Flow in Games: an interactive thesis on dynamic difficulty

Heh, I remember when this first came out. Really good game, though it gets super hard after you become a circly thing instead of a liney thing

isnt this a PS3 launch title on their PSHub?

I don't really trust dynamic adjustable difficulty, especially for games that are challenge-based (aka most games). If the difficulty adjusts, how can you be proud of your achievement?

In my opinion, dynamic AI/dynamic interaction is a better factor to research on..

While I can see how this can be an interesting direction to take AI, I hope we're still left with preset difficulty options. (I belive FarCry had such a scheme) There is something to be said for the sense of accomplish with things like beating Halo on Legendary, etc. If AI is constantly getting weaker when you're getting your ass kicked, you're not really given the opportunity to improve, or come up with a new strategy.
What I would really like to see is expanded difficulty settings. Something like multiple sliders to custimize each aspect of the AI algorithms. You could adjust manually to really customize the difficulty of the game, but it would also be cool to see how the auto-adjustments changed as you progressed through the game.

If you're wondering where you could nab the game fl0w... you can get it here: http://www.filecloud.com/file/3034/flOw+Free+Game/

This guy has the best name ever.

J-E-N-O-V-A!

cue the creepy music with the heartbeat background

I just get that fuzzy warm feeling whenever I'm down to my last heart in Zelda and I crack open a vase and a bunch fall out!

I remember that game. Specifcally, I recall having to use my Graphire to beat it.

I also remember thinking that Chen had the coolest name EVAR.

Currently, I'm wondering why none of the comments have mentioned "Wii" yet. This sounds a lot like what Nintendo claims to be going for, in an indirect sort of way.

This idea strikes me as dumb for video games. It sounds like the "dynamic difficulty" tries to make it so it's easier when you're doing poorly and harder when you're doing well. This sounds like an arbitrary way to eliminate all sense of accomplishment from a game. Isn't part of the fun having difficulty, and then overcoming it, and then dominating it? This idea would really diminish the significance of winning or losing.

The concept reminds me of the way FFVIII's level system worked. As you leveled up, the enemies leveled up as well, keeping a constant parity between your party and foes. This, of course, begs the question -- what's the point of leveling in FFVIII?

i played this game a few months ago after following a digg link. Got old after about 4 hours (total not straight). It had a stopping point,where you couldn't eat anything else, so maybe i need to read the paper to understand why it didn't get that much more difficult.

Wasnt Mario Kart 64 the first (or one of the first) to use this type of thing. I remember reading that somewhere. Mario Kart Rul3z!!11

DDA pretty much equals continuously increasing frustration as you can never actually master the game. It is appropriate (and in fact is really only an extension of what has always been there) in simple arcade-like titles or meditation aids.

I am all for no boredom, but anxiety is a fundamental part of a gaming experience if you want to get the player to feel committed. As are joy, exhilaration, fear, etc... An algorithmic approach won't get you those things.

Personally I found Flow (the game) to be a total bore.

I got to the second last boss the first time I played this game (when you are a circle and must fight the last snake-type creature) but I was never able to actually see him when I was at his level (he was only visible when I was at the level above him), so not knowing what to do I clicked the help link, which took me out of the game and I lost my progress!! NOOOOOOO!

Is there a third form?

"Wasnt Mario Kart 64 the first (or one of the first) to use this type of thing. I remember reading that somewhere. "

Hardly.
It was done way back in the Atari era, when Mario was just a carpenter chasing some ape up a building.
The first game I KNOW had it was Major Havoc, in 1983.


"This idea strikes me as dumb for video games. It sounds like the "dynamic difficulty" tries to make it so it's easier when you're doing poorly and harder when you're doing well. This sounds like an arbitrary way to eliminate all sense of accomplishment from a game. Isn't part of the fun having difficulty, and then overcoming it, and then dominating it? This idea would really diminish the significance of winning or losing."

