haiop
How can I game this information to maximize my rating?
Its good
Interesting, but really useless. Without local time it means nothing. To say most mistakes are made at noon Pacific time, California, is just ridiculous. Most likely this number is so high because people in Europe are deep into their evening. Or in Asia it is the middle of the night.
Mom: How did you lose?
Me: Ahh Yes, The gravitational force of the sun....
It should discuss the errors in each move like no. Of errors on move 21 rather than time
This would be more interesting and meaningful if it was based on the "blundering" players local time and took into consideration the relative ratings or rating difference between the players or simply games where both players were above a certain rating (say 1800 0r 2000 or 2500)
As stated above, this should be analyzed using players local time, on both sides separately as players may be in different time zones.
I do wonder wether the results are statistically significant. If you look at the y axes of these plots, the range where the data varries is very small compared to the average value.
Yes, there are ‘wiggles’ in the data that may mean something but look at the first graph, the changes in error rate are from 13.35 to 13.60 (assuming that that is what is plotted as it seems labeled in Russian and blurry...). This range of variation, ~0.25 is very small compared to the average of ~13.5 and typically that indicates its not significant statistically.
As stated above, this should be analyzed using players local time, on both sides separately as players may be in different time zones.
I do wonder wether the results are statistically significant. If you look at the y axes of these plots, the range where the data varries is very small compared to the average value.
Yes, there are ‘wiggles’ in the data that may mean something but look at the first graph, the changes in error rate are from 13.35 to 13.60 (assuming that that is what is plotted as it seems labeled in Russian and blurry...). This range of variation, ~0.25 is very small compared to the average of ~13.5 and typically that indicates its not significant statistically.
As this is based on UTC, not local players time, it could also be caused by more stronger players playing at that time. Is Lichess skewed towards European players and are they better, i.e. making less mistakes than players elsewhere.
So, to draw any conclusions, more work is needed as they say...