Sports betting - staking plan ideas

[crazybadger]

I've been working on a few sports betting ideas and this current one I've got is about picking correct scores in games. It picks a lot of long shots and because of this it has a very low strike rate but after just a random sample of just over 50 games it shows a very good profit at flat stakes.

This got me to thinking about how to maximise profit from this system. So does anyone have any good ideas for staking plans for sports betting? In particular for low strike rates...the idea of having a % of the bank (like the ones recommended for horse racing systems) seems flawed as there could be 10+ matches at the same day and with the very low strike rate they could all quite possibly lose, meaning the bet next time is significantly reduced.

I haven't started looking at options yet but just curious if anyone has ideas. The stats for the system are:
53 matches
4 wins (7.55%)
$10 Flat stake = +$2472
ROI = 466.43%

The odds of the 4 winners have been... 60, 36, 9.8 and 210. So the point I guess is that if you erase that last one then the profit is much less, so I'm looking for a staking plan that hangs in there for the long haul (which could be very long) but has good bets to take advantage of those prices.

Personally I dont think this is feasible and its just been lucky but I guess we'll see :D

[TheOldhamWhisper]

Whilst %age staking is a good all-round strategy, as you have worked out for yourself, it might not be such a good idea in this case.

Given the very low Strike Rate and the number of losing bets, I think I would start off with a bank that I was not too unhappy to lose and play 100 bets at level stakes. You can (if you wish) add extra staking rules e.g. if the bank trebles, start using a percentage plan instead.

I doubt there's a 'perfect' staking plan for this one (unless there is a distinct pattern in the results?) and as such, simple is best.

[MattR]

CB, is this using the poisson stats? I tried a similar thing last year using the correct scores and like you couldn't decide what was the best way to go about it. I tried using just 1 correct score and also dutching the 3 most likely (lowest pred odds) for each game and a fair few other permutations in between! I had some degree of success for some reason in the Brazilian leagues. One reason I think it did fairly well is that there are not many away wins and I was taking the lowest predicted odds which invariably were home win scores.

The other thing I tried was doing ht correct scores by just using the goals for/against from the first half of teams stats. Again a staking system was a problem to find. I thought about splitting it into 'mini' systems, ie home wins, draws and aways in their own systems. I had a 100/1 2-2 come up which was nice, although the stake was only £2 and I layed a bit off as it became 2-2 at around 37 mins.

There is definitely potential in scores as they have by far the highest odds you can work with in football.

So to get back to your staking question, perhaps level stakes split into home/draw/away score systems?

Another option may be to go for a win target and so adjust the stake depending on the odds.

[Street cry]

You need a big points bank and need to use level stakes. You can increase the bank periodically as oppose to when it grows. I would start witrh abnk of 500 points if this was £5000 lets say then you would obviously be at £10 a point to start out.
Treble the bank and double the points so once you get to 15k increase to £20 a point so you have added safety in to the bank. That would be my approach anyway

[crazybadger]

Level stakes wins! Thanks guys, you confirmed what I had been thinking. I'm still collecting stats for this system so I will see how it goes. If I decide to start betting it I think it will be level stakes and will probably do some statistical anlysis of best starting bank size (in terms of number of units).

Street Cry I like that idea of only double points when bank trebles to add the safety into it. Very nice.

Mattr, it is a poisson based concept but I've been fiddling with it a fair bit. This idea is one of many I've derived from that but is the one with the biggest return so far. I'm only betting 1 option per correct score market and it generally happens to be the big ones.
One idea you might like to try is using your predictions (i.e most likely) from the HT and/or FT markets to pick which scores to back. E.g If the prediced winner at FT is Liverpool then you could try dutching some of their FT scores? I tried it for a bit and while it made profit it seemed pretty crazy (placing up to 7 bets in a market per match!).

Cheers guys.

[MattR]

Good luck with it CB, will be interesting to see how it goes if you post it up here. I'm still testing lots of different ideas on those ht/ft and correct scores, trying to settle on one or two eventually! I'm going to post something up here today that I'm testing from today based on match odds using the goals for/aga and poisson.