Super 15 predictions for week 2(86% last week)

[Raystef5845]

The first week or two can be a bit ragged since I start with last year's ratings and the Rebels are an unknown. Given that, 6 from 7 predicited winners did win (86%) which is a very good start. The system dropped the Rebels rating quite a bit after their week 1 disaster. The computer will adjust upward if they get competitive.

I added more info at the end of the predictions to explain the betting rules that I devloped over 16 seasons. They seem to work, given a 14% profit on turnover.

Good luck with week 2, remembering to bet cautiously for the first 4 weeks.

Ray Stefani

[rlf]

Cheers Ray, I find these useful along with my own information.

Do you have any statistics on the total game points? I noticed every game went under last week.

[Raystef5845]

I went back the last two seasons and about half the total scores were over my prediction and half were under. I'm not sure anything is significant this year other than random chance, I know from past work that over/under on total score is harder to predict based on past scores than score difference. Teams play knowing what the score difference is and they don't care what the total score is. Another issue is a strategy to be more defensive if both teams are high scorers to not let the game get away and to play more agressiviely if both teams are low scorers figuring they don't have much to lose. Tehy resulst bis that its hard to get a finger on over/under totals

Let's see what happens this week.


Speaking of ths week, I posted last weeks results instead of this week's predictions so I have to fix that right now.

Cheers,

Ray

[Raystef5845]

I uploaded a summary of last week's results instead of this weeks predictions. Use this one

Ray