Logic Puzzle Forums (http://www.logic-puzzles.org/forum/index.php)
-   Comments, Bugs and Feature Requests (http://www.logic-puzzles.org/forum/forumdisplay.php?f=3)
-   -   Feature Request - +/- average time ranking (http://www.logic-puzzles.org/forum/showthread.php?t=110)

 thuban51 09-16-2010 07:45 PM

Feature Request - +/- average time ranking

One of the personal stats you have is average time but I feel this isn't as good of a stat as some puzzles will obviously take longer (i.e puzzles with 5 people as opposed to 4) so it skews your average.

I think an easier way of ranking people on time should be by having a cumulative total of second over/under average puzzle time. For example:

Average Puzzle Time: 361 seconds

In this situation i got 73 points for finishing 73 second before the average. If on the next puzzle i finish 56 seconds before the average time then my cumulative total would increase to 129. On the next puzzle i get a -19, then my cumulative would then be 110. And so on and so on.

Doing it this way will give a clear cut number which makes it easier to rank people based on time and puzzle difficulty. This scoring prevents people from getting low average times by only picking puzzles with high completion rates(which can also be deemed as easier to solve).

 pcuser 09-16-2010 08:42 PM

Agreement

Hi,
I agree that your ranking scheme is more reflective of skill. However, I wouldn't want to try and program the site to calculate the stat. Also, I almost never solve 4X4 puzzles and I never solve really high completion rate puzzles. But, I am always at the top of the list for lowest solve time. I'm averaging around 128 seconds per puzzle. You just have to do a lot of puzzles and have a mathematical mind and you can dramatically lower your times.
pcuser

 demeester 09-17-2010 08:29 AM

A problem with this suggestion is that the average changes! So, if you are above average, you get a score. The next person does the same puzzle in the same time, but their score would now be less, because your solve upped the average.

Another problem is that you would need to average these in themselves. If you do 10 at a lot above average it is the same as doing 20 at a little bit above average, or 100 at just a tiny bit above average, so you still have no gauge.

So, you still have no clearcut unless you put all of that in to play.

Also, a score of +10 on a quick 4x4 is actually better than +10 on a difficult 5x5 - so you still have not taken out the skewing, unless you scale score in proportion to the average time.

 pcuser 09-17-2010 04:04 PM

Data validity

The points you make are valid. However, I don't think the average times are updated that often. I've set hundreds & hundreds of new records and have yet to see my name listed. Even if the stats were updated frequently, the averages would even out over time. As for doing 100 puzzles for few points to equal ten with high points, I would submit that it takes far longer to do 100 small ones than 10 difficult ones. As for the 4X4 puzzles and the 5X5 puzzles, nothing about their statistics compares. You can only get about 350 points max on the 4X4 puzzles while you can get about 530 points max on the 5X5 puzzles.
pcuser

 demeester 09-21-2010 12:28 PM

Quote:
 Originally Posted by pcuser (Post 436) The points you make are valid. However, I don't think the average times are updated that often. I've set hundreds & hundreds of new records and have yet to see my name listed. Even if the stats were updated frequently, the averages would even out over time. As for doing 100 puzzles for few points to equal ten with high points, I would submit that it takes far longer to do 100 small ones than 10 difficult ones. As for the 4X4 puzzles and the 5X5 puzzles, nothing about their statistics compares. You can only get about 350 points max on the 4X4 puzzles while you can get about 530 points max on the 5X5 puzzles. pcuser
If the averages are not updated often, then that is possibly worse - although I agree they would flatten over time. But any "NEW"ish puzzles would have to be excluded.

The suggestion was to award points based on a variance from average. So, doing 100 puzzles at 2 seconds below average would be -200 points, as would doing 10 puzzles at 20 seconds below average. This is not saying small or difficult, just saying that even doing the same puzzle types, they have the same points but obviously the 2nd example is a much quicker solver.

I agree that there may be an argument for 2 sets of stats - one for 4X4 and one for 5X5 as they are two different categories entirely.

 demeester 09-21-2010 12:31 PM

Quote:
 Originally Posted by pcuser (Post 436) I've set hundreds & hundreds of new records and have yet to see my name listed.
p.s. I have seen your name. I am yet to see mine, despite setting records myself.

 All times are GMT. The time now is 01:28 AM.