Betting markets.. The Art of Bookmaking

BigCityBanker

I never post here but im a regular on the Poker forum. Anyway just said id post an article here and also post some of the discussion that has originated from it over in the BB thread on the poker forum. The article is linked in my blog. Some ppl here may have views also which they may want to discuss.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If something seems too good to be true – then it probably is! If your first reaction to the thought of being a bookmaker is that it’s an easy old game then I suggest you pay close attention to this article. If you take a look at the odds offered and see that for any given market they are betting to a book percentage of say 150% and think to yourself that it’s money for old rope then the likelihood is that you misunderstand the fundamentals behind the profession.

Before we get too far ahead of ourselves it is important to set out some of the basics. The book percentage or book as it is colloquially known is defined by the Racing Post as the “sum of the quoted probabilities across all horses in a race”. The bookmakers assign each horse in the race a probability of winning, which is represented by the odds. To translate these odds into the book percentage you add 1 to the odds and divide it into 100, i.e. 3/1 contributes 25% to the book, 100/(3+1) = 25%, 4/1 contributes 20% to the book and so on.

In its simplest form the book percentage is designed to serve as an indicator of the level of bookmaker profitability on a given race. If a book is 125% then the bookmaker can expect to make a profit of 20% (25/125) on that race. However, the principle upon which this rests is that the bookmaker can lay each horse in the race for an amount proportional to the amount which they contribute to the overall book. Furthermore, the favourite-longshot bias also dramatically distorts the book percentage.

I will discuss the concepts mentioned above in more detail below but first let us consider the hypothetical situation where the bookmakers aim is not to make profit but rather to provide a service. Consider a five horse race where each horse is deemed to have an equal (20%) chance of winning. In a fair market each horse would be priced at 4/1. If the bookmaker were to accept a bet on the horses’ equal in value to their book percentage, in this case 20% so a bet of €20 per horse, at 4/1 he would have an evenly matched book. He would know that regardless of the outcome of the race he must pay €100 to the holder of the ticket on the winning horse. In this case the bookmaker made zero profit and the betting public as a group also made zero profit.

However, the situation outlined above is unfortunately a hypothetical one. In the real betting world one would expect that one of the horses would be 6/4, maybe another one of them 2/1 and the remaining three horses at 4/1. In this instance the book would be 133%. Again if the bookmaker accepted a bet equal in value to the respective horse’s contribution to the book percentage he would accept a bet of €40 on the 6/4 shot, €33.33 on the 2/1 shot and €20 on each of the 4/1 shots. The bookmaker has a maximum liability of €100 but has generated turnover of €133.33. Irrespective of the result of the race the bookmaker has generated profit of €33.33 on his turnover.

It’s easy right? I’m afraid not! As you will see from above the concept of book percentage is dependent on having an evenly matched book. The bookmaker is at the mercy of laying each horse for the amount which they respectively contribute to the overall book. This evenness of match i.e. the ability to lay horses equal to their respective book percentage is in truth never achieved. This results in a bookmaker having an uneven book and in this instance the reported book percentage is no longer a true reflection of the market and reliance upon it as an indicator of the market is ill advised.

Many people, not armed with an understanding of the fundamentals behind the concept of book percentage are too quick to jump to assumptions that a bookmaker is offering poor or even no value in a given market. They look at the reported book percentage and are very quick to use this as a stick with which they beat the bookmakers for offering punters no edge. From the above example it is easy to see that the book percentage is a weak measure. It is balanced upon the concept of evenness of match and it should be easy to understand that this is almost impossible to achieve.

Another reason why the book is not a true reflection of the market is due to the “favourite-longshot bias”. Bookmakers will tell you that they are unable to lay certain horses irrespective of their price – whether that be 20/1 or 100/1. With that in mind, when pricing a horse they err on the side of caution in case a shrewd stable decides to go for a gamble! As people who want to back an outsider are as likely to back it at 14/1 as they are 40/1 the bookmaker will serve to limit his potential liabilities on outsiders without unduly compromising his perspective turnover on a race and offer outsiders at odds shorter than what they perceive to be their true price. In this instance it is important to consider that the punter, armed with more information than the bookmaker, considers the price available to be fair and accepts the available price.

Let us again revisit the earlier example of the 5 horse race and see how the favourite-longshot bias affected the reported book percentage. We saw how the book was 133%; this included the three horses priced at 4/1. Let’s suggest that the real price of these horses should in fact be bigger, say 5/1, 7/1 and 12/1. This would generate a book of 110%.

