Tuesday 21 April 2015

When April turns to May: what the seat markets imply for the election result

It's time to put the pieces together.  Over the last few days, I've been looking at the constituency markets on a party by party basis, sorting the seats by the implied probability that the bookies have given to each party in turn and looking at what that means for them both in general terms and in specific seats.  I attach links to each of them below (some of the prices are already slightly out of date, particularly following the gyrations in the prices in the Scottish constituencies over the weekend):

Labour



Conservatives 


Lib Dem 


Greens, Plaid Cymru, Respect, NHA and Independents

And new for April, I've prepared tables for the Northern Ireland seats:

https://drive.google.com/file/d/0Bygi8eZw-4q1SXBsTVFWVVRJRmM/view?usp=sharing

I don't propose to give any commentary on the Northern Irish seats, partly because I have no great insight and partly because there aren't that many interesting seats.  Some of the DUP and Sinn Fein 1/100 bets are as sound 1/100 bets as you will find anywhere, mind.
 
Now it's time to put together the overall picture as it stands.

Projected seat tallies

What are the implied seat counts?  Allocating the seats to the favourites (halving them when there are joint favourites) gives the following seat counts:

Conservative: 273
Labour:265 1/2*
SNP: 54
Lib Dems: 30 1/2
DUP 9
Sinn Fein 5
SDLP 3
Plaid Cymru: 3
UKIP: 3
Speaker: 1
Lady Sylvia Hermon: 1
Greens: 1
Respect: 1

*This figure has been adjusted to take account of the intervening seat movements in Scotland since the Labour seat odds table was compiled.

What kind of government would that produce?  Labour plus SNP would be just shy of the magic 326 figure.  Adding in the support of the SDLP, Plaid Cymru, the Greens and Lady Sylvia Hermon and the passive support of Respect would get a Labour-led government control of Parliament.

That would work for UK-wide matters, but the SNP would not be reliable partners for English-only matters.  In those circumstances, Labour would probably need Lib Dem support also.

Nevertheless, the par numbers suggest a rickety Labour minority government.  If it panned out like that, it would be fun.  You can still get that at 7/4 and I've been tipping it since it was 7/1. 
 
What might yet change?

That's all well and good, but we need to think about what are the likely bounds of possibilities.  With less than three weeks to go we can largely confine ourselves to the 1/2 to 2/1 bands.  There will be seats won at longer odds, but we should now expect that to happen relatively rarely.
 
So what's the likely minimum and maximum for any party?  Here are the totals that each party would tally if they took all the seats in which they are priced at 1/2 or shorter (but no more) and the totals that each party would tally if they took all the seats in which they are priced at 2/1 or shorter (but no more):

Labour: 241* - 301*
Conservative: 239 - 305
SNP: 43 - 57
Lib Dems: 17 - 37
DUP: 9
Sinn Fein: 5
Plaid Cymru: 2-5
SDLP 2 - 3
UKIP: 1 - 7
Speaker: 1
Lady Sylvia Hermon: 1
Greens: 1
Respect: 0 - 1
UUP: 0 - 2

*This figure has been adjusted to take account of the intervening seat movements in Scotland since the Labour seat odds table was compiled.

What can we sensibly note?  Well, first, it is now most unlikely that we will see an overall majority.  You can back No Overall Majority at 1/8 on Betfair and that looks like value even at that price to me.  Note that before the latest SNP surge, Labour were priced at 2/1 or less in 322 seats, meaning that their outer bound was within a whisker of the overall majority threshold.  That has receded into the distance.   1/8 on No Overall Majority is some way too long.
 
Secondly, it is going to be very hard for the Conservatives to get to form the next government.  If they don't get an overall majority (or very close), they're going to need the connivance of the Lib Dems.  The Conservatives will need at least 295 seats and that is close to the top end of the range for them.  Laying David Cameron as Prime Minister after the next election looks value right the way to 2/1 at the least.
 
Oh, and the 7/4 on Labour minority government continues to look like very good value, particularly given that the SNP would be an essential component of any majority and there is no prospect of any coalition between those two parties.
But I've plugged all these bets before.  Do I have any new ones?
 
These ranges are surprisingly tight.  You could take the view that the seat prices are too bunched (I have some sympathy with that view) or you could work on the basis that the variations from the norm will cancel each other out.
 
If you take the latter view, you might back 21-30 Lib Dem seats at 5/4 and 31-40 seats at 5/2, both bets with Bet365.  By placing appropriate stakes, you can get effective odds of slightly better than 1/3 that the Lib Dems will get between 21 and 40 seats.  Given the very local appeal of each Lib Dem candidate, it seems unlikely to me in practice that the Lib Dems will get fewer than 21 seats because their chances in different seats are much more loosely correlated than, for example, the major parties. 
 
