Finding the Fun in Folding

[Nicky666]

Nothing fancy, really…

- background of the whole folding thing (short version of what can be read here: http://fold.it/portal/info/science)
- tools with short explanation
- some general info on the scoring (that "least energy" theory)
- some general info on backbone and sidechains
- giving some examples of folding strategies

all that info can be found in the wiki, but so much more can be found there, I felt lost when I took a first look…maybe it wouldn't hurt to use some of that info to make a short "beginner guide" that can be pointed out to people?….maybe links to the more detailed parts part and pages in the wiki could be given at the end of each chapter?

After that, the rest of the wiki could serve as "the big and detailed encyclopedia" :P


Hmmm, maybe instead of talking, I should try and set something up, so you can see what i mean, lol…

[axcho]

That makes sense to me. Where and how should such info be presented to the new players, in your opinion?

[LennStar]

I'm currently writing a text on this thread after thinking about it for two days and I have a nice suggestion I think on that. Basically: include it in the tutorial puzzles.
But I have to work a bit on the ideas, refining, I hope I'm done with it in a few hours.

[LennStar]

I'm writing my thoughts basically from top to down of the thread, and then include what I think about your ideas, so there may be things already said, but that helps if you see that others think the same.

Axcho:
Fun in foldit

I think the most motivating thing is the increase in score at a point where you thought you can't see any way to improve.
Basically its the finding of a better solution, if something you have done has worked.
In the same way it gets frustrating if the score hasen't increased for half an hour.
It is also motivating to improve a teammate's puzzle. Its a bit on the line of „I'm better than you!“ in a friendly way.

(btw: There was an interesting article about game design a few month ago in a game magazin. The most important parts is the constant feedback to the player and rewards. He must see what happened why.
Just ask a Prof on game design about the other points, I would say ;) Invite some students of computer game design, or whatever they are called, to have a look. )

Intro puzzles: They are boring. 2 reasons:

1. You are showed what you have to do, but there is no why! Score increase is NOT an answer on that.
   Write a short explanation to each level in an window: wiggling is the computer trying to fill holes with siedechains or whatever it is. (I think no player really knows this?)
   You can explain in this way, from puzzle to puzzle, how folding works, something that is really needed, I think.
2. You make the move, get the score and than… over. Go to next puzzle. Deadly wrong. That violates the 2 most important rules of game design: feedback and rewards. It is no reward to go to the next meaningless puzzle. (Here thanks to Nicky666 for the base idea)
   Make a „fork“ at the end of these puzzles: go to next, work on current, work on bigger puzzle.
   Work on current puzzle: to see if you can further improve it (not the first few puzzles, too small, but later ones). Work on bigger puzzles: An solo puzzle where the new player can try what he had just learned. Perhaps a puzzle from the first closed puzzles, with the highscore shown from the time of the closing (like in open puzzles), with „rank up“ included. [You should mention that these players used all tools and so got better scores then are possible at the moment in this tutorial]

Also include a point „help“, where is 1. what the player should do 2. a picture of a sucessfull folded version or an guide 3. a video on the complete solving.
I say that because I needed several tries on some of the tutorial levels. I have the impression that they can get stuck in a way that is not solvable if you do only what is expected. [Like the newest one with the flipping: I only succeeded in my sixt try, and there was at least one try where I took he right spot but got stuck (and so tried it on other points of the protein).] If this happens in a tutorial, epecially an early one, there is a good chance that the player says goodbye.

Also for rewards/motivation: What about a nice picture of a protein between the levels with a description of what the protein does, why it is impotant and so on. I find this very interesting, it shows the importance and possible goals of the game. If the player doesn't want to read it, he can skip it, but at least he has seen more than a boring 4-parts protein ;)

Nicky666:
The why do I like folding reasons are more or less correcct for me, except one
„and the fact that you cannot "solve" a puzzle“ that is basically wrong, you can solve it. Thats the goal. I think he means, that you don't know if you reached the goal, but hat is not a good point. I really like games like X3 (worked on a mod ^^), where you can build and build and build… but I think it is very important to know how good the puzzles was solved. The CASP8 results where very motivating for everybody. (I only quote spvincent: „Great news: I fear that this will only encourage me to waste even more time playing foldit than I already do.“) So give feedback (yeah, again feedback :D) on the previous puzzles. If the players where sucessful or not, what was a key point or whatever. A few lines per puzzles only. 5 minutes per week.

Not fun:
manual backbone walks: I think everything that is done repeatedly over a long time gets boring, especially if it doesn't bring points. But that is hard to solve.

Better for newcomers: mentioned it above. Seems Nicky666 sees the same problems.

Second post: „The fact that this whole folding has a purpose is why I allow myself to spend so much time on it.“ Big point. Very big. That's why you should explain the purpose and give feedback on how well the players did. Very big Point!!! Just think of the people who have half a dozen computers running on BOINC projects. They maybe like the teams, they maybe like the statistics (as there are several statistics sites, that seems to be an important point, you can improve on that), maybe… but at least half of them wouldn't even run their one used computer with BOINC if they would not think that they do something important.

Ingame chat: I think you also underestimate that. Folding can be a depressing or just boring thing if you wait for your computer to finish the wiggling, 0.001 point at a time…
Thats when I ( a really silent guy offline) start chatting. Making jokes. Saying stupid thinks. It covers time.
Improve the chat: Make it possible for 1on 1 chat or invited-groups-chat.

Axcho post:
„It seems to me that the fun of Foldit is not so much in finding the "right answer" as it is about finding all the little places where you can get a few more points. It's more like Minesweeper (or Solitaire?) than a jigsaw puzzle.
Right now each of the intro puzzles is designed like a "right answer" puzzle (though not very fun ones, at that) but really, they should be more about collecting points by doing stuff.“
Thats exactly what I think and have talked about above.

