Ishmael Posted October 17, 2009 Share Posted October 17, 2009 From The Times today: Let me tell you three things about Newcastle United. First, they are only the eighth best team in the Coca-Cola Championship. Second, if they were in the Barclays Premier League, we would not expect them to go down. And third, we regard them as favourites to be promoted. How can all these things be true? I’ll explain. The Fink Tank Predictor is a computer model of English football, built by Dr Henry Stott, Dr Ian Graham and Dr Mark Latham. It analyses the past two seasons of goals and shots on goal and weights the results so that the more recent results count the most. The model is designed to help us to provide the probability of different outcomes when teams meet. And armed with this we can simulate an entire league season over and over again. Say, just as a made-up example, that in 10,000 simulations, Ipswich Town came bottom of their division 5,000 times, their chance of finishing bottom would be 50 per cent. Related Links One of the basic building blocks of the model is a team ranking, based on their strength in defence and attack as revealed by their weighted goals and shots. The ranking is topped by the strongest team, Chelsea at present, and they are given the index number of 100. The strength of the remaining 91 teams can then be expressed as a percentage of this number. Thus Manchester United are 94 per cent of Chelsea at the moment. You will be wondering how all the league teams can be in one table, without division markers. This is done simply by looking at cup games and at promoted and relegated teams to weight matches played in each division. Thus the result of a Fulham v Hull City game can be compared with the outcome of a match between Leicester City and Nottingham Forest. When you have the 92 teams’ rankings, you see that there is quite an overlap between divisions. Thus West Bromwich Albion are the best team in the Championship and, at 29 per cent of Chelsea, they rank above eight Premier League sides. Newcastle are the eighth best team in the Championship, although they are rising. They are 24 per cent as good as Chelsea, the drop in standards in the Championship being very slow in a very competitive division. But 24 per cent as good as Chelsea puts them above Birmingham City, Burnley, Stoke City and Hull. How, though, can they be favourites to be promoted? Easily. Their early results have been spectacular, scoring 7.6 points more than the model expected. Blackpool are 5.9 points up on the model. By contrast, Reading are doing particularly poorly (11.6 points fewer than expected), as, it will come as no surprise to know, are Ipswich (13 points fewer than expected). Doing the simulations now adds the points they have scored so far into the calculations. It boosts Newcastle strongly because the division is so close. The grid published below shows the percentage chance of where Championship teams could finish in any position. With a team of a given class, there will be a range of possible finishing positions. There will be variations in a team performance that are down to luck and there will be variation in the performance of all the other sides. Newcastle fans should not get too excited because there is a 40 per cent chance that they won’t gain automatic promotion. On the other hand, there is about a 90 per cent chance that they will finish at least in the play-off places. And once they are back in the Premier League, there is every chance of staying there. Graphic viewed here: http://extras.timesonline.co.uk/pdfs/finktank.pdf Link to post Share on other sites More sharing options...
Dave Posted October 17, 2009 Share Posted October 17, 2009 What a load of bollocks. Link to post Share on other sites More sharing options...
Shak Posted October 17, 2009 Share Posted October 17, 2009 Link to post Share on other sites More sharing options...
Ketsbaia Posted October 17, 2009 Share Posted October 17, 2009 the ranking is actually a realistic looking final table. i always enjoy reading the fink tank! didnt they predict our relegation? Link to post Share on other sites More sharing options...
Pilko Posted October 17, 2009 Share Posted October 17, 2009 What a load of bollocks. Link to post Share on other sites More sharing options...
Dave Posted October 17, 2009 Share Posted October 17, 2009 the ranking is actually a realistic looking final table. i always enjoy reading the fink tank! didnt they predict our relegation? There was only Mike Ashley who didn't tbh. Link to post Share on other sites More sharing options...
Ishmael Posted October 17, 2009 Share Posted October 17, 2009 It's not hard to follow. Basically, for those of you who need this, we are favourites according to Fink Tank to win the league. It also indicates that the even nature of Championship football is helping our cause. Link to post Share on other sites More sharing options...
Ishmael Posted October 17, 2009 Share Posted October 17, 2009 the ranking is actually a realistic looking final table. i always enjoy reading the fink tank! didnt they predict our relegation? Something along those lines, they are always pretty accurate. Tried to look for older Fink Tank articles to see if they did - but they don't seem to exist online now. Link to post Share on other sites More sharing options...
Unlucky Luque Posted October 17, 2009 Share Posted October 17, 2009 Well we are top of the league so blah Link to post Share on other sites More sharing options...
Taylor Swift Posted October 17, 2009 Share Posted October 17, 2009 Future of football analysis, this. (only partly serious...) Link to post Share on other sites More sharing options...
Ishmael Posted October 17, 2009 Share Posted October 17, 2009 Casually dismissing http://www.otf2.com/forums/images/smilies/cheeky.gif Link to post Share on other sites More sharing options...
chicago_shearer Posted October 17, 2009 Share Posted October 17, 2009 If I had to put money on it, I would say playoffs is realistic. You don't need a super computer to tell you that Newcastle are unlikely to go up the easy way. Link to post Share on other sites More sharing options...
