Football Ratings with a Difference

[Sir Puntalot]

I am trying to devise some ratings that reward teams who have been playing well even though their results might not say so and vice versa. It should throw up a few selections that you might generally dismiss just looking at results, but as we know, results aren't always a true reflection of what happened in the match. The ratings are nearly finished, but will need a few games to take proper shape obviously. Each team will be given a "home rating" and an "away rating" rather than a general rating as most teams play better at home or away in general. I'll put the weekend ratings up soon. Even though there has only been 2 weeks results, it should still give a rough idea. These ratings will be for Nationwide teams only as the Premiership imo is too hard to rate, whereas Nationwide teams don't have such big divisions within a division. Possibly Scottish Premier will be added to the ratings too but unfortunately stats aern't available for Scots 1, 2 & 3. I'll update them every week and put them here to see how they go. ;)

[Sir Puntalot]

Re: Football Ratings with a Difference Just found out that I can add Conference matches too. :)

[Guest cjamars]

Re: Football Ratings with a Difference Hiya Pauk Sounds something like the idea Paul Steele's penetration ratings are based on (corners, shots on and off target). He tries to show that a 1-0 match ith 1 shot on target shoudl reflect the fact that maybe tehb oppositiion had all the possesion and chances but did not score. I'll be doing a system based on this myslef so I'll be interested to see how this compares with yours. Needs 6 game settling in period mind

[Sir Puntalot]

Re: Football Ratings with a Difference Hi Cj, Yup, it will need a few games to settle in but I'm prepared to put them up as a guide from now anyway. It really does seem to give a true reflection of how a team is playing as I have messed about with the points system for a while now and I think I have spot on. :) An example would be N Forest 2 Sunderland 0: I rated this as N Forest 53 Sunderland 55 and if anyone saw the match they probably couldn't disagree. Sunderland had a lot of chances but took none, yet still get marked high for their all round play. Another would be Coventry 0 Walsall 0: I rated this as Coventry 72 Walsall 12, a huge difference and rightly so, as Coventry hammered them without putting the ball in the net. :smokin I think this half belongs in here and half in Britain as many people don't look in here and it's intended as a tool for all. Maybe I will add the ratings to the weekend thread for people to have a look at too. :)

[Guest OddsAgainst]

Re: Football Ratings with a Difference People who are interested in ratings systems most certainly look in here. Those who are not, probably don't, so they won't be interested wherever it is posted. A good idea might be to include your ratings when you are discussing the matches in the analysis threads in Britain but post the ongoing details here so anyone interested can study the details easily in one thread. I am sure there will be a lot of interest in (and comment on) this type of system.

[Guest Weststander]

Re: Football Ratings with a Difference Hello Paul, Seperate home & away ratings. i've been all around the block on this one but I reckon you're better sticking to an overall team rating & sticking it into the overall context of the league. I'm not sure if home or away "specialists" persist long enough to be considered anything more than just a naturally occuring fluctuation. Over time most things in soccer regress towards the mean for that particular league.By splitting teams records by home & away you probably end up doing the exact opposite. On a more practical objection,by taking an overall rating you're rating that team on roughly twice as many games over the same time span compared to if you split the games by venue. As for rating how teams should have done compared to how they did,are you finding many games where you're making adjustments? If it's not many you could probably just treat them as outliers. If it's alot aren't you just eliminating possibly a real phenomena that might actually give rise to the price descrepancies that are profitable. It's tempting to think that teams that have squandered lots of chances have been unlucky,but on average only 2 & a half chances are converted per game anyway. W.

[Guest madmick]

home and away I do believe there is a definite home advantage, and a good ratings system needs to take this into account. Interested in your figure about conversion of chances Westander, is that apocryphal or have you done some number-crunching on it?

[Sir Puntalot]

Re: home and away Hi Weststander, This is where I think people go wrong with ratings, they go too deep. You have to have a home and an away rating for each team because there are so many teams that play well at one or another. The question you should ask yourself is this - if home and away ratings don't matter that much, then why are teams priced differently when playing at home than away and vice versa ? I believe it matters a hell of a lot, but we all have different opinions I suppose. :) I would just like to add that my ratings system has 2 parts, it has a results rating based on the performance of the teams and also a predictions rating, although the predictions would need a few games to be realistic.

[Guest Weststander]

Re: home and away Hi Paul, I'm not disputing the existence of home advantage(about 8 tenths of a goal difference between AvB & BvA),just the existence of home or away specialists. I think imposing the league split (50/25/25 or whatever) on a team's overall record gives a better result for the reasons outlined. The teams that do appear for a time to show preferences for specific venues always end up gravitating towards the mean for the league. IIRC the RFO's dedicated home/away ratings were outperformed by the overall ratings. Whatever your views on the RFO,this fact at least is worthy of consideration. JMO. W

[knyd]

Criteria Hi Paul, what are the criteria for your ratings? I'm thinking to start a performance ratings tracking for other European Leagues, and i would be pleased to have your opinion The Will Hill system is: 10 points for each goal scored 3 points for each corner won 5 points for a clean sheet Minus 10 points for each red card received But i don't agree, i would prefer a system like that: +10 for each goal scored +5 for a clean sheet -2 for each yellow card received -5 for each red card received +2 for a corner +3 for each shot on target +1 for each shot off target (for some foreign leagues, as the norwegian one, it can be difficult to find these data, having only the "goal chances": +2 for each goal chance if this is the case) What do u think about?

[Guest OddsAgainst]

Re: Criteria Knyd, if you want to discuss another rating system (your own) then it might be a better idea to start your own thread for it. If you decide to then I will delete these posts.

[knyd]

Re: Criteria No Odds, just a little question to Paul to know the criteria of his own rating system, no need for a new thread. Anyway, if you think my post is not adequate to this thread no problem, you can delete it and i send this message as a private to Paul. :)

[Guest OddsAgainst]

Re: Criteria No problem, it's your choice.

[Guest DJFLEECER]

Paul Steele I know that the purists with nothing better to do will absolutely piss on my bonfire, but here goes: Among Steele's 15 or so systems analyzed, mine beats the rest for homes and aways, and comes second in the draws category. Sit down ! I know this means bugger all unless you know the odds at which these teams won at. After scrutinizing until contracting "spare bedroom arse", I've found that I can scratch out a meagre profit of between 20 and 25% per season - peanuts when starting with the sort of bank that I did. My dream is to market the old system and make a few quid in the process. Some "expert" told me a while ago to forget it as these things are merely a labour of love and would involve setting web-sites up etc..... I just wonder if the inventors of Power Ratings and Penetration Plus ever made any brass out of their hobbies ? If so, Micky wants a slice ! Please excuse the ramblings of a devotee, and a dreamer. I'm just looking for that rainbow - the one with the pot of gold. Relp raggy, MTF

[Guest OddsAgainst]

Re: Paul Steele 20-25% is a fabulous profit. You are talking about yield here? If that is the case, why not proof them here? You will surely win the systems league prize money and once proven you can sell your tips, sell the system or do as Mr Steele did and make money by writing a book about it. First step is the proof though.

Some "expert" told me a while ago

As I was the only one to answer your previous post I must be that "expert". See here.

[Guest cjamars]

Re: Paul Steele A wind up surely by DJFLEECER? Sound like the syntax of a certain West Cork hitman ;) Am in process of analyseing the last 5 years in England with Mr.Steele's Power ratings. Results not encourageing by themselves (which begs the question of course why I'll be useing them this year?

[Guest DJFLEECER]

Re: Football Ratings with a Difference Wind up ? Why would I bother to waste my time and yours by making up a story ? Go and find a dictionary. I have worked on the system for years (called Footy3 MKII) - can't think of a better one yet ! I spend about 3 or 4 hours per week just processing data and re-assessing information in order to generate profit, hoping to find new avenues down which to tackle the enemy. I would love to proof the system at some stage. My worry is that it would no longer be my sole possession. To have it proofed would surely mean that the cat is, in effect, out of the bag ? Obviously some parts of the system are more profitable than others. I intend to change my betting strategy this season. I'll be concentrating on laying aways. On average there are only 2 bets per week, odds of just over 15/1 gets me a very tiny profit - anything else is welcome. The only foreseeable danger with this method is the likelihood of the big underdog being a massive price on the exchanges....I can't imagine Dundee Utd for example, being less than 20's at venues such as Ibrox. We'll have to wait and see. MTF

[Guest OddsAgainst]

Re: Football Ratings with a Difference

I would love to proof the system at some stage. My worry is that it would no longer be my sole possession. To have it proofed would surely mean that the cat is, in effect, out of the bag ?

