Ignore the market and create your own tissue

Betting Strategy / Jack Houghton / 23 July 2008 / 8 Comments

Jack Houghton makes a compelling study of the intricacies of tissue price.

Last week, in response to John's question about selection methods, I banged on about tissue prices an awful lot. To some, it might have seemed as if the question wasn't answered, but the transition from unsuccessful to successful punter is, for most, the transition from picking winners to picking prices. If you're always betting at odds above those you should be getting, the winners will come; as will long-term profits.

How to arrive at an accurate tissue price is a bigger question (and probably what John was interested in), with a variety of potential answers. Early in the days of betting.betfair, Simon Rowlands wrote an excellent piece on it that is well worth a read. But whereas Simon talks more generally about the process, I'll try and focus on some more specific methods.

Broadly speaking, methods can be categorised as quantitative (using lots of data) or qualitative (selectively using data, based on personal expertise); although, in reality, most punters would use a combination of both approaches to arrive at their tissue.

A quantitative, or data-led approach, might be to use handicap or speed ratings. There are a number of freely- or commercially-available ratings, or you can create your own (a topic too large to cover here, but if anyone's interested, post a comment and I can recommend books on the subject). Timeform ratings for a recent four-runner race at Haydock provide a perfect example of how they can be used (Betfair SP is used to be indicative of the price available on the exchange before the off):

I use a complicated sliding scale, developed over a long period of time, that gives the best rated horse a score of 100 and a relative score for each horse below that. This allows me to, more easily, convert the bare rating into a tissue price:

In this example, it shows that We'll Come is the obvious value bet, being available at more than double the price of the equally-rated Welsh Emperor. But then you don't need a complex rating conversion system to tell you this. By seeing they share an equal rating, it is clear they should be the same price. If you're following a purely data-driven approach that is.

Because this race in particular demonstrates the weakness of blindly believing the numbers. The ground was officially described as heavy for a start: conditions Welsh Emperor had shown more aptitude for than his competitors. The favourite had also performed well historically in small-field races where he was able to dominate. With no obvious other front-runners, Welsh Emperor was likely to get an uncontested lead.

These factors, combined with some doubts about We'll Come's fortitude, would lead those taking a more qualitative approach to shorten Welsh Emperor's price and lengthen We'll Come's. In the end, despite these factors, I backed We'll Come. I still thought he represented value. The result? Welsh Emperor got an easy lead and We'll Come wasn't able to throw down a challenge.

Qualitative or quantitative, tissue prices or not, plenty of bets will still lose. But providing your data is robust and you are always backing the value selection, you should win in the long-run.

So if, up to now, you've tended to approach a race by trying to determine the most likely winner, why not try to create a tissue price instead? Ignore the betting market; only looking at the available odds when you have your own idea of the price a horse should be. Then, back the horse with the biggest price discrepancy in your favour.

It might take a while to be confident in the tissue prices you create, but as Simon says in his article: "well nigh every successful punter I have met prices up events at least part of the time and appreciates the concept of value." I couldn't put it better myself. And haven't.

Comments (8)

1. Paul | 23 July 2008

   The logic you state that "If you're always betting at odds above those you should be getting, the winners will come; as will long-term profits". is fundamentally flawed.

   Who determines what odds you should be getting and why are they always correct?

   How you determine the price is also subjective.

   So what you are saying is that if someone subjectively prices up a market incorrectly and you can also subjectively identify those which are priced to high that you will end up with long-term profits.

   The days of tissue pricing is numbered especially when they have absolutely no relevance on a betting exchange.

   There is no substitute for being able to find winners rather than being the master of a subjective exercise in odds determination.

2. Andrew Hughes | 24 July 2008

   I disagree, Paul

   Finding winners is easy. Back the favourite in every race and you'll get plenty of winners.

   Compiling a tissue is just an extension of what every punter does. Very few people would back their selection REGARDLESS of the price and once you start considering the price, then you are into the territory of value, whether you call it that or not.

   Betting is always going to be a subjective exercise, you are pitting your judgement against others and I've found compiling a tissue helps me enormously.

   I thought it was a good article

3. JP | 26 July 2008

   Paul,

   The statement isnt flawed but it isnt "complete" so to speak.

   The fundamentals are correct but it still hinges on how good a judgement the odds/tissue compiler is at this exercise.

   This is just like "seeking the winner" approach punters having to have their judgement good enough to yield long term profits too.

   Someone pricing up a race badly will lose just like someone who's bad at trying to pick a winner
   will lose.

   Although the statement in my opinion isnt complete, the author is trying to get the messgae across "...that based on the assumption that you are good enough at pricing a race/event, the winners will look after themselves in the long term." This is fundamentally correct.
   

4. Jack Houghton | 28 July 2008

   Paul

   Many thanks for your comment and apologies for the tardy response: I've been away since last Wednesday best-manning at a wedding and am only now back in front of a computer.

   I disagree with your comments, obviously, and I think the counter-arguments I would make have been adequately covered by others here. I would also refer you to Simon Rowlands two articles on the subject; including his most recent, which pretty much directly addresses your points, and much better than I am able to!

   But one final point I would make is this. During my career I've been lucky enough to talk to, and work closely with, a number of professional, and winning, punters. And I've met and spoken with even more losing punters. Although they all have different processes and methods by which they reach an assessment of value, the winners, absolutely universally, have all used a "value" approach. There are plenty of losers who use this approach as well, so you're right, you still have to be good for the approach to work; but just because there are losers using the value approach, it doesn't invalidate the approach itself.

