The Attack Bias

Let’s talk about a syndrome that has swept across football betting and many other sports; it is called the ‘attack bias’. The ‘attack bias’ is where as humans we naturally favour attacking teams in sports and therefore our ‘bias’ towards that team blinds us and we think they should beat defensive teams (even when both sides are equal in talent).

This bias is explained as we are programmed to be excited about goals, for they are the holy grail of football, and therefore we favour those that score them above those that do not concede them. As a result we are far more likely to think a good attacking team will beat a good defensive team. Manchester United this season are a great example of this bias, United have in my opinion the strongest attacking line in English football but does this translate into being the best team in the league? Quite simply, no, United sit 9 points behind Chelsea and Manchester City and only one point clear of Southampton. So why are they often favoured despite what the stats suggest?

Take last weekend, Stoke (who were coming off a comprehensive win over Arsenal, despite a late comeback) were at home to Manchester United, United were strong favourites despite being the away side at the notoriously difficult Britannia Stadium. United were odds on despite winning only just over 20% of their away games so far this season, averaging 1.2 points per game during that span.

Why the favouritism?

Well two clear things spring to my mind, United’s global reputation is so big that people just expect them to win regardless of form and also the ‘attack bias’ in games where one team has a much stronger attack the bookmakers will often favour that team regardless of how poor the defence is. Stoke will always be underrated in a market against flair teams, Arsenal were also favourites at the Britannia despite their woeful record there over the years.

This is not just a British phenomenon.

Last May defensive minded and incredibly well drilled unit Atletico Madrid had just wrapped up the league title against moneybags Barcelona. Now they were 1-0 up and in pole position to win the Champions League and deservedly so, Real Madrid their opponents had barely had a sniff until an injury time equaliser and tired legs in extra time condemned Atletico to defeat. Real were 1.5 in the 90 minute draw no bet market and Atletico 3.0, yet Atletico had had the upper hand all season on Real (and continue to do so this season), reputation and the ‘attack bias’ were the reason for those warped odds.

Chelsea beat Bayern Munich and Barcelona on the way to an unlikely Champions League title in 2012. They were dominated in terms of shots and possession in both encounters and were huge underdogs on both occasions but won through in the end with a rock solid back line. But these are just a few examples; let’s see what the data says. This study is taken from The Numbers Game by Chris Anderson and David Sally:

"Between 2001/02 and 2010/11, scoring ten more goals than another team over the course of the season was worth, all else being equal, an additional 2.30 wins, while conceding ten fewer goals was worth 2.16 additional wins in the Premier League."

Very similar numbers with attack very slightly shading it. However, when looking at losing matches ‘scoring ten additional goals reduced defeats per season by 1.76 but conceding ten fewer goals reduced defeats by 2.35 matches.’ That is a much larger margin than the first example. Therefore, goals not conceded according to this study are 33% more valuable than goals scored.

I want to compare this now with my second favourite sport American football. In the NFL 6 of the last 10 Super bowl champions were teams known predominantly for a tough defence. Seattle, Baltimore, New York Giants and Pittsburgh all won the greatest prize in American football by beating the arguably better offences that opposed them (4/6 of those teams were underdogs on the day, ‘attack bias’?). ‘Defence wins championships!’ was the phrase banded about to anyone who played Madden NFL 2005 and our American cousins had caught onto something. It is time football and we as punters caught up.

I am not saying that a good defence is better than a good attack (although stats do seem to slightly suggest this), I think they are both equally valid. I am highlighting our natural ‘attacking bias’ so that we can bare it in mind when placing a bet or trade, if we can learn to let go of the bias then we already have a big edge on most of the people betting in the same markets.