Psychology: Looking Into the Subconscious and Ignoring the Conscious

I ended up browsing on Dan Ariely’s blog. I’ve talked about his views before, but in general he’s trying to use behavioural economics and psychology to show how we aren’t always rational and what potentially can be done to work around that. In reading some of this posts — and in particular some of the times when he answers questions — I was reminded of what I found problematic about psychology in general when I was taking it as part of my Cognitive Science degree (which I’ve pretty much dropped due to life getting in the way): it tries to explain things by generalizations and subconscious motivations instead of looking at individual, conscious reasoning that often really does work to explain a person’s actions. I was tempted to go through some of these as blog posts, but I can’t find a way to link only to the specific posts and the ones that answer questions link to the Wall Street Journal which requires a subscription to read. So, I was going to ignore it, but then I came across one about football that so exemplifies what I was annoyed about that I simply have to reply to it.

In order to keep all the context in, I’m going to quote the whole thing — the graphs probably won’t come through, but I’m not arguing with the graphs — but will reply in-line. The post is by Wendy De La Rosa, Dan Ariely, and Kristen Berman, and is from 2014:

During the last month, the World Cup has captivated the globe, including our team at Irrational Labs. We have watched all 64 games and rejoiced / suffered through each of the 171 goals (not counting penalty kicks). This 2014 FIFA World Cup turned out to be an entrancing tournament: Eight of the “Round of 16” matches went into overtime, four went to penalty kicks, and the final match ended with Germany scoring in the 113th minute!

When we watched the now infamous Germany – Brazil game, we couldn’t help but come up with some interesting behavioral questions. When German midfielder Thomas Mueller scored the first goal against Brazil 11 minutes into the game, many of our Brazilian friends said this is just the start of the game.

And while we all know what happened, we started thinking: Were our Brazilian friends onto something – are there more or less goals and attempts late in the game?

One would stipulate that there is no difference in scoring between the first and the second half. Every goal matters equally, regardless of when it is scored, and players should attempt to score with the same amount of effort and success over time.

This will become important later, but stop and ask yourself about football or any similar sport: does it seem reasonable to you, knowing the sport, that players would try equally hard to score in the first minute as opposed to the last minute in the closest and most competitive games in that sport?

Another hypothesis is that fewer goals are scored in the second half as players fatigue Unlike basketball, where players are often substituted in and out, most of the football players are on the field for the full 90 minutes of play (sometimes 120 minutes if it goes to overtime).

Fatigue can obviously impact defensive abilities as well, as anyone who has watched those sports would know. Again, anyone who actually has knowledge of the sport itself can find a problem with this hypothesis.

Yet another hypothesis is that players score more goals in the second half as they are closer to the end of the game. Motivation research suggests that agents are more motivated as they near the end of their stated objective, whether a marathon or a life altering championship.

I’m really not sure why this works as any kind of hypothesis. Sports are competitive, not solo. The end of the game is really only the end of a stated objective if you’re going to win, not if you’re going to lose. And even in the case of a marathon, the motivation would increase only because you can see the finish line but are also motivated to finish as quickly as possible. In a sport like football, my hypothesis would be that if you don’t think the outcome can be modified you’re more likely to simply slow down and let time run out expending as little effort as possible. Even Ariely et al have to concede that, in general, humans are pretty lazy when they don’t see the extra effort as producing a better outcome.

It turns out our Brazilian friends were right; more goals are scored in the second half! Of the 171 goals scored in the World Cup, 39% of the goals were scored in the first half, 57% in the second half, and 61% in the second half when we include overtime.

Again, this will be more important later, but I don’t think that this would be surprising to anyone who actually really knew the sport.

After learning that players score more goals in the second half, we wanted to know why. There are two ways to increase goals scored: increase attempts or increase skill (measured as goals / attempts). Which one is at play here?

The skill hypothesis stipulates that players are “super humans” who perform best when they are under pressure. Consider German Mario Gotza, a substitute midfielder, scoring the game winning goal against Argentina just seven minutes before the end of the match.

