Balancing Lower Difficulties

[Spectre01]

I've seen a few times people mentioning that cutting monsters for HMP or lower is not the way to go. Would replacing "hard" enemies with lower-tier foes be the preferred solution then? I've seen a few wads that change Cyber turrets to a Manc or something or adjust the number of Archviles to tune a difficult encounter. What's a good way of reducing difficulty without making the player feel like they're not getting the full experience?

[scifista42]

I can think of 2 approaches:

1. Balance your map as well as you can with a high difficulty in mind, playtest it, remember which areas were the most challenging, and redo thing placement in these areas on lower difficulties to make them easier.

2. Balance your map as well as you can with a low difficulty in mind, playtest it, remember which areas weren't as challenging as they had a potential for, and redo thing placement in these areas on higher difficulties to make them harder.

[Deadwing]

For my wad I'm thinking of trying a different approach:

-> UV is the default difficult
-> HMP will feature easier gameplay, with less monsters and/or monsters replaced with weaker version (HK -> Imps, for example). However, ammo and health would be reduced so resources management would be still important.
-> HNTR will feature monster balance from HMP but with resources from UV, so player will not need to worry about ammo and health management.

[Cacowad]

Let's make an example:

You have a room with a pillar in the center, ammo is considered generous enough to allow a couple of misses for monster, an example of difficulty curve could be:

-HNTR: a revenant and a couple of imps, the revenant is the priority target, depending on the position of the imps you can use them as a shield for the revenant (with the possibility of an infight) or place them at the sides of him the prevent the player to strafe freely around the pillar

-HMP: tree\two revenants and a meat shield (imp\pinkie), the revenants are still the priority targets, and you still have your low target in the form of the meat shield wich can still cause infight and soak up some damage for you, while still pose a reasonable threat to you.

-UV: tree revvies and an arch vile, now the archvile is the priority target and the revvies are the (very agitated) meatschields. The difficulty of the encounter is dictated on where the monsters are placed around the room, are the revenants blocking the way? is the archvile behind the pillar or on the side? is he teleported behind you?

You may notice that the setup is the same in all the cases, what change is the threat level of the monsters used in relation of their positioning.
Having a similar experience for each encounter is indeed possible imho, as long you are not replacing a mancubus for an imp and then wonder why the player is suddenly able to walk away from the encounter by simpy pressing forward.

About cutting monsters... sometimes strenght is in numbers, and reducing the number of enemy spawned by a generic trap (be it closet or teleport) can make the fight easier as long as you are not cutting an essential monster, let's make another example :

same room as above, same with the ammo, at the start you have a single baron in the room. As you enter in the room, the trap activate:

-HNTR: tree imps and a pinky are teleported in, the baron should be the target, right? wrong, he is super buffed and take a while to take down, so the player may want to switch focus on the teleported monsters before killing the baron. If the player does not have a SSG a regular hell knight can fit the bill nicely.

-HMP: four imps, two pinkies and a revvie are teleported in. This is a bit of a crowd, the baron is still there trying to kill you, but you have to make space first.

-UV: six imps, two pinkies and an arch-vile are teleported in. You got the idea by now, kill the imps to make space, kill the archvile and then proceed to take down the baron.

This encounter follow the same formula as the one before it, you got your hight priority targets (imp\revenant\arch-vile) and your damage sponge\low priority targets (pinkies, baron).
Now, what happen if we remove the initial baron for the lower difficulties? The dinamic of the fight changes a lot, you don't have the same constant threat of the baron on the player, wich is able to position himself freely in the room and enjoy the protection of the central pillar all by himself.

Welp, this went longer than expected, i hope this make some sense.

TL;DR: both reducing the monster count and replacing monsters are valid strategies, as long that the coreografy of the encounter remain intact.

