Wry? Always. Me

Why Always Me (henceforth WAM) came back with a couple of comments on my last post. They are rather lengthy, so I have broken them down (and corrected poor spelling for clarity) and added comments at each point.

Thanks for the reply Cassini, I did have a wry chuckle at your clever headline too!

I am an amusing man. People are always laughing at me.

OK first things first, you were correct to say my 33% to 50% example is totally unrealistic. Think I must have been using one of those 'inflation tools' that certain system sellers market. But to give an example from real life, it was possible to back Huddersfield on Sunday at 3.80 and they kicked off around 3.10 on Betfair which works out at 22.55%. Now I know this sort of movement is not very common but a 10% edge can be found a few times a week. I do however accept it won't be much more than this, to get more turnover you have to look at other sports too I think.

I am usually correct - you could have saved your keyboard. The example of Huddersfield Town on Sunday needs a little clarification. Was the 3.8 available at the same time as the Betfair price was 3.1 or did the price shorten on Betfair during the day from 3.8 to 3.1? The distinction is critical. If it was the former, then this was an outstanding arbitrage opportunity, and you need to make the most of it because you will soon find yourself limited or shut down completely by such a generous book. If it was the latter, then unless you had the ‘true’ price at 3.1 at the time you backed at 3.8, then you didn’t have an edge. The ‘true’ price at that time was 3.8, and you got lucky. The price could just have easily have moved against you, and it’s important you understand that a fortuitous price move is not an edge.

Not quite sure what the comment on increasing turnover is about to be honest, other than stating the obvious.

That post you did on working out your edge was really important and very helpful - it really gets to the heart of the matter.

Always happy to help. There's a reason why this is the world's number one sports trading blog. The blog that gives.

Turning to your draw systems, I believe you previously said your average advantage was about 3.25/3.11 which I make to be just under 5%. Is this a direct comparison with the Betfair kick-off price or a theoretical estimate i.e. Betfair is wrong on the particular match and your 3.11 is the 'true' odds?

What I am getting at is how do you measure your system's performance? I see the two draw systems are up about 10 units on 300 selections against a theoretical 15. However draws could have been more prevalent this season (the Premiership is 30% for example) or you could have had more winners than might be expected.

These numbers continue to change each week as I add more and more matches to the spreadsheet. As of last weekend, there were 1,525 matches (out of 4,770) that qualified as an XX Draw (Classic or Extended) selection, of which 474 finished as draws. The average draw price for these selections is currently (this season) 3.55 (the price is not necessarily that at kick-off, since I record the price I am matched at and have better things to do that monitor prices at kick-off) and the average ‘true’ price is 3.22. The average edge is thus 10.2%. Now this is not to claim that every match selected offers this edge, only that from a reasonably large number of matches with a specific profile, more than expected finish as draws. Results are measured using ROI%, which allows for a direct comparison with other systems and selections.

Obviously there are differences, sometimes significant, between leagues, and some seasons will see a higher than average overall number of draws, while others will be lower - it’s kind of how averages work!

Some examples of league differences - the Classic selections in England are at an Implied Price of 3.08 while in Italy they are 4.13, but Italy's Extended selections are at 3.08 with Spain the worst at 3.21. 

WAM points out that there are more draws than usual in the EPL this season, but a few years ago the draws were under represented. These runs do not impact the average draw prices as I have written before. They would eventually were the runs to continue, but the draw frequency is remarkably steady.

Cassini when I look for a rational argument for backing the draw from you I can't find any. So far all you have said is it's 'easy to back' and 'unfashionable'. To me this sounds like just the sort of guff you call other system sellers and bloggers out for. It's really disappointing.

If you think for a minute about how markets are formed your 'unfashionable' argument collapses. 1X2 markets are formed from Asian Handicaps, which have the draw in them, so the draw is not in fact unfashionable at all. Also if you are saying the draw price is out 5% that means the 0-0, 1-1 and 2-2 are out too, as well as the U1.5.

I am not suggesting that it is rational to back any draw, only that the draw in historically similarly profiled matches have hit at an implied price of 3.22 and that the price typically available on the draw represents value. If anything is disappointing, it is that people can’t seem to grasp this very simple point. You misquote me on the ‘easy to back’ – what I actually wrote was “If you find it easier to find this edge on the Home team or the Away team, then that’s great, but I’m not sure it’s consistently that easy”, although backing the draw does happen to be one of the easier bets. Finding value in football is seldom ‘easy’, but with the draw rarely below 3.2, I find it easy-er to find value here consistently than on other football bets.

For the 'unfashionable' reference, I’ll defer to Derek McGovern who knows far more about betting than I ever will. He wrote, and I quote:

“ in a match in which theoretically all three potential outcomes - home win, away win, and draw - are equally likely, the true odds should be 2-1, 2-1 and 2-1. But such odds give bookmakers no profit margin, so instead they will offer, say, 6-4 the home team, 6-4 the away team, and 11-5 the draw (the draw being the longest price because bookies know that few punters ever back it, something that will be explained later).“

and also

"Punters are creatures of habit. When they have a bet on a football match they are seeking an allegiance. They do not want to see a match finish in a draw. Consequently few punters ever back the draw, a point I will pick up on later in the book". 

Why WAM thinks the 12X markets are derivatives of the Asian Handicap markets is puzzling. Presumably he has some evidence to back up this claim? I would certainly be interested to see this study.

Being a cynic, what's to stop me using the same sales pitch and setting up my own draw system? Hey the draw is easy to back and I'm a counter-trender. All you have to do it seems is pick the biggest draw percentage games, with a few bigger prices thrown in for luck, get out a prayer mat and hope for some positive variance for a couple of seasons! Charge £20pm per client - TYVM.

