A new FUT-blog has seen daylight: Under the strange name of futingfeefa.com, an old “aquaintance”, Arlington69, is looking for “proof” that scripting, handicapping and momentum exists.
As you may recall, we reviewed some of his work earlier here on FUTfacts. It is fair to say that we weren’t blown away back then.
In one of his recent blog posts, Arlington69 claims to have data confirming that “having more shots on targets reduces the chance of scoring from those shots. Having fewer shots on target increases the chance of scoring from the shots.”
So, if you shoot too much, EA will intervene and make your shots less effective in order to prevent you from getting too far ahead of the opponent.
We were of course curious to see whether Arlington69 finally managed to deliver the proof he promised us. It was time to let his latest piece undergo one of our famous fact checks.
The purpose of Arlington69’s latest experiment is to determine whether the shooting accuracy is reduced when a player has more shots on target. In a sample of 1141 of his own matches, he counted the number of goals scored and the number of shots on target. He divided his matches into the three groups below and calculated the average conversion rate (i.e. the percentage of shots on target which ended up in the net) for the matches in each group for him and his opponent.
|Match group||Percentage of shots on target scored (own)||Percentage of shots on target scored (opponent)|
|Less than 6 shots on target||41 %||44 %|
|Between 6 and 11 shots on Target||35 %||39 %|
|More than 11 shots on target||32 %||30 %|
What you should notice in the table is that the conversion rate drops as the number of shots on target increases. This time, Arlington69 also did some statistical significance testing to beef up his conclusion.
The observations above has lead Arlington69 to conclude that having more shots on targets reduces the chance of scoring. A conclusion that, if proven true, indeed could appear to support the assertion that the game uses some kind of rubber banding to keep matches even.
There is however a tiny problem.
In scientific terms, the purpose of Arlington69’s experiment is to test whether there is a causal relationship between (1) having many shots and (2) having a low conversion rate. But in order to test that, we need to establish what the conversion rates would look like for our three groups if everything was completely normal. This is what scientists refer to as the “null-hypothesis”.
Arlington69 perhaps thinks that the null-hypothesis is so obvious that there is no reason to present or discuss it. Nevertheless, it is clear throughout his post that his experiment relies on the assumption that the conversion rates under normal circumstances would be the same for the three groups of matches.
But as already mentioned, there is “tiny” problem – with that null-hypothesis: The aforementioned assumption doesn’t apply to real football matches(!)
Using a sample of 1572 matches played in England (EPL, Championship, League 1 and League 2) in the 2019-20 season, we repeated Arlington’s experiment.
The results can be seen below, although we note that we allowed ourselves to present the raw data rather than obfuscate it by dividing it into three arbitrary groups.
As clearly seen below, the same “mysterious” trend that Arlington69 found in his FUT data exists in real football as well: The more shots on goal, the fewer goals per shot on goal.
|Shots on goal (Home team)||Average conversion rate||Matches (n)|
It is difficult to pinpoint exactly why teams tend to experience lower conversion rates in matches where they take more shots. One explanation could be that there in practice are two ways to obtain a lot of shots on goal in a match:
- The hard way, i.e. those matches where the team plays well and therefore creates a lot of qualified chances — and typically scores more goals.
- The easy way, i.e. those matches where the team just takes a lot of shots from all sorts of angles — and typically scores fewer goals.
By definition, what we call “the easy way” only will occur in matches with many shots, meaning that it will occur more often in Arlington69’s “Between 6 and 11 shots on goal” group than in the “Less than 6 shots on goal” group.
Hence, it is inevitable that matches with more shots on goal on average will have a lower conversion rate. This of course not only applies to the English football league but also to FUT.
Nothing in Arlington69’s experiment suggests that a rubber banding mechanism or similar exists in FUT.
Having said that, one of course could argue that the absence of evidence isn’t evidence of absence.
But the broader idea that player stats are “dumbed down” when you go ahead in a match has been rejected earlier.
You may recall chemgate, i.e. the discovery of a glitch whereby certain chemistry styles didn’t work as intended. One of the things we learned back then was how to test whether a player has certain stats in game. For example, certain skill moves are reserved for players who have minimum 86 agility.
Following that discovery, some redditors sat out to test whether stats really were reduced depending on the scoreline as claimed by momentum believers. And as stated in one of our earlier posts, the conclusion was crystal clear: The scoreline has no impact on player stats what so ever.
The most important thing we can learn from Arlington69’s experiment is perhaps that it by its design tells us something about why people believe in things like momentum: Part of the reason is without doubt that many FIFA players have a limit understanding of football.
Arlington69 expects to be able to maintain the same conversion rate when he has more shots. And when he plays the game, it doesn’t work that way.
But rather than question the validity of his expectation, he immediately concludes that something sinister must be going on with the game.