Keir Adams, SBTech: Pricing betting’s most unpredictable market

Keir Adams, SBTech: Pricing betting’s most unpredictable market

Bookmakers will have their political credentials put under the spotlight once again, as the UK prepares for a general election on Thursday 12 December. Following year-on-year political upsets since 2015, Keir Adams, Head of UK Sports at SBTech, spoke to SBC News about pricing betting’s most unpredictable market.

___________________

SBC: How do you price an election? And how do you price in particular this election?

KA: Politics has been turned on its head in the last 10 years and both technology suppliers and operators must work extremely hard, round the clock and across a vast array of sources to achieve the right results when it comes to political betting.

You must always try to keep a clear head and not get carried away with one source of information such as a poll, or something someone says during a campaign speech or debate, for example.

Pricing is getting tougher and tougher with so much more information and data available and to monitor. Overall, bookmakers have got it wrong for the last few big events.

Many betting companies looked at London as a guide for how Brexit would fare, but London is a melting pot and was always going to vote to remain. What they failed to do during the referendum was look everywhere else, such as the north of the country and the pro-Brexit towns and villages.

This election will be much tougher to price and impossible to automate the entire process. It will also largely be decided around what tactical positions MPs and voters are taking. For instance, the Brexit party has agreed not to stand in Conservative areas, while the Lib Dems have a pact not with Plaid Cymru and the Greens not to stand against them.

Also, in previous elections, we have been able to see how each constituency and how their MP has voted for or against Brexit, for example. Now we have a case where the process has been delayed for so long that MPs who are against Brexit still occupy seats where most people voted to leave European Union with these MPs going against what their constituents wanted. This makes it even tougher to model.

We began offering constituencies markets very early and, aside from suspending them overnight as its very hard to keep up during that time period, we have kept them open.

SBC: What information and data do you look at and consider?

KA: Political betting today uses a multitude of sources. The backbone of our work and outcomes are generated by our advanced automated platform. The model here is very good at predicting which party will gain the most seats as well an overall majority, however matters become much more complicated when it comes to the 650 constituencies in the United Kingdom.

Here it comes down to trying to keep track of every local issue, and it is local knowledge that helps to find value. It’s very tough to stay on top of all 650 seats and all the issues for each constituency. Countrywide issues that affect us all such as the NHS and Brexit are areas we can track, but if something local happens in Sunderland, for example, that has upset people there, it’s difficult for us and the bookmaking industry as a whole to stay on top of everything.

Beforehand, we were able to use data from previous elections and how seats have changed over time. Elections and outcomes were far more straightforward in the past. You would have safe seats, such as Sunderland that may lose some ground at times, but it would take a miracle for them to change hands. Nowadays things are very different.

We look at every data point and every factor and piece of research that can influence the various outcomes of the election, as well as TV debates and polls. Polls have been used and relied on for many years but the last few elections and referendums across the UK and US have been very unpredictable, and polls have often been wrong. For that reason, we review them but take them with a pinch of salt at times.

We combine automated systems, research and polls with external experts such as final analyst Martin Baxter who runs the political forecasting website Electoral Calculus and uses complex mathematical modelling to predict outcomes. In six out of seven general elections from 1992 to 2017 they have correctly predicted the party which won the most seats (except 1992) as well as the party which won a majority, or the outcome of a hung parliament, in four out of seven (1997, 2001, 2005, 2010).

We also track the market and other bookmakers such as the Betfair exchange and spreads to see where the money is going and which direction people are heading. On the exchange we can see how much money is matched and where people are trying to get on for certain selections.

We also mustn’t forget social media which today has more of an influence than ever. 10 years ago, this was a very different story. 

SBC: What do you think the outcome of the vote will be?

KA: The market is currently expecting a definite but slim Tory majority of around 50 seats. The picture is changing all the time but we’re reasonably certain at this point that the Conservatives will win the election and achieve a majority.

Many bookmakers have been bullish on the Lib Dems winning around 40 seats, but this has now dropped to the 30s. The Brexit Party is more likely to get zero seats than five, but we have them at a likely one to two seats as they could receive support from disaffected Tory supporters.

SBC: Can you predict any upsets?

KA: The North of England could see some potential upsets. Richard Tice, the Brexit Party chairman, is standing in Hartlepool where 70% of the population voted to leave the EU. There may well be a very close three horse race there.

Other Northern constituencies such as Bolsover where Labour MP Dennis Skinner is running and where he has held that seat since 1970 could also see a shock. This is the fifth strongest Labour seat, but many are predicting the Tories will win this seat by a few percentage points. Not so long ago he had a 90% chance of winning, bit this is only now 60%. People there could oust a former miner in a mining town. They are voting for change and with their heads, not their hearts.

There is also good value to be had on the Labour party winning less than 200 seats with Labour down to 200 from 230 in two weeks. Things have changed and are constantly moving with Tories and Lib Dems winning seats against Labour, and the latter fighting a strong Stop Brexit campaign. With those factors Labour may find it difficult in a few weeks’ time.


Source: SBC News