robkor Posted February 20, 2008 Share Posted February 20, 2008 Quote Link to comment Share on other sites More sharing options...
happygooner Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions :welcome to pl Looks interesting. Good luck with it:ok Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Re: Power Rating Soccer Predictions Thank you goon, specially for the great contribution to this forum. Quote Link to comment Share on other sites More sharing options...
happygooner Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions Thank you goon' date=' specially for the great contribution to this forum.[/quote'] I'm not sure I provide much of a worthwhile contribution at the moment but thanks anyway:lol Quote Link to comment Share on other sites More sharing options...
lunatism Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions hey robkor! :welcome to PL!!!! nice to have another system shared on the forum, i wish you the best of luck with this! i presume you are using the data from football-data for your past analysis? or is there another source? Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Re: Power Rating Soccer Predictions Yes lunatism, you are right, it's football-data. Quote Link to comment Share on other sites More sharing options...
lunatism Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions since shots' date= corners,... are hard to gather, I’m forced to stick to English Leagues only i have had exactly the same problems in the past when looking for such data, and unless Goon does take me up on that offer to travel the world jotting down those particular stats, it seems it will never materialise! But as ever,we can hope (GO ON GOON...do it for PL)! :hope how many selections does your past data average per weekend? what was the start date and end data of you past data? i know you said that you'd post some selections later, i was wondering what kind of information you'd tag to that (like win percentage that your system calculated for that selection, etc)...or would it just be the selections? Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Re: Power Rating Soccer Predictions lunatism, I bet approximately 76% of all bets that is around 35 selections per week. I tested the data from season 2000/01 until today. Currently I'm working on selections. They will look something like this: Birmingham-Arsenal:selection 2-stake 0.78Units 0.78 means that System1 indicated 0.25Units, System2 0.33Units, System3 0.20Units. If you wish I will post the probabilities for all the 3 systems. The given selection Birmingham-Arsenal is for example purpose only. Quote Link to comment Share on other sites More sharing options...
lunatism Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions i'll be keeping an eye ;) on this one, looks interesting ....just wonder - would you be able to provide stats (and graph) for each league in addition to the overall one you have produced? what are the average odds you get for a selection aswell? (and the average odds per league?) :) don't let the questions bother you - i'm just interested :ok Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Re: Power Rating Soccer Predictions Here are the selections for this weekend. I guess I'll need a good start here if I want to keep the bank safe for the next weekend's bets:hope. One stupid question: How do you guys edit the tables? Is it possible to edit borders, backfills, etc... Your tables, Lunatism, look pretty cool! I dont know how to do it, so I post images:unsure. Quote Link to comment Share on other sites More sharing options...
lunatism Posted February 20, 2008 Share Posted February 20, 2008 Re: Power Rating Soccer Predictions One stupid question: How do you guys edit the tables? Is it possible to edit borders, backfills, etc... Your tables, Lunatism, look pretty cool! I dont know how to do it, so I post images:unsure. well..for starters, whenever i try and do it in firefox, i get screwed up lines of code all over the place. so i'll use firefox for writing messages, but when putting up tables i ALWAYS use IE explorer. Secondly, i just do all the fiddling (beautifying if you will) in excel, and then it's simply copy/paste, but as i said, only seems to work in internet explorer, and i couldn't be bothered to do screenshots for each table and then post to photobucket and then put links on here.....i'm lazy :) after i copy the table into the "post box", i just make the font smaller and then click the button...simple:ok Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Quote Link to comment Share on other sites More sharing options...
robkor Posted February 20, 2008 Author Share Posted February 20, 2008 Re: Power Rating Soccer Predictions Thank you lunatism, for explaining your table editing. I'm using Firefox so I guess there's the problem. Quote Link to comment Share on other sites More sharing options...
lunatism Posted February 21, 2008 Share Posted February 21, 2008 Re: Power Rating Soccer Predictions no problem...hopefully it's something they will fix in the next release - i am aware of an IE plugin that's available for firefox, but i've never tested it with posting tables, so i couldn't tell you if it works or not, but it is an option i guess (?) Quote Link to comment Share on other sites More sharing options...
robkor Posted February 21, 2008 Author Share Posted February 21, 2008 Re: Power Rating Soccer Predictions I've just installed the plugin for IE and it seems to work!:clap Today I was thinking about my staking strategy and realized that maybe I should avoid bets with stakes under 1.00Units. If you look at the chart on the 1st post you can see that >1.00Unit stakes are the most profitable, meaning that all bets between 0.0 and 1.0Units are loosing bets, although the loss is not big. I was thinking of calculating the optimal cutoff for the minimal stake for each league and for each outcome (1X2). But these cutoffs would slightly change from bet to bet as the data sample increases. I don't know if this strategy seems too complicated for you guys, or maybe limiting minimal stakes to 1.00unit is good enough. What do you think? Quote Link to comment Share on other sites More sharing options...
robkor Posted February 22, 2008 Author Share Posted February 22, 2008 Re: Power Rating Soccer Predictions I've just calculated the optimal cutoffs that maximize profit. From now on I will post all selections the same way I did, but mark differently those with more value. I will only update this values if I notice a larger discrepancy. OPTIMAL CUTOFFS FOR min.STAKE LIMIT 1 X 2 Premier League0.9 1.6 1.4 League 11.1 1.6 1.1 League 21.3 1.4 0.9 Championship1.4 1.6 1.9 Here are some charts for the coparison. Quote Link to comment Share on other sites More sharing options...