There is absolutely nothing to stop you from doing this, but good luck with it because you are again failing to understand that you really need a proven edge to attract discerning investors. Someone selling ‘guff’ like this, as you put it, is certainly going to be highlighted on this blog.

The above shows the total number of matches across all leagues. The percentage that end in draws is 26%, an implied price is 3.79. The results of the XX Draws across all leagues show 474 winners from 1525, implied price of 3.22. You need to do more than select draws at random.

Blindly backing the biggest draw percentage games will not be profitable in the long term, and throwing in a few bigger prices for luck will just add to your losses. Praying has never worked since humans first roamed the planet (more than 6,000 years ago in case you didn’t know), and never will work, although it is said that the secret to successful prayer is to pray for something that will happen anyway. Positive variance might go your way for a while, but probably not over four seasons and 4,770 selections, and a proven track record is something most potential subscribers will look for. Feel free to send your selections in to this blog for inclusion in the FTL table.

Cassini the reason I am having a pop like this is if you go to FB Elite or wherever it's basically done on trust. The guy is saying to you 'I have a private method of gaining an edge, trust me here is my record'. You either buy into it or you don't. But what you appear to be saying is 'buy my system because the markets are wrong', yet I've never seen any evidence from you to back this statement up. Which is irritating.

Pop away. I enjoy the discussion, and it makes blogging so much easier, and you can compare me to Football Elite any day. Football Elite was the first and, so far, only football service I subscribed to, and before doing so I looked at Matt’s track record. His results looked good, and I was interested to see his selections and understand some of the reasoning behind them. Unfortunately, results have been disappointing over the last couple of seasons, which may be a temporary blip, or it may be that the edge is eroded.

Matt is a professional punter and relies on selling his service to generate at least some of his income. I am not, and I am not selling a system. I do have selections available to anyone who is interested in them, but the spreadsheet that contains the data that generates these selections is not for sale. Actually it probably is, everything has a price, but unlike the selections, it wouldn’t be cheap.

As for the reason being ‘because the markets are wrong’, all I can say is “oh dear”. No service would ever be profitable if the markets weren’t wrong! This is how we find an edge - we need the markets to be wrong! 

WAM has much to learn.

He states here that he has never seen any evidence, (it was there all the time – irritating indeed - but he appears to have found it by the time he makes his follow-up comment, even if doesn't understand it all:

A couple more things.

I've been over the Prospectus and it shows an estimated profit of some 71 units for a previous period. But what I noticed is that a lot of profit came from U2.5, which if you are backing the draw is a correlated bet. Really you are double counting, if the draw comes in the U2.5 will come in most of the time.

The prospectus makes it quite clear that the Match Odds draw bet and the Under 2.5 goals bet are correlated, and explains why. It also breaks down the results of backing both bets individually, so you can single count or double count as you wish. It’s a matter of personal choice how anyone plays the selections.

Also your table is quite misleading. You occupy the first three places but your system as a whole is in loss! Either you are competing or you're not, how can it be right to split them up in this way? If you want to split them you could do it outside of the table.

I think my table is quite clear. I track five bets for each of the three categories (Classic, Extended, Bundesliga Extended). In addition to the Match Odds Draw and the Under 2.5 goals markets, I have this season included the Under 1.5 goals, Under 3.5 goals and Half-Time 0-0 markets. My thinking was that these should also prove to be profitable propositions, but as you can see from the current table, the results from these are all over the place. Backing the Under 1.5 market for the Extended selections shows a loss of 30.16 points while the same market on the Classic selections is up 6.52. For the HT00 bets, the opposite is seen - the Classic selections are down 11.92 while the Extended are up 8.11. 

It would be ‘quite misleading’ were I to combine these five bets together, and unrealistic since I doubt that anyone backs all five markets on each selection. How can drilling down and reporting on individual components be misleading? Individual bets should be reported individually so it is clear where the profits or the losses are happening. Obfuscation would be the result of combining bets, not the opaqueness we are looking for.

I suspect that it’ll be a while before these new market numbers settle down, but the lack of liquidity of some of these markets could well prove to be the death knell for these more fringe bets. I was actually surprised at how low the volume was on these markets.

My draw selections should be judged primarily on just that - the draw results. The other results are broken out so that we can all see how they perform. There is no requirement for subscribers to back five markets on every selection, but thank you for suggesting how I should run my table!

The prospectus makes it very clear that backing draws is not going to make you rich in a hurry. It generates a pretty good number of bets with an overall profitable ROI%, and is ideal for those who want to generate a little extra cash or churn some money to reduce the Betfair Premium Charge or the potential Premium Charge. I know a few people who use the selections for just this purpose.

The prospectus also makes it clear that there is no guarantee that profitability will continue, but with the average price on the draw unlikely to shorten, then unless there is a fundamental change in the way that matches of the profile my spreadsheet identifies play out, then I remain cautiously optimistic.

One problem with a profitable selection method that becomes well known is that the value can soon disappear. It’s a problem that Matt of Football Elite is all too aware of. He recently explained in an email that he would not reveal all the reasons for his selections, but he may have revealed too much information already.

I couldn’t do it for ‘my’ selections if I tried. They are my spreadsheet’s selections, and I have no idea when I enter the teams in each week what numbers will come flying out the other end.

Here is the FTL table for this week, a little later than usual, but I’ve been busy writing this blog post! No significant movers this week, (a very limited EPL schedule didn't help) - the only one moving up or down by more than three places was, talk of the devil, Football Elite, who dropped four spots to 24th.