To answer this question, we compared a team’s skill in the entire game to a team’s skill in the last 15 minutes of a game. Our analysis showed that there is no statistical difference in skill when you compare these 15 minutes. This is consistent with an interesting study done by Dan Ariely and Racheli Barkan, where they studied the shooting percentages of “clutch players.” Clutch players are NBA players who are widely regarded as “basketball heroes who sink a basket just as the buzzer sounds.” As it turns out, clutch players do not become better basketball players as the pressure increases in the last few minutes of critical games. Basketball players, like our football players, do not increase in skill towards the end of the game.

Except that we don’t generally claim that “clutch” players are players that suddenly get a burst of skill in pressure situations to make those shots. Rather, we tend to claim that they are players who do not collapse under pressure, and so maintain their focus and ability and so are able to hit that shot at the buzzer. So this reasoning about clutch shooting is incorrect; no one posited it was ever about a nebulous increase in skill.

That being said, in a number of sports you do have players who are pressure players and do better when the pressure is on than when it isn’t. They tend to focus more and so make less foolish mistakes. So there do seem to be such players that we can show both statistically and through observation that are indeed more “skilled” when the pressure is on. Marcus Stroman in baseball is a prime example of a player who pitches far better when the game is meaningful — at least to him — than when it isn’t. At times, he’s seemed bored when it doesn’t matter but is very focused when it does.

So if it’s not a question of increased skill (% conversion), it must be a question of effort (number of attempts). Given this finding, we decided to analyze the number of attempts made by players, and whether they increase as the game wears on. Which team is attempting the most goals at the end of the game? One hypothesis is that the leading team increases their attempts as they are more confident and have strong momentum behind them. The other hypothesis is that the trailing team would attempt more goals at the end of the game because the cost of losing is more salient to them.

Depending on the circumstances of the game, either can be true. A team that is dominating and has a clear lead that should win them the game, they probably would be able to ride that momentum and generate more offense, especially if their opposition is disheartened by the fact that the game is out of reach. This is one of the reasons against “running up the score”, because if a team is out of the game and having a tough day it’s only dedicated pride that can stop the other team from rolling right over them. If the game is close, however, then the trailing team is more likely to be desperate to score and so will be trying to press to score the goals they need to win, which isn’t a motivation for the team that’s leading.

Again, just watching a sport regularly will give you these scenarios, with the commentators specifically pointing out the reasoning. Ariely et al will not actually look at that to come up with their explanation.

We can see loss aversion playing out in golf green. According to researchers, Devin Pope and Maurice Schweitzer, “Golf provides a natural setting to test for loss aversion because golfers are rewarded for the total number of strokes they take during a tournament, yet each individual hole has a salient reference point: par (the typical number of shots professional golfers take to complete a hole).” After analyzing PGA Tour putts (the last shot before a hole), they noticed that golfers are extremely loss averse. Golfers make putts for birdie (one shot less than par) significantly less often than identical-length putts for par (getting to par). The researchers estimate that this loss aversion costs the average pro golfer about one stroke per 72-hole tournament, and the top 20 golfers about $1.2 million in prize money a year.

This is stated in an incredibly convoluted way, but here’s what I think the gist of it is: golfers miss shots for birdie more often than shots for par, even if you control for length of putt. This is attributed to loss aversion: the idea that we, subconsciously, pay more attention to losses than to gains.

In golf, there are only two ways to lose ground wrt your competitors. Either they shot under par on a hole, or you shoot over par on a hole. Thus, any time you miss par, you lose ground against your competitors, but shooting over par doesn’t necessarily cause you to gain ground (they could make a birdie as well). So making par is almost always more important than making a birdie. So, in those cases, we can presume that the golfers are often taking more time to line up their shot to avoid actually losing ground. The only exception to this is when a golfer desperately needs birdies to catch a golfer ahead of them … but then this becomes a pressure situation which can cause misses. That the average pro golfer loses one stroke to missing a birdie that they might have made if it was for par seems pretty reasonable to me: early in the tournament, they’re likely to be more casual about birdie putts as they aren’t as important as par puts, and later when birdie putts become more important that’s also when there is the most pressure on them to make them, more than there was for those par putts early in the tournament. You don’t need to appeal to any kind of generic “loss aversion”, but instead to normal psychological processes specific to the game and the situations they find themselves in.