Similarly, you can back the Conservatives to get 251-275 seats at 7/2 with Ladbrokes and 276-300 seats at 13/8 with William Hill, giving effective odds of slightly worse than 4/6 that the Conservatives will get between 251 and 300 seats - place the bet with William Hill first because it's more likely to be limited.  Conservative chances are much more correlated than the Lib Dem's chances, but this still looks like a good combination bet to me.  I've put some money on both of these combinations.
 
The same combination is available for Labour (251-275 seats at 7/4 with William Hill and 276-300 seats at 2/1 with Ladbrokes) for effective odds of slightly worse than 4/9.  The reward doesn't look quite good enough for me, but it's not far off either.
 
For a quirky bet, Coral offer a UKIP vs Plaid Cymru match-up.  The 7/2 on the tie looks like fair value, given that both are favourites in three seats.  I'm on.
 

Monday 20 April 2015

Other parties and independents: the battlegrounds in April 2015

In the last few days, I have looked at the betting odds in each constituency from the perspective of Labour, the Conservatives, the Lib Dems, the SNP and UKIP.  But they aren't the only ones fighting the general election and they aren't the only ones who have decent hopes of winning seats.  At present three more parties hold seats in Parliament (currently holding five seats between them) and a fourth is led by someone who did so until 2010.  All four of these parties will be contesting some seats at the next election where the bookies deem that they are in the mix, and a couple of independents have made their presence felt at the bookies.  What are their chances?

Here are the relevant tables of target seats for each of them.


Plaid Cymru:



Respect:



National Health Action:


We have not yet exhausted the list of parties contesting the election.  Candidates for the Ubuntu Party, the War Veteran's Pro-Tradition Through Family Party, the Wessex Regionalists Party, the Cannabis Is Safer Than Alcohol Party or the World Peace Through Song Party (to name a few) have yet to set the constituency markets alight, and time is running out.

To recap, these tables are compiled on the basis that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (otherwise known as betting opportunities).  The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.

I last looked at the outsider parties in November, and my post can be found here:

http://newstonoone.blogspot.co.uk/2014/11/other-parties-battlegrounds-in-november.html

There are still only 12 seats where the minor party or independent is rated at 10/1 or shorter: six Plaid Cymru seats and targets, three Green seats and targets, one Respect target (won at a by-election), one National Health Action target and one independent.  

We can look at these 12 seats in some detail.

The Greens

I'd backed a couple of longshots for the Greens, but all but one of these have failed to prosper.  




As I said yesterday, not all longshots can come off. 

The Greens now have three serious prospects: Brighton Pavilion, Bristol West and Norwich South.  Last time round, I observed that in Lord Ashcroft's first opinion poll in Brighton Pavilion, he had not prompted for the Greens.  He rectified this in his subsequent poll in December, resulting in a 1% Labour lead being converted into a 10% Green lead.  The 4/9 on the Greens here looks great value and is well worth topping up.  I'm not expecting Caroline Lucas to be defeated now.

We have had one poll in Norwich South without prompting for the Greens, which resulted in them being second but some way adrift.  I still wonder what difference prompting would make.  5/1 is probably a value loser bet.  I'm on this at 25/1 and don't feel the urge to top up, but it's not a stupid bet.

I've been sceptical about the Greens' chances in Bristol West from the outset.  At one stage you could have backed them at 100/1.  I wish that was one of the longshots that I was on, but it isn't.  They're now in to 4/1 and though there has been no polling in this seat, they're clearly in contention.  The three parties in contention are all progressive, which means that the two left of centre opponents of the Lib Dems are likely to knock each other out.  I'm on the Lib Dems in this seat at 8/13 and I still like this bet at evens.  I'm not topping up further though.

 
Plaid Cymru

Plaid Cymru, unlike any other parties in this list, are sure to retain at least one seat and in all probability at least two.  Their third seat, Arfon, looks like a cliffhanger, while the bookies assess them as having a decent chance of taking Ceredigion, an outside chance of taking Ynys Mon and a faint hope of taking Llanelli. 

This is the most recent Wales-only poll:

http://blogs.cardiff.ac.uk/electionsinwales/2015/04/15/the-latest-welsh-political-barometer-poll/

Plaid Cymru appear to be polling at least as well in Wales as they did in 2010, but even so there is a relative swing away from Plaid Cymru to Labour, its main rivals in most of its target seats, because Labour has improved more.  Incidentally, no one is talking about any possible UKIP gains in Wales, but they're polling as well in Wales as in the UK as a whole and I'd guess that their appeal in the areas where Plaid Cymru is strong is as weak as their appeal is in Scotland.  If I were looking for an under-the-radar success for UKIP, Wales is one of the areas I'd be looking at.