With the above in mind a punter should be aware that the value prices are often to be found at the front end of the market as these horses odds are not unduly biased. This is why you will see the outsider’s trade much higher on Betfair than on-course. In general Betfair layers price horses and not a book so in order to get a taker at the offered price they need to offer odds which reflect more closely the true odds of the horse, in other words they need to offer unbiased odds.

Hopefully you are beginning to understand that there is far more to bookmaking than setting odds which appear unfavorable to the punter and sitting back to reap the rewards. Successful bookmaking requires skill and judgment. As mentioned above evenness of match is crucial to the use of book percentage as an indicator of bookmaker profitability. Not only will bookmakers fail to have an even book, they will very often have markets in which they have been unable to lay some of the horses.

We will again look at the rather closed example of the five horse race. Lets say that the bookmaker was only successful in finding customers to back three of the five horses. The 6/4, 2/1 and one of the 4/1 shots. Now if the bookmaker lays each horse evenly he has a liability of €100 on turnover of €93 – hardly an ideal situation and in practice more extreme situations are regularly found.

This market behavior is often displayed in large field non handicap races. Take a 25 runner maiden hurdle or a 20 runner flat maiden. If you examine the market in these races you will often see that only the front two or three horses in the market are “wanted” by the punters. This leaves the book in a sorry state and whether or not bookmaker profits on the race is as a result of their judgment call on which one they want to get in the race. Despite a huge reported book percentage of sometimes close to 200% in these races the bookmaker is often running a judgment book and fighting a hard battle against the punters.

To look at a recent example of book percentages lets examine the Nassau Stakes run at Goodwood in which Peeping Fawn defeated Mandesha and Light Shift. This race had eight runners and a reported book percentage of 113%. Market activity centered around the market principles and Oaks first and second of Peeping Fawn and Light Shift, and also on the excellent French filly Mandesha priced at 2/1, 3/1 and 7/2 respectively. Others in the race ranged from 6/1 to 80/1. The book of the first 3 was 76% and all were well supported. This looks like a lose-lose situation for the bookmakers before you even take account of the sneaky each way wagers on the placed runners. As mentioned above this is a real example of where bookmaker judgment is crucial to profit.

Hopefully this article will have served to whet your appetite for an understanding of the bookmaking profession. If nothing else it would pay in general for people to be less hasty in jumping towards conclusions of a lack of value in betting markets. Remember there are two sides to every story and to be able to tell the story properly you need to understand them both!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Here are some views from AJ on the topic..

ArmaniJeanss wrote:

I sense a hint of sarcasm, but I'll proceed nonetheless .

OK, in your last example, why doesn't the bookie increase the price of the 2 he hasn't laid to 20/1. OK some 'insider' or value hunter may come along and put €5 on them at that price, but then the bookie will have a fully laid book at 103% (not an ideal % but surely better than an incomplete 93% book).

Generally I think Irish bookies bet to appaling %s and use the same stuff in your blog as an excuse. However most times it is not a valid excuse. How come competitive 16 runner €50K handicaps will still have 160% books?

I've seen in at the dogs as well where the Thursday night yobbos are randomly putting their 20s and 50s on the dog they think looks the cutest or reminds them of the dog they had as a kid or something... this is a perfect situation for a bookie yet still they will have the nerve to go 5/2, 3/1 4times and a token 7/2 'outsider'. If the same market was transferred to Wimbledon or Newcastle dogs then the bookies would bet to 115%-120%, but in Ireland the bookies take us for mugs (I don't blame them for it if the punter is stupid enough by the way and they can get away with it).

As for the situation with 25+ runners in maidens and maiden hurdles with only the top2 fancied, they have these races in England as well, but the most the overround will ever be is about 150%, 200%+ is regular in Ireland.

Now I accept, as you've described, that in an individual race the overround% isn't an accurate guide, however long term over 1000s of races it will be an extremely accurate guide. Variance balance outs over a period of time, there will be plenty of 'skinner' results where an unbacked horse scoots up. The overround% for all types of races is far higher in Ireland than UK, therefore either UK bookies are too charitable or Irish bookies are complete rip-offs. Take your pick.

AJs,

P.S. In a token attempt to keep this on pokertopic, I got AA last night and it stood up.

#####################################

here are my responses........

