Do Bookies Need to Balance Their Bets?

Balancing moneyThe question of whether bookies want to, seek to, or even need to balance their bets, or balance their book, is an interesting one which gets to the heart of how bookmakers look to make money and, on the other hand, how punters should try to.

In this article, we will start at the very beginning and explain what we mean by balancing bets (or books), and we will also look at how betting sites might try and achieve this and, indeed, if they do try to. We’ll also explore the concept of taking a position and look at some famous examples of where a bookie has done exactly that.

What Is a Balanced Book?

Euro bank notes

To understand what is meant by a bookie balancing their bets or keeping a balanced book, we need to understand more about how a bookie makes money. Obviously us punters do, at least occasionally, land winners. A bookmaker cannot simply hope that nobody bets on the right horse, football team, golfer or anything else. So, instead, what they do is offer odds that are lower than they really ought to be.

£10 Bet at Odds of Evens Example

If, for example, two football teams, supported by two friends, are in a cup final and are believed to be absolutely evenly matched, both have a 50% chance of lifting the cup (succeeding overall, rather than winning the game in 90 minutes). If you were betting with your mate, you might “have a tenner on the game”, by which it would be understood that whoever’s side wins, the other would give them £10. This is essentially a £10 bet at odds of evens, struck by one friend with another. Evens are the fair and proper odds for an outcome with a 50% chance of success, which means you could, if such odds were offered by a bookmaker, bet £10 on one team and £10 on the other and break even no matter what happened.

Bookies Add in a Margin

However, if you tried to such a wager with a bookmaker you would find that the odds were not evens. Both teams might well be priced at the same odds but rather than being the fair price of 1/1, or 2.0 in decimal odds, you might be able to back both at somewhere between 10/11 and 4/5. This essentially creates a margin for the bookmaker, because now if you backed £10 on both sides you would not break even, but rather you would lose.

Indeed, this situation, where a bookmaker has, for example, taken £10 of bets on one side at 4/5 and £10 on the other at the same odds, is a balanced book. Their liability is the same no matter who wins and it is less than the total stakes they have received, meaning that no matter what happens in the cup final, they end up making money.

By offering lower odds than they “should” the bookie creates a margin for themselves, this is sometimes referred to as the house edge, overround, vigorish (shortened to vig), or bookmaker’s margin. Technically, some of these terms mean slightly different things but they are often used interchangeably.

How the Overround is Calculated

The overround is calculated by adding together the implied probabilities of the odds that make up the market. To calculate the implied probability you divide 1 by the decimal odds. Returning to our cup final, if both sides are priced at evens then both have an implied probability of 50% (evens = 2.0 and 1 divided by 2 = 0.5, or 50%). You add these together to give the book figure.

If the book is over 100% then the bookie has an overround. If we imagine a betting site offering 4/5 on our two finalists, we would see that:

  1. 4/5 = 1.8 in decimal odds
  2. 1 divided by 1.8 = 0.56 = 56%
  3. 56% + 56% = 112%
  4. Overround is 12% (or 11.11% depending on rounding)

Book Doesn’t Need to Be Perfectly Balanced

In reality, a bookmaker is unlikely to ever achieve a perfectly balanced book, where no matter who wins they make exactly the same amount. On even relatively small markets they accept thousands of pounds of bets, whilst on the largest ones they may accept seven- and even eight-figure sums. It is simply unrealistic to expect them to be able to have those wagers perfectly spread in direct proportion to their odds but they do not need this to be the case because their overround is generally big enough to allow for some imperfections.

Let us return, yet again, to an evenly matched cup final with both sides priced at odds of 4/5 to lift the trophy. For the sake of simplicity, let us then say that the bookie accepts exactly £100 of bets on the market and consider what happens based on four different splits of that cash.

