The Obbvio Prediction Score (OPS) formula is:
Where:
Accuracy = C / T (ranges from 0 to 100%)
- C = Number of correct predictions
- T = Total number of predictions
Prediction Difficulty PD = 1 - O (ranges from 0 to 100%)
where O = Average odds (probability of winning) of correct predictions
Here are some example values of PD for different Odds values
Difficulty (PD)
Odds (O)
70%
30%
50%
50%
30%
70%
- Statistical Significance (ranges from 0 to 100%)
- T = Total number of predictions
- k=100, ensuring that S(T) increases gradually and prevents players with very few predictions from having inflated rankings.
Here are some example values of S(T) for different T values:
T (Total Predictions)
S(T) (Statistical Significance)
1
1.0%
5
4.9%
10
9.5%
20
18.1%
30
25.9%
50
39.3%
70
50.3%
100
63.2%
150
77.7%
200
86.5%
300
95.0%
500
99.3%
Here are some example values of Obbvio Prediction Score (OPS):
Total Predictions (T)
Correct Predictions (C)
Avg Odds (O)
Accuracy (A)
Difficulty (PD)
Significance (S[T])
Score (OPS)
10
6
75%
60%
25%
18.1%
25.4
25
14
65%
56%
35%
39.3%
35.1
50
28
55%
56%
45%
63.2%
46.1
100
55
50%
55%
50%
86.5%
53.2
200
100
45%
50%
55%
98.2%
57.4
500
240
40%
48%
60%
99.3%
60.3
1000
470
35%
47%
65%
100%
63.2
Check WHY OPS is a better ranking method than just counting correct predictions