puzzle picture
470: Revisiting Puzzle 153
Status: Closed


Name: 470: Revisiting Puzzle 153
Status: Closed
Created: 10/19/2011
Points: 100
Expired: 10/28/2011 - 02:00
Difficulty: Beginner
Description: We are revisting old Foldit puzzles so we can compare results to see how useful the recent additions to the game have been. Players will NOT be able to load in any previous solutions for these puzzles. The fourth Revisiting Puzzle goes back to June 2009: http://fold.it/portal/node/986227
Categories: Overall, Prediction

Top Groups

1Void Crushers10,881100
2Anthropic Dreams10,85983
4Trailer Trash10,82156

Top Evolvers

Top Soloists

Need this puzzle? Log in to download.  


Joined: 12/07/2007
Groups: Contenders
Improved scores

It seems that the scores for these "revisited" puzzles are quite a bit higher than the scores in the original puzzles. Assuming the scoring function hasn't changed in any way, any thoughts as to why?

spmm's picture
User offline. Last seen 2 weeks 2 days ago. Offline
Joined: 08/05/2010
Groups: Void Crushers
3 things

IMHO these three things would have an influence:

1 Greater individual and collective experience in developing solutions. Including exposure to more protein types and shapes

2 Improved scripts, plus old and newscripts used more effectively (see 1)

3 Faster personal computers, allowing for more experimentation to arrive at a solution

Joined: 05/09/2008
Groups: None
we have noticed the same thing

...which is why we find these puzzles very interesting.

The question is also: is it just the scores that have gotten better (you have all gotten better at optimizing the Rosetta energy function) or are the solutions getting better (closer to the correct answer) which would also explain better scores (lower energies).

Obviously we will need to run a fair amount of these puzzles before we can come to a statistically significant conclusion, but we all agree these results are very interesting (and we hope these puzzles are old enough that they aren't too boring for those of you who played them already).

spmm's picture
User offline. Last seen 2 weeks 2 days ago. Offline
Joined: 08/05/2010
Groups: Void Crushers
'are the solutions getting better'

'are the solutions getting better (closer to the correct answer)'
It would be helpful to have that question answered for each puzzle, soon after they are completed and to compare the two. Is that possible or desirable?
It may be that in some sections solutions are very similar and in other sections quite different, but perhaps still not the 'best' answer.

Joined: 05/09/2008
Groups: None
I realize how frustrating that must be,

but the reason we don't post the solutions (or a Quest to the Native) just after a puzzle closes is that we often want to give you the same puzzle in a different way:

for example the Exploration Puzzles or using Electron Density or adding constraints.

The only way we can gauge the impact of constraints, exploration, or using Electron Density is if we first have a control.

I completely realize that from your end it may seem like we aren't giving you enough information or feedback, and for that we do apologize.

We do try to provide as much information as we can for unsolved cases (since we obviously don't know the correct answer and want to make sure you have all the help you can get) but since those cases are rare we use the standard puzzles to gauge how we can improve the game to help you fold more efficiently.

spmm's picture
User offline. Last seen 2 weeks 2 days ago. Offline
Joined: 08/05/2010
Groups: Void Crushers
fair enough thanks

understand, and I was certainly wondering what we might come up with for this puzzle if we were given it as an exploration or density puzzle.

User login
Download links:
  Windows    OSX    Linux  
(10.12 or later)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Social Media

Only search fold.it
Other Games: Mozak
Recommend Foldit
Top New Users

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