[Alfonzo]

It's an important process that's easily botched if you're not willing to put in the hard yards. The key focus is to ensure that the entertainment doesn't crap out whenever you drop down a level - an overlooked rule. Easy does not have to translate into boredom. Here's the low-down:

Decide where the fun stems from. Is it the cyberdemon in the middle of the room piling pressure onto the player and giving him the run-around? Good - keep that cyberdemon. Provide health, powerups... whatever you think's necessary to have the less able players get by and have fun while they're doing it. If it's still too hard having tested the encounter, consider replacing the monster(s) with something that's less testing but still satisfies the same gameplay setup. In our example that might be a baron, although frankly that's a massive step down and isn't a one-for-one swap what with the lack of splash and aggressiveness and so on. The gulf that opens up on the ammo ledger is accounted for by the lower-skilled player's presumably missing more, and in any case... well, it makes sense to have a few shells left over if you're playing on HNTR versus UV.

Add an item; spare an imp. Keep the body count and the gameplay setup as close to the most entertaining iteration as possible. Feel free to remove monsters in areas where space is at a premium. Don't remove trash gibbables and other filler encounters if they're only there to be blown up in idle corridor combat; blowing things up is fun, and turning a cacodemon into an imp in a one-dimensional corridor has no effect on difficulty - just time. If your map is made of nothing but such encounters, then even the overall balance through provisions, once again.

Don't neuter your map; don't remove its teeth. Just give the player a sleeve for the beast to bite into.

[kuchitsu]

Can we assume that a weaker player is interested in the same kind of fun? Maybe they would rather just walk around, appreciate the scenery and mow down easy targets without having to deal with some pesky cybersniper all the time. Perhaps a player's abilities somewhat dictate his preferences as well, which would mean that lower skill levels should cater to other play styles instead of being more accessible versions of UV.

[40oz]

Some quick and dirty solutions I do to make something easier:

If powerful weapons are accessible on the map, such as rocket launcher or super shotgun, make them easier to obtain. If you gotta go half way through the map to get to it, flag those weapons as hard and put new ones closer to the start area flagged as easy/medium.

Flag stimpacks as hard and put easy/medium medikits next to them.

Change green armors to megaarmors, or make armor items available earlier on.

Change some revenants to hell knights or cacos

Change some chaingunners or shotgun guys to imps. If shotgun guys drop in some extra shells too.

If stronger weapons are available, dot a few extra rockets or cell charges around.

Be careful with changing monsters up too much. More monsters rarely = more hard. A thousand imps and two hundred imps is relatively the same difficulty if you have enough space to move. I definitely wouldn't change a cyberdemon to a mancubus. A major quality of the Cyberdemons difficulty is his HP consuming a ton of ammo and time. There aren't many monsters who possess that trait so make megaarmors, or soulspheres or a more powerful weapon available on easier skills instead.

[Alfonzo]

kuchitsu said:

Can we assume that a weaker player is interested in the same kind of fun? Maybe they would rather just walk around, appreciate the scenery and mow down easy targets without having to deal with some pesky cybersniper all the time. Perhaps a player's abilities somewhat dictate his preferences as well, which would mean that lower skill levels should cater to other play styles instead of being more accessible versions of UV.

I wouldn't recommend trying to have your three difficulty settings cover so wide a range of potential preferences on the off chance that someone hasn't been reading the text file, or in an attempt to please everyone and no one, thusly. For the game being advertised to suddenly give way to a completely different experience the moment the player decides he's not up to the task would be perplexing, if not dishonest.

Having said that, the catering towards other play styles precisely is an interesting idea. I've toyed with the possibility of designing a level which has its three things settings designed around opposite philosophies in this respect - a tyson vs. plasma vs. your traditional "thoroughfare" approach, for example. I wouldn't know where to start with the required layouts and so on, but it would make for an interesting experiment.

[Bauul]

In an ideal world, the different difficulties would be balanced through a careful use of alternative enemies, different number of enemies, different amount of health, different amount of ammo, and different placement of health and ammo, and different rate of collecting weapons.

