Give a non nul bonus for informative filters

Case number:699969-2009833
Topic:General
Opened by:Bruno Kestemont
Status:Open
Type:Suggestion
Opened on:Sunday, June 14, 2020 - 13:44
Last modified:Wednesday, September 2, 2020 - 18:02

In last puzzles, you don't reward several filters like Core Complex or Buried Unsats.

However, the information is there (number buried polars found)... but we cannot use it for scripts !

Suggestion: always give a very small (non null) reward for these, so that we can give it a higher weight - if we want, - in scripts.

Example:

I have a solution with 8 buried polars (no score change) and another one with 15 buried polars (no score change).

Manually, I can select the right one.

With scripts, I can't.

If there was a very small reward (even 0.001 pt per buried polars), my script could select the best one (max of buried polars filter bonus) even with no impact on the global score.

(Sun, 06/14/2020 - 13:44  |  3 comments)


Susume's picture
User offline. Last seen 1 day 15 hours ago. Offline
Joined: 10/02/2011

This is a brilliant idea. It allows the players to choose how to prioritize the information you are making available without giving up the power of the player-developed algorithms that scripts represent.

Joined: 09/24/2012
Groups: Go Science

Same for binder metrics to measure the DDG, SASA, and Shape Complementarity.

Just give a very small "score" which is correlated with the metrics themselves. We are mainly interested in raising these metrics, whatever the real units ans filter weight given by scientists as "best default".

Optionally you give us the possibility to retrieve these numbers "on demand", without counting them in the puzzle score (if this option can save compute time - I understand that some filters are more CPU consuming han others).

With recipes (or in hand fold), we can activate/desactivate some slow filters in order to save time. The only problem if you let these very slow filters always running is "psychological" for handfolding beginners. It's easy to explain that disabling some of these filters speed up the computer.

bkoep's picture
User offline. Last seen 4 hours 29 min ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

I like this idea!

Ideally, it might make sense to have separate script functions for GetFilterResult and GetFilterScore. But for now I agree that it would be easy and effective to award a very small bonus for binder metrics that are turned off.

Ultimately, we would like for the binder metrics to be a part of the score (with a significant bonus).

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