A new market on Commons seats
January 29th, 2005
An interesting new market on how many seats each of the main parties will win has been launched by Skybet. The structure is similar to the one that Bet365 recently had but with the seat options at levels more in keeping with current forecasts.
The markets make a good alternative for those wanting decent prices but are not willing to take the risks or get involved with the complications of spreadbetting.
Labour seats
7/4 0-340
15/8 340-365
6/4 365+
Conservative seats
5/4 0-190
2/1 190-205
15/8 205 +
Lib Dem Seats
7/4 0-65
6/4 65-70
5/4 70+
With these bets we like to look for value at the open-ended bottom or top ranges. With the Tories and Labour the mid-ranges are very much in line with current spread prices. With the Lib Dems however the latest spread prices are above 70 seats. If there is a value bet here it is the 5/4 on Charles Kennedy’s party getting more than 70.
No doubt some smart Politicalbetting users will find ways of combining these bets with the spread markets as Jon did so well two or three weeks ago.
Mike Smithson
MessageSpace Advertising
You’ll get more allegations of bias, methinks… I should’ve thought that 7/4 was an attractive price for the LDs to get 65 or less, i.e. 13 or fewer net gains over 2001 - my sense of the seat-by-seat discussions is that the balance of opinion is for around 10 to 15 net gains…
Interesting the spread markets are very hard on Plaid - seems to bear out the pessimistic tone of Cymru Mark on Ynys Mon and anywhere else for that matter.
IA - My Lib Dem observation is based on what punters are doing not my personal leanings. Sporting Index have the party at 72-76 seats; Spreafair have them at 71 - 73.4 and IG Index are at 70-74 seats.
I bought the Lib Dems at 58 & 60 in June and sold them at 72 in October. Nice money.
Indeed IA the fair price for less than 65 cannot be greater than that for 70+ without being free money against the spread markets.
Can you elaborate on the last comment John?
5 Yes Chrisco - the term “Free Money” has a certain appeal! I’ve been trying to work out what Jon is saying but I can’t see it. Hopefully he’ll come back.
It is tempting to buy the Tories at 190-205 and Labour at 340-365. Why? If you use Baxter the Tories are only 5 short of 190 and Labour are only 10 above 365. When you then consider that all the seats the Tories are tipped to take are seats with majorities of under 3,000 then one thinks that they will presumably take more (the only side issue is will the Lib Dems take more of the Tories than Baxter’s predicts). Even if the Lib Dems took another 5 seats off the Tories, it still doesn’t seem that unrealistic to think that the Tories will take at least another 10 on top of Baxter to get to 190. However for them to get past 205 would require a performance I’m afraid that the polls do not show. Even CCO seems to be setting there sites on only 40 seats (of which some are Tory seats to be defended vs the Libs).
As for the Labour bet I’m even more convinced there is good money to be made on it. Again if we take Baxter Labour are only 10 above 165. If we then take Tory and Lib Dem targetting then 10 seats does not seem like a lot, especially when one remembers that in 97 seats like Newark and Chesterfield (with majorities of over 4000) were stripped away by the Tories and the Lib Dems. The other good thing about the Labour bet, is the margin is 25 seats, which means they would have to lose an extra 35 seats on top of Baxter for one to lose, which at the moment looks unlikely. Any Comments please?
Ok here goes…
There is a subtle relationship between the seat range markets and the seat spread markets. In a fair market (i.e one in which all of the probabilities implied by the odds adds up to 1) the mean number of seats “expected” by that market can be calculated in a reasonably obvious way… say the LD’s are 20% likely to get 51-60, 30% likely to get 61-70, 30% likely to get 71-80, 20% likely to get 81-90 then the mean value assuming uniform probabilities within those ranges is
0.2×55.5 + 0.3 x 65.5 + 0.3 x 75.5 + 0.2 x 85.5 = 71.5
unless I’ve cocked it up.
This number must be the same as the number implied by the spread markets otherwise there is an arb. Say the spread markets were 75 middle, then by betting an amount of money enough to win the average for each range you could reconstruct a bet that pays you the same as the seat number, but the cost would be only 71.5 (try it if you don’t believe me). So you would do that, sell the seats at 75 and pocket the difference with no real risk.
The relevance of this is that if <65 were implied to be more probably than > 70 you would have to assume some very odd things in order to get an average at the spread seat market average… hence the (almost) free money.
I suspect that will not be entirely clear but you get the idea.
This will become more relevant if Betfair stop being useless and start quoting some seat range markets as theirs will almost by definition be “fair”.
Andrew 7. If you just take the latest Yougov poll(and here I’ll incur the wrath of Graham) and put the numbers into Martin’s calculator you get Lab 352: CON 207: LD 57. Add just a touch for LD targeting and tweak it with some tactical unwind seats and that 352 Labour could easily drop below the 340 mark where you can get 7/4 with Skybet.
This is too far out to make a bet - there are 13 weeks to go - and the market sentiment will stay mostly with Labour. So there is no point betting now.
The Tories have had a boost with the immigration policy but how long will this last?
The big issue is the economy and the link to house prices which, for the time being at least, looks good for Labour. But if house prices start to drop and the spectre of negative equity comes into play then anything could happen.
House pricess couldn’t fall that fast - it is a very sticky market which makes the decline when it does come that much more drawn out and painful.
By the way, the new Communicate Research Poll is up on there website. Unfortunately I can’t download it off this computer. But someone else probably will be able to.
I’ve just posted full details of tomorrow’s Communicate Research poll in a new main article.
IA at 1: which are the seats you think the Lib Dems would/might gain?
IG Index are at 70-74 seats.
Implying an expected 72, neglecting skew.
Skybet -
7/4 0-65 - 36%
9/4 65-70 - 31%
5/4 70+ - 44 %
or 0-70 67 % 70 + 44 %
Either IG (and other spreads), or Skybet are wrong….or both…
you either back the 70 + @ 5/4 or sell @ 72 (Sporting Index 72-76 )
…or both. Pseudo arb imo.
John, It’s a fairly new market for Skybet and my guess is that very few punters have so far placed bets. I only found it because it was referred to by someone who posted on one of our forums. It is not on the betting odds search engines so its probably only been up a couple of days.
The odds will change quickly and my guess is that the LD price will come more into line with the spreads.
agreed Mike.
In the pricing up stakes I’d go
Skybet 20/1
spreads - N/O
Mark [14] - I don’t pretend to the expertise necessary for a “bottom up” approach. But no one here seems to think that the LDs can take more than a handful off Labour or make a net gain of more than a dozen from the Tories even if their vote goes up substantially.
Because we are approaching a three party contest (more accurately, three different two-party contests simultaeously) I would not place any money at this stage because we are still relying on tips from the gallops, as it were, rather than seeing the runners and riders in the paddock…
My comment was based on a gut feeling that, were I making the book, I would have been less generous on a low LD number than 7/4. I appreciate that bookmakers react principally to money laid, the whole idea of the site is that punters can get it wrong!
IA - if you have 7/4 (36%) 0-65 value then the spreads are off to the tune of at least 5 seats.
John - I’m sure that either the LD spread price will move downwards or the Skybet price will move upwards. These things usually level themselves out.
Interesting echos of “Here’s how you’d vote if you thought we could win” - with LDs on 37% with a big majority in that case. I stick with my view that LDs will actually beat Labour in some of the seats where they are far away but clearly the challengers. Particularly Islington S, Hornsey, Cambridge, but even Oxford E, Durham and a number of Scottish Seats. That’s in addition to romping home in the obvious marginals.