ArmaniJeanss wrote:

I sense a hint of sarcasm, but I'll proceed nonetheless ..

no seriously - im open on this. I completed my dissertation in my MSc on this subject so im used to having holes picked in it by now.

ArmaniJeanss wrote:

OK, in your last example, why doesn't the bookie increase the price of the 2 he hasn't laid to 20/1. OK some 'insider' or value hunter may come along and put €5 on them at that price, but then the bookie will have a fully laid book at 103% (not an ideal % but surely better than an incomplete 93% book)..

The 5 horse race isnt a good way to look at this as I said it is a rather closed example. There was a 5 runner conditions hurdle in Tipp about 4 years ago with an OR of 104%. I use the 5 horse race example for illustration purposes only and I have further expanded on it on the example of the 25 runner maiden hurdles where in truth you couldnt lay 15 of the horses at 1000/1 if you tried. You see its all well and good to say that increase the two he hasnt laid to 20/1 and accept a fiver. Markets do not operate in that fashion in practice. If you want to get into economic theory on the topic im sure you could prove it would work just like you can prove the EMH. In short demand is not elastic to price fluctuations on the outsiders in general - in practice the €5 punters demand to back a horse at 12/1 is the same as his or her demand to back it at 20/1. In practice liquidity doesnt exist to support this market activity at the extreme of the book.

ArmaniJeanss wrote:

Generally I think Irish bookies bet to appaling %s and use the same stuff in your blog as an excuse. However most times it is not a valid excuse. How come competitive 16 runner €50K handicaps will still have 160% books?.

you see a punters definition of a competitive handicap differs greatly from a bookmkaers. Over a meaningful sample size OR/R was 3.3%*16 = 53% so that book is still about average. You still cant lay half the field in a 16 runner handicap. In short evenness of match can not be achieved - ever! This is also relevant to your first point on the topic above.

ArmaniJeanss wrote:

I've seen in at the dogs as well where the Thursday night yobbos are randomly putting their 20s and 50s on the dog they think looks the cutest or reminds them of the dog they had as a kid or something... this is a perfect situation for a bookie yet still they will have the nerve to go 5/2, 3/1 4times and a token 7/2 'outsider'. If the same market was transferred to Wimbledon or Newcastle dogs then the bookies would bet to 115%-120%, but in Ireland the bookies take us for mugs (I don't blame them for it if the punter is stupid enough by the way and they can get away with it).?.

This is a different kettle of fish and you really need to take this up with ted Hegarty tonight in Shelbourne park!! In simple terms I think this is easy to understand. 1. the big punters will still back their chosen dogs at the prevailing price armed with what I would classify as a more perfect information set than the bookmakers. 2. The yobbos will still back there dogs irrespective of the price - they are not price elastic. Market activity dictates that the bookmakers actions are acceptable in this environment. From a business perspective the bookmakers would be foolish to offer greater odds in a market which operates in this fashion.

ArmaniJeanss wrote:

As for the situation with 25+ runners in maidens and maiden hurdles with only the top2 fancied, they have these races in England as well, but the most the overround will ever be is about 150%, 200%+ is regular in Ireland.

liquidity at the outer reaches of the market in betfair allow this in the english markets but not in the irish ones. The favourite longshot bias is less prevelant their due to their ability to hedge (it is still prelevant though). The irish bookmakers do not have this ability. I would counter that 200%+ is regular in Ireland. Id have to try and pull my database (i lost it when i wiped my harddrive on my laptop but im certain i have it backed up somewhere - i just cant remember where) but 200% is rare and extreme. I can remember that the largest reported % in my database was 248% in a Navan maiden hurdle and id be pretty certain the bookied did their stones in that race.

ArmaniJeanss wrote:

Now I accept, as you've described, that in an individual race the overround% isn't an accurate guide, however long term over 1000s of races it will be an extremely accurate guide. Variance balance outs over a period of time, there will be plenty of 'skinner' results where an unbacked horse scoots up. The overround% for all types of races is far higher in Ireland than UK, therefore either UK bookies are too charitable or Irish bookies are complete rip-offs.

This anomoly as you have described is explained in my paragraph above - it is largely due to their ability to hedge.

Something which is crucially important but which i havnt got into in the article is liquidity. On course liquidity which ultimately drives the market is negligible. The interaction between highstreets and oncourse is such that the large % of the liquidity remains off course and has no impact or little impact on market direction. This again is an anomoly in market theory.

