Money in chess

[Jonathan Edwards]

Albeit with a small sample size (n=50), and probably missing plenty of variables, the statistically significant variables are the total prize fund (p<0.01) and the number of titled players (p<0.02). Either players entering opens care about the number of titled players; or the number of titled players is acting as a proxy for another factor.

Or your sample size is misleading small.
Disraeli would approve.

Perhaps. Although I mentioned the small sample size... t-statistics are meant for exactly this - small sample sizes! In fact, when d.f.>30 (here it is 49), it converges to standard normal distribution.

Anyway, surely it's quite intuitive that titled players have a positive effect on the number of entries in an open.

Hi Jonathan. Interesting idea. Even if you had more data, how could you draw any conclusion regarding cause and effect though? Could be the other way around - presumably the titled players would care about the number of entries, prize money etc. Most likely there is an interdependence (with most people acting on their own experience of how big the tournament usually is) making them impossible to separate just using stats.

Hi Walter. That's the reason for using regression analysis - determines causation and not correlation. Of course, more data and variables would increase the power of the test. It would be possible to include lagged variables from the previous year to account for how big the tournament is; or even use panel data, but it would take such a long time to find data for the past 10 years or so.

What did you get for the possible effect of the grading prizes? Although drawing a conclusion on this would be open to the same objection in that a larger entry with more titled players would create more of a need for the grading prizes.

The effect of grading prizes was slightly positive (but statistically insignificant). Interesting point that there may be simultaneous causality. I'd love to be more rigorous if it was worth the time. It was more just to demonstrate that encouraging titled players to enter congresses might be a good thing.