Amount Wagered
on Team 1
Amount Wagered
on Team 2
Winner Bookie Outcome Overall Result
£50 £50 Team 1 Pays out £40 on Team 1
Wins £50 on Team 2
£50 £50 Team 2 Pays out £40 on Team 2
Wins £50 on Team 1
£52 £48 Team 1 Pays out £41.60 on Team 1
Wins £48 on Team 2
£52 £48 Team 2 Pays out £38.40 on Team 2
Wins £52 on Team 1
£55 £45 Team 1 Pays out £44 on Team 1
Wins £45 on Team 2
£55 £45 Team 2 Pays out £36 on Team 2
Wins £55 on Team 1
£60 £40 Team 1 Pays out £48 on Team 1
Wins £40 on Team 2
£60 £40 Team 2 Pays out £32 on Team 2
Wins £60 on Team 2

If you are among the small number of people to be excited by maths there are a number of interesting things to note here.

The second thing of note is that, no matter how much is wagered on either side, the average outcome is the same. If the cup final was played twice under exactly the same conditions with exactly the same odds and wagers, and Team 1 and Team 2 both won one each, the bookmaker would make a total gain of £20 each time.

In other words, their average win for the game is £10. This figure is created because at 4/5 they will pay out £40 per £50 of winning bets, rather than the £50 they would pay if the “fair” odds of evens were used. This means that if they have faith that their assessment that the game is 50/50 is correct, it really doesn’t make any difference how much is wagered on either side. If all £100 was bet on Team 1, they stand to either make £100 if Team 2 wins, or lose £80 if not.

If that 50/50 assessment of the probabilities is correct then on average the bookie will come out on top in the long term. Whilst they may lose on some markets, the wins, which should come with equal frequency, will be bigger. Indeed, because their margin is of a decent size, even if that 50% judgement regarding the probability of each side triumphing, they still have room for manoeuvre.

Indeed, if the correct analysis of the probability was 55% versus 45%, the bookie would still expect to come out on top. If they lost £80 on 55 occasions but won £100 on 45, they would still end up in the black to the tune of £100. So, as we can see, bookies definitely do not need to balance their books entirely. That said, doing so, or at least getting close, means they have a smaller risk of bad results affecting them.

Life and sport are unpredictable and even a 1/10 favourite will get beaten from time to time. Equally, even if a bookie assesses teams as being equally matched, variance can mean that rather than results going one way around 50% of the time and the other way roughly the same number of times, they might go against them 60%, 70% or even more.

Such runs might only impact a bookie in the short or medium term but even so, they will lead to losses and this is a situation that, by and large, the bookie would rather avoid. What’s more, the combination of an unbalanced book and an incorrect assessment of the odds can lead to more serious losses.

How Do Bookies Try to Make a Balanced Book?

Up arrow, down arrowThe main thing a bookie does to try and balance their books, or get them reasonably equal, is to change their odds. Let’s say that 70% of punters are backing Team 1 in the cup final, with only 30% backing Team 2. This may be down to the fact that a lot of tipsters are predicting a win for Team 1.

Alternatively, it might be that punters are simply making the wrong call. It could also just be natural variance, with it just so happening that at a given bookmaker punters are tending to back Team 1 whilst at another site they are favouring the other side. Last of all there is the possibility that the bookies are wrong: this isn’t a 50/50 game that could go either way but, in fact, Team 1 really are the better outfit.

Reducing the Odds of One Team Increases the Price of Another

No matter what the reason, once a bookie is in that situation, if they choose to try and create a more balanced book they will reduce the odds on Team 1 and increase the price of Team 2. This should encourage more punters to back Team 2, whilst at the same time reducing the liability on any further wagers they accept on Team 1.

Other tools available to them include extra marketing, for example promoting the odds on Team 2, or using special offers or free bets to encourage more people to bet that way. They could also use their own onsite pundits to tip Team 2, or ask media connections to do the same, though this could be viewed as shady practice. Ultimately though by far the most potent weapon at their disposal is their ability to change the odds.

Alternative to a Balanced Book is to Take a Position

Choosing a side

In modern times, bookies have increasingly looked to keep reasonably balanced books (or at least attempt to). Bookmaking has become more of a science than an art and now online betting sites use big data and computing power to try and create a system where they cannot lose. Any market with an overround that is even reasonably balanced should allow the bookmaker to come out on top and that is the approach almost all sites now take.

