#Brexit me Timbers

Would you have seen Rehagel’s Greece going all the way to win it in 2004, beating the Portuguese home team twice including in the final? Would you have seen Dortmund breaking down as it did in Klopp’s last year with the BVB? And would you have seen Claudio Ranieri being sacked from a flailing national Greek team only to strike back coming out of nowhere with his Leicester to win the Premiere league? And if we can’t figure out the determinant’s of success and failure of such simple small firms whose business model (11 guys in shorts trying to put the ball in the net) is as elementary as it is  then what chances do we have to predict the socioeconomy?

We are now 12 days away from the UK referendum, an event which can shiver the timbers of the European experiment, and we have no way of knowing what the outcome may look like. So where might we turn to get a glimpse of the immediate future? I know of four possible ways to do so:

The bookies

The return of placing one £UK on Brexit has been fluctuating between 2 and 3 £UK with about 56 percent of the wagers being placed on remain and 44 percent on leave. What do we learn from this? If you believe that when a sufficient number of people put their money where their belief is  then they tend to collectively perform well in the Keynesian beauty contest then you might infer that the outcome of the UK referendum will be to remain in the EU. There appears to be a general belief that the bookies “know” but they have been known to have failed spectacularly before.

The pollsters

The Financial Times Brexit poll tracker currently averages the poll outcomes to 45%-43% for stay with 11% undecideds with recent trends favoring the leave vote whereas The Economist currently has it at 43%-42% for leave with 12% undecideds. What do we learn here? A significant and potentially decisive share of voters refuse to reveal their decision or don’t have one yet. Polling does not tell us much although there does seem to say that among those that reveal their preferences on the average we see a small but diminishing lead for stay. I have computed that among all polls if you write both leave and stay as a function of the undecided share then the variation of stay correlates better with the undecided vote than the leave does. Also a drop of 3 percentage points in the undecided vote appears to assign 2 points to the stay vote and 1 point to the leave, averaging out among all polls.

Google Search

Leave related searches are ahead of the stay searches although the difference has been dropping. The searches are still exploratory in nature and do not identify voter intend. An exploratory search is “reasons to leave the eu” or “reasons to stay in the eu” whereas “time and date of the leave rally” or “time and date of the remain rally” would most likely be made by voters who want to voice their corresponding opinion. Depending on the method we seem to see a lead of the leave by anything between 52 and 55 percent.


There is a fourth parameter which can potentially help the leave vote and hence could help us predict that direction. As we approach the finish line the remain campaign fails to formulate a positive vision for the remain option waving the horrors of leaving at the voters. This may very well backfire. The same may happen with European and national officials trying to intimidate the British voter. This might very well be fueling the brexit campaign as we are approaching the finish line.

All in all I thin it is too early to tell what will happen. In the case of the Greek referendum the vote turned to No in the last 2-3 days and it turned in part as a reaction to scaremongering. Whatever the case we should all be bracing for the shivering timbers of the EU ship.

This entry was posted in Big Data, data, Social Media and tagged , , , . Bookmark the permalink.