I cannot say I am satisfied with the results from this research. I definitely expected more from Gauss less square approximate method. I see that there is profit, but I expected more. Even so, 17% pure profit from 31 matches with betting on under over in basketball is not bad.
Here are the stats.
| % L + | %i L + | G L + | Gm L + | Avg. L + | Avg.m L + |
| 61.3 | 54.8 | 41.9 | 58 | 45.1 | 38.7 |
If we start talking in profit percentage, best results shows my old method of betting, where you betting under or over depends from all previous matches for both teams. The profit there is 17%. However, as I learned from previous NBA regular season this method is going to be less profitable at the end of the season, because bookies start to level under