However, this is a lot of effort, so it kind of depends whether you're willing to put the time in to get it right.

If you find difficulty settings utterly tedious (and after all, mapping should be fun) then simply setting fewer monsters to appear at lower difficulty does more or less work ok. Start with UV - put in all the enemies you plan to and just enough health and ammo to cover them, and then start deselecting enemies.

The end result is HNTR players will end up with loads of ammo and health, but it's meant to be an easier difficulty setting anyway (plus they're more likely to take a fireball to the face, so need a cushion). Same deal for HMP, although they'll have more enemies to contend with.

Yeah it's not the most elegant approach - three Mancubi on UV vs. two on HMP vs. one on HNTR isn't exactly a nuanced approach, but it will work most of the time. Make sure you play-test it a lot though to check you haven't neutered the approach altogether.

[baja blast rd.]

Alfonzo's advice is great and is consistent with my own philosophies.

Something to keep in mind is that HNTR and HMP should be fun for you too, if you design around UV, just with a more aggressive playstyle. (So you need to be a versatile player. Probably worth it to dabble in max running.) If you add difficulty settings purely to cater to a wider audience, and you don't really plan on ever playing your maps on them outside of testing, they are probably going to be missing something, even if you balance them "well".

In the past I used to do things like replace the more difficult turretted 'trons with clusters of imps so that monster counts would stay similar on HNTR and HMP, but, eh . . . four imps don't look so imposing, so maybe I'd use two shotgunners instead? It's always really important not to just nerf setups, but to think about how they look. Setups shouldn't look "obviously nerfed" unless you want that. It might help to learn how to make setups that are imposing but easy.

I'm very much a mapper who designs HNTR and HMP for myself too. I feel like in the ideal map, lower difficulty settings would be easier but also a somewhat different (and not worse!) experience, with lots of different nuances in monster composition and gameplay, such that I'd find interest in replaying the map on all difficulty settings, depending on what I felt. Another strong approach is simply to make HNTR and HMP a nerfed UV as far as the experience goes, so for example a typical player on HNTR might get the same experience as an above-average player on HMP, who gets the same experience as a pretty good player on UV. Both approaches have their pros/cons but I prefer the first.

[Vorpal]

Typically I aim for UV to be actively trying to kill you, utilizing enemies for their strengths in their roles, and some element of resource management (though I try to avoid "survival horror" resource management).

HMP I try to tune as if it was doom2 UV - large amount of excess ammo/health, good amount of fodder to mow down effortlessly, and some strategically difficult enemy usage.

HNTR I try to make it so even grandma should be able to flip that exit switch. My calibration is usually to use keyboard with no use of running, and find none of the secrets. After I can beat it like that, remove maybe 50% more of the remaining enemies and add even more ammo/health and voila, granny mode.

[dobu gabu maru]

Like rdwpa said, all difficulties should be fun for you to play. How I personally dictate the difficulties in my maps is thusly:

UV is 5/5 intensity, will be difficult even if you know the map.
HMP is 4/5 intensity, will be difficult if you don't know the map. This is my ideal setting for most players.
HNTR is 2.5/5 intensity, might be difficult in a handful of instances, but is designed to see you through to the end.

I don't think it makes sense to have such a wide gap between the difficulties—for instance, I think ensuring that grannies can beat HNTR is perhaps going too far (that's what skill 1 is for!)—but as long as the gameplay remains interesting, anything can work.

[Steve D]

This is one of my favorite topics, as rileymartin knows, so thanks for starting this thread.

A lot of great ideas in here, but my method is exactly what Bauul describes. I call it The Power of Threes. It's basically a linear reduction UV Lite approach. I always design maps for UV, and in doing so, for the most part I add monsters in multiples of three. So there might be 3, or 12, or 144 monsters of a given type in a specific location. For HMP they are pared down by a third, and HNTR by two thirds. This may seem to make HNTR too easy and boring, but because I tend to use a high monster density, there's actually still a lot of action on HNTR and the "flavor" is the same as UV, just not so many monsters.

