The Grand National Overround Debate

[Tal]

I've been thinking about this since it was mentioned yesterday online. There are lots of related points to consider and discuss, so I'm going to do my best to cover those I understand, those I don't and the ones I think I get that I will be shown I'm wrong about shortly. This won't be a short post, so bear with me. It'll more likely be a few posts.

First things first, the basics for those who have no familiarity at all with how odds work. Feel free to skip ahead if you know this.

The odds of a thing happening is another way of saying the probability of it happening. If Tottenham are 2/1 to beat Chelsea, that means, if they played three times, they would lose twice and win once. It has a 1 in 3 or 33.333% chance of happening. In the same game, Chelsea might be Evens, which is another way of saying 1/1. This means that, if they played twice against Spurs, they'd win once and fail to win once. So, a 1 in 2 chance of winning, which is 50%. The draw could be 3/1, which is a 25% chance of happening. This all seems reasonably plausible and the Saturday coupon will show similar prices for lots of games.

Here's how the bookie pays for the shop, the lights, the heating and the smiling lady printing the tickets: those are the only three things that can happen in the match (win, lose or draw). If you add the percentages up, you should get 100%, but you don't. You get 108.333%. This is the bookies' cut, known as the Overround. It is their insurance policy in case any of the prices are wrong but also, if they are right, it represents theoretical pure profit, much like in roulette or blackjack. You can lose on one spin, but a million spins is happy days.

Now that we punters have things like Oddschecker and are generally more savvy in our lives outside gambling about shopping around for the best price, it isn't hard to start shaving some of that 8.333% off. Let's say another bookie is 6/5 on Chelsea, for example. Suddenly, you're down to 103.5%. If you can find 7/5, you have a perfect 100%. Anything better, and the market actually gives YOU leeway. The house is behind and is, in betting parlance, Overbroke.

Football is a three horse race: only three results can happen. There are literal three horse races, too and the prices work the same way. If a fourth horse is added to the equation, the chances of each of the first three horses winning the race is reduced logically, because a fourth result could happen. Even if that horse is rubbish, it COULD win, so there has to be a price for it, even if it's 100/1. If the other horses have a lower chance of winning, that should lead to their prices being adjusted slightly to reflect that, else it just means that the bookies get an extra 0.999% added to their overround; their edge. So what if the horse was actually better than the other three? Say it had a 50% chance of beating these three. There's 50% that has to be eaten up, so the other prices have to get bigger to reflect the fact they're now less likely to win.

Typically, the more horses running in a race, the greater the bookies' edge is. I'm not sure why that is, but that does seem to be the case. Perhaps it is because there is a greater level of uncertainty from the bookies about the correct price of each horse. Bear in mind that a horse in a 20 horse race at 9/1 is being given as having a 10% chance of winning. If it actually should be 7/1, there's 2.5% missing and unaccounted for in the house's edge and, if there are four or five horses they have got wrong, that can become quite expensive, especially on races where they might take a few quid or where, as the prices are logically a bit bigger, they could end up with a bigger exposure, so perhaps they increase the safety net accordingly.

That doesn't seem hugely fair to the punter, but it's the status quo.

Ok, so to yesterday. The Grand National had 39 runners and, when the starter set them off, the bookies had an overround in excess of 160%. This means they didn't have so much a safety net, but they barely were off the ground. It's hardly an anomoly, either; the Grand National overround has been well over a third for a decade or more. 

More follows...

[BigAdz]

Its like the stupid kid in the maths class standing up and teaching everyone how to add up, like he was the only one that knew.

[Tal]

I'm fairly sure not everyone who posts on blonde understands this stuff.

I'd prefer it if people who do really understand this stuff actually contributed on the subject, but, here we are and I'm writing about it.

Again, more follows...

[BigAdz]

Cough......sound of own voice.....cough.


Carry on. x

[RED-DOG]

I'm fairly sure not everyone who posts on blonde understands this stuff.

I'd prefer it if people who do really understand this stuff actually contributed on the subject, but, here we are and I'm writing about it.

Again, more follows...

Well I don't understand it, but I enjoyed reading about it.

[Chompy]

Spotlight writers are told to bet to 2% per horse when doing the betting forecast, so 120% in a ten-runner race for example.

However, you'll never see anything much above 140% and I never go above 130%. There's no point if you want to do anything semi-accurate, as the 'Fair/'Daq over-rounds will never be much above 100%.

I'm not sure what the over-round was at best prices for the National yesterday morning, but 165% is a disgrace.

The Fodds wankers know that 9 out of 10 once-a-year customers won't take an early price and that it doesn't matter if they screw them over, as they won't see them again until the same day next year.

[Karabiner]

I think that one of the starangest things to happen in recent times was when they changed the way of calculating SP so that it's far more like the worst price available than the best on the off.

How and why the bookies were allowed to get that one through is beyond me.

[Tal]

(I have some scribbled notes I'm typing up, so bear with me if there's overlap, but by all means keep chatting)

Something weird happened yesterday, just before the flag went down. Weird to those of us who hadn't really noticed this before at least.

For weeks, we've heard about Tony McCoy, the great jockey, retiring at the end of this season, so this is his last Grand National. Lots of people who don't bet much or don't bet much on horse racing will follow the jockey they've heard of, the trainer they've heard of, the name they like, the grey horse or something equally mathematically illogical but satisfying. Fine. So, the bookies were taking a lot of bets on McCoy's ride, because this year was his last and the world and his wife wanted to have the betting slip the day he won his last race.

