Script only puzzles

Case number:699969-995597
Topic:General
Opened by:jamiexq
Status:Open
Type:Suggestion
Opened on:Sunday, July 21, 2013 - 15:47
Last modified:Thursday, July 25, 2013 - 22:57

If we are going to have hand-folding only we should also have script only folding puzzles - this would be equivalent imho. If you are going to handicap the script folders with hand-folding only, then the reverse is fair as well. With regular puzzles, the hand-folders cans still use their comfort skill of handfolding.
Why not also have some script only puzzles to equally handicap the hand-folders as the script folders are experiencing? Just a thought

(Sun, 07/21/2013 - 15:47  |  19 comments)


Joined: 04/15/2012
Groups: Beta Folders

Because the Devs aren't looking for just scripts. That's what Rosetta is for, right?

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

Script only folders are not the same as Rosetta - They are making determinations on what scripts to run - how long to run them -modifying them to the particular need of a particular puzzle

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

And by that logic, Why then allow any scripts at all on the handfolding only puzzles? Why allow GUI scripts?

Joined: 04/15/2012
Groups: Beta Folders

Because they are generally small scripts that make your life easier. Change band length, finding the best settling technique, etc. And yes, I get your point, but the Devs were not looking to handicap non-hand folders. They are looking for results. They are not looking for the results you see from just scripts, to my knowledge. Please remember that this is not just a game here, but a science tool. Try not to abuse it too much.

Joined: 09/21/2011
Groups: Void Crushers

The first thing I would do on script only puzzles is making tools to simulate the handfolding. :-)
A pity I would miss the cut tool.

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

Wow - didn't realize suggestions were abusing the system. Nor using scripts were abusing the system.
This is an interesting perspective AsDawnBreaks.

Joined: 05/21/2010

Abuse? I think this is a bit overkill. Funny, but overkill.

Joined: 04/15/2012
Groups: Beta Folders

Not quite what I meant. I'm just saying try not to think of it just like a mindless game where only the points matter. The #1 reason this exists is science. Entertainment comes at #2, but exists as a side-effect of the #1 goal. Granted, it is entertaining, and it's a great thing, but not the whole point.

Joined: 04/15/2012
Groups: Beta Folders

And scripts aren't bad all the time. The devs aren't looking for just scripts. They're looking for designs, and they're approaching that using hand-folding. If they needed just scripts, I'd think they would use it.

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

So then the question becomes:

"What feedback system is there, for the player, to know if they are producing a good design or a bad design?"

The only feedback system I see is score. So then, if we make the following assumptions about score:

1. Higher the score, better the design - then player would use any and all tools to get highest score.(It think we have all seen examples of high scoring badly designed proteins)
2. Score is not indicative of a good design.
3. Score may or may not be indicative of a good design.

So given either assumption 2 or 3, then what feedback does a player ultimately have to know if they have a good design or not?

If a player has no reliably accurate feedback system on the value of their design, how does that player then learn?

They most likely don't. They most likely look to the only feedback system available to them - their score and rank based solely on score.

Joined: 04/15/2012
Groups: Beta Folders

True. And this is why hand-folding was implemented, and why they are continually trying to improve the scoring function and the usability of our tools. Score does have some credit, but it isn't 100% accurate, so they try to make it better. If they weren't having the issue with scripts giving false local minimums (and causing us not to think about our design and actually do some designing), they wouldn't use had folding.

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

This assumes there cannot be just as many false local minimums with handfolding, that script users are not thinking just as much about their designs and that handfolding only puzzles cannot also have no-thought awful designs and still score well

marie_s's picture
User offline. Last seen 2 years 5 weeks ago. Offline
Joined: 05/18/2008
Groups: None

You have other feedbacks than score.
When the native is published, the information is written on the wiki so you can see if your protein was close to the real protein.
For example, with the puzzle 709, 715 and 741.
Susume and Spdenne were very close of the result in 741.
On 709, some results are closer to the native than other : Brow for exemple.
You can watch your solution to compare with the pictures of the natives or the 3D you will find on PDB site.

jamiexq's picture
User offline. Last seen 10 hours 16 min ago. Offline
Joined: 07/31/2011

so sounds like banning scripts is next logical move. It will be interesting to see how many players stay after that happens.

