Backing both teams to score during QPR v Liverpool | Betting Stats
The bet BTTS, which stands for 'both teams to score' is very popular at the moment with the betting public. However, the chances of winning such a bet are smaller than you might think. The average number of goals per game this season is 2.86. In effect, this means one goal roughly every 32 minutes.
Waiting 32 minutes for a bus to arrive would probably quickly become rather boring and frustrating.
The media and some tipping websites have tried to play on the aversion to such boredom by treating betting as a social event like going to the pub with your friends. Many of us would not dream of walking into a bookmakers but betting on a game from our laptops, with the "trigger bet" coming from our social network's "advisory service" may seem like a far more attractive proposition.
I have been researching the biggest "social tipster" and they already have 9000 clients and 23 000 Facebook friends. This is what they do:
They set up a site and get a PR company to promote the idea of "sharing information and winning together", as when we do things in life, like watching our team, we want to feel part of a group.
Looking at my timeline on Twitter I see a recent surge in the amount of tipsters advising punters to take a gamble on BTTS and I thought that this was worthy of investigation. I found out today that the leading tipster in terms of clients advised QPR v Liverpool BTTS.
I am not going to go into predictive goal expectation modelling as I do not want to bore you. Again, though, it is this aversion to boredom which the tipster with 9000 clients relies upon. There is no need for such tipsters to give a rationale for a bet so long as people discuss it during and after the game. This so-called 'un-marketing' means that the betting company uses its own clients to say how brilliant it is, increasing the customer base through word of mouth, which in turn repeats the process over and over again.
As a marketing graduate, I feel that this idea is the best sporting advisory service model I have ever seen and I applaud the company behind it, only wishing that I had had the idea first.
Now let's look at the idea of BTTS in practice. If you had punted for BTTS in today's QPR v Liverpool game you might have been feeling very confident after one minute of play because Liverpool had scored, meaning that you would only need a goal from QPR in the remaining 89 minutes plus injury time.
Academic research has shown that when the first goal of a game is an away goal and it is scored 0-35 minutes, this increases goal expectation with a bias towards the home team scoring more then expected before the game started. In simple terms, if your model decides there will be 2.6 goals and there is an early away goal then you should expect more then 2.6 goals full time. In fact, in away-team early-goal games this season, we see an FT goal average of 4.07, which is a 30% increase upon the average.
The percentage of BTTS games when the away team scores first between kick off and 20 minutes is 90%. So why did QPR not score? The answer is simply that they did not have one shot on goal, with enough strength, despite a statistically significant expectation of a goal. If it was a boxing match then it would have been stopped at HT.
In conclusion, times are changing with more data available and the media seizing on of the idea that betting is about sharing and using data to make informed decisions. My warning is that the decision is only as informed as the provider of the data. Next time you are in the pub or on Facebook, and your friend says that he or she was unlucky not to win on their BTTS bet, remind them that zero chances and zero near misses (as QPR had today) has nothing to do with luck.