Why 90th minute goals are frequent in FIFA

Share

A recurring complaint is related to goals scored in the 45th or the 90th minute. As illustrated by the two quotes below, a lot of people believe that these goals are part of a sinister plan to change the outcome of our matches:

“Why is it there are so many dramatic 45th and 90th min goals? Yes one could argue that you have injury time and are therefore stuck on that minute for a while longer. But its plain too see that whilst you get ever closer to this minute the pressure and momentum ramps up.”
(user Angry_Gooner on reddit.com)

“But in the 45th minute. i think there’s a script to enable bullshit in ,added time and against silver/bronze teams.”
(user justlur on reddit.com)

But are these guys right or is there perhaps another, much less sinister explanation to the evidently high frequency of stoppage time goals?

How frequent are 90th minute goals?

Let me be clear about one thing: In the following sections, you won’t see me dispute that 45th and 90th minute goals happen far more often than other goals. I find it beyond dispute that they do, although te evidence supporting that assertion is somewhat vague. In my 2014 survey, participants responded to how often they experienced decisive goals scored in the 90th minute. The most common answer back then was 5/10, i.e. in every other match.

How often are matches decided by 90th minute goals?

The motive…?

It is only fair to ask people, who truly believe that EA scripts those 45th / 90th minute goals in, what motive EA possibly could have. Although I have seen a lot of people attempt to answer this question, the responses fail to address the key issue: Why those exact minutes?

Personally, I find  it hard imagine that a person’s inclination to buy FUT packs is dependent on when goals are scored. That connection simply seems too far-fetched.

And as for the good old idea that EA scripts our matches to make them more even and help bad players out, I quite frankly have a hard time seeing how goals happening in the 45th minute rather than the 52nd minute could be a solution to that aspiration.



So, I believe it is fair to say that we so far miss a credible motive for EA to script 45th and 90th minute goals in. I can’t rule out that there is a motive, but I find it very hard to imagine. And that brings us on to the other option: That EA doesn’t script those goals in but simply allow them to happen.

The longest minute

Humans have a tendency to speculate about possible conspiracies before having ruled out the natural explanations. In this case, there is a very obvious natural explanation, provided you are familiar with the football rules.

A notable detail here is that when people talk about “45th / 90th minute goals”, they actually mean goals scored while the clock is stopped at 45:00 / 90:00, i.e. during stoppage time. Stoppage time is added due to stoppages during ordinary time. Referees add stoppage time in order to ensure that the match lasts 90 effective minutes of playing time.

As an example, let’s say that the fourth official shows the “5 minutes” sign as below. What this sign actually means is that at least 5 minutes will be added, but the referee may add extra time if new stoppages occur during stoppage time itself. In FIFA 19, where the clock keeps running during stoppage time, it is very common that extra time is added during stoppage time.

4th Official Phill Dowd shows 5 minutes extra time to be played

If the referee adds 5 minutes, we have to assume that only 40 effective minutes were played during ordinary time. Following that, 5/45 = 1/9 of the effective playing will take place in stoppage time.

It follows, that if “the 90th minute” aka stoppage time is 5 effective minutes long, it contains 300 effective seconds, meaning that ordinary minutes only contained 53 effective seconds on average. If we assume that the probability of scoring is constant throughout the match, the chances of scoring while the clock is stopped at 90:00 will be 5.7 times higher than the chances of scoring in any single minute during ordinary time.

On top of that, some people have a broader definition of “the 90th minute” than stoppage time. Goals scored in the 89th minute will figure as 90th minute goals on the match report shows after the match. When you include goals scored in the 89th minute, the “90th minute” all of a sudden has become a very long minute.

However, the 90th minute being considerably longer than any ordinary minute isn’t the only contributor to the excessive number of goals scored during stoppage time.

How the scoring chances increases towards the end of a half

In real football, goals aren’t evenly distributed across the duration of the 90th minutes. In reality, the goal ratio increases throughout the match and of course particularly towards the end of either half. The closer you are to the end, the bigger the chance of scoring during the next minute.

The average number of goals per minute in a match as a function of the time (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0047678)

In the EPL, the goal ratio between the 30th and the 45th minute is 70 % higher than the goal ratio between the first and the 15th minute, and during the first rounds of EURO 2016, nearly 1/3 of all goals were scored in the dying minutes of the match.

Image (c) Telegraph.co.uk (http://www.telegraph.co.uk/football/2016/06/17/euro-2016-on-course-to-break-record-for-late-goals-scored-but-wh/)
Image (c) Telegraph.co.uk

A number of factors contribute to late goals being more frequent in real football: Psychology and fatigue imply that players make more mistakes. Many teams apply different tactics during the last minutes of a match. In particular in a cup-system tournament like the EURO, many teams will attack more aggressively, because every point matters.

I think it’s reasonable to assume that FUT is subject to the same effects. After all, all the needed ingredients are present:

  • Humans, who are able to become nervous or mentally fatigued.
  • Virtual footballers, who will perform worse and make more mistakes due to fatigue.
  • The ability to make tactical changes, where you take a bigger risk in the dying moments.

With regards to the psychological and tactical aspects, it’s worth noticing that not only FUT Draft and tournaments but also FUT Seasons contains a large proportion of knock-out matches.

As explained in our article on the difficulty of promotion and relegation matches, the combination of a very brief season and three different point thresholds implies that anything between 2 and 9 matches per season are deciding promotion, relegation or winning the title. In our article on decisive matches, we assess that most players in fact will experience that significantly more than 50 % of their FUT seasons matches are potentially decisive.

A notable fact to consider in this context is that even if player 1’s match may be indecisive, player 2 may be having his final chance of securing promotion. Hence, if 50 % of player 1’s matches are decisive and 50 % of player 2’s matches are decisive, an estimated 75 %  of all matches will be decisive to minimum one of the involved players, meaning that minimum one of them has a reason to go all out attacking to claim that final, decisive point – or – become nervous towards the end and concede a stupid goal.

Many goals = many goals in stoppage time

A third factor which influences the occurrence of stoppage time goals in FUT is its overall goal ratio.

In a recent 8-0 win, I managed to score twice in “the 90th minute”. Was this scripting? According to some scripting believers, EA scripts matches to make them more even. But clearly, my 7-0 and 8-0 goals didn’t make the match the least bit more even. Considering that I was playing against an opponent who lost 9 out of his 13 FUT seasons matches this year, this hardly was a case of the game helping a noob out against a better player. My example illustrates the inevitable fact that if there are many goals in general, there will be many goals in stoppage time – and in any other random range of time you may define.

A relevant, real life example of this is Liverpool’s 4-5 win against Norwich on January 23rd 2016. The match contained two, potentially decisive 90th minute goals. According to BBC’s detailed commentary, Bassong equalized at 90 + 1:57, whereas Lallana decided the game at 90 + 4:19. The match finished at 90 + 5:33. A notable detail is that the two goals scored in the 54th and 56 minute actually came closer to one another (1:40 in between) than the two goals scored in “the 90th minute” (2:24 in between).

Liverpool4-3

In an earlier article, I counted the average number of goals across 400 FUT 15 matches, which is a fairly large sample. I found that the average was 5.5 goals per match. An average of 5.5 goals per match is extremely high when compared to real football. Among the five big leagues, the Bundesliga has the highest goal ratio with 2.82 goals per match.

At 5.5 goals per match and 1/9 of the goals scored during stoppage time, we should expect to see .61 goals scored in stoppage time per match, assuming that the scoring probability remains constant throughout the match.

Of course, the ratio of goals scored in stoppage time will vary depending on the number of goals and the number of added minutes. The table below shows the expected frequency of stoppage time goals with different combinations of added minutes and goal per match ratios. What you should notice in the table is that even if we look at the more conservative section of the table, stoppage time goals still happen frequently.

Minutes added per half
+2 +3 +4 +5
Goals/match 3.5 0.16 0.23 0.31 0.39
4.5 0.20 0.30 0.40 0.50
5.5 0.24 0.37 0.49 0.61
6.5 0.29 0.43 0.58 0.72

Expected number of goals per match scored in stoppage time

When I collected data for the previous mentioned article, I also counted the average number of goals per match in a sample of 500 matches. The most common ratio of goals per match is between 6 and 7. In fact, 70 % of all matches contained 5 goals or more. Hence, the majority of players should expect goals scored in stoppage time in about every other  game – but for natural reasons.

Goals per match
Goals per match

Why 90th minute scripting is a myth but bullshit goals aren’t

The point I have made above is first and foremost that the very frequent goals scored in stoppage time are accounted for. There is no residual of unexplained goals, which in theory could be put in on purpose. And quite honestly, no one has been able to explain why EA would want to put those goals in so chances that they did are very slim at this point.

There is no such thing as 45/90th minute scripting. It’s a myth!

Having said that, the facts listed above tell us something else about the goals we score and concede in FIFA:

As stated, FIFA has an extreme goal frequency. With 5.5 goals (FIFA 14) in a match which lasts just 12 minutes, we end up with 0.45 goals per minute, whereas the average Bundesliga match has a frequency of 0.02 goals per minute. We are not even close to realism here, and that is hardly a coincidence.

If we got the chance to ask them, I think EA would confirm that the goal ratio is intentional. FIFA probably wouldn’t have sold 12 million copies a year, if the goal ratio was at a more realistic level.

So, while rejecting the claim that EA has a particular preference for 45/90th minute goals, I find it beyond dispute that EA are “at blame” for making it relatively easy to score and hence inevitably also to concede goals. After all, they are in this business to entertain us – not to bore us to death.

So, “bullshit” goals are real, bullshit goals happen during stoppage time – but scripting doesn’t exist.

%d bloggers like this: