Musings on difficulty curve

[PrimeJunta]

I've been getting my teeth into a D&D classic I've neglected, namely, Temple of Elemental Evil (yes, with the Co8 fixpack). It's got me thinking about the difficulty curve in games of this type.

 

But first, brief impressions. I'm not very far in, just cleared the Moathouse and arrived at Nulb.

 

The combat, when it starts, is seriously good. Far and away the best in any D&D game I've played. Tactics matter. Also, I've rediscovered the phalanx.

 

Man is it pretty. No other party-based D&D game I've played comes close, including the latest iterations of NWN2. This makes me really salivate over PE; if they can make it as pretty but at full-rez on my 27" iMac, woo-hoo. Everything is pretty. The monsters, the lovingly-crafted avatars, the cloaks, the animations, the environments. Everything except the UI which looks like it's a development placeholder, which it probably is. Nice sound effects and music too.

 

Plot? What plot? Never mind. I do like meself a bit of plot though, and this would've been way better if Troika had put some thought into it. Plot-wise ToEE is like porn: it's only there to provide an excuse for the action.

 

And difficulty.

 

This is actually really problematic. As in I don't know what my opinion is about it anymore.

 

The game starts out incredibly punishing, just like BG and BG2. It took me a quite a few tries to get my party of first-level delicate flowers to level up. I started playing on Ironman because I'm an inveterate save-game abuser, but that was clearly a bad idea. Ironman clearly requires either godlike D&D skills or metagame knowledge, remembering what's about to be sprung on you around the next corner. So after about a dozen tries I gave up and started one in normal mode. Much better.

 

Then it gets easier. Like, a lot easier. The difference between level 1 and level 2 is huge. By level 3 I was barely dying/reloading at all. I hope the difficulty will ramp up again as I move into the next area as this is getting kinda boring.

 

I still think that kind of punishing initial difficulty is a bad idea. It'll just get lots of people to give up in frustration. If I hadn't had so many similar games under my belt, there's no way I'd have the patience to learn to get the stupid party up to level 2. A game that has enemies that one-hit kill you, and doesn't in some way let you know that you're about to go somewhere really dangerous, is poorly designed IMO. Fortunately we haven't seen much of this lately so it's unlikely PE will be done like this.

 

On the other hand, that kind of difficulty curve gives a really tangible feeling of progression. Splatting that giant + bear that obliterated my party in, like two or three rounds when I first came across it felt like a real achievement.

 

I think treading that line has to be one of the hardest things in game design. Go too far on one side, and it's dull as dishwater, a mechanical, repetitive grind of easily splatting forgettable enemies. Too far on the other, and it gets way too frustrating to be fun. What's more, since you're designing it and know it inside and out, you'll have to rely on playtesters to find out exactly how hard a game you're really making. It's also an area where IMO most games of this type fare rather badly. Mask of the Betrayer was pretty good I thought. Fallout was pretty good (Fallout 2 not so much.) Most others fall fairly far from the sweet spot in one direction or the other.

 

I wish there was a patent answer to this question. I don't think there is though. I do hope Obsidian gets it right.

 

(For those curious, I'm playing with a party of two fighters with a couple of levels of barb for rage and fast movement, one ranger with a level of rogue to get Open Locks and Disable Device, one cleric, and one druid. No arcane casters on purpose; they're especially useless at low levels and this is low-level D&D. Once I got those sad sacks to level 2 it's been working out pretty well.)

[Jarmo]

I've made a few notes of my playthroughs on gog.com forums, but yeah, you pretty much nailed it.

Especially how darn pretty the game is.

 

The difficulty kind of spikes all over the place, but the first levels are the most painful ones.

A couple of horribly impossible battles later on.. and ones that might turn out real hard without area effect spells...

 

Overall though, the game tends to get easier as you progress.

Bad on one hand, but on the other hand, I can't think of any other game that gives you the same kind of feeling of having gained power and skills.

After the hard time with the first ogre you meet, it's rewarding to meet 4 of them at once and simply annihilate them in a few strikes.

