Actual competitors for the global score

Case number:845818-996160
Topic:Game: Social
Opened by:Bruno Kestemont
Opened on:Friday, October 25, 2013 - 09:46
Last modified:Thursday, November 14, 2013 - 18:13

On puzzle 797 I largely underperformed but I was surprised to get 80 points. It appears that it's thanks to the large amount of players (809) for this puzzle, probably students, most of them only opening the puzzle not getting more than 1 point.

Advantage: It's a good incentive to all of us to try to get more players. It gives a warm unattended reward for most good players.

Inconvenience: In popular puzzles, it's so easy to get many global points that there is a risk of under-performance on the top side. Less popular puzzles risk suffering a small number of competitors. This all is not optimal for Science.

Suggestion: In the scoring system (Points = Max(1, RoundUp(((1 - R/N)^5) * X))), only counting the number N of players that enhanced the starting position (getting more points than the starting points). I mean real competitors, not the number of players N who opened the puzzle.

Motivation: The aim of scoring system including the number of competitors N is to reward good ranked people amongst a large number of competitors. But 1 soloist point players are not really competing. They opened the window to see the puzzle or to load another player's solution to evo.

(Fri, 10/25/2013 - 09:46  |  7 comments)

Joined: 04/15/2012
Groups: Beta Folders

Yep. I think it would be good to include scoring range (highs, lows, mids)/difference in the global score calculation myself.

Joined: 09/21/2011
Groups: Void Crushers

Actually only the ones that outscore 1 shake, 1 wiggle sidechains and 2 wiggles should count.

Joined: 12/06/2008
Groups: Contenders

The current scoring system isn't broken.

There is almost no one who opens a puzzle who has no intention of playing it. Conversely, if a puzzle was opened, nearly everyone opening it intended to play it to win.

We have no idea why anyone doesn't score well, and therefore have no right to say they aren't "real competitors". Kicking their efforts out of the ranking system because of their inability to score well simply says to those people "your time and effort here doesn't matter".

Joined: 09/21/2011
Groups: Void Crushers

I didnt mean to say those would not get points. But for the scoring formulae it would be good to have a cutoff point to distinguise between players that have only taken a very brief look and those that have actually played it. I have seen people in global who started in big ones without having done a <150 puzzle or even without finishing the tutorials and having no clue what to do.

Joined: 04/15/2012
Groups: Beta Folders

Exactly. And it's not like they might really work it, when you can just shake and wiggle to jump 50 ranks.

Joined: 09/24/2012
Groups: Go Science

Personally, I often open a puzzle only with the aim to evolve. Open, open/share a team member solution, evolve. But I get 1 point on my soloist.
Other reason: to look at the puzzle. When I see it'll ask too much time (like a de novo) or if it does not inspire me, I might change mind and go to another puzzle.

Joined: 09/24/2012
Groups: Go Science
Status: Open » Closed

ok I close this, there are enough discussions on this topic.

See this one I like:

and the madde 's list just to recall it exists ad could be done by Foldit team in the list of achievements:


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