On Auto-Dynamic Difficulty In Videogames

Thanks to Game Matters for its discussion of the problems with difficulty levels in videogames, as the weblog, authored by 3D Realms' Scott Miller, talks about why "games should only rarely allow players to set their own difficulty level." Miller argues: "One of the most common ways games sabotage their potential to appeal to larger numbers of players is by being too difficult... Practically everyone designing games nowadays is a hardcore player with elite skills. It's therefore easy for game designers to misjudge the difficulty of their own games." He describes 'auto-dynamic difficulty', related to Max Payne, as "...a few variables that rate the player's ability, and the player's rating (completely internal to the game) determines the damage that both the player's weapon delivers, and the enemies' weapons deliver against the player." Miller ends by pointing out: "If a player completes your game, they are much more likely to buzz about, spreading the word that it was a great game."

Bull

[Mork29]

I've found very few games which are to difficult to beat on the lowest difficulty setting. A good example of why a user should set it is JK II: Jedi Outcast. I played it on easy the first time for the challenge of the puzzles, and then upped the difficulty the second time to challenge my skills. I get to play the game twice, but for different purposes. More bang for my buck. Not only that, but you souldn't make a game more beatable to get buzz. The point of a game is the challenge. It's not to make it easier and easier until the person can get through the levels. As long as game makers make sure that their "easy" setting is truly easy, you should have no problem. Let the user decide how difficult it should be. Some people want to get through a game without dying, others want to have to restart a level 100 times to truly feal that they earned the next level (masacists are weird....)

Re:Bull

[Metroid72]

The point of a game is the challenge

Many years ago when the only thing I used to do was to play videogames and go to school, I would have agreed with that comment.

However, now after having a job and having "actual things to do" I realized that the point of a game is just fun. The best game for me now is the one that you can pick up quickly and maximize the fun.

I've played challenging games recently, an interesting example is Ikaruga, the game is challenging, but it "gets easy on you" as you play (opening more lives, continues, etc.). However an extreme is F-Zero GX; the game is beautiful, but unless the only thing you do is play F-Zero GX for many days, you won't be able to beat it.

But anyway.. that's just MY opinion. Remember.. for every taste, you have colors.

Re:Bull

[Godeke]

I have to agree. I used to be "hard core" about games: give me a challenge for my dollar. Now that I can sneak a half hour in on alternating days, if that, I just don't have time to perfect my skills.

As an example of a game that rocked on the difficulty scale: Ratchet and Clank 2. If you ever bump into a difficult patch, go buy a new weapon or upgrade one of the existing ones, and things get easier. If you are rocking, you can keep going, if you are not rocking, the game hands you powerful toys to bring th

Re:Bull

[nekura]

Jak II is not made by the same developers, though the two share technology used in the games. Ratchet and Clank 1 and 2 were done by Insomniac Games, and Jak II was developed by Naughty Dog.

Re:Bull

[Datoyminaytah]

> the point of a game is just fun.

This is true, but to competitive people the challenge of "beating" the game IS the fun.

But, like you said, tastes vary.

JKII's an okay example of why "bosses" are lame

[ianscot]

One of my huge peeves with games is the whole "boss" thing. JK II has a couple of "Bosses" that are just plain meant to be "Oh, you got killed again, please reload" stopping points. Galak Fyyar's armored outfit is just dorky, and takes away from the Star Wars atmosphere of the rest of the game. Might as well have been Doctor Robotnik.

My two bits is that it seems like this design concept is way long in the tooth by now. I have two ten-year-olds who don't play games until they've done their homework and pra

Re:JKII's an okay example of why "bosses" are lame

[kisrael]

I can't disagree more.

I find bosses to be some of the most compelling and interesting parts of a number of games, unique parts of every game, whereas everything else kind of repeats.

Most everything else is just lame running around and beating up so so, no challenge enemies.

This doesn't apply to all games: GTA:VC is a new breed of game where the world is interesting and flexible enough that even the small stuff is interesting, and "bosses" are just barely enhaced regular people with heavy weaponry.

Re:JKII's an okay example of why "bosses" are lame

[ianscot]

I find bosses to be some of the most compelling and interesting parts of a number of games, unique parts of every game, whereas everything else kind of repeats... Most everything else is just lame running around and beating up so so, no challenge enemies.

You make my point, from where I'm sitting.

If the imagination of our game designers is limited to "Filler... filler... filler.... Now let's make a really hard boss by giving him extra special armor and lots of hit points," that's the problem. Making the

Re:JKII's an okay example of why "bosses" are lame

[kisrael]

If the imagination of our game designers is limited to "Filler... filler... filler.... Now let's make a really hard boss by giving him extra special armor and lots of hit points," that's the problem. Making the "bosses" harder and harder to compensate for a total lack of imagination is just sad.
Well, for one thing, I'm not into boss fights being gratuitously difficult, just enough so they might take a few tries and force a strategy to be developed.