Joined: 04/15/2012
Groups: Beta Folders

Not all scripts are bad. People just tend to over-script.

Joined: 09/21/2011

741 was close because the scoring works on electron density, and the cloud is "real time" feedback which i think jamie is looking for. the other puzzles had a couple that were close to the native, but not the very top scorers if i remember correctly.

now, the native gets published after the puzzle expires, so you cannot adjust your protein after the fact. but if you find researching published natives and pdb, and then in some savant like way making an anal yses of the variations between your protein and the native and building new proteins that are each one getting closer to the native than the last with these feedbacks, a realistic and enjoyable gaming solution... who am i to argue.

but about the suggestion, i don't think that would work great from a science perspective. i'd play it if they don't make me play it again with a handfold only release (although my endgame handwork is pretty good). these double releases give me a headache.

Joined: 11/05/2010
Groups: None

Double and triple releases give me a headache too. I still have yet to figure out what natural and un-natural solutions mean. Except in my mind un-natural is what we force a solution to become with our handfoldings and natural is the basic mathematical results of scripting. I'm sure I'm wrong by someone's perspective, but for the most part, I've given up caring what others think.. I like your idea Jamie and since I dislike negative voting so much, I just have to vote it up, besides I see it as being in good taste, compared to the very bad taste that is left by not allowing scripting only puzzles. GUI scripts in handfolding puzzles should be disallowed. We seem to be wasting a lot of breath and typing on issues with wiggle, which to me, has taken us far out on a tangent away from the goals of Foldit. And I, for one, would like to see a stop to it. I know the kind of stop I would like, but I'm also pretty sure that I'm never going to see it. Foldit has become a very strong reminder of the V.A. Administration system. Hurry up on the puzzles and ' wait to the death for many ' for any decent fixes to the client. Now the devs in their wisdom had added another complexity to the client with a choice between wiggles. So how will we be able to compare results and what will it all mean?? I'm not qualified to say. I wonder if the brute force answer of disposing of wiggle altogether might not be a viable alternative. We need goal oriented scripts, not new renditions of the ancient wiggle oriented scripts that proved so lackluster in the past and are being proved that way again and again. So in other words we need more variation and wiggle does not create variation, it instead creates similarities between solutions. I want Foldit to be FUN again.

brow42's picture
User offline. Last seen 1 day 2 hours ago. Offline
Joined: 09/19/2011
Groups: None

GUI scripts in handfolding puzzles should be disallowed.

If you don't care what others think, then you probably won't read this. But I think you are completely wrong about GUI scripts. In almost any other place, "GUI Scripts" would be called macros. They are literally a record of what clicks a human would make, and makes them. That's what determines the limited options of commands. That's why they grind to a halt when you minimize on windows. Macros are added because a lot of clicks are really repetitive.

No GUI scripts means either fusing by hand, or not knowing what your fold "really" scores. I WANT to know what my fold scores. But I am never, ever, going back to fusing by hand because moving that slider back and forth is the OPPOSITE of fun.

Maybe Foldit should have no-score puzzles. I have no idea how they would turn that into a game. It would certainly increase diversity.

brow42's picture
User offline. Last seen 1 day 2 hours ago. Offline
Joined: 09/19/2011
Groups: None

P.S.

1) I thought jamiexq was being sarcastic.

2) I've never known what you meant by goal-oriented scripting. You have a particular dislike for what you called wiggle-based scripting. Well, we humans certainly have abstract goals when we play Foldit. But we have to achieve those using the client. The client, and the Rosetta engine, has only one goal: find the lowest energy for a given sequence. You won't achieve that without wiggle.

"Fast. Good. Cheap. Pick any two." It applies to wiggle. It applies to everything.

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons