Mar 26th – Apr 1st +$11,243.11

You may also like...

6 Responses

  1. Skg says:

    AFL was painful this week. I placed a much smaller bet on the St Kilda line (one of the few times I’ll go against Jonno’s picks), and it was painful seeing Hawthorn’s inept second half send me deep into the red and then see the Saints come back and win what should have been enough to cover Hawthorn’s losses.

  2. Steve says:

    The model does love St Kilda and their lines always go down to the wire, I remember some half point wins and losses on them last season. So far 2 from 2 this year

  3. Skg says:

    It really does, last year the model was slow to react to them but eventually yielded some big wins. I based my decision to not back St Kilda on how machine learning approaches usually work – there’s going to be a learning parameter the magnitude of which can sometimes make models slow to large scale changes in a team. GWS’s win over the Swans made me feel that the team had changed dramatically enough over the season gap that the model was probably going to underestimate them for at least the first few rounds, and we kind of saw that. GWS looked to be the better team until mid way through the third, when they started to run out of legs.

    It’ll certainly be interesting to see how both the bookmakers and the model pegs them over the early season. There’ll be plenty of value to be had if the model is able to get a more accurate picture of the teams evolution than the market, it’ll just depend on how sensitive the learning parameters are.

  4. Steve says:

    I had the same thoughts too, but the model rated GWS a big line bet over the swans which won easily, so if anything, its rating would have gone up massively after round 1, the fact that St Kilda was still a large bet was good enough for me.

    I also know that if I had to choose between my thoughts and those of the Sportpunter model, I know which way I would go ever time.

    I have noticed it before with the Sportpunter Tennis model, it took forever for Raonic to be rated correctly and now it is rating Nadal lower then he should be.

    These aren’t perfect models, but they are the best I have come across. Hopefully the bookmakers keep getting it wrong on GWS and all the other teams.

  5. BetPreviews Thomas says:

    What happens when you get to a 100K profit? Any plans to increase your stake size?

  6. Steve says:

    Nope, it’s just a little goal I have been chasing for the last 4 years. The current stakes are more then enough for the coming years. P/L range is from -200k to $500k.

Leave a Reply

Your email address will not be published.