Many games have too main components: fighting, and explor

Max Payne?

[Fizzl]

Oh, Max Payne has auto-dynamic difficulty?

Infact, I was suspecting it. I'm in the process of playing Max Payne 2 through, and indeed it seems that on a third to fifth try of one particularly nasty spot I suddenly miraculously got through it even thou I felt I got a lot of hits.

Which is good. I hate games where I have to endlessly reload to get past some point. ...Half-Life's end comes to mind. I hated it and actually went through the final encounter with cheats on for the first time. I tried it some 10-20 times without them thou.

Re:Max Payne?

[Soul-Burn666]

What is utterly annoying in some games, is not that the game is too hard, but the fact load times are between 30 secs and minute sometimes! It's very frustating.

I think this auto-difficulty has made my MP2 play harder. I am a perfectionist when it comes to games, I don't like getting hit too much and if I do, I just play it out without care, cuz I don't mind dying. And then, when I finish a section well enough, the enemies in the next one become even tougher and tougher.... even on the first maps!

Re:Max Payne?

[BladesP9]

Now this is exactly the kind of thing I don't want going on. How do they expect players to improve at their game if they dumb down the game to the players level? I understand making it so the player can enjoy the game - and believe me I'm a big fan of entertainment over challenge. However I also think that the mark of any well-designed level or game is that every time you lose you've at least gotten a little bit further than you did on the previous try. If the game auto-adjusts the level on me and makes it

Re:Max Payne?

[Prior Restraint]

How do they expect players to improve at their game...?

They don't. They expect players to shell out money for it, have fun for a while, and then vow to buy the next game to come out of that studio the minute it hits the shelves.

Re:Max Payne?

[F34nor]

Exactly.

But what vested interest do the programers have for making players better? None. Improvment comes from LAN parties, internet gaming and spending hours and hours playing the game, in short it is self indugence on the part of the gamer.

The counter example might be Morrowind with the expension packs. There improvemnet adds to the next experience as well. Its not like you get to keep all your weapons from Quake 2 to Quake 3.

So do we want to make games that really make games better players? I look at

Re:Max Payne?

[Anonymous Coward]

Probably not.

If you read the article, you would see that Miller advocates making difficulty changes between levels, not during.

Max Payne 2 Auto-AI doesn't work well.

[ScorpiusFan]

I just finished Max Payne 2, using the first-person mod, and I found the AI to behave "unfairly" after the first few maps. The AI reached a point where it would: 1) Always rush you; 2) Once an enemy character appeared through a door or around a corner it would target you in less than a second. The whole immersion factor of a shooter game was totally lost at this point. The game became more of a puzzle, trying to determine the best place to stand while defending against multiple enemies rushing towards y

Unreal Tournament

[a_peckover]

Unreal Tournament supports this, increasing or decreasing the skill of its bots depending on how well you are playing. It's a good way of measuring your own skill, rather than just putting it on the highest setting and getting frustrated.

UT does, however, change difficulty a bit too quickly. It's easy to get a few frags in front on Novice and suddenly find yourself on Godlike.

Re:Unreal Tournament

[GigsVT]

UT 2003 does this??

I played that game for hundreds of hours, and I can tell you a novice bot will never become godlike, no matter what.

There might be a little latitude within a range, but it's a lot more limited than you are implying. ...unless you meant some other version and not 2003.

Re:Unreal Tournament

[Enfors]

He said Unreal Tournament, not Unreal Tournament 2003. And I can verify that he's correct. There is a setting in Unreal Tournament that you can enable, called "auto-adjust difficulty" or something to that effect.

Re:Unreal Tournament

[GigsVT]

Interesting that they took it out then..

and it's in UT2K3, too

[waaka!]

You can turn on the same feature in UT2K3, too, just not when you're running a multiplayer game. In Instant Action mode, it's in the Game Rules tab, which seems like a bizarre place to put it (i.e. away from the rest of the bot settings). I know in 2K3, at least, there's a limit on how many levels the computer is allowed to promote the AI--someone who initially chose Experienced bots won't get Godlike ones just because they're kicking too much butt.

Re:Unreal Tournament

[E1v!$]

I think the original Unreal Tournament (to which I expect you are referring) does a very good job of skill evaluation, and the increase in botability is 'just right'.

I still remember the day it went Godlike..... *sigh*

A better solution...