When done right, it keeps you on the edge. Always challenged, but never overwhelmed.
"Dominating" a game seems to mean that it's become a cakewalk. That's precisely what robs me of my sense of accomplishment.
I don't want a game to yield. If the challenge has let up, it darn well better be because the game is over. If I wanted it easy, I'd grab an Action Replay.
That's part of why I like the concept of adaptive difficulty.

Of course, if you want to dominate... the game will cap out somewhere. You just have to surpass it. :P

The big problem is most games that do it... do it BACKWARDS. Ever played a shooter, started doing well, gotten your ship powered up, then died and lost all your powerups? That's a form of adaptive difficulty, just one that's gone horribly wrong. The game gets easier as you do better and harder as you do worse.



As far as tracking if it's in low or high gear...
If you're getting jumped by one or two boxers, it's in easy mode. If you're getting assaulted by a force of 50 ninjas armed with rocket launchers, it's probably safe to assume it's up in the hard area.
And the grin on your face after comboing their bazookatana-wielding faces into the pavement is all the reason you need to keep your skill up. It's a far more satisfying battle than the "easy mode" one.

If you integrate the difficulty variation into the game setting, you could even track the current challenge level with a HUD element.


For a good example of the concept in action, try the NES game Zanac some time(or the PS1 sequel Zanac X Zanac, if you can play imports). It's a fairly good demonstration of the idea. Partially because it's not very subtle so it's easy to see what's going on, partially because there's just not a lot of samples to compare.

It's VERY clear when you've made the game made at you, because it escalates challenge mainly through more ships and more bullets.
As a nice touch it also changes what enemies are coming at you as you change weapons, in an attempt to exploit your current weaknesses. So you can't just grab the forcefield and laugh with impunity.

It has some rather severe rules for increasing difficulty, and only 2 ways to reduce it(death being the main one), but balances that out by handing out 1-ups like popcorn(score-based, at a very low point threshold). The madder the game is, the more target-rich the environment, and thus the more 1-ups you earn, right? :)

Me and some coworkers were just discussing this the other day, in regards to Oblivion, which scales enemy difficulty and dungeon rewards to your level.

For a game like Oblivion, it diminishes a lot of the accomplishment -- not because the game is too easy (which it quickly becomes) but because when you level up your character, it almost doesn't matter, because the world levels up with you.

In its current state I have no love of DDA.

A game should be balanced and challenging. It should make us play better, improve, or think outside the box. There's a massive sense of satisfaction knowing that you've had to attempt a tricky bit in a game.

Knowing that the game is adjusting it's difficulty to suit the player is wrong in many ways.

At what point does it adjust? If I get to a tricky point and fail the first attempt, how much 'easier' will it be the next time?

Will there be a point where if I fail enough I'll actually succeed!?

Where is the challenge in knowing that instead of playing against cunning AI I'm actually playing against a coded sub routine which states that the game should always be 'just enough' of a challenge but ultimately doable.

Also, this impacts stuff like XBox Achievments. How can you have DDA and also enforce a rigid measure of players comparative ability.

Game AI needs to come a long way before DDA is viable. In its current form (i.e. in Oblivion) DDA is just dumbing down the experience.

I played flOw and I thought it was really neat, but I am totally against DDA. Mario Kart pissed me off because no matter how well I was doing I always got hit with a lightning bolt or a blue shell from some jerk who couldn't get out of last place, and I usually lost my spot in first place because of it. DDA does nothing to help people who are good, or want to become good at video games. It just lessens the challenge and thereby the effort required to win. If designers want to add DDA to their games I vote that they make it an option.

Didn't we learn our lesson from watching Jordy ask the holodeck to create an opponent that could defeat Data?!

