Discussion about this post

User's avatar
mukul's avatar

More like the conclusion that progress doesn't have any major pattern. With computers maybe initial rise is getting at younger age. Would be cool to look for this pattern on more number of players of each generation ( maybe looking at average)

Expand full comment
dboing dboing's avatar

Since the statistics of the ELO type are only making sense and depend on the variable that is the set of players and the set of games played, in "their" pool, I wonder how that floating variable, in looking over life span durations, would complement this research here.

The demographics of the pool. In fact I have difficulty or igrnorance about how the pool, of players/games driving the ELO statistics, evolved over time. The demographics dynamics surely would have some effect. But then the question of mixing across rating in the events also driving the statistics, might become a factor to consider. The tiering.

I think some of those questino might use database of full poppulation like on lichess, where one can filter non tournam,ent tiered rated game events from the ones tiered..

that could serve as some kind of control group. The round the clock and more reliable random pairing mixing policies there might offer some conparison signal with the more tiered structures population mixing of OTB.

just ideas. I might have misconception and ignorance about OTB facts that my questions might tie into.

Expand full comment
1 more comment...

No posts