[Filik]

...is to make those parts of the game that you have to solve fairly easy, while still adding lots of extras with varying difficulty (just don't fall into the trap of rewarding the good players with items that makes it even easier for them...instead focus on fun but useless rewards)
Personally I think it is bad that the player has the option to solve everything , so a few impossible or near impossible spots should be added as well, just to teach the player that they aren't supposed to go exploring every cranny of the map, but instead focus on their mission.
Then again, some hardcore players will never give up until every single resistance is dealt with, however little they have to do with the mission...

Re:A better solution...

[gl4ss]

well.. in System Shock you could change difficulty to your liking. don't like puzzles? turn puzzles to ultra easy! hate cyberspace in it? turn it to ultra easy! hate combats? turn that to ultra easy!

like challange? turn 'em all to max!

Re:A better solution...

[danaris]

I generally subscribe to this school o f thought. I mostly play RPGs (on consoles, mostly Final Fantasies), and I consider the best reward to be a cool FMV sequence. It gives you no advantage, but it's also not a worthless item, like "proof of Nemesis" in FFX (which I've never gotten). I *hate* doing a hard sidequest and getting nothing but a worthless item for it, whether it's a certificate-type item ("you finished X sidequest") or just a moderate-to-boring regular game item.

FMVs are a lot of fun (I love watching them, anyway, and I'm shamelessly extrapolating to the rest of humanity), and it feels like you've gotten something at least somewhat worth it, but it doesn't give you any advantage over those who didn't complete the sidequest. That's not to say there can't be sidequests that get you useful stuff that makes other parts of the game easier; I think there's definitely a place for those, too.

However, I do think that the main game bosses/puzzles/whatever should get more difficult, significantly so. (speaking in Final Fantasy terms here because it's what I know; substitute whatever is appropriate for your favourite genre) Though it's reasonable to have a few sidequest bosses more difficult than the final boss, the final boss should definitely be tougher than all the previous regular-game bosses, and most of the sidequest bosses, unless there is a specific, given reason for him/her/it/them not to be.

Well, there's my game-related rant for the week. Someday, I'll make my own games, and probably not follow any of my own advice! ;-)

Dan Aris

Re:

[account_deleted]

Comment removed based on user account deletion

Re:A better solution...

[danaris]

Final Fantasy IX follows your suggested pattern more; the difficulty ramps up over the course of the game; the final dungeon is particularly nasty, and the end boss is just horrible. Nice closing FMV as a reward, though :-)

That's one reason I like IX so much ;-) I like VI very much, too, and though there aren't FMVs, of course (at least, in the original version, though there are opening & closing ones in the PS1 remake), I've always thought that its difficulty is perfectly balanced. If you just go

Re:A better solution...

[ziggles]

Playing the game should be a reward in itself. My idea of fun is not to overcome challenges to gain rewards (and FMVs aren't much of a reward for me anyway). A game should be fun because it's fun to play. I find that Nintendo almost always designs their games with this school of thought. And although they're not always successful, they seem to be one of the few developers left even trying.

Re:A better solution...

[MilenCent]

Please do not shamelessly extrapolate to cover my opinions. Have at least a little shame. And try to extrapolate less, there may be children present.

Heh, anyway, I consider FMV to be a source of much evil in gaming today so of course I disagree with you. Useless items, likewise, are annoying. More interesting are useful things, other advantages, score (or experience) bonuses, new characters, hidden story brances, new areas to explore, alternate endings, and T-Shirt offers (like many old Atari arcade ga

Re:A better solution...

[Carnildo]

Personally I think it is bad that the player has the option to solve everything , so a few impossible or near impossible spots should be added as well, just to teach the player that they aren't supposed to go exploring every cranny of the map, but instead focus on their mission.

And when you do so, rest assured that someone, somewhere, will figure out how to do the impossible. After that, it will go from being a major accomplishment, to being the definition of "good".

Case in point: When Civilization II c

3DRealms eh?

[Chilles]

This is just preparation for a press release later this week about how Duke Nukem Forever will be delayed a bit more because it has become clear that an auto-dynamic-difficulty systems is essential for a game of that type.

On ADD itself:
I think this stuff might work a bit for some games, but generally it would suck. How much sense of accomplishment would you get from completing a game that you knew just tuned itself down to your level? Of course the general public won't know about this so they'll think they

Re:3DRealms eh?

[Decado]

I agree wholeheartedly. The nice thing about choosing a difficulty is that you get to decide how much challenge you feel like, if you can only afford one game a month then you can play em hard to make the experience last. With ADD your only option is to play it clean through, knowing all along that the enemies are handicapped to accomodate you. That sucks.

Re:3DRealms eh?

[irc.goatse.cx troll]

The solution is really simple:

Dificulty: [Auto]
Easy
Medium
Hard
Nightmare

Your comment violated the "postercomment" compression filter. Try less whitespace and/or less repetition. Comment aborted like a fetus.

