already have difficulty flags, so you could switch them up. In the meantime, monsters in the closet will be jiggling and with only a teleporter on one end - especially an end opposite where the player is - it'll take longer for the closet to empty out compared to a teleporter with exits on both ends. So perhaps THAT's the idea, to slow the rate of monster entry into an arena.

 

3 hours ago, Aurelius said:

Not sure if the same intent, but I personally use this to keep closets tight. Early on I noticed that if you just remove closeted monsters on lower difficulties, the closets become "loose" and monsters spend more time wandering in the closet than perhaps originally intended (especially true with big closets), which slows down throughput. That's not always a bad thing for lower difficulties, but if you specifically want quickly dispersing closets, this is a good way to keep things consistent across difficulties.

 

For quickly dispersing closets, I'd add in more lanes, with monster blocking lines and more teleporter destination sectors. More lanes also makes it easier to adjust for difficulty.

[Aurelius]

1 hour ago, Stabbey said:

For quickly dispersing closets, I'd add in more lanes, with monster blocking lines and more teleporter destination sectors. More lanes also makes it easier to adjust for difficulty.

Of course, but the sort of lane control you describe is difficulty-agnostic. My post was a direct response to rd's example of adjusting closet size specifically based on difficulty, and was in no way meant as a general method of speeding up closet output.

 

Plenty of good insights on all things closet in NIH's post from a few years back.

 

[fai1025]

here's what I do:

1. place monster for HMP

2. if I'm feeling lazy I remove some pesky enemies, if i have time I downgrade them

3. for UV I add more pesky enemies

4. I almost never touch resources as a balancing tool, but I'll change mega sphere to blue berries for UV.

[DoomGappy]

I'm really glad how many replies and points of view this got. Pretty good insights and very good practices discussed here. 

[Kinetic]

I usually predominantly add more health/ammo for HMP with some slight monster changes for super dangerous enemies, then on HNTR I keep the health/ammo from HMP, add a little bit more, then go heavier on switching up the monster composition. My idea was that HMP is UV but with a bit more room for error, and then HNTR is just overall a lot easier both in terms of the opposition and your own resources.

 

Solo-net is a difficulty that I just make absurd. If I'm making an easy map ala Scythe e1, on solo-net I'll turn it into Italo or Sunlust or worse. Hard, but still doable and fun for a very select few.

[fishy]

Normally I make my maps for UV.

I playtest them until I can beat them consistently, and once I get there I refine enemy placement and ammo/health balancing until I feel like it's a fair challenge.

 

Then for lower difficulties I normally change chaingunners for shotgunners/imps, archies to revenants, barons to hell knights, etc. If I have many enemies of one type, I keep the encounter the same enemy type but with just less demons.

 

I don't even bother with easy since the damage/health adjustments do all the work for me pretty much.

[Trigsy]

This topic brings up fond memories. During the first D&D campaign I participated in, I suggested that the DM add monster closets to create chaos, and he expressed concerns about balance issues. I'm certain that I had recently learned that DOOM was partially inspired by Id's founders' own D&D sessions.

[TheHambourgeois]

Others have said a lot of good stuff, I think in general there's a few general pointers to remember:

Keep the big set piece stuff in, even if its hard. Both the Cyberdemon and the Spider Mastermind are in the easy version of Gotcha, and there's the same number of archviles. I think that was a good decision.

Different "genres" of mapping will have different methods. In incidental combat wads, you can do more resource starvation and add more and stronger monsters at higher difficulties. In slaughter, you probably want to if anything replace monsters with weaker enemies that function similarly, maybe reduce some of the high target priority monsters, and mostly give the player more power on lower difficulties. Make sure you don't get in the way of the playstyle.

Personal approach here, and I tend to favor a scythe-y mapping approach generally, is that I try to have the monster count be 60%/80%/100% for skills 1,2/3/4,5. I usually do not replace bit nasty monsters like cyberdemons, though maybe reduce the count where there's multiple. I do a lot of substitution (replace half the shotgunners with zombiemen, etc. on skill 1, make sure to put more shells down). Then add a few more big powerups on easy, and maybe beef up some of the ammo drops and health pickups in big encounters on skills 1, 2, and 3. I admittedly take a kind of quick and dirty approach because I don't always have time to playtest thoroughly other than skill 4, but I have gotten feedback that this generally results in a fair experience on HMP.

[QuaketallicA]

I think Ultra-Violence is easiest to do first, since you know where everything is and tested the map yourself a bunch, so whatever the most fun, ideal sweet spot lies, that's UV. Then HMP aims to be just as satisfying as UV, but for someone who doesn't yet know what to expect or where the secrets are. So to be as punishing as you can, while still not dying and having to reload, without relying on secrets for health/powerups.