I'm making a game where the difficulty is adjusted as you play, becoming more difficult when you win, and easier when you lose. The idea is to keep the game challenging for any player, without barring players from getting further in the game.

What do you think of dynamic difficulty in games, either as a player or as a designer? Do you think it would feel like less of an accomplishment to beat a level, knowing it got easier after the first couple times you died on it? Do you think you'd prefer having more control or knowledge of the difficulty settings?

Console games use this system. Some players hate it (generally, I would not try it on PC, players here are quite knowlegable with google and everything and would loathe you for it). The more you go into the realm of mobiles/consoles and young kids/casual players the more acceptable this mechanic is.

well. ireally like this question. its a very new approach in artificial intelegence in games.i try explain as much as i can and you can find more in internet.

as we all know immersion is most important touchstone to measure that a game is good or weak. if you make the player to play the game more hours so you immersed him much more. if the game is very hard and he loses every times so its a hard game for him or if its very easy it will not satisfy the veteran player. most of games ask you which level of difficality you want to play but most of the time player cant understand the word hard means how much hard and... modern games let you change level of difficality in middle of the game or call of duty in modern warfare series had a training mode in first level of game that measures you ability and suggested suitable level of difficulity or in maxpayne 3 the game gave you one more health pill.but most of the time players choose false level of difficulity.

but you can do much more. its all about your creativity and your own algorithms but you have to know one important thing. you have to make player feel that he/she is doing a very hard and important work and avoid to make him feel that you are giving him chances. for example in a shooter game you can just write a code that if health of player is very low, just add health kits on the way he is dynamicly or make less shots end to him or in a strategic game when player is losing, the enemy uses cheaper tactic. these like you fool the player but its you art to not make the player fell that way.

I prefer something in between, not fully dynamic difficulty, but not "difficulty settings" either. (I think both are easy to miscalibrate, and have the potential to insult the player when miscalibrated.) What I like best is adaptive difficulty as a result of in-game player choices, where the consequences of that choice are clear to the player. Particularly when difficulty is encapsulated in a conceptual metaphor that doesn't itself have a value label. For example:

Difficulty as a place: This is definitely the most common conceptual metaphor for difficulty in games, where easy and hard difficulties are encapsulated as easy and hard places, and choosing the difficulty therefore uses the metaphor of traveling. That's a natural human metaphor; accomplishing things in the actual world is easy and hard depending on location. If you want an easy game, you can stick to easy "places", if you want a hard game, you can stick to hard "places".

This doesn't necessarily mean making more levels; you could have same or similar levels be duplicated with different difficulties, just represented as different places. For example, even if you're intending to have a linear game with levels 1, 2, 3, etc., you could implement this metaphor by giving them a grid-like map of levels to choose from where the horizontal axis is the level and the vertical axis is the difficulty. Traveling "east" gives you new content, traveling "north" gives you the same content adapted for greater difficulty.

I'm planning this for a game of mine that I need to expand to have more "levels" to challenge more-skilled players. I'm not going to label the difficulties, in part because players are proving to have really strikingly different natural aptitudes for the in-game task; I don't feel like telling the player that what they find "difficult" is what other players find "easy". Rather, the level map will represent the difficulty of the level -- which corresponds to the availability of resources -- by representing harder levels as more arid, sparse, or harsh. Players can spend as much time as they want having lower-challenge fun in the more lush areas, and venture out into the more-difficult outback whenever, and to whatever distance, they feel like.)

Difficulty as an item: You can also have things that you carry, or abilities that you can turn on and off, that either make the game harder for a greater reward or easier for a lesser reward. (For example, Bastion had the shrines, or whatever they were, where you could make enemies harder but drop more rewards.)

Difficulty as a consequence: There's some kind of action you can perform in-game that changes the world to become easier or harder. (For example, in Dishonored you always have a choice between lethal and non-lethal takedowns, and things become more difficult the more people you kill. In... I think it was Shovel Knight, you can go around destroying autosave checkpoints.)

Actually, there are two other things to consider, rather than going for dynamic difficulty. The reason so many games of almost every genre now have EXP mechanics is to mitigate exactly what you're concerned about (locking out less-skilled players from seeing content that you want them to see), by letting player pursue easier challenges as much as they want in order to make further challenges a bit easier.

The other is that lots of games for "casuals" are actually quite difficult, sometimes; it's just that they tend to make failure states as smooth as possible. That is, they make it easy to jump back in and don't shovel a lot of judgment on the player for failing.

I'm making a game where the difficulty is adjusted as you play, becoming more difficult when you win, and easier when you lose. The idea is to keep the game challenging for any player, without barring players from getting further in the game.

What do you think of dynamic difficulty in games, either as a player or as a designer? Do you think it would feel like less of an accomplishment to beat a level, knowing it got easier after the first couple times you died on it? Do you think you'd prefer having more control or knowledge of the difficulty settings?

Well here's the first complication, and it's a big one: "Challenging" means different things to different people. For example, some people are frustrated by a single failure, while others don't feel challenged unless they are failing almost constantly.

Another issue is that dynamic difficulty adjustment directly incentivizes the player to play only as well as they need to to not fail (rather than playing at the best of their ability) which is antithetical to the idea of "challenge."

Also, some people don't just feel less accomplished if the game becomes easier, they feel downright insulted.

as we all know immersion is most important touchstone to measure that a game is good or weak.

Without any solid data to back that up, "we" don't know that at all; so let's not spread speculative opinions as facts, please.