What “Take a Position” Means

The alternative to aiming for a book that is sufficiently balanced to guarantee a win for the house is for the bookmaker to “take a position”. In essence, this amounts to the bookie backing their judgement about who will, or sometimes who will not, win any given event. Reverting to our cup final for what may or may not be the last time, if a bookie firmly believes that Team 1 is by far the better side and will lift the cup, they may be happy to build a very one-sided book.

Instead of offering both teams at 4/5, they may offer Team 1 at 4/6 and Team 2 at the much bigger price of evens. This results in a similar overround as pricing both at 4/5 (it is slightly lower at roughly 10% here) but means that they may well be the best odds available at any bookmaker on Team 2. If they are convinced that Team 1 will win they are happy not to get the money wanting to back that option. Anyone at all concerned about value would surely back the 4/5 or better at another site, rather than the 4/6 available with our bookie.

On the flip side, anyone wanting to back Team 2 can now get evens rather than 4/5, so this type of punter should be flocking to our bold bookmaker. Bookies will have different approaches to how they go about doing this, with some, for example, deciding how much of a hit they are prepared to take if Team 2 does win. Whilst the bookie may be happy to take a position and might be hugely confident in their judgment, no bookmaker is stupid enough to believe that any team or player has no chance at all.

Outcomes Compared Amongst Different Bookies

As such, they may set a limit in terms of how much they are prepared to lose, or alternatively may have a split in mind in terms of how much they want to have on both possible outcomes. Let us compare how their outcomes may compare to a rival bookie who opted for – and achieved – a perfectly balanced book.

Outcomes Position Puncher Pete Balanced Bookie Bob
Odds on Team 1 4/6 4/5
Odds on Team 2 Evens (1/1) 4/5
Total Staked on Team 1 £20 £50
Total Staked on Team 2 £80 £50
Outcome if Team 1 Wins +£66.60 +£10
Outcome if Team 2 Wins -£60 +£10

Obviously the figures used here are far smaller than we would see at any major online betting site. For a cup final the total wagered might be 1,000 times bigger (or much more), meaning that the bookmaker prepared to put their neck on the line might be taking on a potential loss of £60,000. However, if their judgement was proved right, they would be in the black to the tune of almost £67,000 (based on a total stake of £100,000).

In contrast, the bookie who favoured a balanced book can sit back once the game kicks off. The result matters not to them because whoever lifts the trophy, they win the same £10,000. In essence, when a bookie takes a position they are also betting on the outcome of the game, albeit from the advantageous position of having the odds in their favour (again, thanks to the overround).

Taking a position can, and often does, involve wanting one or more participants not to win. For example, in a horse race, the bookie may feel there are some horses that are almost certain (note “almost”, because as already said, nothing is 100% guaranteed) not to win. They may be happy to establish a book where, for example, two outsiders are offered at large odds and if either win, they will lose overall. However, should any other horse in the field pass the post first, they will yield a handsome result.

It might not always be an outsider that the bookie believes they can take a position on either. It is not uncommon for a bookmaker to believe that a favourite is too heavily fancied by punters and has been backed into odds that they are now happy to lay. A bookmaker may construct a book in such a way that should the favourite deliver they lose quite heavily, whilst any other horse hands them a very positive result.

Taking a Position Is Risky

As mentioned, when a bookie takes a position, they are essentially having a bet themselves, albeit one with odds more favourable than a punter would typically get. And as all punters should know, having a bet can be a risky business. Perhaps the most famous example of a bookmaker sticking their neck out and backing their judgement came when Frankie Dettori landed all seven races on Champions Day at Ascot in 1996.

Gary Wiltshire

Frankie Dettori

Frankie Dettori (Paul Friel /

That was a terrible day for bookies in general but it was especially bad for one small bookmaker in particular. Gary Wiltshire lost over £1m that day, far more than he could afford. He said he lost around £1m on the last race alone and it took him four years to repay his debts, a feat only achieved by working every single race he could and selling all his belongings!