   One thing is for sure though, there are thousands more losers who are only interested in finding winners/losers, with seemingly no regard for price. One of my huge frustrations with newspaper tipsters is that they rarely back up their selections with a recommended price. Are they seriously suggesting their tip would still be a wise bet if it were 1.01? This to me epitomises the whole value debate: everything has a price, in any market, where it represents value.

   What I was attempting to show in the two-part article was one particular approach to assessing value: by using ratings. I certainly don't claim it to be the only way, but I thought it worked as an illustrative example for those who may want to improve their approach to betting but weren't sure how to do so.

5. Simon Rowlands | 02 August 2008

   "...you can create your own [ratings](a topic too large to cover here, but if anyone's interested, post a comment and I can recommend books on the subject)..."

   If no-one else will ask, then I will: what books are these?

   I know of not one book that deals with compiling form ratings credibly and of only a few that deal with compiling time ratings, and they are flawed to varying degrees.

   But I may be missing something.

6. Jack Houghton | 03 August 2008

   You're right Simon - to some extent - but I think my use of the word "recommend" is the issue. When I used it, I really meant I could point people (thinking about this for the first time) in the direction of a couple of things that would get them started compiling their own ratings.

   The books that got me going - in the period 1999-2002 - were Beyer on Speed and Nick Mordin's two books: Betting For A Living and Winning Without Thinking.

   I can already hear gasps of disbelief - especially at the Mordin books (I know he's ridiculed in some circles) - but, although full of imperfections, they certainly gave me a starting point that explained the process of creating ratings that I was able to change and develop as I played with them more and realised their flaws.

   Much of that development also came from asking stupid questions of people who'd done it longer than me. People like you Simon! (by the way, thanks for answering my latest question on the subject) And of course, I was lucky enough to be surrounded by some people at Betfair - the likes of Steve High - who pushed me in the right direction.

   So, to better write the sentence you quote: "if anyone's interested, post a comment and I can give people the titles of a couple of books that cover the subject to some degree that will get them started on the way to creating their own ratings."

   Be warned though... My own experience is that it took about 6-7 years, with lots of false starts and lots of money lost, before I was able to create and effectively use ratings to turn a regular profit. Much of that was to do with ill-discipline in their use, and brighter folk would no doubt get there quicker, but it's certainly not an easy process.

7. John Walker | 11 October 2008

   This concept of 'Value' being the MOST important thing is a bit suspect.

   Agreed, 'getting value' is important and if I find a runner (say at 6/1) with a chance of winning, then if I believe its price should be shorter, I have value. I have also found a value-horse that is likely to win; or at least place.

   There's great satisfaction in backing an 6/1 winner, when you know the horse should have been a 4/1 shoot to begin with.
   This is where your own tissue is invaluable. You don't back a horse just because your tissue says it is overpriced. You back it if you know it also has a genuine chance of winning.

   Conversely, you can back a horse at 16/1 and watch steam in; rubbing your hands in glee. Only to see it get beaten at say 8/1. Where's the value in that?

   Yes, value is important, but it is value ONLY when the horse has a realistic chance of winning. That's how you make your profit. Backing such horses allows you to win more for lower stakes, but the horse doesn't win because of its price. It wins because it is good enough and you have to read a bit of form to find that out.

   That's how I see it.

   regards
   John

8. JP | 12 October 2008

   John,

   The realistic chances of a selection winning are part of the tissue compiling process. So, if you calculate/decide that a horse "should" be 33/1 but is on offer at 100/1, then the odds you are receiving imply worse chances then you believe the selection has and therefore, value is obtained (assuming your judgment is correct of course.)

   Theres three things (other then obtaining value on each bet) that you have touched upon here. All are very real but none of them (on their own or in any combination) actually invalidate (or damage) the importance of obtaining value.

   1) Value betting should be viewed over the long term. It is long term profit that the value bettor is seeking, not short term (i.e. the next race) vindication/satisfaction of getting a winner (even though that is obviously enjoyable!) This is where the "winners look after themselves" bit comes in. In fact, a true value bettor who is good enough, wont/shouldnt even care what happens to his bet after its placed as its part of the overall approach.

   2) Personal tolerance. We're all human and we all have our own limits. Mathematially speaking, when it comes to value betting, these limits are inefficiencies but it doesnt make the tolerance problem an unreal one. For whatever reason, some people will not like (or tolerate) placing a bet on a selection that has a relatively low chance of winning (pyschologically, this ties in to the short term satisfaction "trap") even if they believe the odds do actually represent value. This is tricks of the mind and makes up part of the discipline required.

   3) Money Management. With a long term approach assumed, the value bettor has one more important thing then actually making a profit - making sure he doesnt lose the lot so he cant bet anymore! Its all very well getting 5/2 on something that "should" be 6/4 but theres not point putting all your money on it just because its value as your own tissue is implying its 60% likely to lose! Again, money management kind of ties-up and interlinks with points 1 and 2 above.

   As far as I know, this is the overall approach that needs to be cracked. No good being the best money manager if you cant judge value in the first place. No good taking a long term view of your bets if you cant manage your money, and so on. The thing is, none of this lessens the importance of making your bets only when the odds imply a value chance compared to your own prediction.