Comparing Systems

[bairn1]

Hi there, I was just wondering what people's thoughts were on comparing betting systems. i.e. If you've got a few systems which produce different ROIs, different number of bets, etc. do you have an easy way to tell which is best? I know a lot is down to personal preference (e.g. how risk averse you are), but say you have three systems that show up the following results on back-testing. Which would you say is best and why? System 1 113 bets £479 profit to £10 level stakes ROI = 42% Avg Odds = 2.87 Strike Rate = 51% System 2 763 bets £1,506 profit to £10 level stakes ROI = 20% Avg Odds = 3.3 Strike Rate = 36% System 3 259 bets £730 profit to £10 level stakes ROI = 28% Avg Odds = 3.46 Strike Rate = 30% I'm guessing there's no particular best / worst out of these, but I'd be interested to hear a few opinions of which they prefer and why. Many thanks.

[winningbin]

Re: Comparing Systems That was also my question ! ;) I will choose System 2, why ? Because there are up to 763 bets and I believe that more bets will show more exact system. At System 1, just 113 bets, and I am not sure that it can bring the high Strike Rate in future. In System 2, I guess the amount of matches is enough (and better than others) to ensure a system run well ! :ok

[sodabrab]

Re: Comparing Systems Hi Bairn, System 2 definitely. Simply because you'll have more money in your bank account at the end of each year.

[ofcourseitwas]

Re: Comparing Systems easily System #2. That wouldn't happen to be your aways system would it? ;) System 1 does not have nearly enough picks to say it will be profitable in the future. Only 113 picks could be pure variance. System 3 has a fair amount of picks and would rank 2nd imo. However, that profit could drop off over the next 500 picks and end up with the same or worse yield that is already proven in System 2 (20%).

[Matthew]

Re: Comparing Systems Are these all over the same time period? What you have listed is nowhere near enough info to decide if a system is to be followed. Season by season breakdown - is it consistent? How sensitive is it? Odds breakdown, are you reliant on longer odds in lower leagues that are cut much quicker than shorter odds in higher leagues. If you randomly sampled your backtested data choosing 300 of the 763 matches are the stats you posted consistent. I'd repeat this 1,000 times to get a sense of how likely it might be to be a loser over a certain time period. All well and good if it makes profit over the 763 bets but I'd bet a lot of money you'd pack it in if you were down after 300 in the real world. With a strike rate that low losing runs are a concern. What is the distribution of losing runs? Winning runs? I'd look at all that first and if it passed, and is over the same time period, it is all about absolute profit so #2 for me. A high ROI won't lead you to bet more per selection I imagine so you need to choose the one that your bank size and personal feelings on risk can handle.

[bairn1]

Re: Comparing Systems Hi there, Thanks for your replies. These are helpful and basically confirm what I thought.

[piginabush]

Re: Comparing Systems the only important figure is the profit, number of bets/ROI etc etc dont really matter as long as the profit is in positive figures :clap

[Lardonio]

Re: Comparing Systems Information missing to make an informed choice (such as length of testing for each system) but if I had this selection of systems to choose from I would not choose, I would use all 3.

[Fearthainn]

Re: Comparing Systems Like others have mentioned you would need a longer period of testing, so on that basis alone system 2 would look like a good choice. Personally I pay more attention to ROI rather than the total profit. A higher ROI will make me feel more confident about a system, and if you want more profit just up your stakes. I prefer less selections too.

[AndrewKent]

Are these all over the same time period? What you have listed is nowhere near enough info to decide if a system is to be followed. Season by season breakdown - is it consistent? How sensitive is it? Odds breakdown, are you reliant on longer odds in lower leagues that are cut much quicker than shorter odds in higher leagues. If you randomly sampled your backtested data choosing 300 of the 763 matches are the stats you posted consistent. I'd repeat this 1,000 times to get a sense of how likely it might be to be a loser over a certain time period. All well and good if it makes profit over the 763 bets but I'd bet a lot of money you'd pack it in if you were down after 300 in the real world. With a strike rate that low losing runs are a concern. What is the distribution of losing runs? Winning runs? I'd look at all that first and if it passed, and is over the same time period, it is all about absolute profit so #2 for me. A high ROI won't lead you to bet more per selection I imagine so you need to choose the one that your bank size and personal feelings on risk can handle.

Totally agree with Matt is saying here, more information is needed here before you can make a choice like that. One thing I have learned in the last 9/10 months of learning to bet is that there is no quick and simple approach. It's a combination of small factors that make the difference when you put them all together. For example Matt mentioned above winning/losing runs. I would be inclined to ask the question about trend. Is #2 giving a consistent return or is this the ROI for the system in general. I suppose the analysis in many respects is like the bet itself. If you only look at a football match scoreline, that doesn't tell you who the better team was, you have to dig deeper. In analysis, I'd never go with the quick high level figures, I'd want to dig deeper. Happy hunting though, it's a lot of fun once you get into it!