Caruana - Who'd have guessed?

Sep 4, 2014, 3:00 AM |

The chess world is watching in awe as Fabiano "the fabulous" Caruana is crushing in the Sinquefield cup. With a perfect score after seven rounds, his achievement is already historical. Who would have thought this would happen. Well, as it turns out, yours truly predicted something like this (albeit in vague terms) almost two years ago. Based on expert knowledge of chess? No. Based on careful analyses of Caruana's games? No. Based on statistics!

In this post, I wrote the following.

I messed around with some FIDE rating data for the (currently) top 10 GMs. I found that the rating progress can be approximated with a logarithmic function based on the number of rating periods after a player breaks the 2200 limit. I also found that the natural variation of the rating is about +/-120 points from the base line given by the function

  R(t) = 2200 + 150*ln(t)

In the image below, we see how the younger players in the top 10 develop compared to the "average".

As we see, Aronian and Radjabov do not really fit the bell. What is interesting is the fast development by Carlsen and Caruana. Based on this, Caruana is a guy to expect more from in the future.

I haven't checked to see if the model is still valid, but I think it's kinda cool to see that my prediction holds, although it is only based on the rating history of the players.