Max Payne & ADD

[lxt]

Of course, the one thing that does make Max Payne different is the fact that once you have completed the game, you unlock a harder mode of difficulty - so although many people can be happy that they've completed the game and so spread the word, not as many people can say they've completed the game in New York Minute mode. It means that everyone can 'complete' the game, but only the best can fully complete it.

Re:Max Payne & ADD

[Clover_Kicker]

Max Payne wasn't a good enough game to slog through twice.

Re:Max Payne & ADD

[Inf0phreak]

When you read ADD as "Attention Deficit Disorder", that comment becomes a bit funnier :)

eye tou groove

[dario_moreno]

has a "dynamic" mode which seems to change between the easy, medium and hard difficulty levels.

You bet!

[odorf]

Mario 1 was a hard game to beat on the system, but once you did it you felt like the king of the world and had to get all your friends to play it so you could brag that you beat it and they couldent:P

bad way to scale difficulty

[*weasel]

By changing the inherent rules in the system halfway through the game? please.

the last thing a gamer wants to see is a shot that used to kill a bad guy suddenly not killing bad guys anymore. give the bad guys bigger guns, grenades, cover, backup -- something like that. don't ruin the verisimilitude because you have no imagination.

scaling difficulty is fine - but assess it between 'missions' and adjust those for skill for chrissakes and don't change the physics of the game and try to masquerade that as 'difficulty'.

and imo, when a game scales difficulty it should be akin to GoldenEye for the 64. On easy maybe just making it from point A to point B is enough to complete an area. But on 'hard' there should be more stringent requirements (no alarms, rescue a prisoner, steal some data, assassinate a general, destroy a depot, etc, etc).

having to alter the physics should be the first clue that your AI and design aren't capable of being challenging in the first place.

Re:bad way to scale difficulty

[hymie3]

and imo, when a game scales difficulty it should be akin to GoldenEye for the 64. On easy maybe just making it from point A to point B is enough to complete an area. But on 'hard' there should be more stringent requirements (no alarms, rescue a prisoner, steal some data, assassinate a general, destroy a depot, etc, etc).

Yes! This is the way to do it! Tie Fighter had something that was similar. Within the level (this is from memory, I could be wrong, and if I am, this is the way that it *should* have been), you had a goal to accomplish. Kill all of the X-Wings, lose at most one wingman. But then there were "optional" missions within the mission that were more difficult. Kill all of the X-Wings *and* all of the A-Wings *and* don't lose any wingman. (and then there were "secret" objectives like "capture, don't destroy the shuttle").

You didn't gain anything extra by doing the extra crap other than getting promoted more quickly or becoming a super sekrit Emperor drone with a sekrit tatoo on your arm.

Design the mission/level so that most players can complete it. Add additional subtasks/goals that are optional, but greatly increase the difficulty of the game.

Thief:The Dark Project also had this feature.

Re:bad way to scale difficulty

[Torgo's Pizza]

Tie Fighter probably isn't the best example. The entire Fighter series suffered from "complete the objectives or the game ends syndrome". It was all or nothing. So even if you couldn't complete jsut the basic objectives (which on escort missions was almost impossible at times) you couldn't advance.

Tie Fighter is and was an excellent game in all respects except for this one. Players would point to other space sims, most notably Wing Commander which had a branching storyline based on success or failure.

Re:bad way to scale difficulty

[CFTM]

That's not entirely true, I remember one of the Wing Commanders that had a split tree. If you beat a mission you went one way, if you lost it things got more desperate and you went another (until eventually you go boom). Maybe I'm on crack though.

Re:bad way to scale difficulty

[Torgo's Pizza]

No, that's right. WCIII probably had the best tree since you were fighting for the existence of the human race. But it was possible to totally biff one mission and then win enough to get back on the "winning" tree. This is something all of the XYZ-Fighter series wouln't allow you to do. It had a linear mission tree that you *had* to complete to advance in the game. That was just a game design decision that many didn't like. Doesn't mean that it was wrong to do. Just that players didn't generally like

Re:bad way to scale difficulty

[kaisyain]

Because everyone knows that in real life the guys who are wearing red vests take exactly 3 shots from the MP-5 to kill.

Personally, I think it would be a welcome relief to find that all bad guys of a given type don't take exactly the same amount of damage, that I can no longer memorize enemies' damage capacities and thus know I only need to shoot X twice before turn to shot Y three times. Instead you shoot people/things until they fall down/die or your risk being killed in return.

Or knowing I can take exa

Re:bad way to scale difficulty

[*weasel]

while I'll agree on principal to your desire to see less of a game's mechanics as obvious as they are ... changing the rules mid-game is still not a good solution to either problem (static difficulty or visible mechanics).

