Backing the draw?

You can tell I've had a little bit of spare time today!

Greg asked another question which I thought was quite interesting. The crux of his query was, how much profit/loss would you make from backing the draw in all these games that are defined as value?

My initial reply is in the comments section to his question but simply, I expect it to be loss making as if there is an edge backing the team I think is value, there can't be an edge backing the draw as well! Clearly, in a short-term period, you can make money backing the draw and I thought this was probably the case over the recent period.

Now I've had time to look at it and analyse it a little, the recent results are worth highlighting.

Firstly, backing the draw isn't a way to make money based on my ratings. That isn't to say that they can't be used to make money as I could backfit my data to try to maximise the draw returns but based on the fittings at the moment, you will lose if you solely back the draw.

However, the results from this season are different to the preceding years. In the first half of this season, things were in line with the long-term and backing the draw would have lost you 185pts from September to December. Nothing too unusual there.

In the period from 1st January to 11th April (today!), you would have made a profit of 188pts!

Looking back, I can't see a spell like this one during the past 4 and a bit seasons. I'm fed up going on about the draw and clearly, this all just fits in with the DNB analysis and so on as it's all the same data just cut a different way but before making any decisions about how I use the systems next season, it's important that I don't get too caught up with what has happened in the last few months. Yeah, it hasn't been pretty and the systems have made no profits in 2011 but then again, it's unlikely this run is going to continue for much longer based on the previous 13,000 bets the systems have had. Either that or I've developed a system to help me win the pools!

Here's the numbers for those geeky readers.

Replies to a couple of comments

I had two interesting comments over the weekend regarding the results on here. As I said in my earlier post today, I've deleted the comments by accident as they were in amongst some moronic anonymous comments but I did read both comments!

The first comment asked me what the returns would have been this season had I just suggested backing every team that is value. I've graphed the P&L over the season for the traditional H/A betting, DNB bets and DC bets.



There have been 672 games that had a team that is described as value. By value, I mean where the odds at the time of betting are greater than the % chance that they will win the game according to my ratings.

The profit for traditional H/A betting is 15pts. This gives a return on investment of 2.2%.

The profit for DNB betting is 33pts. This gives a return on investment of 4.9%.

DC betting is break-even.

Clearly, these results are slightly disappointing at face value but please remember that this includes all value bets. Hence, the purpose is to simply create a profit or at least, break-even. As soon as I start to filter these bets, I can increase the profits substantially by only playing the higher value bets.

If I was being judged on this measure this season, I would have had a filter in place to only pick out value bets once a set criteria has been met. Hence, I would drop all low value bets (system 6 and system 21) and only play higher value bets as it would reduce the number of bets and increase the returns. I'm not being judged on my ratings though, I'm being judged on my systems and therefore, I don't need to worry too much about this aspect.

All the above shows is that you can't just back every team that appears as value, regardless of how much value you think is achieved by backing the team. If you have a team with 50% value appearing on system 8 and you have a team with 10% value appearing on system 6, then clearly, you would want system 8 bets to be better than system 6! Hence, that's what the systems do.

The other thing worth considering is the shape of the graph. H/A betting was above DNB until Christmas (roughly 400 bets) but since then, DNB has kept going up and H/A betting has been in freefall. This is driven by the number of draws that have been appearing in the value bets since Christmas.

In the first half of the season, there were 85 draws in 339 bets. A draw % of 25% which just so happens to be the long-term % for draws in the ratings.

From Christmas, 333 bets and 107 draws. A draw % of 32%. The extra 22 draws or so that you wouldn't expect have eaten away our margin with the ratings after Christmas when backing H/A. If you are covering the draw, as the graph shows, you are still able to make money after Christmas.

Anyway, I hope that answers the question that was posed.

The second comment asked for a breakdown of pts won on each system this season by month. I'll keep it nice and simple and not worry about ROI etc. as it wasn't asked for. The table below has all the info you require by month. I've also shown the cumulative view of the season as it tells the story pretty well I think of how the systems have done this season!

Weekend Update

Before I get on with the results, a quick comment re. a few anonymous comments this weekend.

I've had a couple of anonymous comments from morons over the weekend who clearly can't count up if a system makes a profit or a loss as they appeared to be suggesting it had been another shocking weekend on the footie. I don' mind people posting rubbish if it's been a bad weekend but when the systems make a profit on the weekend, please give them a break!

Another weekend on the footie and yet again, I'm left scratching my head about what exactly is going on with the footie.

The bare facts are 17 games, 9 wins, 5 draws and 3 losses. Resulted in a profit of 3.6pts if backing each game or a return on investment of 21%. The systems took these 17 games and turned it into a profit of 8pts from 58 games. ROI was 13.8%.

As I've been saying recently, I'm tracking A(0) and A(0.5) for my bets also. For those who prefer the more traditional terminology, it's DNB (Draw No Bet) and Double Chance (Win/Draw or +0.5 handicap)

The DNB produced a profit of 3.4pts for an ROI of 20% from the 17 games and DC was a profit of 2.5pts (14.7% ROI).

Interestingly, when you look at how DNB and DC did across all systems, the systems turned the profits from the 17 games into losses. DNB made a loss of 1pt from the 58 games and DC made a loss of 8.5pts across the systems!

Quite frightening the damage the systems caused at the weekend but again, it was caused by a few well picked teams losing. Yeovil as I said on Thursday were picked by every system and they lost away at Hartlepool. Bournemouth were also well fancied and they got turned over at home. Hence, when two of the most picked teams lose, then it is going to suppress the returns of the systems massively.

This is again reflected in the multiple bet systems where they all suffered losses (as usual!) and continue on a massive downward spiral.

It's hard to know what is going on at the moment with the football. All of the systems are really struggling since Christmas and as the season starts drawing to a close, I can't see it getting much easier.

We've discussed the seasonality that comes with football betting in the TFA forum a few times (mostly to do with other tipsters who appear to struggle after Christmas) and I'm afraid that these systems might well get labelled with the same accusation. It's hard to understand why they've not performed over the last few months but I'll look into this more during the Summer.

I've got two comments to answer from weekend posts but unfortunately, I deleted the comments as they were in amongst anonymous comments and I accidentally ticked them all to be deleted. One was asking for the returns of backing every team that appears as value and another one was asking for the pts profit by for each system by calendar month this season. I did mean to publish them but when the comments are anonymous, I tend to delete them more often that not!

I'll answer both of these later today hopefully as it won't take me long.

Weekend Bets

Another busy weekend with 17 different games this weekend.

One game stands out this weekend as a clear value bet and appears on every system which doesn't happen too often. Yeovil are the team and at 2/1, they are clear value against Hartlepool based on the recent form of both teams.

I've had a quick look at Fink Tank out of curiosity and they make Yeovil more likely to win this game than Hartlepool. Hence, 2/1 seems decent and won't last I suspect.

Looking quickly, Hartlepool have lost their last couple of games at home and have no wins in their last 4. They drew 1-1 last time out to stop the rot but looking at the stats, Swindon dominated the game with more than twice the number of shots on goal.

On the other hand, Yeovil have won 3 of their last 4 games and in the defeat, they easily beat Bristol Rovers on shots but just couldn't score. They've won their last 2 away games, scoring 8 goals and beating teams in the top half of the table.

In their last 6 away games at bottom half sides, they've won 3, drew 2 and lost 1.

It should be fairly close between these two but my ratings favour Yeovil. Hence, at 2/1, it's an obvious value pick which explains why it appears on so many systems.

I think it's interesting when I look at some games and why the ratings have found value but I try my best to not look as being honest, I struggle to justify most of the bets!

The Football Analyst Bets for 8th-10th April

System 6
Date League Game Selection Odds Bookies
08/04/2011 League 2 Macclesfield v Accrington Accrington 2.40 Tote
09/04/2011 Championship Derby v Coventry Coventry 3.44 Pinnacle
09/04/2011 Championship Ipswich v C Palace Ipswich 1.82 Pinnacle
09/04/2011 Championship Swansea v Norwich Swansea 2.25 SJ
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Bristol Rovers v Exeter Exeter 3.00 Blue Sq
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 League 1 Oldham v Charlton Oldham 2.50 SJ
09/04/2011 League 2 Cheltenham v Rotherham Rotherham 2.50 Tote
09/04/2011 Premiership Bolton v West Ham Bolton 2.25 VC
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals
09/04/2011 Bsq Prem Gateshead v Histon Gateshead 1.57 WH
10/04/2011 Premiership Aston Villa v Newcastle Newcastle 4.72 Pinnacle

System 7
Date League Game Selection Odds Bookies
08/04/2011 League 2 Macclesfield v Accrington Accrington 2.40 Tote
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals
10/04/2011 Premiership Aston Villa v Newcastle Newcastle 4.72 Pinnacle

System 8
Date League Game Selection Odds Bookies
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes

System 9
Date League Game Selection Odds Bookies
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes

System 21
Date League Game Selection Odds Bookies
09/04/2011 Championship Swansea v Norwich Swansea 2.25 SJ
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Bristol Rovers v Exeter Exeter 3.00 Blue Sq
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 League 2 Cheltenham v Rotherham Rotherham 2.50 Tote
09/04/2011 League 2 Lincoln v Gillingham Gillingham 2.25 Ladbrokes
09/04/2011 League 2 Northampton v Bury Bury 2.50 Corals
09/04/2011 League 2 Port Vale v Southend Port Vale 2.05 WH
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals

System 22
Date League Game Selection Odds Bookies
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals

System 23
Date League Game Selection Odds Bookies
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes

System 6-21
Date League Game Selection Odds Bookies
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Bristol Rovers v Exeter Exeter 3.00 Blue Sq
09/04/2011 League 2 Cheltenham v Rotherham Rotherham 2.50 Tote
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Championship Swansea v Norwich Swansea 2.25 SJ
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes

System 6-22
Date League Game Selection Odds Bookies
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals

System 7-21
Date League Game Selection Odds Bookies
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes


System 7-22
Date League Game Selection Odds Bookies
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20 Corals

System 8-21
Date League Game Selection Odds Bookies
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67 WH
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.80 Ladbrokes

System 8-22
Date League Game Selection Odds Bookies
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 3.00 Ladbrokes

System 16
Date League Game Selection Odds Bookies
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 2.75
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.73
09/04/2011 League 1 Oldham v Charlton Oldham 2.50
09/04/2011 Premiership Bolton v West Ham Bolton 2.25 Corals
Any 3,4 and all 5 = 16 Bets
System 17
None

System 18
Date League Game Selection Odds Bookies
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 2.75
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.73 Corals
Any 2 and all 3 = 4 Bets

System 19
Date League Game Selection Odds Bookies
09/04/2011 SPL Aberdeen v Hibs Hibs 3.20
09/04/2011 League 1 Bournemouth v Tranmere Bournemouth 1.67
09/04/2011 League 1 Hartlepool v Yeovil Yeovil 2.75
09/04/2011 Premiership Tottenham v Stoke Tottenham 1.73 Corals
Any 2,3 and all 4 = 11 Bets

System 20
None

More of the same in midweek

Just bringing the latest results up to date on the blog.

2 games in midweek, 2 losses. 4 system bets, 4 losses.

One thing that's quite clear now is that whatever worked in the first half of the season on systems 21-23 isn't working in the second half of the season. System 23 was hailed as some sort of freak when it hit 9 winners in 10 games earlier this season. Unfortunately, it has followed this run with a historical worst run of 9 wins in 36 games at average odds of 7/4.

It is hard to explain how a system can make 16pts in the first half of the season and lose 15pts in the second half of the season.

System 22 is even worse in a sense as it reached a peak of 38pts profit on the 18th January and has since lost 23pts from this peak.

I find it hard to understand what has changed in football from the middle of January! Clearly this period from January to now hasn't followed the statistical 'norm' in the sense my ratings can't quite cope with what is going on at the moment.

I can point to a couple of things (too many draws and terrible performance of low priced away teams this season) as reasons why the ratings haven't performed since Christmas but it isn't really an underlying cause. The effect is is that there have been too many draws and a terrible performance but I'm not sure what has caused it this season.

I've obviously got lots of plans for the summer to look back at my first season and see what has gone right and what has gone wrong. I know which sort of bets I'm comfortable backing and which I'm not comfortable backing and I'll be bringing these changes into the systems from next season. I also know that I need to take more chances on marginal value teams I suspect as too many bets at 7/2-5/1 qualify as marginal value bets but get dropped due to the low probability and increased losing runs these games add to the systems.

Anyway, I could go on and on about things I need to tweak this summer but I need to get to the Summer with the betting bank in one piece and this month is doing its best to destroy that idea!

