Most bookmakers offer a tempting line, generally between 14/1 and 30/1, on both teams to score in both halves (hereafter BTTSBH) in a given game. Given the generous odds on offer, I wanted to take a look at how often this occurs, and what factors may be important, using fairly simple data from football-data.co.uk.
I’ve used information from the Premier League, from 1995/96 up to but not including this latest round of fixtures (beginning October 25th). This gives 7300 matches for us to look at.
Out of these 7300 matches, BTTSBH occurred 375 times, or in roughly 5% of the total matches. Based on this very basic stat, the odds should be around 20/1.
Lets split this up by season. How stable is this 5% figure?
I’ve included data from 2014/15, hence the large % of BTTSBH occurrences this season. After 80 matches of this season, there had been 10 of these matches, accounting for 12%. This is unusual within this sample of seasons, and I will return to this later.
Are there any teams who are particularly likely to be involved in these matches? The chart below shows all teams who have spent 5 or more seasons in the Premier League since 1995/96, with the bars representing the total BTTSBH occurrences each team has been involved in, divided by number of seasons. So, in essence, the number of BTTSBH occurrences per season.
Of the current Premier League teams, Stoke appear marginally more likely at 0.8 over 7 seasons, with Sunderland pretty high too. Birmingham averaged one BTTSBH occurrence in each of their 7 seasons in the Premier League. The differences on show here aren’t really that great though.
Finally, is there a particular time of year when these types of matches my be more likely to occur? I had a theory that they may be more likely on the opening day, and the final day, of the season, and wanted to check. Here are the total number of BTTSBH occurrences by month.
Everything is within a pretty similar range here, bar two months. The first is December, where historically there have been double as many BTTSBH matches as in August, October, January, February, March, and May. The second is April, but the increase isn’t that great compared to December.
Whilst it doesn’t appear that any particular team has been historically more likely to take part in these types of matches, there may be some factors that influence teams to score and concede more freely around December. Given the close proximity of games around the Christmas period (a team can play 3 games in 7 days around this time), it may be that this leaves teams more open at the back, resulting in more of these BTTSBH matches. A team will typically play more games in December, so you would expect the rate to be a little higher compared to other months, but probably not quite this high.
Its also worth pointing out that at this point in the season, there have usually been on average 4 BTTSBH matches, based on our sample. As mentioned earlier, there have been 10 already this season. Whether this reflects some broader trend, or whether we will see a great deal of regression to the mean, remains to be seen.
So, if you’re going to take up these odds at any time, it may be worth taking a look at the fixtures around Christmas, and chance your hand. With the odds around 20/1, its definitely worth a try, and this season may be the season to do it.