ELO progression by year, 2006-2021

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

You are doing excellent work!

Here is the average rating gain over the last 15 years for every 5 years. And the results are telling.

Average Yearly Gains for 5 years.

2007 to 2011 - 41.2 Elo
2012 to 2016 - 34.8 Elo
2017 to 2021 - 41.0 Elo

And I argue this point all the time on my channel. As people think Chess engine progress has slowed down over the years. As chess engines are playing near perfect chess.

My testing, and the stats say otherwise.

mwyoung wrote:

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

You are doing excellent work!

Here is the average rating gain over the last 15 years for every 5 years. And the results are telling.

Average Yearly Gains for 5 years.

2007 to 2011 - 41.2 Elo
2012 to 2016 - 34.8 Elo
2017 to 2021 - 41.0 Elo

And I argue this point all the time on my channel. As people think Chess engine progress has slowed down over the years. As chess engines are playing near perfect chess.

My testing, and the stats say otherwise.

The testing only showed that in the past they did not play near perfect chess.
I see no tests that prove that today engines do not play near perfect chess.

I believe that in the past the best engine could beat itself often with time handicap of 1:10 and today it is not the case.

I do not know if future engines are going to beat the engines of today at long time control.

I believe that it is possible but I saw no proof for it.
I believe that most draws are with no losing mistakes but the engines do not play in the right way to cause the opponent to make losing mistakes.

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

Excellent!

Why are there two columns?

TheSelfImprover wrote:

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

Excellent!

Why are there two columns?

SP - Single Processor
MP - Multi Processor [ 4 cores ]

mwyoung wrote:

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

You are doing excellent work!

Here is the average rating gain over the last 15 years for every 5 years. And the results are telling.

Average Yearly Gains for 5 years.

2007 to 2011 - 41.2 Elo
2012 to 2016 - 34.8 Elo
2017 to 2021 - 41.0 Elo

And I argue this point all the time on my channel. As people think Chess engine progress has slowed down over the years. As chess engines are playing near perfect chess.

My testing, and the stats say otherwise.

You will also like (I suppose) the current extensive study I am doing measuring draw rates and diminishing returns.



You see the elo dropping and the draw rates increasing every doubling of time control. Especially the numbers of SF14 and Dragon 2.5 imply that progress becomes harder and harder for search and that NN or NNUE becomes more and more important.

Admin wrote:

TheSelfImprover wrote:

Admin wrote:

ELO progression by year, 2006-2021

http://rebel13.nl/misc/stats.html

Excellent!

Why are there two columns?

SP - Single Processor
MP - Multi Processor [ 4 cores ]

Thanks for clearing that up.