Plaid Cymru are only competing seriously in a handful of seats, so I don't regard even Wales-wide opinion polls as particularly informative.  We have to dig deeper.

The politics of Arfon and Ynys Mon are very much personality-driven.  I'm not particularly familiar with the personalities, so I'm not betting.  Plaid Cymru have suggested that the rise of UKIP has been helpful to them in Ynys Mon and Llanelli:

http://www.walesonline.co.uk/news/wales-news/labour-losing-votes-in-droves-8943801

This seems quite plausible.  Certainly, the Labour response in that article is not exactly carefully reasoned.

Ceredigion is a fascinating seat.  Mark Williams has very much followed the Lib Dem tradition of building up a personal brand rather than rely on the party to secure him votes.  He has a large majority.  Will it be enough to survive a collapse in national Lib Dem support?  He has probably been helped by his main rival having got bogged down in a slanging match with the no-hoper Labour candidate. This has been very entertaining, but can only benefit the incumbent:

http://www.telegraph.co.uk/news/politics/plaid-cymru/11522401/Plaid-Cymru-candidate-compares-the-English-to-Nazis.html

http://www.walesonline.co.uk/news/wales-news/chavs-sickening-england-flags-words-9018453

I'm still not betting on this seat, but the 4/7 on the Lib Dems is starting to look very tempting.

Llanelli, on the other hand, does look straightforward to me.  The 1/8 on Labour has to be value, even if UKIP might be drawing support from Labour.  On the other side of the fence, an independent is standing who has a proven record of drawing support away from Plaid Cymru:

http://www.walesonline.co.uk/news/wales-news/maverick-independent-sian-caiach-contest-8386444

I have put something on Labour in this seat.

Respect

Respect have only one short-priced chance: Bradford West.  Their three longer shots are also all brought to you by the letter B, but I'm not tempted by any of them (in two of them, Bethnal Green & Bow and Bradford East, they aren't even standing, which makes the odds on them look a little uninviting).
 
Last time around, I thought that Labour were value at 1/3 in Bradford West.  Out of spite to a gambler, they promptly imploded in an orgy of in-fighting.  Unsurprisingly, Respect are now favourites in this seat.
 
Out of the chaos, Labour have ended up with a candidate with a backstory that appears inspiring:
 
 
But this looks like the dirtiest constituency campaign in 2015:
 
 
It seems that the Respect machine is well-oiled, contrary to my previous impression.  Their favourite status is probably justified.


National Health Action

As I noted last time, this is a new party which was founded by Dr Richard Taylor, the former MP for Wyre Forest who had sat as an independent campaigning on health for two parliaments.  It is standing in 12 constituencies but the only seat in which it is considered to have more than an outside chance is Wyre Forest, where Dr Taylor is again standing.  He would need a 2.6% swing to retake the seat.  In 2001 and 2005, the Lib Dems stood aside for Dr Taylor, but there will be a full complement of parties this time around.  We have had one poll from Lord Ashcroft in this constituency, but unsatisfactorily he did not prompt for National Health Action, which must be an error given the history of this seat.
 
I had previously backed UKIP at 25/1 and NHA at 5/1, and have now closed out by backing the Conservatives for a near-guaranteed profit.  If I were starting from scratch I'd find the 6/1 on NHA very tempting.

Independents

As in every general election, there are independents standing up and down the country.  Most will vanish without trace.  But sometimes they challenge hard and sometimes they win.

In East Devon, Claire Wright is apparently putting together a decent show.  She's a local councillor who apparently has a good local support base.  Her national politics, so far as one can discern them from her website and manifesto, are green-tinged:

http://www.claire-wright.org/
https://8c78c03a-a-62cb3a1a-s-sites.googlegroups.com/site/realzorro1/CWrightJanFeb2015Leaflet.pdf?attachauth=ANoY7covzq0edyv-AzqBuWTJjpoVLjPAT3tV6J6ldRnIU0GMfKaDvMeM9_p4rT7GOQuO12NMjEo2jKf8kcmPqpxBmJO5L9BiGAMREs8qGYL4eqw9j9X_StCOCV6SWY1yebO0wyOK81jsxbLvYsCxX5VtdPtslIzHdqU7N4xVBHbkxto6OTxZjw1WQiE2eKEt2I8bJJAfX0ppIZLjclKqCf_he6hYLTRahAe0XNFaCu-KfpxUspE19Hk%3D&attredirects=1

Notably, no Green is standing against her.

The $64,000 question is whether she has the organisation to back up a constituency-wide campaign and whether the Lib Dems' past strength here can be built upon by a different party.  I anticipate that the 1/6 on the Conservatives is good value.  But this is one bet I'll pass in the absence of more local knowledge.

