Puzzle size and speed / mix

Case number:699969-2000477
Topic:General
Opened by:Skippysk8s
Status:Open
Type:Suggestion
Opened on:Monday, March 9, 2015 - 02:07
Last modified:Monday, March 16, 2015 - 14:46

It is taking all night to run a program that might take an hour on ordinary puzzles for the symmetry puzzles like 1059 and apparently 1060 (I didn't even open that after looking at comments on vet chat). Most of us are down to a single client and still struggling. Even fast machine people are having problems.
We need to find a way that these puzzles are workable. Puzzlers can't even try alternatives due to work speed. I think it would benefit everyone to rethink if there might be a way to achieve science objectives and have a puzzle that more than 20 people can actually work.
In the meanwhile, can we mix in some smaller puzzles? that way those of us nurturing novices as well as those of use with smaller machines can still play. We would hate to lose the next great player due to frustration, and many of our vets want to work on puzzles but lack adequate capacity.
Watching new season of House of Cards....
Thanks
Skippy

(Mon, 03/09/2015 - 02:07  |  2 comments)


Joined: 04/24/2014
Groups: None

Thanks for this feedback, Skippy! I've brought it up with the team to see what we can do on this front.

Joined: 09/24/2012
Groups: Go Science

With the combination of puzzle length and number of filters, it's possible to evaluate how heavy a puzzle is.

1) You could inform the players by giving an info like "Puzzle demand for CPU: high, medium, low", or "estimated response of the puzzle: slow, medium, fast".

2) Never post more than 1 heavy puzzle at same time

3) Always keep a fast puzzle in the list

4) Independent period management for fast, normal, slow puzzles cycles: you might decide to give us 10 days for slow puzzles (10 days are then needed before we get the next slow one).

5) Maximum (group) points could be different in order to encourage groups to invest in slow puzzles (e.g. 80 to 120 max points for fast to slow puzzles). Why groups? Because a group is less subject to resource limitations. For soloists, I suppose it's not fair to give more points to puzzles not everybody can play.

6) Concerning solo rankings, you might decide to add a new category "heavy puzzles". The puzzles in this category would be like a bonus for other categories. I mean, if a category (e.g. Design) takes the best 8 scores out of last 10 puzzles, it would be changed like this: the best of (the best 8 normal puzzles out of the last 10 normal puzzles) and (the best 8 puzzles out of the last 10 including heavy puzzles).

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