Rugby Champs week 5, Currie Cup round 8, Pro 14 + other rugby
- Brent Graham
- Topic Author
- Offline
- Posts: 4761
- Thank you received: 1046
Chat with me on Twitter - @BrentGraham
Please Log in or Create an account to join the conversation.
- Kirky
- Offline
- Posts: 396
- Thank you received: 98
I’m also a newbie to the forum and not good enough at beating the bookies (grrr) for the confidence to contribute. But I do keep a stat model on all the leagues I participate in on Superbru, and I have about a 75% success rate there, indicating that I‘m doing something right. I also do appreciate your courage in asking for advice.
My method is:
Using a spreadsheet, I rate the teams on their success week by week on a 20 scale (20 = good, and the bad teams can be as low as 15). I keep it as simple as I can and when making my predictions I’ll choose the winner based on the rating, then look back over each teams performance ie. log position; home or away; size of their scores, the rating of the team they won or lost to on the day, etc. any factor that can mitigate. Ideally where possible, I’ll consider key player exclusions, injuries, etc. Also once you get a feel for the squads, you gain understanding of their depth – like the All Blacks have a bottomless pit of top players to call upon.
Also, if a 17 rating beats a 15 rating by 20 points (as expected), then they’ll remain at their ratings. But if a 17 beats an 18, then I’ll re-look at their past games to figure out why I got it wrong, or if they simply had a blinder, etc. But the likelihood is that the 17 will be up-rated to 18, and that the loser could be down-rated.
My Super Rugby spreadsheet (running for about 7 years now) is a bit more sophisticated, but recently (Premiership, Pro 14, Mitre 10) I’ve simplified it as above, and it’s working pretty well in terms of picking the winners.
There are too many variables: who the ref is; on-field injuries; foul play; intercepts; etc. for any more statistical accuracy (IMHO) with rugby, hence my reason for simplifying things, although a true statistician would include these variable into their calcs..
If you’re interested, wattsapp me your email address and I’ll mail the spreadsheet to you to do what you want with. Mobile: +27 (0) 82 929-2118.
In my experience, statistics can be as sophisticated (usually is) as rocket-science, but my preference is always simplicity, which works well for me.
Careful with the punting, there ain’t any poor bookies.
Good luck, cheers,
Mike (Kirky).
Please Log in or Create an account to join the conversation.
- SirCharlesZA
- Offline
- Posts: 18
- Thank you received: 11
Please Log in or Create an account to join the conversation.
- gareth_
- Offline
- Posts: 9
- Thank you received: 2
Please Log in or Create an account to join the conversation.
- gareth_
- Offline
- Posts: 9
- Thank you received: 2
1 underdog definitely doesn't seem right at the moment. See how it goes I guess
Please Log in or Create an account to join the conversation.
- Kirky
- Offline
- Posts: 396
- Thank you received: 98
Cheers,
Kirky
Please Log in or Create an account to join the conversation.
- gareth_
- Offline
- Posts: 9
- Thank you received: 2
Please Log in or Create an account to join the conversation.
- gareth_
- Offline
- Posts: 9
- Thank you received: 2
I still haven't received the spreadsheet. Would it be okay if you could try again please?
If I still don't receive it I can give you a Gmail address to send it to. I'm guessing you got my txt okay?
Please Log in or Create an account to join the conversation.
- Mgqwesi
- Offline
- Posts: 35
- Thank you received: 27
Mgqwesi wrote: So no New Zealand until the final (if they make it that far) next year at the World Cup. Springboks to triumph at 37/4 seems like good value to me at this rate of progress.
Johan Erasmus, Siyamthanda Kolisi and the rest of the team, thank you and thank you once again, tonight we eat nice. Not all of it though, I need keep some to contribute to Rassie's statue.
Please Log in or Create an account to join the conversation.
- Bookie Basher
- Offline
- Posts: 290
- Thank you received: 89
Please Log in or Create an account to join the conversation.