Now, the bookies were in a situation where they risked a big loss if that horse won. There was talk of £50m bandied about. This is commonly the most profitable race of the year for the bookies: hundreds of millions of pounds placed on horses for no reason but the colour of the silks. What do you do? Let it ride and hope McCoy doesn't win or shorten the price a bit and encourage some money elsewhere to limit your exposure? The bookies pretty much always choose the latter and that's what they did.

Shutthefrontdoor was never regarded as having a 10% chance of winning, but that was his price near the off. And yet, just minutes before they began racing, the price dropped again to 6/1. Why?

[Tal]

There is a logical difference between the bookies on the course and the big corporations that advertise on the TV.

The TV bookies have had bets placed in shops and via apps for weeks and weeks. They have deep pockets and like to publicise when they have had to pay out to a little old lady from Ipswich. They also have accumulators and that was the problem on the first day of the Cheltenham festival, when three horses from the same, Irish trainer won the first three races. The fourth was a big odds on shot and the likes of Ladbrokes and Paddy Power were facing all sorts of trouble if Annie Power managed to clear the last hurdle, given lots of folks had her in a juicy acca. Their price for Annie Power would logically be a bit shorter than Billy Gilbert, standing there near the winning post. He isn't facing a monster acca, so even though he is losing money - probably a lot of money for him - and even though he might reflect that a bit in the price, he probably won't go as far as the big bookies.

It makes sense, therefore, that the starting price for a lot of bookies for Shutthefrontdoor would be much shorter than its actual price; they were desperate to get it beat, but they were also desperate to get people to bet on other horses and reduce the chance of a monster loss.

But the Starting Price - the official starting price - is generally (I assume it still is, as it was historically) calculated on the bookies at the course. They pick a sample of twenty or so of the bookies and that determines the prices that come up when you watch on TV.

So, why did THEIR prices go in to 6/1? They weren't facing that massive exposure. If you've ever been to a race course, you'd see there's always a chap on a laptop behind every bookie, who is watching the betfair exchange and can, if required, lay some of the money they are taking off. Now, the betfair price at the off was still in double figures, which takes us back to the question...why did the bookies on course drop their prices?

Here's a more sinister point to all of this: the vast majority of people who bet on this one race a year will take the starting price. Some won't know that you can take the price that it is at the time they have a bet, some will know but will be frightened that it might go out and they miss that bigger price, some will know but genuinely won't care. When the SP went from 8/1 to 6/1 right before the off, every one of those people lost out.

Did Bert Wilkinson at Aintree win when that happened? No. But the guys on the telly did.

(one more and i'm done)
 

[Tal]

In 2010, Tony McCoy won the only Grand National of his career, on a horse called Don't Push It. For most of the reasons already explained, it was the most backed horse in the field, so the bookies had a very rough day.

As the Guardian explained in the aftermath, the big bookies approached the course guys to lay off some of their exposure just before the off. The punters who had a bet on the Tote on the day were paid out at 14/1 and they got a much better deal than anyone who backed at the starting price, because that suddenly turned into 10/1, having been twice that in the morning.

http://www.theguardian.com/sport/blog/2010/apr/15/grand-national-starting-prices

In that year, Ladbrokes said 40% of their shop punters took the starting price. FORTY PERCENT!

It's suddenly massively in their interest to lay off some of that exposure with the on-course bookmakers. They can now massively influence the amount they pay out. They don't just cover some of their losses by betting on the result themselves; they force the price down that they have to pay the other winners out at. It's worse than that: they are betting on a horse to win at a higher price than they are paying other people out at.

And let's not forget that they are only losing money on 1 out of the 39 horses running and all those people who backed McCoy yesterday lost their money. Let's also not forget that the bookies had a 60% edge to start off with.

Some might argue that it's how a market works and that it is all just good business sense. Maybe it's what has always happened.

Does it seem right? Should the big bookies be more prepared to gamble? The rest of us were, after all.

OK, over to you, including the bloke who backed the winner

[TightEnd]

I think that one of the starangest things to happen in recent times was when they changed the way of calculating SP so that it's far more like the worst price available than the best on the off.

How and why the bookies were allowed to get that one through is beyond me.

when did this change please?

at the start of the racing day 7-10 rails firms are nominated as the "pool" of SP setters for the day by the agent, i think...so the off course boys know where to go to "hedge"/shorten prices

in practice how did the process of setting the SP change?

[AndrewT]

I did think it was bonkers when STFD was 8/1, and trading at 13 on Betfair. It then went 7/1, 6/1 in the blink of an eye whilst drifting to 14.5 on BF, which was you could get on the off.

[GreekStein]

Its like the stupid kid in the maths class standing up and teaching everyone how to add up, like he was the only one that knew.

Cough......sound of own voice.....cough.


Carry on. x

Apologies if I'm taking these posts the wrong way but they seem rather bitter and spiteful.

I am enjoying this topic Tal and your posts on it. ty

[Marky147]

Greekstein, meet Adz


FWIW, I don't think they're intended to be either of those, but Adz is no Jimmy Carr...

[Tal]

Thanks, Greenstein, but it's just Adz's humour. I've grown used to it.

He's alright...for a gooner.

I am looking forward to his actual views on this stuff, though, as a regular and savvy horse racing punter, who backed a horse that arguably should have been one of those pushed out by the bookies, thereby giving him a higher return if he wanted to press on the day.