[Kjaamor]

 No arcane casters on purpose; they're especially useless at low levels and this is low-level D&D.

 

...until you get Fireball, and then what was left of the difficulty drops right out.

 

I've owned ToEE twice now, and both times I've suffered similar feelings. Really liked it at the start, but then it drops out fast. It doesn't help that if you're playing anything other than lawful/neutral evil (or chaotic neutral, I guess), you run out of quests by about a third of the way in.

[PrimeJunta]

As an afterthought, Arcanum is a pretty good example of a game that manages to fail in both directions. As Chris Avellone has demonstrated, it's impossibly difficulty if you roll up a squib – which is very easy to do unless you already know what to expect and how the game works – but once you find out about Harm it becomes ridiculously and tediously easy. It's kinda amazing that they managed to fail so spectacularly. Real shame too as the game does most other things really well.

 

@Jarmo, if you were referring to my choice of not having arcane casters, I hope my cleric+druid will get the job done. They do have area-effect spells, just not  so many that deal direct damage. But area buffs/debuffs/status effects are really effective. I've already made use of Bless, Bane, and Entangle actually. Entangle is pretty neat with the reach weapons that let my characters stand outside the radius and poke at stuck things with impunity.

[Iucounu]

For a group of only level 1 characters, the beginning is indeed quite hard in TOEE, especially when you get ambushed by numerous tough monsters while travelling, or while resting, it's almost guaranteed that someone bites the dust. I also once tried Ironman, and died relatively quickly when a couple of giantspiders ambushed me while resting. My whole party failed their saving throw to their net ability, and was then finished of one after another.

What is especially annoying there is that the monsters always spawn directly next to the first character, not with some distance to the party like in the IE games, so it's hard to protect your more frail casters.

 

However, you can make the game much easier for you by simply hiring one of the many level 3 and level 4 mercenaries in town. Actually, right at the beginning one of them stumbles into you, a competent fighter with over 40 hitpoints, and hight AC. That guy could probably do the first few quests all by himself. 

 

If your party consists of only few charcters, there might also be the possibility to level up without having to fight, simply by solving the quests available in town.

 

BTW, arcane spellcasters are far from useless even at low levels. The level 1 spell Charm Person is golden (as are most enchantment spells), and helped me out on many occasions.

Edited October 9, 2013 by Iucounu

[PIP-Clownboy]

The Moathouse is a serious kick in the ass if you're not prepared. 

 

I really hate the opening village though. Bunch of boring fetch crap that you sort have to slog through in order to get that first level or two. The game becomes amazingly better once you reach the TOEE actual. Puts Watchers/Durlags to ****ing shame, to say the very least. 

[Althernai]

The level 1 difficulty issue is a generic problem with literal translations of D&D to CRPGs. At higher levels, the damage dealt to characters before death is a function of many dice rolls so you quickly run into the central limit theorem and can strategize accordingly. At level 1, the battle can be decided in a few rolls so there's much less strategy -- the outcome is much more likely to be based purely on luck. The "difficulty curve" is a side effect, the real problem is that this is a lousy mechanic probably intended to be mitigated by a human DM. Since Project Eternity is not using D&D, I don't see why they can't avoid the root of the problem and give characters enough health and/or stamina for the central limit theorem to be applicable throughout.

[PrimeJunta]

I didn't actually do any of the fetch crap. I just heard about that meadow and the moathouse and went exploring. I leveled up by splatting those ugly frogs outside the moathouse, skeletons on the meadow, plus a bit – but not all that much really – camping for wandering monsters.

 

I think there would be a quite a few ways to make an exciting first-level D&D game without having to do boring first-level crap. I have in fact run a few myself. They can even have some combat. It just doesn't have to be mortal combat. And naturally there are any number of fun non-combat challenges that you can throw at your players.

 

The first serious adventure in my campaign involved holding back a nighttime attack until the town guards showed up, followed by a two- or three-session ghost story/murder mystery. It would have been theoretically possible for someone to get critted and die in the combat due to no fault of their own, but it would've had to be really unlucky, and it didn't happen.

