Fact check: Does the momentum swing when a player takes an early lead?

Share

The claim that going 2-0 up will swing the momentum in favor of the opponent has haunted the community for ages. In a recent Reddit-post, Arlington69 claims that he has evidence supporting it. He presents a statistic showing that players, who go two goals up in the first 30 minutes, score fewer goals in the rest of the match.

This naturally caught our interest, so we decided to investigate Arlington69’s evidence more thoroughly to find out whether he really did have the long awaited holy grail of evidence confirming momentum.

The claim

At the center of Arlington69’s claim is a data set containing all sorts of facts from 984 of his own matches. We already discussed and used his data in earlier articles.

Using his data, he made the observation that players, who are minimum two goals up after the first 30 minute, on average see their goal ratio drop from 12.74 minutes per goal in the first 30 minutes to 30.61 minutes per goal in the remaining 60 minutes. In the same time span, their average opponent’s goal ratio went up from 0 goals to one goal per 34.24 minutes.

According to Arlington69, this observation proves that the game actively makes it more difficult to score when you go 2-0 up early in the game.

There are however a couple of Enterprise-class problems with that conclusion.

Criticism

The first problem is that Arlington69 confuses correlation and causality. Players, who went two goals up early on, did indeed score fewer goals later on in the match on average. Arlington69 asserts that going two goals up triggers a momentum effect, which causes a drop in goal scoring. But this corresponds to saying that storks brings the babies because the average fertility rate is higher in countries with more storks. Thus, we definitely can’t conclude that the goal ratio dropped because these players scored 2 early goals.

And as we will show next, there is a very natural explanation for the drop in goal ratio.

Momentum is everywhere

Below, we repeated Arlington69’s experiment, but unlike him, we also repeated it on other 30 minute sections of the matches in his dataset. For example, we looked at matches where either team scored 2 goals between the 42. and 72nd minute and calculated the goals-per-minute ratio inside and outside that timespan-

Selected 30 min section Avg. goals-per-minute inside 30 min section Avg. goals-per-minute outside 30 min section
00-30 min (Arlington69’s test) 0.15 0.06
42-72 min 0.15 0.07
14-44 min 0.15 0.08
60-90 min (incl. stoppage time) 0.17 0.07

As seen above, the goals-per-minute ratio is ~.15 inside the chosen 30 minute section no matter what 30 minute section we looked at and ~.07 in the remaining 60 minutes.

So, if we apply Arlington69’s logic, we can conclude that EA has a momentum mechanism which predicts that someone is going to score 2 goals after the 60th minute and therefore reduces his goal ratio up until the 60th minute. And lo and behold, they also invented a momentum mechanism which predicts that someone is going to score 2 goals between the 42nd and the 72nd minute and reduces the goal ratio before and after that timeframe.

But EA’s magic capabilities are not restricted to FIFA matches. We also repeated his experiment on a set of English Premier League matches. Like Arlington69, we created a sample of all matches, where either team was two goals up in the first 30 minutes. We found that the average goal ratio inside that timeframe was .07 goals per minute, whereas it fell to .01 goals per minute in the subsequent 60 minutes.

Selection bias

Either, momentum exists in the EPL or something is wrong with Arlington69’s method. We lean towards the latter option, and here is why:

When you create a sample of EPL-matches, where either team goes minimum 2 up in the first 30 minutes, you inevitably also get a sample where minimum 2 goals were scored in the first 30 minutes. That’s a lot when the average EPL match contains 2.7 goals, and it inevitably means that only .7 goals on average will be left for the remaining 60 minutes. Therefore, the goal ratio for the team which went 2 up will appear to have dropped.

The same logic obviously applies to FUT as well: When you select matches where 2/3 of a team’s goals are scored in the first 30 minutes, you inevitably will see that 1/3 of it’s goals are scored in the remaining 60 minutes.

The drop in goal ratio observed by Arlington69 is a product of selection bias, not momentum.

Conclusion

Simply put, the goal ratio is lower in the remaining 60 minutes because Arlington69 has picked a subset of frontloaded matches and not because EA has fired the momentum booster.

Still, a lot of people believe that going 2-0 up invokes momentum swings. We already provided plenty of evidence demonstrating that this isn’t the case, and Arlington69’s experiment certainly doesn’t change that conclusion. But it may illustrate why people believe such things in the first place.

Perhaps, one of the reasons is that they fail to understand the full implications of the fact that the distribution of goals in FIFA matches isn’t even and that your goal ratio isn’t constant throughout the match. You may go 2-0 up early on and still lose. There is no law of nature dictating that the goal ratio from the first 30 minutes should remain constant through the match, unless big, bad EA intervenes. Apparently, there are people who just don’t understand this.

%d bloggers like this: