Whopper's Over/Under system - thoughts please

[Guest Whopper73]

I have been looking into over/under 2.5 goals as this gives only 2 results. Much more appealing than the win/draw/lose set of results as draw is pretty random (and thus essentially annoying and pointless!). From this seasons data on 2133 European football matches I have a system that picks out matches using some set criteria to determine a result. From this result the system will pick out matchees to bet on. In this case it has picked 750 matches to bet on, with a prediction rating of 52.33%. Will give you value if you can get odds of 1.92 or above (possible). However if I adjust the measurement a little and bet on the prediction is below 2.5 for under but abd 3.5 for over then I get accuracy of 57.2% (197 correct from 343 predictions). This will give value of odds greater than 1.75. Has anyone else noticed the accuracy drop over when predicitng over2.5 goals, or is this a peculiarity of my system? I have noticed that most odds for this range from 1.8 to 2.1 (give or take a little) with a few going as low as 1.7. It would seem that 57.2% would be good enough for value with disciplined betting! I will of coarse be back testing this over this summer. I just had some time at work today to make a start on it! Any thoughts are welcome.

[Guest OddsAgainst]

Re: Whopper's Over/Under system - thoughts please For me, any system has to produce fair odds for the outcomes. Otherwise you will have no way of knowing if each individual bet represents value. I also suspect, but cannot verify, that layers make subjective adjustments to the odds for individual games and that may make a difference. Odds can and do go as low 1.5 for the over, (even lower in extremes). You really need the odds for each game for back-testing.