New system to encourage making more solutions for scientists.

Case number:699969-2011214
Opened by:nspc
Opened on:Wednesday, February 17, 2021 - 21:28
Last modified:Sunday, February 21, 2021 - 02:39

The "Move Limit" recently used in new design puzzle 1956, adds an important constraints to the many already existing constraints and difficulties in this kind of puzzle.

It prevents strategies, such as starting one solution from another for exemple, or the use of some tools consumes too much move, etc.

I have a suggestion in place of "Move Limit" to encourage more solution shares :

We could simply have a system that attributes points to players, when the solutions they share are scientifically validated.

It can be those that are selected to be tested in the lab, but it can be an other criterion (just a validation like "interesting solution" by a scientist for exemple).

This system can be independant from the last puzzle. It is just an other system that attribute points later.

This would not add any constraint for the creation of a solution, and would encourage making several.
And also, instead of making only the best score, players would do more interesting solutions for scientists.

(Wed, 02/17/2021 - 21:28  |  4 comments)

jeff101's picture
User offline. Last seen 3 days 14 min ago. Offline
Joined: 04/20/2012
Groups: Go Science

The Feedback below seems related
(a different way to achieve the same goal)

bkoep's picture
User offline. Last seen 1 hour 15 min ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

I really like this idea! We definitely want to encourage more solution attempts per puzzle, and the Move Limit is an imperfect mechanism for that goal. Your suggestion of attributing points for every unique "valid" solution is much more direct.

We have some ideas about how to do this, and hopefully we'll be able to implement something similar in the near future. Most likely, this would be kept entirely separate from the current ranking system that awards points based only on the best scoring solution from each player. We do like certain aspects of the current ranking system -- but it was developed with an eye toward protein structure prediction problems. In recent years Foldit has shifted away from structure prediction, so it's maybe not surprising that the points system could be revised.

I think you're right that we can do better to reward multiple solutions for each puzzle. Ultimately, this is a better measure of each player's contribution to actual research goals.

Joined: 12/06/2008
Groups: Contenders

It ain't broke, so it don't need fixin'.

There's nothing wrong with the current system of scoring a puzzle and assigning a rank based on that score. The existing system rewards players for how well they meet the criteria for a puzzle. The scientists know what's important, and the scoring system (and the more recently introduced metrics) shows how well a player can design a protein to meet those requirements.

A separate leaderboard for keeping track of the number and or quality of "acceptable" solutions could encourage players to submit more solutions. But don't change something that's worked for over a dozen years.

infjamc's picture
User offline. Last seen 20 hours 5 min ago. Offline
Joined: 02/20/2009
Groups: Contenders

Another possibility to consider is the creation of new Foldit achievements.

(There's an existing "The Write Stuff" category, but it's limited to being part of a published paper-- and "scientifically valid" solutions may or may not necessarily end up in a journal article.)


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, Boehringer Ingelheim, RosettaCommons