[Stun]

Plot? What plot? Never mind. I do like meself a bit of plot though, and this would've been way better if Troika had put some thought into it. Plot-wise ToEE is like porn: it's only there to provide an excuse for the action.

Yep. This is what happens when you take a Greyhawk Module and strictly translate it to a video game. The Pen and paper version of TOEE was created as a bare-bones. Gygax's intention was to allow the DM to flesh out an epic, politically based storyline using the various plot figures and factions in the module.

 

But Troika either didn't get the message or else just didn't care enough. Their idea of "fleshing out" was to bombard the player with dozens of the most boring, moronic, fragmented fetch quests ever.. So what we basically got was a computer version of the module, page by page. Utterly faithful rendition of it, and little else, and therein lies the problem. As for the game itself, had the graphics and combat not been so breathtakingly brilliant, the game would have been an embarrasing failure.

Edited October 9, 2013 by Stun

[Jarmo]

I didn't actually do any of the fetch crap. I just heard about that meadow and the moathouse and went exploring. I leveled up by splatting those ugly frogs outside the moathouse, skeletons on the meadow, plus a bit – but not all that much really – camping for wandering monsters.

 

I think there would be a quite a few ways to make an exciting first-level D&D game without having to do boring first-level crap. I have in fact run a few myself. They can even have some combat. It just doesn't have to be mortal combat. And naturally there are any number of fun non-combat challenges that you can throw at your players.

 

Co8 modpacks added content comes with velkwood bog, a pretty decent short baddie hunt involving a bunch of kobolds and goblins to whack.

Intended to be done instead of the fetch&carry stuff from the village.

 

That and all other added content comes from talking to the village smith. 

[Sabotin]

Yeah, I think it's more of a problem of the linear leveling system in DnD.

 

At the start, the dice rolls are like 90% of the whole number you use somewhere. Later you at least have a chance for things to even out (and even later the dice are almost unimportant).

 

Also if you look at it this way: going from lv1 to lv2 is a 100% increase, going from lv10 to lv11 is only a 10% increase in levels (but by then you get more powerful stuff per level). If I remember right most games are aware of these issues, starting you off at level 3 or something after the tutorial.

But I don't necessarily agree with this, because often it feels like that's just something you want to "get over with", while I think the low levels should be there so that a player gets some sort of base game play (positioning, function of classes/abilities, timings, etc.) which is then upgraded and expanded as you level up. On the other hand, at these fragile levels you're just too weak to test anything without reloading a lot and if it's made that much easier then there's no need to try anything other than headbutts.

 

It's also probably pretty subjective how the difficulty should be. Some people don't like to waste lots of time on a random lv1 enemy (me), while others may want to try dozen times to best it (me 10 years ago). The first time I played BG I must have reloaded a dozen times for that stupid assassin at the Friendly Arm Inn, to kill him before he could one shot someone from my party with his 3(!!!) magic missiles.

[mcmanusaur]

Unfortunately players' and characters' growing ability to handle challenges as they become accustomed to the game and their yearning for a sense of achievement and progression can be opposing forces in RPGs. While an old-school arcade game might be able to get away with continually becoming more difficult over the course of the game, I don't think that really suits the RPG design personally because save-scumming too often tends to break flow and immersion. That said, you can get away with a short "easy learning" period at the beginning of the game, which is probably preferable to the situation you describe. Given both character progression and also the progression of the player's abilities over time, it can be quite a struggle to prevent a game becoming easier over time. However, most games seem to either settle into a somewhat consistent difficulty trajectory or alternate between more and less difficult segments, relying more on increasing complexity than increasing difficulty to maintain the player's interest.

 

Personally I have no reservations about playing on easier difficulties in RPG's myself because it serves what I hope to get out of the genre, which focuses more on stuff like immersion and exploration over gaining a sense of triumph from gaming the system.

[Lephys]

The difficulty really doesn't need to stray at all, relative to your toolset. It's kind of like puzzle games. If you get a new tool, then a new need for that tool presents itself. It can still be pretty dynamic, but, if you've got 3 tools at level one, the difficulty should be pretty moderate as long as you're considering the use of all three tools. If you've got 25 tools at level 5, then the difficulty should remain the same, so long as you're not still just trying to use the same three tools. You don't have to require players to react even faster, or use an even smaller selection of tools in a given situation. You just have to engage them with things that say "Hey, this is pretty tricky when just tackling it blindly, so you might want to puzzle out, for just a brief period at least, what types of things work the best, and which tools to avoid using in this particular situation."

 

It's easier said than done, I know. But, the idea I'm trying to get at, specifically, is that you don't need to increase the relative amount of player skill involved (reaction time, micromanagement, etc.) as the difficulty goes on. You don't need to have someone master juggling, THEN stand on their head while they juggle, blindfolded. You just need to let them get a little used to juggling 2 things, then add a third. Or change what the objects are. Not add a bunch of extra things to worry about simultaneously. "Okay, now you're going to juggle, AND count to 1,000 in increments of 13. GO!"

[PrimeJunta]

@Lephys, I kind of disagree about that. Valve does exactly that extremely well. They introduce new tools little by little, and every time after you get one you get a situation where it's useful, and then it builds on those to create more complex situations. Half-Life 2 did it. Portal did it. And I found both kind of monotonous and grindy.

 

I much prefer varied difficulty. Perhaps my ideal curve is a relatively easy start that introduces the mechanics and pulls me in, followed by troughs and spikes, with the possibility of going into really tough areas by choice -- but with the game clearly communicating that those areas you're about to go in are dangerous and you might maybe want to level up a bit first.

[Osvir]

Hmm...

Early Levels:
"Oh **** an Ogre!"

Mid Levels:
"HAH! ONE Ogre!? Feel my blade!"
- Two more wild Ogre's appear -
"**** THREE Ogre's!!"

I'd also wish to chip in an idea about "Melee Engagement" that I'm surprised I haven't brought up. But something to make combat more interesting for both Player and AI alike would be to add a slight "ToHit" bonus the more targets hit the same target.

Basically, if you fight 1-on-1 against a Goblin, you'd have normal "To Hit" chance and "Resistances" etc.etc. all that Deflect, Reflex and so on would be the same. But as you go up against 2 Goblins instead (1-on-2) you could start seeing some invisible penalties. Your character has more trouble Deflecting 2 characters than 1, and reacting to both of them also becomes harder, not to mention Psyche should get a little bit lowered as well.

Additionally, if you are fighting 2-on-1, you should be at an advantage, and your Psyche goes up (Team effort), Deflection becomes way easier and Reflex (reaction) becomes much easier, because if one of your guys slips up, the other can react to it.

This would make the game potentially more balanced and more tactical. In terms of difficulty it'd make 3 Ogre's more terrifying than a single one at later levels. It would at the same time make Ciphers (for instance) a viable option as a second attacker in melee combat. A Fighter can take care of two of the Ogre's for a while (because he has abilities which allows him to deal with several enemies at once, maybe he wouldn't get any penalty for having 2-3 enemies attacking him at once) whilst a Monk or Barbarian can handle the 3rd Ogre. The Cipher can then sneak up on the 3rd Ogre and gain a bonus to attack and everything because there are 2 attackers.

I don't know if any of the IE games had any of this, but sometimes in the IE games it felt as if it didn't, and sometimes it felt as if it did. It gets silly when you have 4 characters hitting a single enemy, and almost all dice rolls miss. Shouldn't it be easier to hit a target if you group up on it? Regardless, I believe it could spawn some pretty interesting AI play, for instance, the 3 Ogre's grouping up and doing a sort of focus fire on one of your party members.