80 Posted October 17, 2009 Share Posted October 17, 2009 So... make a load of predictions. Find them to be quite significantly wrong and then factor reality into those initial predictions in order to create more predictions, which history suggests will also be significantly wrong. I see. Ashley will definitely hang on to us now AND sell the entire squad because we are DEFINITELY staying in the Premiership for more than one season. Link to post Share on other sites More sharing options...
Mike Posted October 17, 2009 Share Posted October 17, 2009 Does that simulation have a witch that randomly injures players after a string of decent performances? Link to post Share on other sites More sharing options...
Ishmael Posted October 17, 2009 Share Posted October 17, 2009 So... make a load of predictions. Find them to be quite significantly wrong and then factor reality into those initial predictions in order to create more predictions, which history suggests will also be significantly wrong. I see. Pfffft. The teams are ranked in order of how good they are, modelled on previous results (dating back 2 seasons) with more weight on more recent results. For instance, according to the simulation, Chelsea are ranked as the best team in England. This does not mean that the simulation finds them at this moment the most likely to win the EPL, but that they are the best team in the country. Similarly we were ranked as the 8th best team in the Championship. Again this does not mean that the model now predicts we will finish 8th. Bare in mind that our terrible form from last season counts towards our ranking. Also bare in mind that they have explained how they can rank all the divisions simultaneously in the article. Now taking the simulation of how good all the clubs are, they then factor in current results to determine how well teams are doing against the model. We, according to the model, are over-performing against expectation and not that the model is wrong. By factoring in recent results - basically adding into the equation the points we have won that the model didn't expect - it finds that we are favourites. So AND THIS IS THE CRUX OF THE ARTICLE even if we continue performing as the 8th best team we still have a high chance of winning the league. We are the most likely team to win the league. Also, we have a 90% chance of making the play-offs. Seriously, it's not hard to follow. Link to post Share on other sites More sharing options...
Keefaz Posted October 17, 2009 Share Posted October 17, 2009 Doesn't take into account so many other crucial factors, especially in our case. Link to post Share on other sites More sharing options...
Teasy Posted October 17, 2009 Share Posted October 17, 2009 You will be wondering how all the league teams can be in one table, without division markers. This is done simply by looking at cup games Link to post Share on other sites More sharing options...
Guest Phil K Posted October 17, 2009 Share Posted October 17, 2009 You will be wondering how all the league teams can be in one table, without division markers. This is done simply by looking at cup games Pompous middle class pr*cks - what else do you expect ? The same arrogant twaddle appears almost weekly in the Grauniad and others. Then you have the cesspit moron version of the above, from the Mirror (that some of our more dopey and gullible contributors on here grin gormlessly like a disney cartoon hick from the 1930's era and say "It's...duh...tongue in cheek...huh, huh, huh...") We'll get more and more - and it'll get worse if it looks like we may actually go up. Link to post Share on other sites More sharing options...
Mowen Posted October 18, 2009 Share Posted October 18, 2009 You will be wondering how all the league teams can be in one table, without division markers. This is done simply by looking at cup games Pompous middle class pr*cks - what else do you expect ? The same arrogant twaddle appears almost weekly in the Grauniad and others. Then you have the cesspit moron version of the above, from the Mirror (that some of our more dopey and gullible contributors on here grin gormlessly like a disney cartoon hick from the 1930's era and say "It's...duh...tongue in cheek...huh, huh, huh...") We'll get more and more - and it'll get worse if it looks like we may actually go up. Fuck me sideways. I quite like the Fink Tank, obviously it's miles from exact but it's really interesting to look at applied modelling in this context and seeing how the model changes and improves. Link to post Share on other sites More sharing options...
Guest Phil K Posted October 18, 2009 Share Posted October 18, 2009 f*** me sideways. I quite like the Fink Tank, obviously it's miles from exact but it's really interesting to look at applied modelling in this context and seeing how the model changes and improves. Well, matter of opinion, of course. And I do have a problem with the national press. Maybe many don't (or can't ?) see it, but the broad sheets are as petty against us as the tabloids, with less excuse. Link to post Share on other sites More sharing options...
Alberto2005 Posted October 18, 2009 Share Posted October 18, 2009 What a load of fucking gash Link to post Share on other sites More sharing options...
Guest Phil K Posted October 18, 2009 Share Posted October 18, 2009 What a load of f***ing gash Reading your past comments again ? Tut tut don't be so self-critical... Link to post Share on other sites More sharing options...
Guest Dr. Richard Kimble Posted October 18, 2009 Share Posted October 18, 2009 Read "The Punters Revenge" 1987 http://books.google.co.uk/books?id=JNk9AAAAIAAJ&dq=punters+revenge&printsec=frontcover&source=bl&ots=R1WmcMWi0o&sig=JNwmUM9Lh50cTp44d8QpuQhlmgE&hl=en&ei=FlnbSpioKZSd4QaL16H1Bg&sa=X&oi=book_result&ct=result&resnum=10&ved=0CB4Q6AEwCQ Its a golden oldie but still the best book about sports betting ever written. Most of these guys using these modelling systems will have it as their cornerstone. Link to post Share on other sites More sharing options...
Mowen Posted October 19, 2009 Share Posted October 19, 2009 BTW they tested the Fink Tank against the bookies in its first season. Whenever there was a significant difference of opinions (i.e. bookies offered better odds than they should have done according to the model) they stuck x amount of money on where x was dependent on how big the difference of opinion was and they made a lot of money out of it. Link to post Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now