Sunday 19 April 2015

The UKIP battleground in April 2015

It's time to look at the constituency betting markets as they currently stand from UKIP's perspective. So as at 13 April 2015, here's the table of all the seats where UKIP's chances are ranked by the bookies as 16/1 or better, ranked in ascending order of implied probability:


To recap for first timers, this table is compiled on the basis that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (otherwise known as betting opportunities).  The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.  

This is how things stood at the beginning of November 2014.

http://newstonoone.blogspot.co.uk/2014/11/the-ukip-battleground-in-november-2014.html

At that time, UKIP were soaring.  Since then, their prices have plateaued or even waned.  UKIP are favourites now in just three seats (down from five in November) and priced at under 2/1 in just seven seats, (down from nine in November).  UKIP were priced at under 10/1 in 42 constituencies in November, and that has now reduced to just 34 constituencies.  Some of this is explicable by prices in general becoming more competitive, but not all of it by a long way.  Whether or not UKIP have peaked in the polls, they have peaked at the bookies.  I shall consider later whether that is a correct reaction.

The decline in UKIP's fortunes has been exclusively as against the Conservatives.  In November, UKIP were priced at under 10/1 in 13 non-Conservative constituencies, exactly the same number as today.  UKIP have actually shortened their price in the one Labour constituency, Great Grimsby, priced at under 2/1 over this period.

UKIP do seem to have slightly come off their very highest points in the national polls and even the most ardent kipper would have to concede that they are not advancing further.  Some of the previous prices had been predicated on the possibility of UKIP continuing to improve.  That has not happened and their election campaign has not caught light.  The dreams of the most excitable kippers look set to be crushed.

That means in turn that UKIP look most unlikely to take anything more than a handful of seats.  Recent constituency polling seems to confirm that (click to enlarge):



UKIP will be particularly disappointed in the polling in Dover, a seat where they had entertained outside hopes but where they seem to be coming a distant third.  They might well have hoped to be polling better in Cleethorpes and Dudley South too.

This means that UKIP's chances for the election are probably all gathered in those few seats where they are short-priced and the rest can probably now be forgotten about.  There is still the chance of an odd shock under the radar somewhere, but if it happens, it will be an odd shock not part of a clutch of victories.

As someone with a collection of UKIP longshot bets, this is disappointing, but I can hardly be simultaneously thrilled with the strong hope of success of longshot bets on the SNP and glum about UKIP longshots looking likely to be heading for the ashtray.  That is in the nature of longshot bets: they don't all come off.

I have called UKIP wrong at every point, discounting them before they surged, then thinking the surge would continue when in fact Rochester & Strood marked the end of that period of progress.  So I hesitate to give any advice at all, and you should discount my thoughts heavily given my track record.  But here goes.

Where UKIP are still short enough to affect the price of the others in the market without having a real chance of success themselves, there may be opportunities worth backing elsewhere.  So I am looking at those seats where UKIP are priced between 4/1 and 10/1 with that in mind.  I've backed the Conservatives in Folkestone & Hythe at 1/4 and Wyre Forest at 8/15 (closing out a three-way bet for a guaranteed profit).  I backed the Lib Dems in Eastleigh at 4/7 the other day, as I mentioned when I looked at the Lib Dem seat markets.  And I've backed Labour in Walsall North at 1/3 and in Newcastle-under-Lyme at 2/7 (as I mentioned when I looked at the Labour seat markets).

What of those seats where UKIP remain in with a shout?  My view of those is that they are seriously murky and that no one really understands how they are going to play out.  I am therefore proceeding with extreme caution in this final phase.  The 1/9 on UKIP in Clacton is a decent 1/9 shot though, as 1/9 shots go.

Saturday 18 April 2015

The SNP battleground in April 2015

So far I've looked in turn at Labour, the Conservatives and the Lib Dems.   And now it's time for my look at the constituency betting markets as they currently stand from the SNP's perspective.  But as in October, my original plans have had to be slightly modified.  I had planned yesterday to show all the parties as they stood as at 13 April 2015, to have a nice neat record of fully interlocking constituency odds tables.  But the prices in Scotland have changed so much in the last two days (thanks, Lord Ashcroft) that to base a post on the prices then in place would be pointless and misleading.  So I've had to wait for the prices to settle a little then update the table to take account of these, and this has been reworked several times today.  Still, I can for the first time confirm the truth of Camus' observation that Sisyphus was happy.  The prices used in this post are correct as of today's date.

Anyway, here is the table of all the Scottish seats, ranked in descending order of implied probability of the SNP taking them, as determined by the best odds available at the bookies:


To recap for first timers, this table is compiled on the basis that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (otherwise known as betting opportunities).  The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.