Why? Posting your selections does not reveal how you make them. Have a look at the various system threads in this forum. Can you work out how they make their selections?

The only foreseeable danger with this method is the likelihood of the big underdog being a massive price on the exchanges....I can't imagine Dundee Utd for example, being less than 20's at venues such as Ibrox.

Don't see a problem there when the true price for them at Ibrox is closer to 33. You have made three posts now with reference to this system and told us nothing except that it makes a fabulous yield of up to 25%. Such posts are of llittle interest to anyone and many will suspect they are just idle boasts. The time has come to show what you can do, otherwise just exactly what is the point of your posts?

[Guest DJFLEECER]

Re: Football Ratings with a Difference I was looking for advice, I'm not in this to dish out idle boasts. You're right about one thing though, it's time to put up or shut up. I think now is the time to shut up and disappear into the ether. Hopefully I'll find a forum where folk don't see you as a challenge to their self-elevated guru status. We're all looking for the answers, it's just that some of us don't mind it when others are successful in their own right.

[Guest OddsAgainst]

Re: Football Ratings with a Difference Before you go, could you possibly point out exactly where in your two post above you are asking for advice?

[Luddite71]

Re: Football Ratings with a Difference Very interested in Paul's ratings system. I personally believe that home advantage has to be taken into account even if two separate ratings are not maintained. I developed a simple system some years back whereby each team had a single rating but the home teams rating was then increased by 1-5 points dependant on current home performances. The system itself gave each team a rating from 0-100. 100 points being the pinnacle of club football which IMO was then (and still is),the great AC Milan side of the late 80's early 90's. A rating of 100 therfore would place that team as an equivalent to the great Milan outfit. It was possible for the rating to exceed 100 points however due to the work load involved assessing each teams performance game by game i only did the ratings for the English Premier. I think i perhaps over complicated the criteria for updating the ratings which often had to be done twice a week. I would be interested to know what people think are the relevant stats in each match for updating performance!! ratings

[Guest cjamars]

Re: Football Ratings with a Difference DJFLEECER I merely offered the idea that you might be a certain other poster posting under another name. If you’re not then mea culpa. You have however made some pretty bold claims that really by themselves mean nothing other than cheap titillation. QUOTE Wind up ? Why would I bother to waste my time and yours by making up a story ? Go and find a dictionary. Fail to see why I need a dictionary. Lots of people (including myself at various times) post wittingly and unwittingly complete and utter drivel on this forum so why should you be any different. I can’t remember a single contribution of yours that would warrant anyone taking anything you say other than with a very large pinch of salt. QUOTE After scrutinizing until contracting "spare bedroom arse", I've found that I can scratch out a meagre profit of between 20 and 25% per season - peanuts when starting with the sort of bank that I did. I’d love some of that meager profit. Obviously you’re a big time player that such returns mean so little.:D QUOTE I think now is the time to shut up and disappear into the ether. Hopefully I'll find a forum where folk don't see you as a challenge to their self-elevated guru status. We're all looking for the answers, it's just that some of us don't mind it when others are successful in their own right. Not a single person here would claim to be gurus. This is a thread with some seriously talented people that try and share their ideas without necessarily killing any golden geese. Many turn a regular profit. I’m sure they’ll all miss you as much as I will. Enjoy the ether..just please stop sniffing it before you post ;)

[Guest cjamars]

Re: Football Ratings with a Difference Hi Luddite Most ratings systems do use a Home rating advantage as far as I know. Steele uses a .2 goal head start for Homes in his main system rating (POWER). Unfortunately from what I can see his ratings don't seem (on their own anyway) to move one closer to a purely mechanical system to beat the bookies. I like the idea you talk about of an absolute point (AC Milan being yours). Getting to it is the problem. The quality of opposition, the time the team plays , the game itself all play an impact. On a geam by game basis I'd offer goals scored shots on target shots off target corners pass completion % tackles won % I notice Paul (Ross) has cards in his but I wouldn't use that myself to be honest as I think cards are rather irrelevant until they are red and they tend (generally anyway) to be random events.