The first six winners came in at 2/1, 12/1, 100/30, 7/1, 7/4 and 5/4 respectively, as Frankie would ultimately land a seven-fold acca of over 25,000/1 that, quite simply, nobody thought was possible. As the hugely popular Italian jockey landed winner after winner the bookies’ losses mounted up and by the time of the last race people were so convinced that Frankie would win that his mount was backed in from 12/1 to 2/1 favouritism.

Few serious racing people thought Fujiyama Crest had a chance carrying top weight, including Dettori himself. The Italian said afterwards “Never in a million years did I expect Fujiyama Crest to win. I was very relaxed because I just couldn’t possibly see the horse winning.” The money coming in for the horse was purely down to people daring to dream and what the jockey had done before. It was not a reflection of the horse’s chances of success and so bookmaker Wiltshire saw a fabulous opportunity.

Punters queuing up to back a 12/1 shot at odds of around 2/1 is something a bookmaker can only dream of. Even if 12/1 were the horse’s true odds, without any overround, it could have been expected to have around a seven or eight per cent chance of winning. The chance for a bookie to lay it as if it had a 33% chance of success, at odds far, far shorter than they “should” have been, was too much for Wiltshire (and others) to turn down.

The bookie noted that the horse was “without a win for a year, was out of form and was a 12/1 shot”. He had actually fared well at the meeting overall, despite the success of Dettori, saying “Believe it or not, I was in front after five races. I (lost) a few quid in the sixth, but it was nothing to worry about. I knew we were going to get it all back and more because Frankie’s horse in the last was certain to start at (very short odds)”.

As other bookmakers sought to back Fujiyama Crest, in order to offset their anticipated losses (another way to help build a balanced book), Wiltshire gobbled up their stakes. He says he laid one bet of £40,000 with Coral at 7/2 and also that all, or almost all of the cash he took on Frankie’s seventh winner of the day came from other bookmakers rather than the public. In the end he lost £1.4m on Fujiyama Crest, all because he took a huge position about a horse where he saw bags of value.

Of course, one man’s very bad day over 25 years ago at Ascot is far from the only example of bookmakers taking a position. Despite the move to a more computerised, technological and risk-free approach, bookies still regularly take a position to varying degrees. Where it tends to happen most frequently and most famously is around big events with very well-backed favourites.

Well-Backed Favourites Make Balancing Books Impossible

If there is a horse like Red Rum in the 1970s or more recently Tiger Roll, the massive volume of money that pours in for it means that it becomes virtually impossible for bookies to balance their books. Bookmakers may well end up with a negative position on it (should it win) and invariably it is when a big favourite wins that you hear stories of “bookie bloodbaths” or “massacres” or however the tabloids choose to report it.

Bookmakers are happy to take such a position because they know that the odds they have been offering are well below what the probability of success would imply. They may well lose some but they will win more overall and, what’s more, a favourite like Tiger Roll in the Grand National is hardly a sure thing. For the Irish horse’s second success in 2019 he was sent off as the 4/1 favourite, meaning the bookies expected him to lose the race around 80% or more of the time.

Of course, he won and estimates of what losses the bookies might sustain were around the £200 to £250m mark. There have been similar losses reported when a number of big favourites have won all of the major races at the Cheltenham Festival too.

Three Lions

Perhaps the biggest loss ever might be yet to happen though. Few betting events come bigger than the football World Cup and few regulated betting markets are as big as the UK, with England representing the majority of that. Every four years there is a huge amount of money wagered on the Three Lions finally ending their seemingly never-ending “years of hurt”. Much of this cash is put down through patriotism, optimism, blind faith or some combination of them all and punters backing England tend not to be overly concerned about the odds or the value.

This means the bookies can get away with offering bad value on the Three Lions and still attract plenty of custom. They know that the English public tends to overrate the national side and so they are prepared to take a position on this, especially given they can do so at such bad (from the punter’s perspective) odds. When it comes to England the bookies have been winning since 1966 (as an aside, it is widely reported that Fred Done used winnings from backing Sir Alf’s boys to set up Betfred!) but the bookies can’t win forever… can they?!