The linear reduction only happens with monsters. Health and ammo is exactly the same as UV. Maybe an HNTR player would like a Soulsphere instead of a Medikit before facing that Cyb, but if the Soulsphere isn't there on UV, it isn't there on HNTR, either. At least they'll have tons of ammo. ;)

I resorted to this method back in the late '90s when I was testing a map I'd already gone through many times on UV and had it pretty well down. I had already used some variant of Troublesome Monster Theory to reduce the hordes on HMP and HNTR. In particular I had removed a lot of dangerous Sergeants. I had also switched around the health and armor. When I playtested HMP, I was shocked to find the map was impossible to beat. I had removed too many Sergeants and now there wasn't enough ammo from dead hitscanners to kill all the heavy meat. That's when I decided to leave the base level of health and ammo the same, but to reduce the monster count in a linear way for all types, in order to avoid unintentional ammo starvation. I had an extremely demanding job in those days and simply lacked the time to even playtest each difficulty setting, so linear reduction was well-suited to my needs.

FWIW, I've never heard anyone complain that HNTR was too easy or too boring.

I disagree with Bauul in one sense; linear reduction is in fact the most elegant way of setting difficulty. There are 3 basic difficulty settings, so what could be more direct, simple and elegant than assigning 1 more monster for each higher level of difficulty? The approaches suggested by scifista42, Alfonzo, 40oz and Cacowad are not more elegant, instead, they are more sophisticated.

Something worth noting is that UV, HMP and HNTR are terms with no set, objective meaning, they are always relative to the mappers and maps one encounters. I always play a Demonologist map on HNTR, for example, because I know that UV is too much for me. It's not that UV is too hard, it's that it's too hard for me. I've seen it a million times that players who lack the skill to tackle the hardest maps on UV do it anyway and then complain that it's too hard. They might find that same map far more enjoyable if they go down the difficulty scale instead of being stubborn. Then again, I can say that my nastiest maps are more difficult on HNTR than the majority of maps I've encountered on UV. Unlike Vorpal, I don't support a Granny Mode. ;)

That said, the next question is how much time can you reasonably expect a mapper to devote to sophisticated approaches of encounter design for all settings? How many people actually play HNTR? How come they don't show up in playtesting threads? Why cater to them at all when we're busy people and this is a hobby, meaning we don't get paid for this? Romero got a Ferrari. We won't. ;D

Then, as Kuchitsu said, we can't always know what an HNTR player is looking for. They are not a monolithic block of people. Eris Falling plays HNTR even though, as his performance in the DW Ironman League shows, he's clearly capable of playing UV. Maybe he just wants a more relaxing stroll most of the time. I've also seen players on YouTube who clearly use only the arrow keys and get wasted by simple imp attacks. They have no choice but to use lower settings.

I like rdwpa's idea that we should like our own maps on every setting. On those rare occasions when I've played my maps on HNTR, that has usually been the case, especially in the nastier maps where UV is aimed way above my skill level.

I do plan to use some of the ideas espoused here by scifista42, Alfonzo, 40oz and Cacowad. In particular, I'll probably focus more on leaving in larger numbers of low-tiers in order to keep the monster counts on lower difficulty closer to those of higher difficulties. However, I'm disinclined to replace the tough monsters with weaker monsters, as opposed to my usual practice of just reducing their numbers. I want those Revvies and Mancs in there since otherwise it's going to feel wrong to me if I use Imps instead.

It's an evolving thing. ;)

[Pinchy]

Over time my methods have changed, but here's what I currently use:

Easy > Medium > Hard

Monsters

Hitscanners

Former Human > Former Sergeant > Chaingunner

In some cases a wall of Former Sergeants is more dangerous than Chaingunners because they can potentially all fire at the same time/very close together and create big burst damage, preventing your escape. They lose power at range, though.

