On global points

Case number:671076-986220
Opened by:infjamc
Opened on:Sunday, May 31, 2009 - 14:20
Last modified:Sunday, April 17, 2016 - 01:55

According to the FAQ page, global points are assigned as follows:

Points = Max(1, RoundUp( 1 - (Rank - 1)/(NumPlayers - 1) )^7 ) * 100)

Indeed, the formula is used for calculating group scores and soloist scores. However, it seems to me that the number of players from the soloist competition is used for calculating the evolver and all hands scores as well. For example, take a look at an earlier puzzle, "136: Cell Adhesion" ( http://fold.it/portal/node/985906 ):

- With 42 groups, the group score correctly decays to 1 at 20th place [(21/41)^7 = 0.00925].
- However, notice that the evolver scores and all hands scores are awarded the same way as soloist scores, in the sense that you get the same number of points if you end up with the same rank. As a result, the lowest all-hands score is a 3, and the lowest evolver score is an 84.

Or take the recent "150: Rosetta Decoy 12" for example ( http://fold.it/portal/node/986180 ):

- 37 groups, with the group scores calculated correctly
- 247 soloists, so 121st place is the cutoff for 2 points
- 20 evolvers, but the scores are calculated as if there were 247 players. For example, I expected my 2nd place result [obtained by local wiggling a teammate's solution to the next full point] to yield a score of 69 [(18/19)^7 = 0.685] rather than the observed 98 [which would make sense if it's calculated as (246/247)^7 = 0.972].

==> Obviously, the problem affects everyone across the board, so it's at least being "equally unfair" to everyone. But this also means that, in the most extreme scenario, one can always get free evolver points simply by asking a teammate to uploading the starting configuration, opening the "solution," and the pressing wiggle. Why? Because even a last place in the evolver competition will yield more than a single point simply because there are always fewer evolvers than soloists in a puzzle.

Bottom line: Is this a bug, or was this intended? If it's indeed a bug, I would not mind about updating to the correct scoring system starting at a recent puzzle if re-calculating scores for every puzzle since 2008 would be too much work.

(Sun, 05/31/2009 - 14:20  |  6 comments)

Madde's picture
User offline. Last seen 2 years 50 weeks ago. Offline
Joined: 05/29/2008
Groups: Void Crushers

By the way, there is a Global Points Calculator (ODS-File for OpenOffice) in the German wiki:
(And yes, to calculate Global Evolver Points you have to enter the number of soloist players.)

LociOiling's picture
User offline. Last seen 3 hours 29 min ago. Offline
Joined: 12/27/2012

Very slow reply, but the formula used in the spreadsheet:


should be


In the MAX function, the argument "1" ensures that 1 point will be awarded.

admin's picture
User offline. Last seen 41 weeks 2 days ago. Offline
Joined: 11/10/2007
Groups: vi users
Topic: General » Server

The point raised is quite right, and yes, the best solution is to simply recalculate the scores for all the puzzles with evolvers. This will lead to changes in points, and it may even lead to changes in evolver rankings, although such variations will probably be few.

Joined: 06/17/2010
Status: Open » Closed

Scoring fixed. Closing.

infjamc's picture
User offline. Last seen 1 hour 16 min ago. Offline
Joined: 02/20/2009
Groups: Contenders
Status: Closed » Open
Type: Question » Bug

Unfortunately, it seems that the problem still exists. Look at Puzzle 428 ( http://fold.it/portal/node/989822/show_improver_players ) and 427 ( http://fold.it/portal/node/989815/show_improver_players ) for example-- evolver points are still being awarded using the denominator for the soloist competition.

infjamc's picture
User offline. Last seen 1 hour 16 min ago. Offline
Joined: 02/20/2009
Groups: Contenders
Status: Open » Closed

Closing the old case for now as there was a discussion last year to revamp the scoring system for the evolver competition.


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