These prices are still in a state of ferment.  It is quite likely that the best prices in many of these seats will be dramatically out of date in a few hours (the bookies have had their prices pushed all over the place in the last two days).  But I had to draw the line somewhere.

It is worth comparing with the position on 13 April so that you can see just how quickly things have changed:

https://drive.google.com/file/d/0Bygi8eZw-4q1TGkwMmRGbHNNSTQ/view?usp=sharing

What has caused this turmoil?  This:


Out of the blue, Lord Ashcroft released another batch of Scottish constituency polling.  It showed the SNP doing extremely well everywhere (other than the very special case of Berwickshire Roxburgh & Selkirk) and apparently even better than it had been doing in January and February when Lord Ashcroft last looked at the Scottish constituencies.

As a result, the bookies appear to have concluded that it is game over.  The SNP are now 5/6 (the bookies' evens) or better in 50 seats, favourites in a further four seats and joint favourites in a further two seats.  The SNP are priced at 1/2 or shorter in 43 seats.  

Are the bookies right?  Essentially I think that they are.  I expect that there was an element of book-balancing in the face of money backing the SNP and there are arbs all over the place.  There may be a late swing, and as a matter of caution you might want to knock a few points off the SNP's constituency polling which is relatively untested and bearing in mind that the pollsters overstated the Yes vote in the referendum by about 3% in the end.  But there is nothing in these polls to suggest that the SNP's appeal is waning and with only three weeks to go, time is running out for Labour to pull it back.

This leaves me with the emotional problem of whether to start hedging my bets.  I've done a little of this to cover my stakes in East Dunbartonshire and Edinburgh North & Leith at what I considered favourable odds, but I've decided not to in general.  I only gamble what I can afford to lose, so I don't need to insure the losses.  Indeed, most of my bets in the last two days have been on the SNP - this latest batch of constituency polling is seriously good news for the SNP.

Are there any bargains?  On the SNP side, all the 1/2 bets look good to me (Ayr Carrick & Cumnock, West Dunbartonshire, Paisley & Renfrewshire North and Paisley & Renfrewshire South), as does the evens in East Renfrewshire - 9% is a healthy lead in a constituency poll and nothing in the detail suggested that tactical voting was going to save Jim Murphy. The 11/10 in Rutherglen & Hamilton West looks good - the swing of 22.4% required would be comfortably exceeded by the average swings in the constituency polling in nearby seats.

On the Labour side, I'd want local knowledge before backing them anywhere.  I suspect that the scope for tactical voting would be greater in the Edinburgh area, but the swings that the SNP requires there are smaller too.  So until I get more information, I'm not really betting on Labour anywhere.

These latest polls are grim for the Lib Dems.  So I'm not making any more bets on them either.

Next up, UKIP.

Friday 17 April 2015

The Lib Dem battleground in April 2015

This week is my pre-election round-up of the seat markets, and I'm ploughing through the parties one by one (metaphorically speaking).  On Wednesday I looked at the constituency betting markets as they currently stand from a Labour perspective.  Yesterday I looked at the Conservatives.  Today it's the Lib Dems' turn.

So here's the table of all the seats where the Lib Dems' chances are ranked by the bookies as 16/1 or better, ranked in ascending order of implied probability:

These figures are up to date as at 13 April 2015.

To recap, this table is compiled on the basis that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (otherwise known as betting opportunities).  The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.
The overarching picture for the Lib Dems has deteriorated decisively since November, the last time I took a comprehensive look at the seat markets.  They are longer than 5/6 in 31 constituencies that they currently hold (up from 24 in November), 5/6 being the price that the bookies use for a 50:50 shot (they have their margins to make).  

In June, the Lib Dems were priced at odds of greater than 2/1 in ten of the constituencies that they hold.  By November, that had risen to 14.  Now, there are 21 such seats.

Despite the Lib Dems facing many more Conservative challengers than Labour challengers, the favourites in these 21 seats is exactly evenly divided between Labour, the Conservatives and the SNP.  This represents a strengthening in both the SNP's and (to a lesser extent) the Conservatives' position since November, while Labour have stood still.  

The position is getting worse for the Lib Dems at the other end of the spectrum too.  In June, the Lib Dems are rated at shorter odds than 1/2 in 22 seats.  By November, this had fallen to 19.  Now, there are only 14 seats where the Lib Dems are thought to be a better than 1/2 shot.

So why are the Lib Dems doing so much worse on the seat markets now?  In part, as with Labour, it's a consequence of the SNP's surge in Scotland.  But in the main, it reflects the failure of the Lib Dems to recover in the polls in the run-up to the election.  A recovery had been priced in and it hasn't happened.

