Soccer totals analysis

Since going live on the 14/09/09 we have made 2482 bets on totals for a very small 0.2% ROI profit – basically break even. This therefore is a great reason to have a look at where the sportpunter totals model has gone good and where is has gone bad. The analysis of these bets is shown below:

From the first table it shows that overs and unders have been equally profitable or maybe equally level. There doesn’t seem to be any major pattern with reference to better odds to bet. Perhaps a case can be made for betting with larger odds ( > 2.15), however the small sample size and the face that betting at odds of 2.10 to 2.15 negates this argument.

Having a greater minimum overlay definetly helps here. With a minimum 17.5% overlay, 500 bets were made for a profit of 6.4% ROI, and with a minimum 20% overlay 334 bets were made for a 5.9% ROI profit.

Interestingly, and maybe quite expected, most of these profits were made betting unders than overs. It has been the case in many sports, especially for example, baseball or ice hockey, that overs are bet more than unders and hence a bias occurs in the odds. It will be interesting, and we will look at it, to see if such a bias occurs here.

Overs with a minimum 17.5% and 20% overlays resulted in a -1.9% ROI and -1.6% ROI respectively, whilst Unders with the same overlays resulted in a 11.2% a and 10% ROI respectively.

Keeping in mind that the above unders bets were only 313 bets for 17.5%,  and 212 bets for 20% minimum overlays. Hence about 10% of the original number of bets.

Whilst restricting yourself to only 10% of the suggested bets might mean a leaner betting season, it could mean greater profits over the long term. The small sample size is something also to keep into account.

I wouldn’t completely change your betting strategy just yet. An analysis of the potential bias, and perhaps the bias in the greater or smaller leagues, would be well worth looking at first, which we will do.

Also interesting is the lines. Overs are profitable at 2.25 line’s and below. (505 bets at 4.5% ROI), whilst unders are profitable at the line of 2.75 and above (541 bets at 3.9% ROI).

Clearly to me this shows the market overcompensating for recent matches. I’ve seen this happen a lot in other under/over markets like baskeball and ice hockey, and it seems to be happening here. It could be just one high or low scoring match that makes the punters get on board, and the odds are moved to much in one direction for just the one match. This over compensation, gives us a good edge, and we will look at the bias in this in more detail in the future as well.

In conclusion, it would be great if overall the totals model recorded slightly better results, but perhaps restricting yourself to higher overlay under bets, as well as perhaps over bets at smaller lines, then one can obtain a greater profit.

To sign up to the soccer model, please click here.