How loss aversion can make football more exciting
It’s a familiar sight now – pantomime injuries, blind throw-in takers, and very deliberative goalkeepers. All ways to waste time, or in managerial parlance: ‘slow the game down’. Amongst fans, its known as the dark arts, but is perhaps most aptly described as ‘killing the game’ – and unfortunately, it might just be doing that.
An insult for neutrals and fans alike – during Arsenal vs Newcastle, a top-of-the-league box office clash, the ball was in play for just 53 minutes. This was first against third, an in-form Arsenal against a Newcastle side with supposed genuine ambition. The final cracker of the festive period. But Newcastle came with a plan: kill the game.
The frustration amongst fans isn’t new, and there have been recent attempts to defend against these dark arts. Refereeing officials in Qatar limited the impact of time wasting, where we saw stoppage times frequently hit double figures – more accurately reflecting the amount of time the ball was in play. Others have called for a more revolutionary approach – adopting a rugby-style stop-clock which pauses time for certain stoppages (although this is not without controversy).
However, there seems to have been less exploration into the flip side – rather than disincentivising negative play, how do you incentivise more attacking football. And what can cognitive psychology tell us about how to do this effectively.
It isn’t revolutionary to adjust incentives when trying to encourage attacking football. Pre-1981, a win would secure you just 2 points, whilst it was still 1 point for a draw and none for a loss. The ‘Three points for a win’ was thus introduced to greater incentivise more attacking play.
In the five seasons before the switch, there were an average of 133.0 draws per season (in what was then the first division); in the five seasons after, there was an average of 113.4. Of course, any causal analysis on what the impact of this switch was – particularly on the game today – is nigh on impossible, and there would have been many effects beyond the number of draws. But it does seem to suggest that changes to the incentivisation structure can impact how football is played.
Just two years before this intervention, two psychologists, Daniel Kahneman and Amos Tversky, formulated Prospect Theory – a behavioural model that describes how people make decisions under risk. Unlike more classical models of decision making – the foundations of prospect theory were rooted in data on how humans actually behave.
A key cognitive phenomenon that the value function demonstrates is loss aversion – we are more sensitive to losses than equally sized gains. In the words of Tversky & Kahneman: “losses loom larger than gains”. Would you take a gamble with an equal probability of either gaining £20 or losing £20? Evidence suggests that people would avoid this gamble until the potential gains are twice as large as the potential losses.
Loss aversion would explain why the original ‘two-point for a win’ system led to a negative style of play: when drawing, if attacking led to say a 50% chance of scoring and gaining one point but also a 50% chance of conceding and losing one point – then the loss would loom larger than gain, psychologically encouraging the protection of the one point you have over trying to gain another.
What loss aversion can also account for is our risk preferences. When making gains – we are risk-averse: we have a tendency to prefer a sure gain over a speculative gamble, even if the average outcome – or expected value – is higher in the latter.
Consider the following…
Most people would select option A, despite the possibility of a higher reward and higher expected value in option B, to avoid the risk of ending up with nothing. In footballing-terms, we prefer to hold on for a point, then risk going for three and possibly leaving with nothing.
But what the loss averse value function also explains is how our risk-preferences switch depending on whether we are dealing with losses or gains. When confronted with potential losses, we are risk-seeking: we prefer a risky gamble over a sure-fire loss.
Let’s reconsider our choice experiment from before...
Most people would select option B. In an attempt to avoid a sure loss, people become risk-seeking and take the gamble, despite option A having a higher expected value. The switch from gains to losses has led to a reversal of our risky behaviour.
So, what next? If you are concerned by the negative style of play today – would a four-point win prove better? Perhaps not. Draws shouldn’t be overly punished – watching a team press for a last-minute equaliser is good fun, and it would hugely encourage the winning team to ‘park the bus’. More importantly perhaps, fewer points awarded for a win means a closer title-race and relegation battle. A subtler change is needed…
What about changing our reference point: if instead of starting on 0 points, gaining three for a win, 1 for a draw, and 0 for a loss – we started on 114 points and lost 3 for a loss, 2 for a draw, and retained all our points for a win. According to prospect theory, we would be more inclined to seek the risk of losing more points, for the chance of dropping no points at all.
Imagine. It’s the 70th minute, you are on track for a point away from home against an evenly-matched side – are you going to risk losing that point for a chance at three? How about now, it’s the 70th minute you are on track for losing 2 points away from home against an evenly-matched side – are you going to risk losing three for a chance at losing none at all?
The two scenarios are equivalent. All that has changed is the reference point – how an outcome is perceived in terms of gains and losses – but subtle framing effects can lead to big changes in decision-making.
Prospect theory would predict that, for some teams in certain situations, the loss-frame would be enough to encourage them to go for the win instead of otherwise holding on for a draw. If that’s what you are after – a prospect theory-informed loss-framed point system may be your answer.
Comments