Petra;

here's the summary of the chi-square test info.

 AV. PROB.	O	E	(O-E)^2   (O-E)^2/E
0.968941364	86	83.328	7.13        0.08
0.848842105	36	32.256	14.02       0.43
0.759181818	28	25.053	8.68	    0.35
0.650666667	12	11.712	0.08	    0.01
0.553842105	12	10.523	2.18	    0.21
0.450434783	9	10.36	1.85	    0.18
0.3458	        15	8.645	40.39	    4.67
0.251043478	6	5.774	0.05	    0.01
0.146112903	16	9.059	48.18	    5.32
0.038680641	5	3.01709	3.93	    1.30
					
			X2 =	12.56	
			with 10 degrees of freedom

I evaluated all his predictions, took the probability for each and logged hit or miss.

Then I sorted the list into order by probability, grouped it by tenths, took averages of each group and multiplied the counts by that to get expected values.

Chi-square is straightforward after that.

If there's a better way, let me know.

Roger