Bloody Kids

Don't you just love them?

I was doing OK today, on the dogs at least, when my son stopped by and almost immediately started fighting with his sister. This completely ruined my concentration and, over the next handful of races I lost all the profit I had built up during the day.

I tried to recover some of it in the evening, but didn't get anywhere and finished the day a few pennies down.

So, not much to write about today but I thought I'd take the opportunity to revisit this issue regarding dead heats.

I'd been thinking about that email I got from Bet365, and after checking what the P&L was in my Betdaq account for that race and it is clear that the response from Bet365 is complete waffle.

If you recall, they said, "The stake is split proportionally according to the number of relevant Dead Heated participants. In this case a bet on a selection at 4.0 would only return your money at a rate of 2.0."

That first sentence clear states that the STAKE is halved in the event of a two way tie. This, multiplied by the odds at the time of the bet gives the return. Fair enough, that's what I thought happened. However, as I mentioned yesterday, this sentence is, in my view, contradicted by the next one.

Now, let's consider the lay bet that I placed with Betdaq.

The stake was £52.55 @ 4.3 giving a liability of £173.42. If we apply Bet365's first sentence rule and divide the original stake by the number of dead heat participants we get a liability of £26.28 * (4.3 - 1) = £86.72. Minus of course the other half of the £52.55 which was lost, giving a liability of £60.43, which is what the P&L for that race on Betdaq was.

You won't be surprised to learn that if we divide the ODDS by the number of runners involved we get a liability of £60.43. i.e. £52.55 * (4.3/2 - 1),

So, irrespective of methodology, the returns are the same and should come as no surprise. It is of course completely irrelevant from a punters perspective who has placed a bet on a runner to win. The same return is achieved.

Clearly though, the bookies are going to use whatever method benefits them, and by using the odds rather than the stakes, they avoid paying a free bet to yours truly. Why don't they just come out and say it instead of beating around the bush?

Bet365 managed to live up to the reputation of help desks worldwide, as being staffed by inept people. I simply asked for confirmation of my understanding of the rules, with a yes/no answer being sufficient, yet their first response was to simply repeat, verbatim, what their Terms & Conditions say. Excuse me, I've read those. I just want confirmation!

When I pressed them on the matter, they replied with that tosh I've already cited, which, at first glance, confirms my original understanding, then seems to contradict it.

So, there we have it. Although not confirmed, the betting industry divides the ODDS by the number of dead heat participants and not the stake as I originally thought, simply for their own benefit. Having cleared that matter up, I cannot help think that that is a strange way of doing it from a probability point of view.

Are they really saying that, at the start of a race, the probability of your runner being involved in a two-way dead heat is TWICE that of winning the race outright? It's THREE times more likely to be involved in a three-way tie than winning outright?

Now, I'll be the first to admit that it's been along time since I studied statistical theory and probability, but surely that cannot be right? How often does a dead heat occur? Or a three-way tie for that matter?

That seems utterly bizarre to me.

Think of it another way. At that split instant, when those runners' noses cross the line at exactly the same time, the odds of any one of them winning the race outright have SHOT OUT to infinity, NOT shortened.

Not that my ramblings will change the way the betting industry works, but at least it gave me something to write about on an otherwise wasted day.

Update 4th Dec 2008: It's 8:30 in the morning and I realised while lying in my bed last night that I'd made an error in my calculations, but couldn't be bother to get up and fix it. I've adjusted my post accordingly. Apologies if there was any confusion.


Scott Ferguson said...

Hi Alistair,

Different companies and jurisdictions state it differently, but a dead-heat should deliver the same result either way.

£50 at 5 = £250 full payout.

Split the stake in two at full price, and you get £125 returned.

Split the odds and it's still £125.

In Australia, the phrase we use is 'half face value of the ticket'.

IBAS will back you up on any claim there, it's industry standard but somehow they've managed to confuse it quite a lot...

Alistair said...

Hi Scott,

Clearly, however they do it the return should be the same and a straight forward punter won't be bothered either way.

In this case however, Bet365 are choosing the option that allows them to avoid giving a free bet - by manipulating the odds. Something that is clearly wrong for the reasons I've mentioned in my posts.