Flow

Thoughts and responses regarding the research at acousticlearning.com.
Post Reply
KosciaK
Posts: 227
Joined: Fri Aug 12, 2005 8:45 am

Flow

Post by KosciaK » Tue Dec 12, 2006 1:00 pm

Hello!

I've just found a nice game called flOw (http://www.jenovachen.com/flowingames/flowing.htm) but it's not a game that matters here. It's the thesis behind the game. It's about active Dynamic Difficulty Adjustment so player can stay in the flow zone all the time - that means maximal focus and fun. The thesis can be read here: http://www.jenovachen.com/flowingames/thesis.htm

It might be a good idea to incorporate something like that into APB. For now not very good balance between frustration and boredom is the main drawback of the game.

KosciaK
Posts: 227
Joined: Fri Aug 12, 2005 8:45 am

Post by KosciaK » Wed Dec 13, 2006 2:29 pm

Got some ideas about this flow and making better balance between boredom and frustration in APB.
Acceleration factor could depend not only on number of correct answers in row as it is now. Answering immediately after the question means that you have answered with no problem. But correct answer but after lots of repetitions of the questions and comparing to the blue keys indicates that even though you gave the correct answer in the end but it was very hard so it shouldn't add to acceleration rate.

etaxier
Posts: 136
Joined: Tue Oct 04, 2005 1:57 am
Location: NYC

Post by etaxier » Thu Dec 14, 2006 4:41 am

Hey Kosciak, I wrote a long reply earlier, but it was more confused than anything. Not sure how to condense it, except to say "I'm not sure the flow theory applies quite so directly to APB at this point, though there is currently another good practical model for APB (chordhopper)." Perhaps in the future, all the games can be combined into a more complex game that uses flow gameplay elements.

In any case, here's the longer reply:

------------

I bet that would completely change the experience, game-wise, but I have no idea how Chris would implement it, or what kind of programming skills would have to be involved.

I think the 'flow' thesis is: To make a game as accessible as possible, the game designer includes choices in the core gameplay through which the player actively increases or decreases game difficulty. This control gets implemented on several levels, depending on the duration and complexity of the game experience. The more general theory that a game should straddle the line between easy and difficult is already a cliche, having come long before any particular thesis.

The game 'flow' itself implements its thesis on a macro scale by setting difficulty according to the player's progress (earlier levels are easier), and then letting the player choose when to go back and forth between easier and more difficult levels. This creates an open-ended difficulty slider. Of course I don't see how that would directly translate into the current incarnation of APB -- would you let players go from C to G to D etc. whenever they want, or let players go back and forth within the range of already-completed levels...?

In 'flow,' players act through an avatar in real time, and constantly make micro 'flow' choices: you can simply enjoy swimming around, or you can search for passive 'food' to make your fish bigger, or you can actively engage in 'battles' with other fish, and within those battles you can choose your own approach. Again, I have no idea how that would translate into APB. Would you have a little avatar running around, choosing whether to "collect" chords/melodies that definitely have C (you just listen to them passively) or whether to "fight" to find the C when it might not be there? Or would it be more like a slider, where you know that certain encounters have 100% chance of a C, some have 75%, some have 50%, and some have 25%? Would it become more real time by giving the player a count-down for making his choices? Would you let players go back and forth in levels WITHIN a particular pitch, but only go to the next pitch after defeating a boss? I can't think of how to incorporate that in a fairly rigid instructional game, to give the player as many choices as possible while still encouraging or enforcing a play style that results in "maximum learning." Most direct translations of the 'flow' philosophy would require completely redesigned gameplay! Indirect translations of the philosophy (like "not too easy, not too hard") are not helpful, as they incorporate too many possibilities. I suppose one general principle is always fun to think about: give the player as many in-game (not menu) choices as possible.

Of course, if you were redesigning the game, you'd want to start with the smoothest and easiest possible pedagogical progression (not sure how to put that in information theory terms -- minimum new bits per game?), then you'd want to design a game that lets the player actively control the difficulty as part of the gameplay without letting the games get longer than 5 to 10 minutes per game. Sounds a little crazy. I think that for a one-person, non open-source project like this, it's best to just try to recreate the Chordhopper experience, which is simple, direct, and easy-while-still-progressing, which makes for good short games and a positive, painless long-term learning experience (exactly what we want, from a learning point of view).

TS
Posts: 168
Joined: Sun May 07, 2006 4:58 am

Post by TS » Thu Dec 14, 2006 6:21 am

Of course I don't see how that would directly translate into the current incarnation of APB -- would you let players go from C to G to D etc. whenever they want, or let players go back and forth within the range of already-completed levels...?


This could be done for example by letting the player choose how many blue-button-questions he wants to answer. You could answer all six of them, which is difficult, or you could answer only one, and pick the easiest question. Then, if the player answers all six blue-button-questions right, he gets to shoot six missiles. If he only answers one, he gets to shoot only one missile. If he answers wrong, he doesn't get to shoot any missiles, and the aliens advance.
So, answering all six is more difficult, there's a greater chance of getting one of them wrong, but there's also the possibility to get to shoot the maximum amount of missiles. By giving the player an option to answer "I don't know", by leaving some of the blue buttons blank, the player can concentrate on the problems that he feels are suitable for him, but the constantly advancing aliens are still making sure that he must answer the harder questions eventually.

KosciaK
Posts: 227
Joined: Fri Aug 12, 2005 8:45 am

Post by KosciaK » Fri Dec 15, 2006 12:05 pm

etaxier - it wasn't confusing at all.
Active DDA might not be a good idea. It would be hard to implement and might end up playing on too easy levels (to get the impression of how good we are) instead of getting to harder levels that would make the progress.
But improving passive DDA - some mechanism to better adjust the difficulty level depending on player's answers is needed.
One of the most important questions here is deciding what is better for the learning activity - staying longer in the "safe zone" where the main point is repetition of "not so hard" questions so you can get almost 100% accuracy. Or constantly "pushing the edge" - so asking the hardest possible questions.

The flOw is just a representation of one of the possible ways of implementing the thesis so there's no point in direct translation of the gameplay to the APB.
For me Chordhopper is the best designed game in the ETC. First part is relatively easy (first 20 questions) staying in the "safe zone" and the second part is pushing the edge - where you encounter new sounds but you are free to make errors.

aruffo
Site Admin
Posts: 1692
Joined: Tue Dec 14, 2004 12:09 pm
Location: Evanston, IL

Post by aruffo » Fri Dec 15, 2006 12:43 pm

I think that for a one-person, non open-source project like this, it's best to just try to recreate the Chordhopper experience, which is simple, direct, and easy-while-still-progressing, which makes for good short games and a positive, painless long-term learning experience (exactly what we want, from a learning point of view).

I agree completely. The 50-hop limit has been such a plus that I've been trying to figure how to redesign the game without making a direct clone of Chordhopper, although it will very likely be similar. When the blue-button system becomes the main part of the gameplay the game will, I think, become much easier; I'll definitely keep in mind its flow.

I'm going to have to swear off performing next semester; rehearsing Frog and Toad has literally been a full-time job, and this is of course time that I was expecting to use for ETC v6.

Post Reply