AndrewKent Posted January 8, 2011 Share Posted January 8, 2011 Hi folks, I think this is turning into a bit of an obsession, this system malarky! I haven't put a bet on in weeks, having dedicated so much time to learning how statistical betting works! And as masochistic as this is, I've actually enjoyed the headache! So, for my next thread (and I do apologise if I am flooding the forum here but I'm enjoying the chat!) I thought I would take a few steps back to go a step forward. I keep thinking that the systems I have looked at are giving good results, but in truth I have no idea if they are or not as I don't have a bench-mark to compare it against. I therefore thought I would prepapre some analysis that gives me an idea of what % of success I would have from the following: Back Home Back Draw Back Away Lay Home Lay Draw Lay Away If I can get a % of success from each of these measures (across several seasons and divisions) it should give me an idea of how good any future system is. For example, if the bench-mark for backing the home team is 50%, and I develop a system that is 45%, then I know I would have a better chance of success by simply betting blind. Now I know that there are many other factors to take into consideration such as value odds and so forth but I thought this might be a good place to start. What do you all think? Andy Quote Link to comment Share on other sites More sharing options...
Faxy Fowler Posted January 8, 2011 Share Posted January 8, 2011 Re: Bench-Marking I can only offer something that virtually states the obvious. In most leagues the ratios seem to be around 47% home win, 28% draw, 25% away win. The draw percentage seems to vary a bit based on overall number of goals scored - more goals, less draws. We could all just take United, Chelsea and Arsenal games in order to get more than 47% correct, but, to reiterate something from another of your threads, the odds are so useless that you are almost bound not to profit. Quote Link to comment Share on other sites More sharing options...
AndrewKent Posted January 8, 2011 Author Share Posted January 8, 2011 Re: Bench-Marking As far as I'm concerned, nothing is stating the obvious when it comes to me ;) Seriously though, now that I'm learning all these little facts it's building my knowledge base. Until you said its roughly a 47/28/25 split, I had no idea. I figured if I could benchmark each available division in my data source over the last few years it might give me a starting point to work from. Andy Quote Link to comment Share on other sites More sharing options...
Faxy Fowler Posted January 8, 2011 Share Posted January 8, 2011 Re: Bench-Marking I would have thought this is better in the punters tools and general help forum next door. You aren't really demonstrating a system as such, but asking for ideas. Quote Link to comment Share on other sites More sharing options...
Syndicate Posted January 8, 2011 Share Posted January 8, 2011 Re: Bench-Marking Here you have some stats on who percentage of wins divided on match location and league: http://www.soccerstats.com/ Quote Link to comment Share on other sites More sharing options...
sjt91 Posted January 8, 2011 Share Posted January 8, 2011 Re: Bench-Marking As far as I'm concerned, nothing is stating the obvious when it comes to me ;) Seriously though, now that I'm learning all these little facts it's building my knowledge base. Until you said its roughly a 47/28/25 split, I had no idea. I figured if I could benchmark each available division in my data source over the last few years it might give me a starting point to work from. Andy To find some overall stats for previous years (and other divisions), try here. The season summary section on this site provides the data. http://www.statto.com/football/stats/england/premier-league/2010-2011/season-summary It will allow you to see any nuances with different leagues and trends over time - to see if that split still holds. Quote Link to comment Share on other sites More sharing options...
Datapunter Posted January 9, 2011 Share Posted January 9, 2011 Re: Bench-Marking Fairly pointless exercise. Let's say the average win % for home wins is 40% and you manage 50%, it doesn't actually tell you anything apart from a very rough picture. In order to make a profit you still need the average odds of all your winning bets to be above 2.00 Let's say the win % on away teams is 25% and you only get to 20% , it would lead you to conclude you're not on the right track. Hang on a minute, if the average odds you get on your winners are above 4.00 you'll still make a profit. Use the websites already mentioned to get a rough idea, but don't draw any conclusions from the % alone. Quote Link to comment Share on other sites More sharing options...
Datapunter Posted January 9, 2011 Share Posted January 9, 2011 Re: Bench-Marking In fact as "benchmark" idea i'd drop it completely, it has no meaning, only profit matters. Quote Link to comment Share on other sites More sharing options...
AndrewKent Posted January 9, 2011 Author Share Posted January 9, 2011 Re: Bench-Marking @ Datapunter, good point well taken. Thankfully I didn't waste any longer than an hour knocking this up and it didn't tell me anything the websites didn't already. I know this is going to sound stupid but for my learning curve here; I am seeing that EVERY system I am looking at now, I'm being drawn down one route: 1) Run analysis 2) Define my own "value" odds from analysis 3) add "value" odds to analysis criteria I'm guessing that a system is not so much looking for % of success but to be able to define what's value and what's not? for example; If my system was 40% accurate, my value odds would have to be above 2.50 in order to make a profit? Andy Quote Link to comment Share on other sites More sharing options...
sjt91 Posted January 9, 2011 Share Posted January 9, 2011 Re: Bench-Marking @ Datapunter, good point well taken. Thankfully I didn't waste any longer than an hour knocking this up and it didn't tell me anything the websites didn't already. I know this is going to sound stupid but for my learning curve here; I am seeing that EVERY system I am looking at now, I'm being drawn down one route: 1) Run analysis 2) Define my own "value" odds from analysis 3) add "value" odds to analysis criteria I'm guessing that a system is not so much looking for % of success but to be able to define what's value and what's not? for example; If my system was 40% accurate, my value odds would have to be above 2.50 in order to make a profit? Andy The aim of value betting is to try to devise a system which you think can give you true odds, or at least an approximation of them. Where the bookies odds are better than the true odds, you bet. In the long run, you'll win if your system approximates/finds true odds better than the bookies, essentially. If you think the true odds for a team winning a certain match are 2 (50% chance), and the bookies odds are greater than 2, there's value there. Likewise if you think the true odds are 7 (14.23% chance), then odds of 3 are no good to you. It's not so much that you devise a system that's 40% successful and only bet on outside 2.5s, because those ones outside 2.5 are likely to be the ones that are less likely to win, if you see? You'd have to try to have a system that can determine true odds between teams - and building that is the tricky part. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.