I'm talking about the difference between a headshot taking out enemies early in the game, to a headshot being insufficient later on. The game establishes an expectation of verisimilitude when the same tactic has the same effect across many enemies. When that tactic is suddenly less effe

Re:bad way to scale difficulty

[kisrael]

Now that we're well beyond using hand-drawn sprites, I wish we saw more games (are there any?) that didn't just have 4 or 5 basic types that were all clones of each other, but ones where enemies or bystanders come in a wider variety of body types and other variations...

Re:bad way to scale difficulty

[Haeleth]

Now that we're well beyond using hand-drawn sprites, I wish we saw more games (are there any?) that didn't just have 4 or 5 basic types that were all clones of each other, but ones where enemies or bystanders come in a wider variety of body types and other variations...

Isn't Half-Life 2 supposed to be going to do this? I'm sure they were saying something about all the scientists looking different.

Re:bad way to scale difficulty

[Prior Restraint]

GoldenEye for the 64

Thief 2 was this way, as well (didn't play the first one, so I can't speak to that). I agree that's a fun way to do it, and I'd like to see more games try doing it that way.

The other approach I kind of liked was used in Hitman 2. You were restricted to X saves per mission, and X got smaller as you upped the difficulty level (7, 3, 0, IIRC).

Mario Kart anyone?

[kisrael]

It's wasy to take this too far, Mario Kart comes to mind, both multiplayer and single, using somewhat cheesy tactics (weapons and arbitrary speed increases) to keep the field much closer to even than would otherwise be the case.

One problem with systems like that is when players try to scam the system, like purposefully just hanging back in second or beyond to get good weapons but never getting so far back that they can't catch up. Admittedly, that takes a certain amount of skill on its own, but still.

summary ??

[kayen_telva]

make the games easier so our industry can rack up the dough ??

sorry, guess Im feeling cynical this morning.
but it does seem like most gamers I know finish games
very quickly and then move on to something else.
Seems like if they were more thought provoking
instead of run around crazy shooting everything
people would find them more enjoyable and
recommend them

Difficult to say...

[wickedj]

I've played racing games where if I'm dominating the track, the 2nd place driver all of a sudden can go 50 mph faster than me and turn on a dime in order to catch up. I've also seen the effect where if I make a mistake and wreck, the competition slows to a crawl to allow me to catch up. I can understand this in games such as Mario Kart or other arcade racing games but I don't want this in my Gran Turismo or other "real" racing games.

Also, in some of the RPGs I've played, the monsters get stronger as I get stronger. That's ridiculous. I understand meeting new monsters that are stronger but when the little slime you had a hard time with at level 1 is still giving you a hard time at level 20, that's just plain ridiculous. Even worse is that you still get the same xp and gp.

There may be some really good reasons out there to have auto-adjusting difficulty, but for me personally, I don't believe it's that great a feature.

Re:Difficult to say...

[darkmayo]

"Also, in some of the RPGs I've played, the monsters get stronger as I get stronger. That's ridiculous. I understand meeting new monsters that are stronger but when the little slime you had a hard time with at level 1 is still giving you a hard time at level 20, that's just plain ridiculous. Even worse is that you still get the same xp and gp.
"

The Lunar Series has this feature, and IMO it made the game better since you couldn't just level like mad and walkt through the game like a certain incredibly popula

i hate to keep hyping PGR2 but it's so good...

[h0mer]

Project Gotham Racing 2 has the best difficulty curve I've seen. Basic/bronze medals are very easy to get, and give you a good feel for the course. Silver medals are the sweet spot, you won't have trouble if you are good, but it's not a cakewalk. Gold medals will take some retries, and platinum medals are punishment :)

Not to mention that you get to see your Kudos rank on Xbox Live after each course. It was a motivating factor to keep playing because I kept getting higher and higher on the lists.

Difficulties...

[sbryant]

I remember playing Red Alert on easy first, then again on hard. It was more fun for me - I got double the challenge. Did the same with Half-Life. I liked being able to do that.

I can understand what Scott is saying, and I think that a properly implemented ADD will give you this too. A bad ADD will mean that a poor player who got lucky ends up in a situation they can't win, and gets frustrated and gives up.

I had another idea though: instead of just changing some variables (hit points or whatever), what about changing the gameplay? For example, you could change puzzles or add new ones. Eg: remove a box, so the player can't just jump up somewhere - they have to be more creative. You could also add access (eg: remove walls etc) to areas which are hidden to beginners - let them focus on the mission, and send the experts a different (more difficult) way round. How about making better players go off to find a key/card to open a door, but letting the other players through without needing it. Are there any games that already change the maps according to player skill?

The key to really making it work is finding the balance of what to do for which skill level; being able to accurately judge a player's skill is an important part of this. It's a lot of work, and sometimes it's easier to let the player choose their skill level.

