This service has now fully recovered from its bad start and is now in profit to the tune of 4.15 points.
We have been recording results for around 6 weeks and have had 45 selections over that period.
That’s too small a sample to make full judgement on when I still think the method behind it is a bit random. As I said at the start of the trial, it’s hard to think of real valid reason why a team should be less likely to draw a game just because they drew their last regardless of venue, opposition, overall form etc.
However, I have been looking over what results we have recorded and there’s a very obvious “sweet spot” type pattern emerging.
The official figures (just laying each selection for the price obtainable mid morning of the game day) shows a profit of 4 points and a strike rate of around 78%. A look down the list of selections shows a couple of prices that I’d be extremely worried about laying, there’s a good few over 6.00, with one bet at a whopping 11.00.
If we just apply a sensible odds filter to knock out those that are too low (and therefore more likely to lose) and those that are too high (and therefore potentially very damaging) we can massively improve those figures.
If we use 3.75 as our minimum acceptable price and 6.00 as our maximum we end up with 25 selections, 23 of which won. Meaning a strike rate of 92%, a profit of 16 points and ROI of 64%. The average odds figure barely changes from the full method.
I won’t put up with back-fitted cherry picking, so if the “sweet spot” was within too random a price range I would disregard it. But 3.75 & 6.00 are perfectly sensible cut off points, tucked just above 5/2, which is dangerously low for a draw, and maxing out at 5/1, above which gets scary when you’re working with such simple/flimsy selection criteria.
Anyway, all the above is just conjecture at the moment. Here are the actual on-going trial figures:
AV ODDS 4.48