For those that want to look at my previous outings on the Lib Dems, here's my review in October:

http://newstonoone.blogspot.co.uk/2014/11/the-lib-dem-battleground-in-november.html

And here's my June review:

http://newstonoone.blogspot.co.uk/2014/06/the-lib-dem-battleground-in-june-2014.html
 
One of the central questions of the next election is: how far will the Lib Dems fall from the 23% they recorded in 2010 and where will those losses be felt most?  

The Lib Dems seem to be doing particularly poorly in Scotland, if the Scottish only polls are to be believed.  They also seem to be set to be obliterated in seats where they are not in contention.

What of seats where the Lib Dems are in contention?  We have conflicting polling.  The constituency polls for Lord Ashcroft tell a reasonably hopeful story for the Lib Dems, showing their local effort potentially leading to them to hang on in many redoubts.  But a recent ComRes poll of the Lib Dem-held seats in the south west with Conservative challengers tells a very different story, showing a swing of 13%, which would lead to them losing every single seat in the region on a uniform swing.  At the other extreme, we have details of some Lib Dem private polling showing many of their incumbents doing phenomenally well.  They cannot all be right (though they can all be wrong).  Which should we believe?

Lord Ashcroft derives his headline constituency poll numbers from the second of a two part question, where voters have been guided to concentrate on who is standing in their constituency.   His standard questions are as follows:

"1. If there were a general election tomorrow, which party would you vote for? Would it be Conservative, Labour, Liberal Democrat, UKIP or another party.  If 'another party': Would that be, the Green Party, the British National Party (BNP) or some other party - or do you not know how you would vote?

2. And thinking specifically about your own constituency and the candidates who are likely to stand there, which party's candidate do you think you will vote for in your own constituency at the next general election? [Prompts as at Q1]"

ComRes's question did seek to elicit a local response. It read: "Please now think specifically about your own constituency, the issues it faces, the local MP and the different candidates. At the General Election coming up in May, would you vote Conservative, Labour, Liberal Democrat, UKIP or for some other party?"  It did not, however, explicitly lead voters through the same process as Lord Ashcroft.

Lord Ashcroft's approach relies on voters going through the same two stage process that he prompts them to.  If the Lib Dems' electoral machine is running smoothly, that may be a safe assumption.  In seats where the Lib Dems are not so fluent, perhaps it isn't.

So my expectation is that the Lib Dems may overperform in seats where they have the organisation to get out the vote and underperform elsewhere.  Past performance is not always a guide to the future, but it is at least a guide to where the Lib Dems have the organisational skills.

The methodology behind the Lib Dems' private polling is here:

http://www.markpack.org.uk/files/2015/04/East-Dunbartonshire-constituency-poll.pdf

As you can see, it asks the respondents to consider the merits or otherwise of the Lib Dem incumbent and her chief opponent before asking the voting question.  I shall be charitable about the Lib Dems' private polling and say that it may provide a useful guide to them as to how to frame the election to maximise their vote, and releasing such polling may well be useful to them for obtaining tactical votes.  It may also be a hint to the seats where the Lib Dems are well organised: we have seen leaked polls from some constituencies but not from others.  I draw a fairly good inference in the most threatened constituencies where we haven't seen leaked polls that the Lib Dems may well be up against it.

So, what individual seats do I recommend?  Generally, I'm cautious where the Lib Dems are heavily involved because of the general murk about how their local performance.  I don't believe that there are many home bankers unless you have very firm views about which constituency polling methodology is correct.  I don't.

That said, I pay closer attention to individual constituency polls from Lord Ashcroft than usual, because they should pick up the performance of the local MP, which is going to be exceptionally important in Lib Dem held seats.
Given the Lib Dem decline in vote share, any bets on short priced Lib Dem seats is a bet on the quality of that constituency's party machine.  I've put some money on the Lib Dems in Thornbury & Yate at 2/7 because I reckon that Steve Webb's personal vote will see him through comfortably, but I don't regard this bet as being as safe as some of the short priced bets on Labour and the Conservatives yesterday.
 
It may be better to identify misfiring challenges.  The Conservatives in Southport seem to have been no better organised than in 2010, so the 1/2 on the Lib Dems is worth a bet.
 
Alternatively, the murk may mean that you get better value in the 1/2 to 2/1 band than in supposedly safer bets.  I've put bets on the Lib Dems in Cheadle at 4/5 and Eastleigh at 4/7.  The latter price in particular looks too long, with the Conservatives failing at the by-election and UKIP apparently fading outside their top targets at present.

All bets on the Lib Dems are very much AYOR.  These are inherently riskier bets than bets in straight Conservative/Labour fights, because local considerations may (or may not) weigh much more heavily than in other contests and much is not visible from the outside.
 