Does this theory hold true in football? After analyzing the data, we found that during the last 15 minutes of each game, number of attempts made by the trailing team (as a percentage of their total attempts made during the game) increase, while the number of attempts made by the winning team decreases.

Is anyone who follows football at all surprised at this? I mean, you can just watch the game and note that this is happening.

They are then going to try to explain this with loss aversion.

Why is this? We believe we can explain this phenomenon with loss aversion. Loss aversion is the behavioral economic concept that states that we value losses more than we value commensurate gains. The other stipulation in loss aversion theory states that we are risk seeking in losses and risk averse in gains. In other words, our risk appetite increases when we are losing. This phenomenon is known as “risk shift.”

The feeling of loss aversion is heightened in football. Think about the cost of a goal in football compared to the cost of a basket in basketball. Because goals are so difficult to make, the cost of giving up a goal is greater than the cost of not making a goal. Thus, teams are naturally more defensive, focused on avoiding “a goal” or a “loss” much less than “scoring” or “gaining” for most of the game.

Loss aversion is a powerful concept, and we are all susceptible; even world class football players. So to go back to our Brazilian fans, expecting more effort as the game continues is a reasonable expectation. Unfortunately, and as the Brazilian fans found out, sometimes effort is not enough.

So, they claim that it’s the concept of “loss aversion”, tying it to the general cost of giving up a goal vs getting one and all of that … and completely missing that in a close game in sports like football the team that is leading playing more defensively and the team that is trailing pressing offensively is actually an explicit strategy. In a close game, teams that are leading deliberately go into defensive shells in an attempt to do what is explicitly called “protect the lead”. They don’t need to score to win the game, and only need to prevent goals from being scored on them, so they pull back. A big part of this is that pressing offensively can indeed leave a team vulnerable defensively, and so make it easier for them to be scored against, so they want to minimize that as much as possible. The team that is trailing, on the other hand, had to take the chance of being vulnerable defensively because they need to score. Players, then, press forward more often, faster, and from positions where they wouldn’t and don’t come back defensively as quickly to help keep their offensive advantage. And it’s commonly noted in these situations that this puts them at risk for a counter-attack that could ultimately seal the game for the other team, but they have to do so to even have a chance at winning. And it’s also commonly noted that the team that plays to protect can indeed give up the momentum and, if the other team scores to tie it up, can end up not being able to switch to a more offensive mode quickly enough to win the game regardless.

These are not things that observers of the game merely note without explanation. These are not even merely things that they’ve come up with an explanation for. These are deliberate strategies that coaches instruct their players to do. And these strategies being in play would explain the results that they tried to appeal to the generic and subconscious “loss aversion” to explain. It’s this sort of thing that really annoys me about psychology at times and makes me believe that it will have a hard time explaining human behaviour. Yes, we have subconscious biases that might have an impact that we didn’t expect, but you can’t simply rely on those when there is conscious reasoning and strategies that need to be considered, and when individual beliefs and circumstances matter to the outcome. Too often, psychology tries to cut consciousness out of the picture, and when it does so it quite often gets things incredibly and ridiculously wrong.

One Response to “Psychology: Looking Into the Subconscious and Ignoring the Conscious”

  1. Hypotheses vs data | The Verbose Stoic Says:

    […] I’ve criticized Psychology in the past, despite or perhaps because I took a fair bit of it both in my Philosophy courses and having taken actual courses when I was doing Cognitive Science. One of the the things that always bothered me about it was that in an attempt to be scientific it ended up adopting one of the major flaws in at least some sciences: too much focus on extracting conclusions from data and not enough time examining the conclusions logically to see if they held up, both by looking for confounds in their hypotheses and also by looking to see what conclusions actually follow from the hypotheses they’re working on. I pointed out one major confound in a classic Cognitive Psychology experiment, and there’s all sorts of issues with the classic “rotating figures experiment” that are obvious to me because I am incapable of actually rotating figures and so screw up the results any time I take it. […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: