[hutrdvnj]

If we assume that the data is reliable (not hand picked just to make a case). Then let's check.

Two sided un paired t-test.

T-value of 4.9. SE of 56.325

Group 1 mean of 2383 with sd 125.56. Group 2 mean of 2659 with sd 119.16 39.7 SEM.

P value of 0.0001.

By conventional criteria there's about 0.01% chance the differences in the means is due to chance alone. That's 0.0001.

I.e. If you were to play 19 tournaments randomly 10,000 times, only 1 in 10,000 would you perform better on live boards as well as Hans did due to chance alone.