„where the extra points you get beyond the level requirement“
Highscore with the sum of all tutorial levels. (not the extra ones I suggested, but they could be rated in an all-together-highscore, too).

Other suggestions:
Make the achievements more rewarding. Add a nive picture to it. Give the best solver of a puzzle a diploma with his name that he can print out or just make a desktop picture. Give the best team a medal on their team page.
As I said in a feedback, make the rankings more visible for the players: difference to last puzzle, ranking graph… (if you want to see how to not make a rank graph, I can give you a link to my animexx Go-rating, that graph is hillariosly unuseful. Don't make the same mistakes ;))
Thats just gimmiks, but many small things add up.

I thought on a randomizer. A tool that changes sidechains (choose amount) around a sidechain you clicked in random positions. You never know what this brings. Like a rebuild for sidechains.

And one idea I can't remember that I had 2 days ago in the bed. Unfortunately my page that is pinned to the wall at the side of my bed for exact these reasons (that I get the most ideas when I want to sleep and forget them on the next morning) was full from the night before and I hadn't changed it, so I didn't wrote it down. You see, its a regular problem for me with the ideas at night :D
I hope I can remember it again…

[axcho]

LennStar, thank you very much! I totally agree with everything you've said here. Especially about why the intro puzzles are boring, and especially the lack of "constant feedback to the player and rewards. He must see what happened why." You have covered almost all the areas of improvement I see for Foldit, at least in terms of improving the experience for new players. I'd like to write up a summary of it in my own words. I'll post it here and we can all go back and forth on it.

I had started coming up with a plan for completely new intro puzzles, which I was excited about testing out. However, since I am almost done with my time on the Foldit team, I have been advised to focus on wrapping up the various things I've started rather than launching into a whole new refactoring of the design. In the slight chance that I'd stay on for another few months, I'd definitely make this redesign my goal.

But given that this is unlikely, the plan which seems to have the best chance of success would be to later make a 2D protein-folding game in Flash where I can more easily experiment with these better game design practices. Then if I am able to create a successful protein-folding game in 2D, I can hand this to the Foldit team and allow them to model their design after this proven 2D prototype. Though I am not confident that there will be anyone on the team prepared to put their time into improving the game rather than adding technical features. But that's my plan.

[axcho]

I was just talking with Seth (the main developer on Foldit) about the problems with the current intro puzzles and how we might improve them to be more clear, more robust, more informative, and more fun. He suggested that we could start by making puzzles where the objective is to get your protein to match the shape shown by a guide rather than to get past a certain score.

I'm thinking about this, and I'm curious what everyone else thinks of this idea. What might be the advantages or disadvantages, and how it might affect the way players learn or how levels could be designed? I'm thinking using guides as a goal could be a good compromise between the current tutorial-ish levels and more experimental exploratory levels that may be too hard to design at this point. At the least it would provide a stronger foundation for tool-based progression, letting players make use of the tools they've been introduced to in previous levels rather than being forced into using what the particular tutorial level is trying to show you.

The guide-based, rather than score-based, approach would have the advantage that feedback to the player's actions would be much less arbitrary, in that seeing how closely your protein matches the guide is much more informative than seeing whether a single score went up or down. You're moving around in 3D, so you really want to have more than 1D of feedback to guide you. Having a protein guide would help provide a few extra dimensions of feedback. Localized feedback such as clashes, bonds, voids, and exposed hydrophobic markings also provide some much-needed guidance for your 3D actions beyond a 1D score change. I'd like to make use of voids and exposed hydrophobic markings (which needs a shorter name, by the way) to a much greater degree in the intro puzzles as well, for this same reason. You'd still see your score, of course, but it would feature less prominently.

There could still be some levels that use a score requirement instead of a guide. Seth suggested that we make the last level of each levelset a "boss" level where you don't have a guide to follow, and you have to use what you've learned to get a high score. I'd also use the score for the extra degrees of completion you can earn for each puzzle beyond the basic requirement. You could unlock the next level by matching the guide, but to get the "Expert" completion distinction for the level, say, you'd have to get a certain number of points that would require some more sophisticated fine-tuning of the structure.

So it seems that using a guide as the primary goal of each level could make sense. It certainly fits into the common physics game idiom of "get this one thing to the other thing" whether it is an armadillo to a portal, a goo tower to a pipe, or a contraption to an exit. I'll have to think about what such a level might look like.

Any thoughts?

[ZeddicusZulZorrandr]

I like this idea.  There are already a few intro levels that use guides(at least there were), but more extensive use of such guides would be useful.  I would suggest using them for each new tool. I would suggest separate tutorials for each of the uses of tweaks(I don't think there is a good straighten one at the moment), and several different examples of rebuilds, as they are probably the most complicated tools.  A color change of the guide to indicate how close the current protein is to it.

As for suggestions for shorter names for exposed hydrophobics, I think wet willies sounds good:)

[axcho]

Hey, glad you like the idea. :) I agree that Tweak and Rebuild both deserve several puzzles each. Currently the guide becomes more transparent as you get close to it. I'd imagine that a color change would be too distracting. What do you think?

I was just trying to come up with some names for exposed hydrophobics. How about "leak"? Like, the orange part is leaking out instead of being contained within.

<ul><li>clash</li><li>void</li><li>leak</li><li>bond</li></ul>

[ZeddicusZulZorrandr]

The reason that i suggested a color change rather than transparency is because I would think that it would become more difficult to work wit the guide if it became too transparent.  I don't use them often, so I don't know if this is a problem.

[axcho]

I work with guides a lot in the intro puzzles and I would imagine that the reverse would be true.

Would anyone else like to offer an opinion about the guide transparency versus color change?