One other thing occurs to me. Remember Doom's nightmare mode? I don't think you could ever reasonably have something like that with an ADD system, but there are some (strange) people who find it fun.

-- Steve

X-Com did this years ago.

[Teddy Beartuzzi]

'Course, it was a bug, and it accidentally kept setting the game to Easy.

Sounds like a lot of work for nothing to me. Just give the user enough difficulty levels so that they can set it to what they need. I've played a few games where even Easy mode was too hard, and Impossible mode was anything but.

Unless you're trying to hide something...

[Tom7]

"If a player completes your game, they are much more likely to buzz about, spreading the word that it was a great game."

Unless it has a real shitty, anticlimactic ending like XIII.

Baldur's Gate 2

[vadim_t]

It had a really nice system. You could change the difficulty setting while in the game. It changed the amount of damage done by monsters, but low settings reduced the amount of experience you gained.

This was really nice for those annoying times you got stuck in a place. For example in my first game I made the alarm ring in the room near the dryads. Then I got damaged by the traps in the room, and overwrote my old game. Then came the golems, which quickly killed me.

In other games this would have meant downloading a cheat, restarting the game, or perhaps loading a saved game from an hour ago. In BG2 I could just temporarily set the difficulty level to easy, kill them, and set it back to normal.

For me in most games it doesn't happen that it's too hard in general. It's usually too hard in a specific place, because I screwed up, went to the wrong place, or especially in RPGs, had a party that couldn't deal with the enemy. It can be bad luck too, like in Morrowind, where you can be really screwed if you *have* to sleep, do it, and have a zombie wake you up and attack before you're healed.

Re:Baldur's Gate 2

[balthan]

Baldur's Gate 2 also increased the level on the monsters depending on what the level of your party members were. One time through the game I did the quest with the red dragon very early on and werewolves (or some other lycanthropes, I forget exactly.) On a seperate time through the game, I did that quest much later in the game when my characters were much more powerful and instead encountered greater werewolves (a pain-in-the-ass to kill, BTW.)

I though this method of auto-difficulty worked really well si

Payne for your Brain

[100lbHand]

"If a player completes your game, they are much more likely to buzz about, spreading the word that it was a great game."
Or in the case of Max Payne 2 that you can beat it in less than five hours.

Level restart ? Dying ?

[mystran]

Best games are those, where you won't try the same level twice, because the whole game map is generated randomly (up to certain rules ofcourse) and if you ever happen to die, that's permanent.

This doesn't naturally make it any easier to complete a game. Indeed, it makes it a lot of harder. But it makes extracting FUN easier, since you don't have to play the same 25 levels just to discover something new (or replay the same level 99,5 times).

Auto-dynamic difficulty is way over-rated

[TheLink]

I really don't see any advantage with that method. If you can dynamically adjust the difficulty in a game, you can manually adjust the difficulty too. (For some games you can't adjust the difficulty without changing the game - e.g. puzzle games).

Let the users select different difficulty levels. If they want/need it easy let them have it easy.

If someone with high skill wants to run around splattering enemies just for relaxation why dynamically increase the difficulty level so that they can't? Doh.

If your

Unreal Tournament

[shoptroll]

Unreal Tournemant implemented the option that bots would adjust to the skill level the player was playing at.

Also, they introduced the idea that the final boss of the single player "campaign" would be barely beatable by the player. Essentially if you look at the code for the final level you will find a note saying that the final boss is designed to "stay one step ahead of the player", which goes hand-in-hand with the strategy i've heard which is to stay sucky for a while and then ramp up your apparent sk

Re:Unreal Tournament

[Derkec]

Being able to set the settings in UT myself was great for me as well. But, I was fairly hardcore at the time. I had played at college extensively and when I went back home for the summer (no broadband), I thought my skills would rot. Playing against pretty even competition, they did to some degree. That's when I set the bots to their highest level. My game improved dramatically because it had to. I was crisper at finding cover, and more accurate in my shooting. By the end of the summer, I was finding myself

Say No to Dynamic difficulty scaling

[p7]

Dynamic scaling is not the way to go. I am guessing that the majority of gamers play games because they feel they have overcome a challenge. I think this applies to the non hardcore gamer too. Do you really think people would enjoy Dance Dance Revolution if it slowed down to match a players ability and let them hit the wrong pad?? With online play becoming more and more common, I think the dynamic scaling could set you up for a big letdown. Here you go finish the game and think you did really well and then you get online and can't score a frag.

Ramping up difficulty the lazy way...

[Anonymous Coward]

"...a few variables that rate the player's ability, and the player's rating (completely internal to the game) determines the damage that both the player's weapon delivers, and the enemies' weapons deliver against the player."