ArmaniJeanss wrote:

P.S. In a token attempt to keep this on pokertopic, I got AA last night and it stood up.

well done.

masterK

You are forgetting one critical factor in modern bookmaking, the betting exchange. There isn't a single bookmaker today that doesn't limit his liabilities by offloading his bets on an exchange, in your Peeping Fawn example a bookmaker have been backing all 3 market principals on the exchanges to ensure he profited from the race.

BigCityBanker

masterK wrote:

You are forgetting one critical factor in modern bookmaking, the betting exchange. There isn't a single bookmaker today that doesn't limit his liabilities by offloading his bets on an exchange, in your Peeping Fawn example a bookmaker have been backing all 3 market principals on the exchanges to ensure he profited from the race.

i fail to see your point here really in the context of the op and the article in general.

masterK

My point being that bookmakers are now no more than facilitators of bets, they take a bet and hedge it off on the exchange to guarantee themselves a profit. There is very little art in bookmaking any more, almost every bookie on course will be hooked up to Betdaq. It is fairly difficult for any bookmaker who has decent collateral not make a profit, the technology available to them ensures this.

BigCityBanker

masterK wrote:

My point being that bookmakers are now no more than facilitators of bets, they take a bet and hedge it off on the exchange to guarantee themselves a profit. There is very little art in bookmaking any more, almost every bookie on course will be hooked up to Betdaq. It is fairly difficult for any bookmaker who has decent collateral not make a profit, the technology available to them ensures this.

this as a sweeping statement is incorrect. inklings of it are valid. generalisations like this whether they be to betting markets or any other market are for the most part invalid. I think you are failing to consider that hedging existed in the days pre exchanges and also the relative liquidity issues. fwiw the follow up article to this is an interview with a prominent layer whom i hope will open up the discussion on this far more.

Colonel Sanders

Some of the points are very valid, some aren't IMO.

I agree with masterk, there are very few 'bookmakers' left. Most are just glorified traders (especially in england). About the only bookmaker i have respect for as a man who lays to his opinionsand is truely fearless is David Hyland. Most others are:

1. The type hedging onto betfair
2. The 'major' rails bookmakers who's prices are generally even skinnier than those betting to already scandalous percentages.
3. The jokers who, in the words of Barney Curley, wouldn't lay 2 bananas to a banana.

Bookmakers also have other advantages such as:
1. Being able to manipulate markets in their favour. The example a few months ago at Lingfield (can't remember the horses name) where the favourite was backed late from 2/1 into even money by off course bookies in an attempt to rip off off course punters (this was not a simple hedging exercise, there were far bigger prices available on the exchanges if it was just plain hedging). I cannot think of another industry in the world where this type of market manipulation would be tolerated.
2. In the case of off course bookmakers the company that supplies 'live show' prices (SIS) is owned by bookmakers, a conflict of interests which, again, would not be tolerated in most other industries. i have commented on this forum before about examples where a price was freely available on course but never relayed back to the betting shops.
3. Limiting people to what they can/cannot have on a horse. I am not talking about 6 figure sums, i am talking about the likes of boyles who'd be lucky to let you have 25 quid ew on a horse at a double figure morning price.
4. Related to the above not having to advertise off course what they will lay a hore to lose. This legislation is supposedly in force on course in Ireland but not enforced.
5. Some bookmakers (fcuk it I'll just say it, boyles) idea of odds compiling is to punch up page 297 on aertel.

So before you give the 'poor auld bookmakers' line consider the above.

Slightly off topic (when i begin to give out about bookmakers i find it hard to stop) bookmakers also give the poor mouth when referring to exchages. They whinge about them taking business but not making the same contribution to racing. They conveniently forget that the reason the exchange flourished was that big punters couldn't get the amounts of money they wanted on with bookmakers. And what have bookmakers done to address this? Zilch, they still have poor morning prices which they won't lay to good money. The amount of times I hear that clown Leon Blanche on telly saying Boyles cut a horse 'after the big hitters were in early' is laughable. I once backed a horse ante post at the cheltenham festival (the biggest meeting of the year remember) to win the princely sum of just over EUR800. True to form Good old Leon was on ATR a few minutes later saying they'd cut the said horse from 7s into 5s citing 'massive support' in the office I'd just backed it in. I was in that same office for 15-20 minutes either side of placing that bet and no one else placea bet on that horse. Laying to lose 800 quid is massive? Play the if in life indeed.

