This piece was originally published on May2015 – our general election microsite.
In the beginning there was the betting market. Long before anyone worked out the basics of statistical sampling, quotas or past vote weighting, the betting markets proved remarkably accurate in predicting political outcomes.
And in recent years things have only got better. New exchanges, such as Betfair, have eliminated the bookie, and allowed customers to create their own markets. Each punter can choose to either take a bet (the normal method everyone is familiar with) or “lay” one (in effect becoming the bookie).
If the market has many returning punters – if it is “active and deep” – you can have faith in the odds the market is offering. Bias is all but eliminated. And you can then work out the implied probability of an event, like who is going to win the next election. This involves a little maths but is relatively simple to do.
The bookies were an invaluable insight during the independence campaign, when they always heavily favoured a unionist victory. When the polls narrowed in September, the likelihood of a “No” vote fell, but still never dropped below 65 per cent (having been more than 85 per cent in August).
We can now use the bookies to track the implied probabilities of outcomes in May 2015. Here are five things they show which the polls miss.
1. Prepare for a hung parliament
While the polls still have Labour ahead for now, many are adamant it will all turn around for the Tories in the last few months. My response to that would be the market does not agree with you.
In the last few months “no overall majority” has become by far the most likely election outcome. The probability of it is now greater than 70 per cent. Compare that to Labour’s and the Tories’ chance of winning a majority – they are less than 20 per cent.
2. Largest party? It’s 50:50
The “long” election campaign theoretically only just started, but it’s been quietly raging since 1 September. Back then, when Labour had a consistent 3-4 per cent lead in the polls, Labour were favoured to be the largest party by 58:42. After Labour’s challenging autumn and early winter, those odds are now effectively 50:50.
Ladbrokes, the bookies, agrees. They think the odds of either party winning the most seats is 10/11 (52 per cent).  William Hill even slightly favour the Tories: they also put the Tories’ odds at 10/11 but Labour’s slightly lower – at “evens” (50 per cent).
3. No one has any idea what the post-election government will be
Both of the above markets are relatively active and deep so one can be confident about the information contained within the prices. In the more exotic markets you start to get quite speculative, however even here we can see a few interesting pieces of information.
For example the market for the next “type” of government is basically a shrug of the shoulders, with money spread everywhere and “any other” more likely than any specific scenario. Should we prepare for a Tory or Labour-led government? A majority or minority? A coalition involving the Lib Dems? All outcomes rate as equally unlikely.
For anyone thinking the next government is going to be a Tory/UKIP coalition, right now you have a 94 per cent probability of being wrong.
4. The SNP surge is only being taken half seriously
In Scotland we are all mulling over the possibility of an SNP advance after the independence referendum. Right now the party are riding high in the polls with double-digit leads over Labour (predict how many seats they might win on May2015).
This has led to near weekly forecasts that the SNP will sweep all of Scotland. The markets are, as yet, unconvinced. The SNP currently hold six seats. Polls suggest they could win the vast majority of Labour’s 40 seats and all of the Lib Dems’ 11. May2015 recently looked at why they might struggle to win more than 20 Labour seats. The bookies agree – Ladbrokes thinks they will win 25-26, adding 19-20 to the six they currently hold.
The bookies think the SNP will win 25-26 seats, not 50-plus.
The exchanges offer less precise odds but concur – they suggest the SNP will to more than 12 but fewer than 35. If we look to the polls, something revolutionary is going on in Scotland. If we look to the markets, the SNP is surging but will still be smaller in 2015 than Scottish Labour are now. Wait for the markets to move before taking the latest apocalyptic poll too seriously.
5. Ukip are not expected to win many seats
The other big unknown in May is Ukip. They are consistently polling around 15 per cent. They now have two MPs, and Lord Ashcroft’s polls have suggested they are challenging in another half a dozen or so.
But the markets suggest they will only win between three and six seats. What should you trust? The betting exchanges offer a crowd-sourced level of analysis of what is expected to happen in May – only a mouse click away. The more you get into analysing odds the more you will start to understand their power (and limitations).
The information is even updated second-by-second, so you needn’t wait for the latest polls to get an idea of who is up and down.  The combined odds are slightly greater than 100 per cent because of the “over-round“, which is how bookies make money.
May2015 will keep you updated with the latest odds – on which party is expected to win the most seats and most votes, who is favoured to be the next PM, and what the next government will be, from now until May.