What a stupid way to set the difficulty in a game!
I don't know about ya'll, but there's nothing I hate more than shooters which ramp up the difficulty by making you shoot everybody 5-6 times before they drop dead.
I mean please, talk about the lazy way out.
Give me smarter enemies, or mo

Challenge != Fun

[Psykechan]

It's important for developers to remember that challenge does not equal fun. I personally think that difficulty levels should default to the easiest level possible and gamers looking for a challenge could manually set them higher.

That's not to say that ADD couldn't be implemented properly. One of the best examples of this I have seen is in "Sly Cooper and the Thievius Racconus" for PS2. If a player died multiple times on a given level, they would be given a lucky horseshoe that would allow them an extra hit before dying. It made the game much more enjoyable for my casual gaming friends.

On the flip side, "Mario Kart 64" (N64) had the worst ADD as enemies would always be a few mistakes behind you. It didn't matter if you performed average or godlike, a few slip ups and you would be overtaken. I referred to this as "cheating opponents" and will not even think about buying MK:Double Dash until I know that this "feature" isn't in the game.

If Auto-Dynamic Difficulty can prevent frustration, then it is good. If it causes frustration, then it should be left out.

Re:Challenge != Fun

[Disoriented]

Mario Kart Double Dash does not have the "feature" you're complaining about.

To compensate, the overall game is harder at the 150cc and Mirror difficulty levels (so be warned).

The way things ought to be

[spyrochaete]

Skill level adjustment is not a linear argument. It depends on the type of game. While auto-adjustment may make sense for an arcade-style game like Max Payne (where you do the same thing throughout the whole game) but it is unwelcome for simulators (racing, sports).

What I really appreciate (thanks, KoTOR) are games that let you change the difficulty level at any time. I also appreciate games that offer you hints (Popcap's Bejeweled, Sierra's Phantasmagoria) to keep the pace from lulling.

Indeed, the purpose of games is entertainment! Some people are willing to spend 8 hours every consecutive day until a game is finished, while others would prefer to spend 3 hours a week. Neither party should be penalized. I'm sure these demographics are related to the article I read about the average age of gamers rising ever closer to 25.

Re:The way things ought to be

[Haeleth]

What I really appreciate (thanks, KoTOR) are games that let you change the difficulty level at any time.

Somebody mod this guy up - this is the first comment I've seen here that I agree with 100%.

Let me play on "medium", but give me the option of switching temporarily to "easy" if I run into trouble.

Note that - give me the option. I don't want the game to dumb itself down; sometimes I actually want to try and get through a tough bit by myself. Other times I want a walk through the park. But I should b

REZ

[plagioclase]

Rez [gamespot.com] did this to a certain extent.

Essentially, the difficulty of the boss at the end of the level was determined by how well you;d done leading up to it. It was kind of like a reward for doing well, and added some replayability, because the enhancements the boss would get going from normal to hard were somewhat obvious. Once I knew this was happening, I got a charge from knowing that I had earned the 'super' boss.

It also meant that if you weren't that good yet, you'd stand a better chance of getting to s

Re:REZ

[Rallion]

That reminds me a bit of StarFox 64. You could go and beat the game the easy way, sure. But it was much, much more satifying to take the hard track all the way through, fight a tough battle against StarWolf, and then fight creepy-brain-Andross. Not that that was really all that hard, either.

Needless difficulty is a problem

[Mr. Piddle]

Challenging games are fun, but sometimes they are downright stupid. After having enjoyed a couple Alone in the Dark and Silent Hill games, I decided to try Resident Evil. All I have to say is the first thing I did was to enable the cheat for double saves. As an adult, I simply don't have the time or patience to go through artificially great swaths of a game over and over just to feel like I won't run short, especially given the poor weapon control in that game! I just don't remember any of the other games being so troublesome.

I think it's fine...sometimes.

[Rallion]

Sometimes I don't want to worry about my game being too hard or easy. I don't want to get stuck or a hard part or get bored if I get too good. You know, sometimes. But I think having it as an option is best.

The absolute worst attempt at scaling difficulty I've ever seen is Homeworld 2. Really all it does is takes the force you have coming into that mission, and creates an initial enemy force that will beat it (not only in numbers, but also in ship types -- if you have fighters, they have lots of anti-fig

Difficulty

[Quill_28]

One thing missing from today games, is the suspense of losing.

I remember getting so far into certain games and being nervous about dying, because if I did I had to start over.

Now sometimes it was exteremely frustrating and ruined the game. Other times it made winning that much more fun. nethack comes to mind.

But now with a quick save every minute there is no suspense.

Because of the quick saves though the game makers make the games harder.

Somehow a balance of the two is needed, but I am afraid there wil

Re:Difficulty

[John Gaming Target]

I think Maximo: Ghosts to Glory balanced this perfectly.

