How You Could Have Made a Million on the 2015 UK General Election

Introduction

As you may know, I love numbers and trends, and especially love betting odds to show trends. So I've done some analysis on seeing if the bookies are better at predicting the outcome of major events than the pollsters. With the Scottish Independence vote, the bookies predicted a couple of days before the vote what the actual result would actually be [here]. 

So I've been analysing the 2015 UK General Election, and the bookies were certainly tending towards a Conservative major, but they didn't quite get the impact correct, with David Miliband the favourite to become the next prime minister. For this I did an analysis on 2 May 2015 [here], and we'll analyse to see if we could have made some money from this.

How to make a million on the election?

So let's have a bit of fun, and assume that we could go back in time, and put a bet on, on 2 May 2015 (just a few days before the election). I appreciate that this has been done in Back to the Future II, where Biff went back to bet on the outcome of sports matches, and ended up extremely rich, but let's try and replicate it, without being spotting for creating a time machine.

First we can avoid the short odds of 1/5 for a Conservative majority, so our first bet is going to be on a Conservative Majority. For this we get 6/1! A nice start on the way to gain a million. So let's say we have a ￡1,000 stake, we now have ￡7,000.

Even money bets

Next we'll go for some easy bets (as we don't want to alarm the bookies that we have managed to predict the future). On 2 May 2015, here were the expected number of seats for each party:

• Conservative seats (average): 282.8 
• Labour seats (average): 268.8
• SNP seats (average): 51.5
• Liberal Democrat seats (average): 28.8
• Plaid Cymru seats (average): 3.5
• UKIP seats (average): 2.6
• Greens seats (average): 0.5

Yes. You are reading this correctly ... Liberal Democrats were on 29 seats and UKIP somewhere between 2 and 3, with SNP on a low number of between 51 and 52.

So let's take an even money bet on Conservatives to get more, Labour to get less, SNP to get more, Liberal Democrats to get less, Plaid Cymru to get less, and UKIP to get less. That's six winning bets at even money, and which are interelated in terms of SNP and Conservatives gain more seats against a reduction in the others. So now our odds are:

Total odds = 7 * 2 * 2 *2 *2 * 2* 2 = 448

So we have now won ... ￡448,000 ... not too bad, for just predicting a Tory and SNP landslide, and a collapse of the others.

For the next PM, it was fairly even, with Ed at 4/5 and David at Evens:

• Ed Miliband 1.8 (4/5)
• David Cameron 2 (Evens)
• Caroline Spelman 51 (50/1)
• Boris Johnson 61.9 (61/1)
• Tristram Hunt 67 (66/1)

So let's put a bet on David, and now we are on:

Total odds = 7 x 2 x 2 x 2 x 2 x 2 x 2 = 896 ... with winnings of ￡896,000.

Next we'll take the easiest bet of all ... the turnout. With this the threshold for evens was pitched at the high value of 68.3% (average), so a bet on lower than this, where the turn-out ended up at around 65%, we are now on:

Total odds = 7 x 2 x 2 x 2 x 2 x 2 x 2 x 2 x 2 = 1,792 ... with winnings of ￡1,792,000.

So we've almost made it over one million, just by predicting the Conservatives getting a majority, and by even money predictions on how well each party got on and on the turnout.

Could it have been predicted with a time machine?

The UKIP trend was definitely downwards, from 6.5 seats at the start of the year, to 2.5 on 6 May 2015, so it could have been predicted that it might small even more. For a Liberal Democrat collapse, there wasn't many pointers from the bookies (Figure 2). The only thing was that they were at the top end of their band (24-27.5 seats), so perhaps a little bet on them moving somewhere in the middle could have been a reasonable guess. The same approach for the Conservatives (Figure 3) could be used to identify that they were at the lower end of their seat range and perhaps would bounce back (in a way, obviously, that few could have predicted).

Figure 1: UKIP seats prediction

Figure 2: Liberal Democrat seats prediction

Figure 3: Conservative seats prediction

Conclusions

The bookies didn't quite get the detail right, but they could certain see the Conservative and SNP push, and the collapse of the others. The trend was very much up for both Conservatives and SNP in the last few days for odds, and there was a sense that something was changing. The prediction of the number of seats would have been difficult, but there were pointers to the thresholding of the number of seats that each party were going to get.

So are the bookies better than the pollsters? .. Yes! ... as they have a lot more to lose if they are wrong.

The only three things you would have needed to predict, as everything else will fall into place, would have been:

• The Conservatives were going to get a majority.
• SNP were going to get the vast majority of the seats in Scotland.
• UKIP were not going to do as well as was expected.

With those three things, you would have made over one million UK pounds!

Note ... this is just a bit of fun, and my only interest is in the mathematics of this.