Imp Posted February 22, 2008 Share Posted February 22, 2008 Re: Power Rating Soccer Predictions Robkor....congratulations on an excellent post. Is it possible for you to contact me direct via e.mail? [email protected] Quote Link to comment Share on other sites More sharing options...
robkor Posted February 22, 2008 Author Share Posted February 22, 2008 Re: Power Rating Soccer Predictions No problem Imp:ok. Quote Link to comment Share on other sites More sharing options...
robkor Posted February 29, 2008 Author Share Posted February 29, 2008 Re: Power Rating Soccer Predictions The selections for this weekend... Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 1, 2008 Share Posted March 1, 2008 Re: Power Rating Soccer Predictions Great stuff! Very interesting. I have a question for you robkor, You say you are using 3 systems, did you use multiple regression analysis?? Quote Link to comment Share on other sites More sharing options...
robkor Posted March 1, 2008 Author Share Posted March 1, 2008 Re: Power Rating Soccer Predictions Hi noolnool, and thanks for your interest. No, I don't use multiple regression analysis. I just regress rating-differences to get 1X2 probabilities (or goal expectations in the case of System3). I get very nice curves with R-square=0.99 for 1&2 and 0.95 for X. As you can notice, the model isn't doing very well at the moment, but I'm not worried about it, because it has happened many times before. The highest loss that the model experienced from any point in the past was about -90Units, and losses of -10Units/week can happen very often, I'm afraid. If you take a look at the graphs from my previous posts, you can notice, that sometimes about 1000 bets are needed to come back into profit. At the moment I'm testing it on the random-splitted samples (one for learning and other for backtesting), and after more than 200 iterations it's doing very very well. It's showing profits in almost all odds-ranges and for the majority of the teams. You were probably interested in how the parameters (goals, shots, etc...) were combined, but for the moment I'll keep it a secret. A good estimation of team's performances it essential for the longterm profitability, and I believe that my ratings are accurate enough to beat the bookie's overround. I hope only that I don't empty the bank before I prove anything. Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 2, 2008 Share Posted March 2, 2008 Re: Power Rating Soccer Predictions Thanks for your reply! I fully understand you keep the finer details to yourself. Basically, I wanted to know how you implement the 3 systems as I probably don't understand it completely. Forgive me if I come accross a little bit stupid here, but if system 1 produces a match rating then you can calculate the probabilities for that match. But where then, come system 2 and 3 into play? Quote Link to comment Share on other sites More sharing options...
robkor Posted March 2, 2008 Author Share Posted March 2, 2008 Re: Power Rating Soccer Predictions Noolnool, I simply run the 3 systems contemporary, and the stake depends on the value that each of the systems finds-if all the 3 systems show value, then the stake is higher, of course it depends on how much value each of them indicates. Sometimes they may contradict each other, but I wouldn't care if they all prove to be profitable in the long run. Hopefully I answered your questions. Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 2, 2008 Share Posted March 2, 2008 Re: Power Rating Soccer Predictions Ok, thank you! That's a very smart way of doing things. I recently tried to do a multiple regression analysis with a few systems also based on shots, corners etc. But there really wasn't enough data available. I think it's better to do normal regression and then apply your selection methods. Quote Link to comment Share on other sites More sharing options...
robkor Posted March 4, 2008 Author Share Posted March 4, 2008 Re: Power Rating Soccer Predictions Here are the results for the previous weekend. ...and today's/tomorrow's selections... Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 7, 2008 Share Posted March 7, 2008 Re: Power Rating Soccer Predictions robkor, I have another question for you. Lets say you would have started betting with your system(s) at the start of this season and the results for this season have been good and steady. Then for the next season, would you do a new regression including the new data or would you keep it as it is, since it performed well....? Quote Link to comment Share on other sites More sharing options...
robkor Posted March 7, 2008 Author Share Posted March 7, 2008 Re: Power Rating Soccer Predictions noolnool, the regression parameters udate automatically every time I enter new data. I don't care how well the system performed in the recent past, because more data give more reliable predictions, I think. The system isn't doing good at the moment because it mostly picks the underdogs, which I hope, will turn the things around. If you take a look at my tables and take only bets where the selection was a favorite (odds under 2.50 or even 3.00) then the picture would be much much better...But, I think I'll wait for the underdogs... Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 7, 2008 Share Posted March 7, 2008 Re: Power Rating Soccer Predictions Ok thanks, that is even better of course. Which program do you use for that purpose? Sometimes though, it's good to review the regression analysis "manually" so you can remove any outliers. These especially occur at those extreme ratings that do not have a lot of data. Quote Link to comment Share on other sites More sharing options...
robkor Posted March 7, 2008 Author Share Posted March 7, 2008 Re: Power Rating Soccer Predictions I split the ratings into deciles, so in each range there is equal amount of data. This way I don't need to worry about extremes. Basically the average-range ratings are regressed, not every single data. I simply regress them against %HDA in Excel, using LINEST function, which is a little bit tricky one to use. But anyway, because there is a very large amount of data, the parameters don't change visibly after new data has been entered. Quote Link to comment Share on other sites More sharing options...
noolnool Posted March 7, 2008 Share Posted March 7, 2008 Re: Power Rating Soccer Predictions Thanks again, robkor! I'm not that familiar with the term decile but If I understand correctly you can only have ten different Odds outcomes for each match? 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.