The game was super hard and saving became progressly harder as the game went on. You had to earn Continue Tokens by catching spirirts but every time you continued Death would require more Continue Tokens to let you continue.

And then you could only save in between levels (or choosing full armor) or using your hard earned money to buy game saves (which then couldn't be used on better weapons, armor, or those stylish cowprint boxer shorts).

I've heard M

Re:Difficulty

[Snowmit]

It depebnds on the game. Rainboz Six 3 is supposed to be a high-tension shooter where you have to move caustiously and carefully to avoid dying. In that kind of situation, limited saves makes perfect sense. Max Payne 2 is supposed to be about experiencing the story and creating jaw-droppingly cool gunfights.

Theres not faster way to ruin the sense of fun in that game than by making you play big sections over and over again. So it's a good thing you can autosave.

Difficult

[scabb]

Auto-dynamic difficulty can be great, if implemented properly (Max Payne, Lylat Wars/Star Fox 64). However, when this "difficulty" consists of making a range of asinine and vague "choices" in places - ala Kingdom Hearts and "the sun is setting" - then it sucks.

Good games usually have just the one difficulty setting, although that depends heavily on the genre.

Re:Difficult

[MilenCent]

StarFox's difficult doesn't automatically adjust, instead you complete missions to "earn" harder levels. However, you can still go back and play the easier ones even if you've unlocked the hard ones, you just aren't forced to. In fact, to get a really good score I'd say you have to do this, as taking the warps gets you the chance to earn buckets of points, yet the easier levels are better scoring (especially the Independence Day level) than the harder ones.

Lame.

[Anonymous Coward]

Auto scaling difficulty is lame. I don't want a fake sense of accomplishment. I want to play games like Ikaruga, and Viewtiful Joe, where on the harder difficulty levels the game will stomp your nuts if you make a single mistake when you're playing it.

Why do I like this? Because when you do figure out a level, when you do get in the zone, and play through the level perfectly, you get an amazing feeling of accomplishment. This rush is what gaming is about.

If all I want is to be told a good story and not ha

Let's take an example

[metamatic]

Two games built using the same game engine: "Jak II" and "Ratchet and Clank: Going Commando".

"Jak II" is just way, way too hard. Worse, it's hard in irritating and boring ways, like being prevented from completing a mission by a random traffic jam, or the "Escape from the boardwalks" mission where the game will literally throw an endless supply of guards at you until you force your way through or die of boredom. (That was the point at which I resorted to the cheat codes.) I should point out that I'm no klutz when it comes to games--I'm a pretty good Wipeout player, and I finished Jak & Daxter without needing to cheat. Jak II is just ludicrously tough.

Now contrast with "Ratchet and Clank: Going Commando", which is the game Jak II should have been. The elegance of the R&C game design is that it's automatically self-adjusting without changing the rules--it starts off easy, and if the difficulty ramps up too quickly for you, you can just keep trying for a while. Eventually by killing the stuff you *can* kill, you get enough bolts to buy bigger and better weapons and armor which will let you plough through the nastier enemies. The only potentially frustrating parts are the environment-related traps and puzzles, like the pit of lava at the bottom of a river of lava that took me half a dozen attempts to get across.

The end result is that Jak II was nowhere near as much fun as R&C:GC has been. In fact, even with cheat codes I gave up on Jak II, because the final level seems to dispense with actually allowing you save/continue points, so one small slip and you have to start the entire thing again. Really, I don't know what Naughty Dog were thinking...

Speaking of completing games...

[bottlerocket]

"If a player completes your game, they are much more likely to buzz about, spreading the word that it was a great game."

Why is he worrying about the players completing his game? He and a team of programmers can't finish it [planetduke.com], either!

Good:dynamic. Bad: abuse, MarioKart, skill testing

[MilenCent]

I really like the idea that a given game situation may be different between plays, either randomly, due to conscious player actions, due to his playing skill, or from selectable options. That's the entire idea behind Roguelikes. There are lots and lots of ways to do this that are ignored by most developers. (Perhaps because it'd hurt the market for hint books, bleah.)

But there are many things wrong with auto-adjusting difficulty as described in the article. It's open to player abuse, for one thing.

It

Less Predictability = More Fun

[el_gregorio]

the biggest thing that bothers me is the predictability with which most enemies appear in games.

say i'm playing Ghost Recon or Max Payne, i'm going too fast thru an area, and an enemy caps me from behind a good hiding place. he got me that time, but as soon as i replay the section, i know exactly where he is. i suddenly have an unnatural advantage over my enemy.

but put that same guy in a different spot, and now we're back on even ground: neither of us know exactly where the other is.

at least Ghost R