/zwoop

Hold up, this is a good idea BUT I think FPS and games like WOW online have the right idea. This brings me back to the nes days. You would play a game simply because you want to beat it no matter how hard it was. If you could walk thur the game then it showed skill. Most games that I walk thur have a type of skill that most be praticed. Games that envolve pratice to keep your skill level up seem to be the most rewarding. This is what the industry is forgetting. The story lines and other things have gotten in the way alittle.

You have to wonder why every one would buy a game that lets you train your brain... It envolves pratice... to master or beable to use certain skill. Easy to pick up hard to master. Creativily beating challenges , and a the freedom to express and gamers style of play. These are key things to most players and they don't even realize it.

The only real reason DDA has popped up is because some one had to write a paper... thesis... I have not read this paper, but good game design goes beyond the difficulty level. This type of difficulty level if applied to most games will not increase there value. It's sort of like when you go to a server and you get owned by a quake player... Then go to another weaker server to kill every one up... Well That player is effectively ruining his own experience by doing that. Instead of being proud and feeling accomplished he actauly looses respect for the game. Most gamers most experience that thrill of a challenge and then start to game there way around it. Most addictive games start simple but rely on you feeling like you can do better. If a game changed it's difficulty level on the fly you would loose some of the strategy elements of the game. And really good player will use the DDa to actualy beat the game.

Someone here mentioned Oblivion, and i totally agree. DDA (or dynamic levelling in this case) totally killed the experience for me in that game.

On the opposite side, we have the Gothic series of RPG's. I dare you to venture too far out in the woods before you got your character somewhat beefed up... First time i saw a Shadowbeast coming after me I almost shit my pants! And the feeling of finally being able to take one down... amazing. Thats what I call creating an immersive world, that doesnt feel molded around YOU.

I believe this DDA thing is taking gaming to the complete opposite way of where i would like it to go...

/Martin

My friend sent me this link. I agree with all you have to say about DDA. And I'm totally against changing difficulty based on how player played. Before you read what my thesis is TRUELY about, please don't mix traditional DDA with my work. My thesis actually analyzes why traditional DDA doesn't work, and how to make a design adaptable to many different individuals.
http://www.jenovachen.com/flowingames/thesis.htm

The first thing that came to my mind when reading this was the Saga series. I believe this was incorporated in the Romancing Saga games and the first Saga Frontier. It, of course, was not the first series to implement this, but I recall it being on of my largest complaints. I love the Saga series, and I like games that scale difficulty because you are constantly being challenged.

There is nothing lamer than being forced to backtrack in an RPG through some low level area, and being forced to button mash through pointless 1 sided battles. Why would that level 1 gelatinous cube attack my level 20 warrior. Does he have a death wish? I can almost feel my pixelated avatar roll his eyes.

In these Saga games the scale seemed too steep though. It is actually recommended that you level up only when necessary in these game. If you attempt to power level in these Saga games then you will make the boss fights almost impossible, and all random encounters become a desperate fight for your life. I guess it's all about finding balance.

Also I guess it's a question of intent. I agree with #20 to an extent. Is the game intended to be immersive? Is it suppose to emulate reality? Then scaling difficulty might distract from that goal. If I'm a level 20 warrior then that gelatinous cube should fear me and run away...or at least try and gang up on me with about 20 of his friends. If the cubes gang up on me then the number of encounters should be decreased in the area accordingly considering they are now traveling in packs rather than operating independantly.

If scaling of difficulty is desired in an immersive game like Oblivion than the scaling needs to be explained properly. The cube's stats shouldn't be boosted proportionally to my stat growth, unless these cubes have been training in anticipation of my return. Perhaps changed the cube's model to add little bandanas and tonfas. I also should not have to fight gangs of cubes at the same rate as I did a single cube unless their was some sort of boom in the cube population since I've been there. Perhaps replace the cubes at some point in the game with a more fearsome enemy, but make it relavent to the story. For example, the area has been invaded by some other foreign monster, like Dragon Turtles, who have since killed off all the cubes for sport. A game can be immersive/realistic and scale difficulty. The two are not mutually exclusive.