What of the seats where the Lib Dems are the challengers?  For quite some time, I have regarded many of these prices as way too short.  If voters in these seats weren't going to vote Lib Dem at the height of enthusiasm for Nick Clegg, you need an exceptional reason to think that they're a worthwhile bet to make a gain.  The value for me is almost always all on the other side of the bet.  In the last two posts, I've already mentioned Ashfield and Maidstone & The Weald as seats where the incumbents are good value.  You might well form the same view of Oxford West & Abingdon, Truro & Falmouth and Newton Abbot.

Tomorrow I'll turn to the SNP.

Thursday 16 April 2015

The Conservative battleground in April 2015

Yesterday I looked at the constituency betting markets as they currently stand from a Labour perspective.  Today I'm looking at the Conservatives.  So here's the table of all the seats (as at 14 April) where the Conservatives' chances are ranked by the bookies between 1/16 and 16/1, ranked in ascending order of implied probability:


As I explained yesterday, such a table assumes that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (otherwise known as betting opportunities).  The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.

Here are the milestone seats with the seat counts implied if the Conservatives take every seat in order of probability up that point:

350 Clacton (7/1)
340 Birmingham Edgbaston (5/1)
330 Thornbury & Yate (7/2)
326 Plymouth Sutton & Devonport (7/2)
320 North Warwickshire (3/1)
310 Weaver Vale (5/2)
300 Hastings & Rye (2/1)
290 Morecambe & Lunesdale (6/4)
280 Cheadle (6/5)
270 Great Yarmouth (5/6)
260 Rossendale & Darwen (4/6)
250 Elmet & Rothwell (4/7)
240 South Basildon & East Thurrock (8/15)
230 North Thanet (2/5)
220 Folkestone & Hythe (3/10)
210 Maidstone & The Weald (1/4)
200 Burton (1/6)

5/6, the price on the Conservatives for Great Yarmouth, is sometimes referred to as the bookies' evens (reflecting their need to build in a margin to make a profit), so it seems that the current central implied expected point for the Conservatives is around 270 seats. 

For an overall majority, the Conservatives would need 326 seats.  The 326th seat is Plymouth Sutton & Devonport, for which the Conservatives are priced at 7/2.   As of this morning, you can get just under 9/1 with Betfair on the proposition that the Conservatives will get an overall majority, with good liquidity above 8/1.  The seat markets are by and large for the birds in this vicinity.  I would not consider backing the Conservatives in any seat at these odds without the most outstanding intelligence that the Conservatives were doing far better in the seat than was widely understood. If you think that the Conservatives stand an earthly chance of an overall majority, back that proposition directly on Betfair.
 
As I noted yesterday in relation to Labour, it's less clear how many seats would be required for either main party to have most seats.  Both parties are priced in the constituency markets at 5/6 to get their 270th seat.  Yet Labour are 15/8 on Betfair to get most seats while the Conservatives are 8/15.  These two markets are completely detached from each other.

It follows that there is an indirect hedging strategy, which is to back Labour on the most seats market and to back the Conservatives in the seats which they would need to take from Labour if they were to get around 270 seats.  Such seats would include Worcester, Blackpool North & Cleveleys, Pudsey and Ipswich, for example.  If you back a suitable selection according to taste, you would be unlucky to come out behind and if you were  lucky (or skilful in choosing your Conservative bets) you might win on both sides of some of the bets.
 
Another important milestone seat is Eastbourne (7/4), which is the 295th seat.  This probably represents the practical minimum for the formation of a Conservative government, with David Cameron as Prime Minister.  Yet you can still lay David Cameron on Betfair at odds on to be Prime Minister after the next election.  Another indirect hedging strategy is to do just that and to back the Conservatives in seats like Stroud, Eastbourne, Cannock Chase and Carlisle.  Get the right seats and you might again come out ahead on both sides of the bet.

Obviously for both of these strategies there is the risk of losing on both sides of the bets.  But that would be seriously unlucky (or you would have chosen poorly when you picked the seats in which you had backed the Conservatives).

If you want to play things more simply, just lay David Cameron for next Prime Minister, which is what I have done.  He should be at least 7/4 for that proposition if the constituency markets are to be believed. To be able to lay him at odds against is incredible value.

Or, if you disagree and think that the Conservatives are going to get most seats, look in the 270 to 310 band of seats.  If you are right, this is where you will find the best value on the constituency markets. 

