World Cup madness

I see that they've made the group stage draw (but seemingly not the knockout draw) for the World Cup. I've instantly seen two wildly different takes on England's chances - they are, naturally, the favourites, and are priced in one place at 4/11. Now, yes, vig, but let's call that a 75% chance. In reality it's lower, but let's just round it off to that. Oche's said it should be 1/7 - i.e. that the chances of England not winning are half of what the bookies say. Lendel's said it should be evens - i.e. that the chances of England not winning are twice what the bookies say. That's quite the difference, but whose thoughts are more outlandish?

With no draw, it's hard to say, and it's even harder to know how to model doubles - we have no idea if Littler is running into his perfect format, or whether he hates it with a passion. The latter is possible, I hate hate hate playing pairs. I think the best thing we can do is to set bounds, and say that the best chance the better team has will be the chances of the better player for the better team (likely Littler, but it's close), and then swap the players to get the worst chance, and the winning shot will be somewhere in the middle. Let's assume it was done in draw order, whereby England, as the 1 seed, would get the 16 seed in round one, then 8, 4 and 2. This would give them Finland, Ireland, Northern Ireland and then Wales. Only the last two are fixed - they could end up getting much tougher teams in the first two rounds (although I think more teams seeded in the group stages are fielding nowhere near their strongest teams - but let's go with it.

vs Finland:

Littler v Kantele - 100%

Littler v Harju - 89%

Humphries v Kantele - 100%

Humphries v Harju - 86%

vs Ireland:

Littler v O'Connor - 80%

Littler v Barry - 94%

Humphries v O'Connor - 75%

Humphries v Barry - 92%

vs Northern Ireland:

Littler v Rock - 62%

Littler v Gurney - 86%

Humphries v Rock - 55%

Humphries v Gurney - 81%

vs Wales:

Littler v Price - 63%

Littler v Clayton - 73%

Humphries v Price - 54%

Humphries v Clayton - 66%

Now these numbers are quite, quite telling. Let's chuck the Finland numbers out as both Kantele and Harju are working on very limited data for the smallest sample, although if we do play split the difference, say England win 95% of the time and it's 1/20 in the pricing, it's probably not too far off, it's not worth mentioning. If we look against Ireland - Willie would be expected to beat either Luke somewhere between one in four and one in five times. Keane would not be as good, which is fine - no slight on Barry, but O'Connor is clearly the better player right now. The average of all those is 85%. That's clearly an indication that England are massively favoured, but not a lock. That drops to, say, a 1/6 pricing, pre vig. That immediately makes the 1/7 valuation Oche had look ridiculous, and we're not even at the semi final stage.

In the semis, we get a pair of opponents where one player is doing really well (fun fact - since March Rock has a higher points per turn in my database than Littler over a 50% larger sample), and the other player is certainly not playing badly, and may be a tad underrated. Add all those percentages together and split them and you get to 71%. That's more or less the 4/11 that the bookies (or at least the bookie was quoted) to win the whole tournament. Then we go to the final. No player is rated at less than a one in four chance to win a best of 19 against any other player in the match. That should be the deciding factor in telling you not to bet on England to win this event - they're too short. Way, way too short. Wales ought to be no more than 2/1 to win the final - and this is a known pair which has binked this event twice, so understands the tournament, and are known to be playing well as of right now, with (since March) Price being second only behind Humphries in scoring, and Clayton being over 93 per turn which is clearly enough to be considered at an elite level and he has a Euro Tour bink of late. Combine, say, a 1% chance to lose the opening game, a 10% chance to lose the quarter, a 25% chance to lose the semi and a 30% chance to lose the final, all of which are likely underestimates of England's chances to lose, and you get them at being odds against to win the tournament.

I'm clearly making some simplifications. But in such a unique format, you have to make these sorts of adjustments. I'm going off the data I have. Are England favourites to win? Absolutely. Should we bet England to win? Absolutely not. We should probably do the opposite.