TFA Ratings

I think it’s worth doing a quick post on some of the work that has gone into producing the ratings behind the systems. I’ve steered clear of this on the blog as ultimately, this is my intellectual property and there are a lot of very intelligent people out there who can take what I’ve done and improve upon it I think. I don’t pretend that any of these ideas are original ideas as they aren’t and I’m not ashamed to say that I’ve picked up most of this from other sources in the public domain. However, what I’ve done with the data is different to anything I’ve seen before and therefore, my ratings are unique from that aspect.

Firstly, a bit of background about me to try to fill in how I’ve reached this stage. I went to university to study Mathematics but switched to Financial Economics at the end of my second year. I went on to gain a first class honours degree BSc degree in Financial Economics. My passion was statistical modelling and in particular, regression modelling.

When I went to work for an actuarial pricing team in the insurance industry after graduation, my modelling ability came to the fore and I now manage a small actuarial team within the organisation. I’m very comfortable with SAS, SQL and of course Excel. Data manipulation and statistical modelling is what I do as my day job so to speak.

When I first turned my attention to looking at the football, it involved an initial 6 week process of trying to establish how I would go about building a statistical model from scratch. After a bit of research into what others had done in the past, it became fairly clear that others had put a lot of thought into this and therefore, I didn’t need to build something from nothing. All I had to do was understand what had been done before and improve upon it.

I make no apologies for not referencing my reading material here but I hope you can understand why I don’t want to lead people to trying to understand what I’ve done. In summary, the model would have to:

 Take account of different abilities of teams
 Take account of the fact that teams at home are stronger
 Take account of the fact that more recent form should take precedence over earlier form
 When looking at the results of games and performance of teams, account should be taken of the ability of the teams involved
 Lastly and most importantly, a team’s ability can be summarised by their ability to attack (score goals) and their ability to defend (concede goals)

I then set about seeing what data I could obtain which would enable me to build a statistical model. Importantly, I had to try to find something that could be updated regularly (after each set of fixtures) and of course, it had to be manageable within Excel preferably.

In summary, I captured data going back to 2002 which shows Home Goals, Away Goals, Home Shots, Home Shots on Target, Away Shots, Away Shots on Target for every game in the leagues I was interested in looking at.

I then spent the next few months trying to build a regression model that would allow me to build a rating algorithm for the football games. For those not familiar with this, it is simply using the historical data I’ve collected to calculate the strength of each team when each game has taken place and then when combining the variables in my model together, this creates an equation which can predict the likely home goals and away goals in any game.

I then plug this into a Poisson distribution equation (again, this is common knowledge amongst football system builders and has nothing to do with my ideas!) and this gives us the likely home/win/draw % in any game after a bit of manipulation.

I ran 100,000 iterations to achieve the optimal weightings for the regression equation and didn’t use the traditional GLM method of minimising least squares. Simply, I didn’t want to backfit the systems so much in that I restricted the number of games massively and ended up with a 100% strike rate over a dozen games. I simply worked it to goal seek a return of circa. 20% across the backfitting of the games and maximise the number of games where possible so I got a large enough data set.

This gave me my first rating algorithm and then it was simply a case of testing this with the results in 2008 and 2009 and ensuring that the returns were in line with the backfitted years. As it turns out, the returns were within a few % points of the earlier ROI and my first rating algorithm was built.

What do I do with it now was the obvious question…..

Clearly, backing every team that appears as value is a profitable method (system 6) but I looked at filtering the selections by the amount of value that the ratings determined was inherent in each selection. This created systems 7,8 and 9 as ultimately, these systems purely separate the bets out by the amount of value the ratings define in each game. The ratings can clearly differentiate between the games that appear as value bets.

Once I had built the first ratings algorithm, it was always going to be the case that a second algorithm would follow I think. Simply, when I was looking at the regression model for the first algorithm, I had lots of ideas about other ways to model this and therefore, I always wanted to go back and look at doing something a bit different by manipulating the data a bit differently.

After a month’s work or so, system 21 was built and using the same idea for systems 7-9, systems 22 and 23 were built.

It was purely by coincidence that systems 6-21 etc. were looked at as it was someone else who pointed out that when the two algorithms pick the same team out as a high value bet (system 22/23 and system 8/9), the results appeared to be exceptional. I looked into it a bit further and as we’ve seen this season, the results that can be achieved by doing this are as good as anything I’ve personally seen for football betting.