Projectile

Imp > Hell Knight > Mancubus

Upping power and health with each step.

Strong Projectile

Hell Knight > Arachnotron > Mancubus

Arachnotrons can be more powerful than a Mancubus at range if firing at open areas with no cover, as they create a wall of plasma, the steeper the angle, the harder it is to avoid.

Blockers

Demon > Spectre > Spectre x 2 (multiply x 2 as needed)

Demon > Pain Elemental > Pain Elemental x 2

Imp > Hell Knight > Baron (rare use)

Flyers

Lost Soul > Cacodemon > Cacodemon x 2 (increase as needed)

The Pain Elemental is unique in that it's danger is based on time uncontested, if you can get into melee range, the availability of a berserk pack, chainsaw, rapid fire weapons that can stunlock them or a BFG (which can pierce through/one shot them, so the Lost Souls blocking ability is lost). You can "lose" via attrition or just getting blocked into a corner/sandwiched between the lost souls and high damage monsters. They also counter the rocket launcher well.

Most of the time I just increase their number, which increases the flood speed, but you can also tag teleport Things to adjust the position they teleport into a sector, or in UDMF, pretty much anywhere. This will change the free time they get to spawn Lost Souls, increasing the skill needed to get them under control.

Horde Spawns

In a case where you want to mass one monster, I divide the number down equally: 12 > 8 > 4 as an example. I build wide but narrow monster closets for this. Add additional closets and spawn points to increase the speed of the spawns from one location. The lowest number of horde spawn should still be the minimum amount to be a fun mass to crush through and dangerous, the idea is that the more there are, the quicker you need to apply your shots and the better they need to be, requiring more skill to avoid being overwhelmed or whittled down with hitscans.

Bosses

Cyberdemon > Cyberdemon x 2 with extra items > Cyberdemon x 2

So far I have only used the Spider Mastermind one at a time:

Spider Mastermind with a easily obtained BFG > Spider Mastermind with extra cells > Spider Mastermind.

Items

In cases where you are not able to easily change a monster, typically in a single monster space and consider it too powerful, you can add items to adjust.

Medkit > Stimpack > Nothing

Good for hitscanner or high hit chance fights.

Megasphere > Soul Sphere > Medkit

I use this in long fights with waves where the player is locked in and can only use whats in the arena.

Weapons

These are a special case especially with available ammo types.

I move each weapon one main room or section earlier for easy, so far have not needed to do this for medium as there are enough things I can tweak with.

Design

So far I haven't tried this, but you can have monsters activate things, the simplest is to put a monster in a sector near to a door so it can hear you fire and will move to a wall and activate a trigger in it's room This can activate a door with action 1 as an example DR Door open wait close (also monsters) letting monsters in earlier or access to both sides.

There are more setups that can be done with this the further up the source port level you go, that can make things easier (teleport in items/barrels, expose monsters to barrels/crushers) or harder (teleport barrels near the players shooting area, put monsters into advantageous positions)

Note

There are of course a lot of unique setups required based on the geometry and mechanisms available, this is just a general guide I go by and is useful as a starting point and speedmapping.

[Bauul]

SteveD said:

This is one of my favorite topics, as rileymartin knows, so thanks for starting this thread.

A lot of great ideas in here, but my method is exactly what Bauul describes.

Yep, I think you and I approach this in exactly the same way. Perhaps saying it wasn't "elegant" was the wrong turn of phrase: i meant simply reducing monster numbers by a factor of three does seem pretty one-dimensional compared to some of the approaches outlined here, but the reality is it *works*.

As an aside, it wasn't until I watched a few less-good players (i.e. those who'd choose HNTR or even ITYTD) play a few maps that I realized they consistently need more ammo and health per monster than a better player, as they miss more and take more damage. So having fewer enemies than UV but the same amount of health and ammo actually ends up balancing remarkably well for the lower difficulties.