And in any case even though they complain about the big bad exchanges they use them both for hedging and as an 'early warning system'. Hypocrisy.

The days when bookmaking was an art are long gone never to return.

Rant over, sorry!

cson

From working in a betting shop my view of 'bookmaking' in its modern form is thus; The major chains, Boyles, Lads, Powers etc really do screw over your average punter who'd bet in say 10s and 20s even 50's. They don't have any time for punters backing large or on doubles (**** it, I had a €10 double on a 14/1 and 8/1 and the girl working in Boyles had to call their head office before she would accept it!). A lad I know has an account with one of the firms and having a fairly decent record with them, they always give him reduced odds, say a shop price of 16/1 would be 12/1 to him. The only thing I have in favour of the big firms is that they really lay it on for customers in terms of comfort. trying to keep them in the place.

Bookmaking as an art is dead and gone. Most of your prices are done by the same lads, maybe 6/7 different companies doing up prices which are eerily similar. I wasn't around but I'd imagine back in the day a true bookmaker would do his own prices himself.

I'd echo what colonel sanders said about that horse in Lingfield. Last week in Galway we had a 20/1 horse in the morning backed into a 7/2 SP iirc, (Black Beauty). Came absolutely no-where for a huge gamble like that. All I'll say is that a certain K. Fallon was riding him.

Basically, if I was a Bookie I'd want a guy coming into me with say €300 and leaving with €200. A guy like that will sustain those losses in the hope someday he'll come out with €3000 (and he will sometimes). Imo staff also play a huge part, nice, friendly staff who have a bit of banter with the punters I think definitely help.

Robin1982

For those who didn't know how a betting market works or had never thought about it, this is a well-written, useful article. I think many educated punters would already know much of what you are saying but nonetheless...

Just a couple of things I would like to ask/comment on:

The interaction between highstreets and oncourse is such that the large % of the liquidity remains off course and has no impact or little impact on market direction.

I wouldn't neccessarily agree with that completely. On the average race perhaps the shops would have no interest in instructing money to the on-course market, however there are situations where it takes place.

Many, many punters back at SP. Popular tipsters' selection will often receive significant support from this group, and what Col. Sanders says is correct; it becomes very easy for the shops to manipulate the on-course market quite cheaply. I remember once where some shop regular had a £2 accumulator (or similar) on (at the time of the bet) 4 large price selections at a minor meet, all to SP. His first 3 selections all won. Come the final race, his selection opened at 33/1. The bookmaker shop that took his bet "flooded" the on-course market (quite cheaply) and backed his selection to limit their liabilities. The SP was 2/1. And the horse won. The punter was delighted with his large payout, but in fact, the shop had only to outlay £10k to limit their liabilities by about £400k.

But, true, thats a rare occurance.

As regards, the favourite-longshot bias, what you say is largely true. Most of the written material about this has been based on tote markets, so its not completely directly applicable to multi-participant (bookmakers) fixed-odds markets. In fact, the conclusions over there say that favourites are generally underbet versus their actual success rate while longshots are overbet. The distributed nature of the markets over here means a similar study can't be carried out.

For bookmakers here, the "general" rule is if a favourite wins, they lose money. When Paddy Power's results fell below analyst expectations a few years ago, the main excuse was the above average success of favourites at Cheltenham that year.

Is bookmaking an "art"? I think that my be stretching it. Its risk management mostly, and although markets are far from information efficient, some seem a bit too risk averse. Even when bookmakers "take on" a favourite they believe to be false, I don't think occurs too often and not to a suicidal degree. The market, generally, is smart. By managing the liabilities intelligently, hedging when available and adjusting prices to encourage/discourage punters, I think a suitable medium can be found - as long as one survives long enough to gain the experience. That takes skill.

nice1franko

Interesting article but, as has been said already, trading has replaced a large part of "the art" of bookmaking.

After opening his 100th shop, John Boyle, founder of Boyle Sports, said in a candid interview that they use exchanges.

The bookmakers at the track are no different: they have wifi access and they're not in it for the "art".

So, this part is incorrect:

BCB wrote:

In general Betfair layers price horses and not a book so in order to get a taker at the offered price they need to offer odds which reflect more closely the true odds of the horse, in other words they need to offer unbiased odds.

Good poker players use poker tracker. Good bookies use exchanges.