Solve time list?

[AirRandom]

Wondering how the solve times are stored on the back end... in addition to the "Record Time" shown before and after each puzzle, it'd be cool to see an expandable list of solve times by user after the puzzle to see how you stack up against the other users. Getting a new record time is cool and all, but seeing that you have the #2 time (or beat the time of another user you're usually competitive with) would also be really nice. I'd imagine that all the solve times are stored in a table somewhere because of the statistics shown, so if that's the case, it shouldn't be too difficult or cumbersome code-wise to pull them all up on the solution page.

[redpicker]

I dunno, that sounds a bit data intensive and brings up privacy issues. Why should anyone else have access to my solve times? That also assumes they maintain a record the individual solve times on each player (which I doubt). They probably only record statistical information, other than the player with the current record for each puzzle.

I do think it would improve a player's enjoyment if they could display what percentile you scored. That is, for example, if you achieved the "average" time, one would expect you to be in the 50th percentile (your score was faster than 50% of all those who have completed the puzzle). If you scored in the 95th percentile, than you were faster than 95% of those who completed it.

[genevoise]

Graphing the actual scores could solve the privacy issues while still letting you see for example if the record time was a complete outlier.

[AirRandom]

The percentile is already there, in the form of the position on the bell curve.

Maybe just the top ten scores for the puzzle? As far as privacy issues go, for that case, it seems no different than displaying your user name if you have the record time for the puzzle. Displaying the top ten scores seems like a small extension of that. If you're below the top ten, your score is rolled into the bell curve.

[redpicker]

In your example, what would you say the percentile is?

The bell curve is very deceiving. I don't believe it is showing what you assume it is showing. Have you noticed that the minimum time is always a whole number but the maximum is either a whole number or one half? The average time is always one fifth of the maximum, so it is always a whole number or ends in x.3, x.5, or x.7? The differences between shadings is not equal, so the shadings go not represent standard deviations (so, what do they represent?).

The distribution of scores is not a "normal distribution" (it is obviously skewed) but the bell curve appears to be a depiction of a normal distribution. Statistics are math intensive and most people don't care enough to dig into them much. The graduations of "Very Fast", "Fast", "Average", "Slow", and "Very Slow" are informative, but does it mean that "Very Fast" >80%, "Fast" > 60%, "Average" > 40%, "Slow" > 20%, and "Very Slow" < 20%? That would be the implication, but I am not sure that is correct.

Of course, they probably do not have the records to do this, just like they probably don't have the records of the names of the "top ten" solvers for each puzzle (is that ten shortest times, or the ten highest scores?), so it probably doesn't matter. However, since the administrator has decided to change the scoring algorithm to give an incentive for players to solve longer puzzles, if they had a way to give the percentile of a player's score (and, perhaps, the min, max, and mean percentiles of the player's solve times for the month), it would make it easier to see how much a player was improving.