I already have a book of Conservative relative longshot bets from the last year, and now I'm looking at the value at the other end of the scale.  Redditch at 2/7, Sittingbourne & Sheppey at 3/10 and Crawley at 3/10 all look safe enough to me, and I've backed the Conservatives in all of these.  I'm not expecting the Lib Dems to come close in places like Maidstone & The Weald either, despite rumours of private Lib Dem polling showing them doing well: 1/4 looks like a good price here too.

Wednesday 15 April 2015

The Labour battleground in April 2015

Over the months I have looked a few times at the individual constituency betting odds on party lines, ranking the seats in order of the odds on a given party taking that seat.  I last did this in a comprehensive manner in October/November last year.  Over the next week or so, I shall be giving my final pre-election look at the seat markets.

The idea behind this is not immediately obvious to everyone, so new readers may want to start here:


The central point from that post is that such a table assumes that in aggregate the constituency odds are our best estimate of what's going on, while noting that there may be individual anomalies (more commonly known as betting opportunities). It also makes the heroic assumption that the individual constituency odds - for Labour and the Conservatives at least - are perfectly correlated contingencies (which of course they are not) to establish a handy tool for judging the chances of success for each of the main parties overall. The general idea is that looking exclusively at marginality takes insufficient account of the different nature of the seats (who's second, is there a relevant third player, where is the seat). The constituency odds factor those matters in to the best judgement of bookies and punters. By arranging constituencies by order of odds rather than majority, we can see how many seats gamblers expect the parties to take - or what the odds are that each party will take a given number of seats.

So, here's the table as of this morning:


I've included every seat where Labour are priced between 1/16 and 16/1.  These prices are up to date so far as possible (subject to the inevitable transcription errors etc) as of first thing this morning.  

Here are the seat tallies that would be implied by taking these milestone seats:

210 Cardiff North (2/9)
220 Enfield North (2/7)
230 Plymouth Sutton & Devonport (4/11)
240 Hove (4/9)
250 Bury North (4/7)
260 Ayrshire Central (8/11)
270 Harrow East (5/6)
280 Lanark & Hamilton East (6/5)
290 Stevenage (5/4)
300 Glasgow Central (13/8)
310 Blackpool North & Cleveleys (7/4)
320 Rossendale & Darwen (2/1)
326 Aberconwy (9/4)

I last looked at the Labour position in October here:


What has changed?  In the intervening months, the markets have become much more competitive.  We have many new entrants to the seats markets, with at least 14 different firms quoting on some or all of the different constituencies.  Prices have generally improved in aggregate and there are numerous seats where there is a practical underround (those who wish to seek out low risk double bets may wish to study these lists carefully).  A slight deterioration in the best price for a party may signify nothing more than a new entrant to the market.

But the glacier of safe Scottish Labour seats has melted.  All 41 Scottish Labour seats are far less safe than they were in October.  Some seats that were so safe in October that they were not even listed then are now odds against for Labour.  This has made Labour's task much harder.

Otherwise, the seat markets have moved predominantly in response to constituency polls (and then often not by as much as one might have expected).

Anyway, back to the present day.  5/6, the price on Labour for Harrow East, is sometimes referred to as the bookies' evens (reflecting their need to build in a margin to make a profit), so it seems that the current central expected point for Labour is between 265 and 270. 

For an overall majority, Labour would need 326 seats.  The 326th seat is Aberconwy, for which Labour are priced at 9/4.   As of this morning, you can get 41/1 with Betfair (with good volumes of money at 33/1) on the proposition that Labour will get an overall majority. 

That means one of two things.  Either the bookies' prices on the constituencies are ridiculously mean or the impact of how Labour will perform in individual constituencies has not been properly factored into the overall markets.  The two are not mutually exclusive, of course.

It's less clear how many seats would be required for Labour to have most seats.  You can get 7/4 with Bet365 on Labour getting most seats and do slightly better even than that on Betfair.  You don't start getting prices in the individual constituency markets until you approach the 310 seat mark.  310 would definitely be enough for Labour to have most seats - indeed, 280 might well be enough.  So once again, the most seats market offers far better value than the individual constituency markets if you are looking to back Labour.

So if you want to back Labour, the simplest strategy by far must be to ignore the constituency markets and instead to back Labour in the most seats market (and, if you so wish, the overall majority market).  I've been following this strategy for some time.

However, it is now worth considering whether to place some short term low risk bets on Labour in their safer seats.  2/7 in Newcastle-under-Lyme and 1/5 in Ashfield (where the previous dynamic Lib Dem candidate had to stand down in the worst circumstances) both look good, with UKIP apparently fading in their more distant prospects if other polling is to be believed.  I'm on both.  I haven't detected a Plaid Cymru surge, so I'm backing Labour at 1/8 in Llanelli also.  The rates of return seem well worth the increased risk over bank rates to me.

Tomorrow I shall look at the markets from the Conservatives' perspective.