Tuesday, 4 June 2013

2012/13 Season Review Part Six - Under/Over 2.5 Systems

I think the next 3 reviews are going to be a fair bit shorter than the first 5 reviews. I personally have never liked the Under/Over 2.5 goal systems as quite honestly, I think they don’t have an edge and any edge they may have had is probably eaten away by the fact I’m giving out the bets on a Thursday evening when the markets are sitting with a fairly chunky overround in the English lower leagues.

The last two reviews on the Euro systems are never going to be too meaningful since I’m effectively ripping these up and starting again next season. Obviously, it’s worth looking back to try to see what went wrong last season and hopefully it will help me when I built another couple of algorithms later this Summer but in terms of looking back and setting targets for the future, it can’t be done.

Right, this review centres on the Under/Over systems. These were developed two Summers ago and proofed for the first season in the SBC forum. All I did last Summer was remove all the odds on bets from the Under 2.5 systems as they were badly lossmaking and I kept the Over 2.5 systems as they were in the first season as they had an OK first season. I also decided that it was better to just have one Under system and one Over system rather than try to filter the bets. The systems don’t have enough bets as it is, so trying to filter them is a waste of time.

Season 2012/13 Performance

I think the results say it all here. A 1.6pts loss across both systems combined from 363 bets, with a small profit on Under system and a small loss on the Over system. Not sure what more I can add. It backs up what I’ve said all season that I don’t think these systems have an edge. I will persevere with them for another season to see if there is any improvement but in terms of following these systems to make money, I wouldn’t advise anyone try it!

All Live Results (last 2 seasons)


I was highly critical of the systems in their first season with the lowly ROI but after basically a break-even season last season, I’m starting to think the first season was better than I thought it was! A 3.5% from 836 bets since the systems went live but a system is only as good as it’s peers. These are the blacksheep of the TFA family as I’ve said before and therefore, why would anyone follow these systems when there are a whole host of better systems to follow?

If you look at the ROC figures for the live results, I think this highlights the problems that a low turnover, low ROI system suffer from. Over two seasons, they’ve only grown the bank by 34% and 42.3% respectively and an average ROC growth of 17% and 21.1% ROC hardly set my pulse racing when other systems are looking to double the bank in a single season.

Targets for 2013/14 Season

A target of 3% ROI for each system, couple with the size of betting bank needed and the low turnover suggests that we’re looking at a lowly ROC target of 11% and 14%. Given the poor performance last season, I’m not even sure that’s a target that is achievable and beatable, so I think these systems remain trial systems again next season.

Detailed Analysis

Performance by Season


As I highlighted above, you can see how much better the first season was than last season and the first season wasn’t great either.

Performance by League


With the very small sample sizes, you can see what can happen. Two pockets of data that look amazingly profitable and the rest looks rubbish! If League One and League Two were behaving the same, you might think we were onto something but it is the Over system in League One and the Under system in League Two doing well!

Performance by League and Season


Impossible to draw any conclusions due to the sample sizes but nothing really stands out too much.

Performance by Odds


I thought I would look at this to see if I could see anything that stands out. Apart from the fact that the Over bets with higher odds are really poor performing, nothing jumps out too much. Again, sample sizes are small and it’s too risky to try to draw any conclusions.

No comments:

Post a Comment