[Guest DJFLEECER]

Re: Football Ratings with a Difference Right, Sorry to anyone that may have been offended by any of my bad tempered ramblings, I just get defensive when challenged by the condescending "elite". I'm not for a minute saying that the replies to my previous posts are from this band of people, just that the ones that tend to mock are the ones that are still clueless as to where to look for ideas. In fact, upon reading previous links and posts, there ARE in fact, a number of contributors who know their stuff. I'd like to think that I too know a reasonable amount, system wise. I don't claim to be an expert, just a very dedicated believer in my own system. It has produced the results previously stated, I haven't got the time to locate sites of interest, just to post "idle boasts". My initial scribings, many moons ago, did express the desire for advice. That is still the case. I may indeed take up the challenge of recording my forecasts for all to see, but that, for me, would be a minor irritation. I don't give a hoot if my tips come first or last in any table, and are praised or lambasted by all. You see, the pleasure for me comes in the assessing, studying and the final dream that one day I may even make a living out of this (some dream I know !). All I wanted was lively interaction with like-minded people, some of whom may bring me closer to achieving my ultimate dream. Ideas, advice and experiences, if not clearly indicated in prior postings, are ALL I'm looking for - not an audience for an inflated ego. Mick.

[Guest Weststander]

Re: Football Ratings with a Difference Surprised that there's not more flame type wars on this bit of the PL,given that it's essentially about opinion/fact/perceived fact....& its easy to come across as abrupt in the written word. Anyway this is jmo so feel free to flame away. Performance ratings. If you're looking to create an artificial number for the purposes of spread betting or Hills' type fixed odds performance betting then include everything.Shots,corners,goals,red cards etc. If you want a predictive indicator of future performance then include goals scored,goals conceded....& nothing else. Reason being,if you include shots/corners etc & goals,you're including events that are themselves correlated to each other & that's a problem in itself. At one level you're measuring the same thing over & over. W.

[Sir Puntalot]

Re: Football Ratings with a Difference

I notice Paul (Ross) has cards in his but I wouldn't use that myself to be honest as I think cards are rather irrelevant until they are red and they tend (generally anyway) to be random events.

I don't use cards in my ratings and agree they are irrelevant. :) I will say this, the away teams score differently to the home teams in my ratings, simply because it is harder to play away than it is at home, but the difference in points awarded is not massive at all. ;)

[Luddite71]

Re: Football Ratings with a Difference Thanks for your input Cjamars! I used to update my ratings not based on individual stats within a game but more by giving a value (whether it be plus or minus) based on the last result achieved. eg. A 1-0 away win against a team rated 10pts higher would be better than a 2-0 away win against a team 30pts lower. also the margin of victory/loss had a bearing on the movement in rating. i kept if going for about two and a half seasons before getting bored with it. im thinking of resurrecting it again but i'm not entirely sure how to rate the teams at the outset. Any ideas?? It ended up getting quite complex

[Sir Puntalot]

Re: Football Ratings with a Difference

im thinking of resurrecting it again but i'm not entirely sure how to rate the teams at the outset. Any ideas??

That's the problem most people come across, they try to make it perfect and it never works. Make it simple and logical and it should work. :)

[Guest smudge]

Re: Football Ratings with a Difference DJFLEECER, i hope you stick around and are able to put some of your passion for systems and football betting in general into this board. i know the place can seem daunting at times - i have been posting over 2years now so i should know - but this really is the best place on the internet to really get down to the bare science of football betting! there are some very talented and experienced individuals here whom are worth listening to when offering advice - i'll mention no names - and they are not here to flame you or anybody, they are here for the same reasons as you = the answer! there are also plenty of dreamers [myself included] who will fill you with confidence and encouragement but will not be able to offer and REAL constructive criticism about your system from a mathematical perspective! this forum of the lounge is, by its very nature, the one under most scrutiny - take nothing personal!!! enjoy your stay :)