Different types of enemies could have a "2-in-1" type of deal as well. The Ogre could be counting as a "2 Creature" so having it attack your Monk would naturally give the Monk a penalty to Psyche, Deflection etc.etc. but hitting it with 2 characters would give you normal values, and 3 characters could give you a bonus in this case. 3 Ogre's would, in this case, count as the same value of 6 Creatures. It quickly gets terrifying, even at mid- to late-levels as more and more enemies begin to appear in mobs.

[PrimeJunta]

I don't think the IE games did that, except maybe for morale. The D&D rules certainly don't. What the D&D rules do have are rules on flanking bonuses, flatfooted, and that kind of thing, and at least some of the games we're talking about did implement those. ToEE certainly did.

 

I kinda like the idea though. I just wouldn't like the bonuses/penalties to be invisible. It is a game after all and I prefer games where I don't have to guess what the rules are.

 

Tangent: The Numenera PnP system has an interesting way of dealing with this, by the way -- the GM can have mobs (between 4 and 10 critters, depending on size and coordination) attack as single creatures that are two levels higher and do double damage. Totally unsuitable for a computer game of course, but it's a really elegant solution for a tabletop game, especially if you're talking through the combat and not using figurines. It makes mobs of low-level critters potent threats -- if you have 3 Armor a level 3 critter will only be able to harm you on a crit, but with that rule a mob of them would do 6 damage, which gets nasty fast.

[Osvir]

I kinda like the idea though. I just wouldn't like the bonuses/penalties to be invisible. It is a game after all and I prefer games where I don't have to guess what the rules are.

I'm kind of bothways on the invisibility thing. The rules could easily be explained in-game in a sort of immersive manner as well. In a "it makes sense" kind of way a la "If you go up 2 guys against 1 guy you'll work better together, and the 1 guy will probably get his ass handed to him". But then again, could be nice to get some sort of representation of it as well, but how and where? With all the "dice rolls" of hitting and dodging already, where would this thing fit in without cluttering it up? 1 solution could be a "Morale" meter.

 

The lower the "Morale" meter is, the more penalties your character gets. A sort of feedback on your decisions. If you send 1 Wizard to deal with 2 Goblins, you might see the "Morale" meter quickly go down, but if you send him back, it would replenish itself slowly. If you send in 2 Fighters to deal with 1 Goblin, you could instead see the "Morale" meter go up steadily or stick steadily at its capped neutral value. 

 

Here's an idea for a sort of "Morale Meter":

 

Good decisions make your guys happy, and bad decisions make your guys sad. This is also why I think it'd be more difficult if a sort of "Morale Meter" is invisible. Because if I send in my Fighter into a fight and I see his Morale is dropping (or he's getting some penalties of some kind), then I could just send him back and send something else at the enemies until I see a "growth" in the "Morale Meter". Similarly, if my Morale Meter is going down, then I know that the Morale Meter of the enemy is going up. Well, in this idea it would~ so there's a potential that something like this would give the Player too much feedback on their decisions.

 

1. Danger Level (Red): Risk of panic, you might lose control of your character.

2. Penalty Level (Yellow): Your character is in a disadvantegous position and is getting penalties to defensive and offensive rolls (e.g. 2 enemies vs 1 character)

3. Neutral/Normal Level (Blue): Standard, walking around in the world~ 1-on-1 of equal power level.

4. Bonus Level (Green): 2-on-1, you are doing good decisions and in a good position, hence you get a slight advantage to offensive and defensive rolls.

5. You-Are-Playing-Like-A-Boss Level (Pink w/ Sparkling Stars): You are perfection itself and win every fight, well rested always and never take any damage. Welcome to "Da Boss" Level. Your characters are dominating the battlefield. Jokes aside, this could be a Morale level you get if you can keep your party on a roll by making good decisions generally in the game and almost never go down to "Penalty Level".

 

The only thing that should be affected really would be the rolls, not any statistics per say (So your character isn't suddenly stronger, but just gets better rolls).

 

Example:

On Neutral Level 3, your characters rolls might be between 1-20 as standard. But on Penalty Level 2, your character might see an increase in 1-15 rolls, and 16-20 rolls become rarer. On Bonus Level 4 you might instead see rolls between 6-20 popping up more often, and 1-5 becomes more rare. So it wouldn't mean that you'd get better "damage" output in any way, but the dice gets "adjusted".

Edited October 10, 2013 by Osvir

[PrimeJunta]

I'd just be happy with a combat log that breaks down the adjustments for every attack. If I see "Flanking +2" or "Assisted +2" there, I know that hey, flanking and assistance is kinda cool.

 

I've played RTS's with morale meters. It's a great feeling to be able to make a sudden maneuver which succeeds in breaking the enemy's morale and sends the lines unraveling and the troops fleeing. I'm not sure how much fun it would be in an RPG though.

[Osvir]

Yes, I mean, I don't have much more to say *shrug* your idea seems simple enough (but I'm still gonna dig a lil bit deeper into Morale).

Morale Meters does not have to mean "Run awaaaay!" as we can see in Total War games or the IE games (Morale Checks IIRC). In the IE games the Morale Checks are horrible and chasing down enemies is only frustrating. 

I suppose a better term of what I mean is a "Combat Efficiency Meter" or "Party Effeciency Meter"... or maybe just as simple as a general "Efficiency Meter". So in light of that, it would affect your character's general efficiency. If down at the "Panic Level" it would merely mean your character would perform bad, not that they'd run away. I just think it could aid in the discussion on difficulty (specially on harder difficulties).

About "Panic" in context of this conceptual idea:
If "Berserk" makes your character stronger and uncontrollable, "Panic" would make your character weaker and uncontrollable. Panic could even have some low-chance critical hits that deal more damage; "Desperate Critical Hit" or something, a "more-than-lucky" shot. Topdecking Mechanics if the link doesn't take you straight to the time, it is 18:08-19:00.

[Lephys]

@Lephys, I kind of disagree about that. Valve does exactly that extremely well. They introduce new tools little by little, and every time after you get one you get a situation where it's useful, and then it builds on those to create more complex situations. Half-Life 2 did it. Portal did it. And I found both kind of monotonous and grindy.

 

I much prefer varied difficulty. Perhaps my ideal curve is a relatively easy start that introduces the mechanics and pulls me in, followed by troughs and spikes, with the possibility of going into really tough areas by choice -- but with the game clearly communicating that those areas you're about to go in are dangerous and you might maybe want to level up a bit first.

I believe you took more specifics from that than I intended. I wasn't vying for a very smooth, gradual/stepped difficulty that ALWAYS correlates directly to your newly-expanded scope of abilities. I'm all for varying difficulty. I only mean that, in general, you don't give people 7,000 potential combos of things they could encounter at level 1, and 500 different abilities, or the game's ludicrously overwhelming at first. You have to build up to that. And, secondly, I just wanted to say that the difficulty doesn't need to increase at a steady-yet-exponential pace the whole time. You know, "Now you have to worry about enemies with immunities, AND all enemies have like double the hitpoints, AND you're going to need to be 3 levels higher just to even have a CHANCE against any of them!" Some games do that, and it becomes a bit ridiculous.

 

As the amount of stuff you have to deal with increases, the difficulty naturally increases (more possibilities to deal with/keep up with/overcome, etc.). So, there's no need to make it mandatory that the RELATIVE difficulty increases at a hard rate, as well. That's what I was trying to point out. It's very similar to the aspect of Halflife and other games that you pointed out. You get armor-piercing rounds, and now all enemies can only be harmed by armor-piercing rounds. Basically, it starts saying "these things you've been able to do for a while are now OBSOLETE! You can only use the NEWEST subset of capabilities you've gotten!" It narrows the field of possible strategies and factors, rather than expanding upon what's there.

 

Another related thing is the "your early abilities never gain effectiveness as you go" thing, where you've got a level 1 Firebolt that does 10 damage, and enemies in the 2nd half of the game all have 700HP and a resistance to level 1 spells. Instead of making the Firebolt viable for the duration of the game (even if it's still never the strongest spell at your disposal) and allowing for strategies that utilize it, it's rendered obsolete.