Now, people can read the above and make accusations that I’ve copied Dixon/Coles or Fink Tank and so on but at the end of the day, I’m not saying that any of this is my own idea. I’ve taken ideas developed by others before me and improved upon them and hopefully, can outperform the results anyone before me has achieved. That’s my aim.

2 questions people have always asked me about the ratings this season.

Will this method always work? I’ve no idea. I doubt it myself as if I look at the returns over the past 8 seasons, the returns have basically been on a downward trend. Hence, I don’t believe that using data from 2002-2006 can really help me predict the results of games in 2011 as well as it could in 2009 but I can always build new systems using more data as time goes on. Therefore, at the end of each season, I can build new algorithms with an extra season’s data and go through the same process as I’ve done before this season and hopefully, that will keep me ahead of the game and the bookmakers.

Why do my ratings work better than some others which use the same variables? I don’t know the answer to this one. I’m using very similar criteria for bets as other rating models and shots on goal is my main indicator and this is similar to what others use. My advantage is purely from my backfitting/backtesting I think as I’m quite comfortable with regression modelling and how to use it and I figure that’s probably my competitive advantage when it comes to football modelling.

How can we maximise profits and minimise risk using the ratings? Unfortunately, even though we are 80% through this season, I’m not sure I’ve quite got to this stage of the process yet. Saturday was a good example! I just want to get to a stage where I’ve got a proven way of making a return of 10%+ on the football over a season and of course, I want to do it with the highest number of games, lowest betting bank possible. That way, I can maximise profits and minimise the risk.

Anyway, I hope the above gives a flavour of what I’m doing this season. Unfortunately, this season was always going to be an experimental season and being honest, due to the need for me to tweak the algorithms each season with extra data, the systems will never be the same two seasons in a row. However, the base assumptions will stay the same in each algorithm and therefore, the systems should be as profitable one season to the next I hope.

Weekend Massacre

I'm not sure where to start after a weekend like that.

Firstly, apologies for not updating the blog sooner. I've spoken about this before on here but when you have a bad weekend and need to update your results and numerous spreadsheets, it does seem more difficult to find the time. I know some of it is mental but the truth is, I've just been really busy since Saturday night and therefore, I've really struggled with the football this week.

I've not posted the bets for the midweek yet (was only two games and one lost last night!) but I'll just keep updating the blog as I would usually do. Hence, I'll do the midweek results later tonight when the game finishes. It's Dundee Utd to win tonight for anyone interested.

Well, Saturday will live long in the memory. Of the 12 games on Saturday, only 5 draws and sadly, 7 losses. 57 bets, 57 losses for the systems. Sunday improved things slightly with Fulham winning but it made little difference to the weekend. On the Saturday, the 57 bets were split 33 draws and 24 teams losing outright.

If I’m honest, DNB or DC hardly saved the weekend either with DNB losing 24pts and DC losing 14pts. Hard to take after the run the systems have had in March as March managed to recoup the losses from February. Unfortunately, the weekend past has just wiped out nearly all of last month’s profits in one hit.

There was a bit of talk about it in the TFA forum and things just didn’t seem to go the way for any of our bets. Luton were a big bet for everyone following and after going 2-1 up with 10 minutes left and then 3-2 up in the 91st minute, it’s hard to believe they didn’t win the game! For the record, it’s the 9th 3-3 draw this season for the systems with 8 of the teams throwing away a lead to draw. I hit two of these 3-3 draws in one weekend when both teams lost 3-0 leads!

Leicester lost a 94th minute goal to draw 3-3 away to Middlesboro and teams like Dagenham, Leyton Orient, Rushden were leading and somehow lost the game!

Overall, a nightmare weekend all round and it’s the worst day of the season (so far!)

As someone in the TFA forum said on Sunday, this is a marathon and not a sprint but I do feel like when you lose so many points in the first weekend of a month (and it’s the size of all of last month’s profits!), you do feel like it’s going to be a long month…..

Anyway, we’ll see what the remainder of this month brings. Last night wasn’t much better as I had Bournemouth on 3 systems and they lost 1-0 away from home, so it keeps the bad run going. I’ll update the midweek results later hopefully.

I’m conscious of the fact that I’ve promised a post on the ratings and how I find the bets. It’s partly written but I’ve been busy updating the weekend results when I’ve had a chance, so I’ve not had time to get back to it. I will post it when it’s done though.

Here’s the massacre from the weekend below :(