Also, some people don't just feel less accomplished if the game becomes easier, they feel downright insulted.

Yes, the idea is inherently condescending, and taking control over the difficulty away from the player. This can lead to strong negative reactions (personally, I would never pick up such a game, or feel insulted if I found out a game I played had this feature).

It's very important to have a vision for what kind of gamers you want to make your game for, and then make sure your game design supports that vision. If you want your game to be accessible to all kinds of players, then first of all make sure that your game design is accessible to all kinds of players. If you want hardcore players to play your game, then make sure there is enough depth in the design to make it truly difficult to master, and give incentive to do so (by offering very challenging content, high scores, etc.).

Once your game design supports your vision, you can think about how to use dynamic difficulty to further support your design if that is what you want; or you may find that offering a Hard and Hardest modes is more interesting for hardcore players; or that you want to offer difficult, optional bosses for them; or maybe you want to give your more casual players an option to make a boss battle they struggle with easier; and so on.

A leveling system like valrus mentioned is another good option. One thing newer RPGs by SquareEnix do here is giving the player EXP after battles, but making the actual leveling optionable. That way hardcore players can try to beat the game at the lowest level possible.

There's a few different ways off-hand I'm aware that have been used adjust perceived difficulty:

• Scale Quantities - Increase/decrease the number of elements; e.g. enemies, items, etc.

• Scale Frequencies - Increase/decrease the frequency of elements; e.g. enemies, items, events, etc.

• Scale Stats - Increase/decrease the effects of player actions or actions on the player; RPG like leveling anyone?

• Scale Information - Increase/decrease the context clues/hints provided to the player; e.g. character says "If I only had/knew 'x' I could [blank]", light glints off an object to draw attention, exaggerated movements, hint systems, etc.

• Scale Solutions - Provide multiple avenues to achieve objectives as not everyone is as good at the same tasks; e.g. stealth, brute-force, coercion, etc.

• Hybrid Scaling - Mix and match the above approaches to achieve a flexible balance of options.

A lot of people have tried dynamic difficulty systems with varying degrees of success; generally it has been done by measuring and statistically analyzing effectiveness at various tasks and setting upper/lower limits to trigger parameter changes.

Some examples of this would be varying parameters based on:

• Ratio of damage done vs. received.

• Mean time to complete objectives.

• Number of attempts.

• etc...

However it's really difficult and time consuming (lots of tweaking) to get it 'right', and it's very easy to get it wrong.

Get it wrong it and suddenly it's obvious, obnoxious, intrusive and unnatural to the player; hence the people that loathe the concept itself.

If you do really get it 'right', then probably no one ever truly knows it's there because it seems transparent and natural.

Which makes it difficult to market as a feature... instead it becomes part of the intrinsic qualities of the game experience for the player vs. something you can example to people.

Odds are games you're playing that felt 'right' to players of multiple skill levels already are doing this; even if the developers and designers haven't realized that's what they were doing as they tweaked game systems, added different paths and included RPG mechanics. It's one of those situations that's part science and part art, and most people can't explain it so much as know when it begins to feel 'right'.

My game is a mobile game. I'm tying the score in with the difficulty, so as the difficulty increases, the score goes up exponentially. If you want a high score, you need to win to do well enough to make the game tougher.

Difficulty as a place: If you want an easy game, you can stick to easy "places", if you want a hard game, you can stick to hard "places".

Difficulty as an item: You can also have things that you carry, or abilities that you can turn on and off, that either make the game harder for a greater reward or easier for a lesser reward. (For example, Bastion had the shrines, or whatever they were, where you could make enemies harder but drop more rewards.)

Difficulty as a consequence: There's some kind of action you can perform in-game that changes the world to become easier or harder.

You're right, I should keep those kinds of options in mind. I think players that want a challenge are likely to go for things that feel like an achievement or high score, like Bastion and Transistor. Those things were completely optional, but feels like an accomplishment to say you beat the game with all of them. As long as it's clear the place they're going or the item they're using will make the game more difficult, without feeling like it's weakening the player.

Well here's the first complication, and it's a big one: "Challenging" means different things to different people. For example, some people are frustrated by a single failure, while others don't feel challenged unless they are failing almost constantly.

I'm considering including a difficulty selection that doesn't set the game to an exact difficulty level, but instead just changes the math for how much it reduces and increases the difficulty level. So a player that doesn't like much of a challenge can set it to easy, and the game would calibrate the difficulty so they don't have to lose much for it to get easier, and a player who loves a challenge can keep the difficulty high.

I also have this thought about human potential, like how olympic runners keep breaking records, and professional starcraft players take that game to a place the creators probably never imagined when they made the original. So I like the idea of the difficulty level not having a ceiling, letting players reach their own limit instead of the game's limit. I guess any game with a score board does this to some extent, like through the game being endless, but it can also be done with speed, and other elements that thewayout mentioned.

Have you read up on Valve's "AI Director"?

It forms the heart of the Left 4 Dead franchise, and makes an appearance in several other games (i.e. the top-down survival horror title, Alien Swarm). It is effectively a dynamical AI presence which reacts to the player's progress and status, doling out ammo and health in some situations, and punishing waves of enemies in others.

Even if those games aren't exactly your cup of tea, they are worth checking out for the AI Director alone...

I think I've heard of it, but haven't read up on it. I'll have to check that out, because I'm interested in making something like that someday. Like an RPG game master, helping players to make a story as they beat up on a bunch of monsters.