In a recent post, redditor ruho6000 claims that you can “beat dda/script/momentum” by going “as dumb as you can in the first half”. Interestingly, he argues his case by presenting a small statistic. The statistic allegedly shows that when he deliberately chose not to attack and to give away possession in the first half, his performance improved on certain parameters. As proposed in one of the comments to ruho6000’s post, this would appear to fit with the idea that “the game recognises you are struggling” and therefore will “adjust the ‘sliders’ to help you in the 2nd half“.
We have already made numerous experiments aiming at testing this exact claim. So far, all evidence points in one direction: The game doesn’t adjust any sliders if you trail behind – at least not in online multiplayer game modes. So, did ruho6000 find evidence suggesting the opposite? We of course needed to investigate.
Ruho6000 is under the firm perception that FIFA will attempt to level the playing field if you are trailing behind. The main purpose of his post is to provide evidence that you can circumvent this leveling mechanism by applying certain tactical steps. Also, evidence indicating that you can circumvent such a mechanism could be seen as evidence to confirm the presence of said mechanism, although ruho6000 doesn’t bring that point up explicitly other than arguing that “[i]t’s obvious to me why there is such a dumb difference”.
To test his idea, ruho6000 noted down some of his stats during two consecutive weekend leagues (2 x 30 matches). In the first weekend league, ruho6000 maintained his normal tactic. In the second weekend league, he allowed the opponent to maintain possession and avoided attacking in the first half. Additionally, he started out with cheap silver strikers. In the 2nd half, he substituted his normal strikers on and went back to his normal tactics – whatever that was. It follows from ruho6000’s comments on his own post that he is pretty effective when it comes to defending “I also knew I am a good enough defender that I can just defend a half without giving more than a goal”.
Below, we have presented ruho6000’s results in a table:
|Weekend league 1
|Weekend league 2
(going dumb in the 1st half)
|Save percentage||24% (Burki)||68% (Pope)|
|Finishing (first shot)||87%||90%|
|Conceded goals in total||42||29|
|Conceded goals from bounce backs||18||6|
|Rage quits (opponents)||8||11|
What you are supposed to notice here is the increase in save percentage as well as in conceded goals — including fewer goals from bounce backs. On the offensive side, ruho6000’s finishing became better, although he doesn’t define what he means by “finishing”. Presumably, it corresponds to “shot accuracy”.
Additionally, ruho6000 notes that “This method allowed me to finish elite 2”. In the comments, he adds that “I finished E3 last weekend. I normally finish around Gold 1-E2.”
The first point that catches our attention is something that is absent in the statistic above: Namely goal scoring and win ratio. It appears that we are supposed to comply with the idea that football is all about making saves, conceding fewer goals, better “finishing” (whatever that is) and making the opponent rage quit. As opposed thereto, things like winning and scoring are less important. And as it happens, ruho6000 confirms that he didn’t improve on these parameters when he notes that his method allowed him to finish elite 2 — which is within the boundaries of what he calls normal!
So, a fundamental issue with ruho6000’s claim is that his data doesn’t support it. Nothing presented in his post suggests that he improved or even less so that he “beat dda/script/momentum” by going dumb in the first half.
Furthermore, when someone decides to “keep stats of [their] weekend league games”, but then decides to present only those data that appear to confirm whatever beliefs they were having, it inevitably gives off the impression that the remaining data perhaps told a different story.
With small samples like these (we will get back to that later), you often get results pointing in all sorts of directions. Obviously, that isn’t an open invitation to pick those data points that support your desired result and ignore the rest. Rather, it’s a clear indication that your experimental design is crap.
Going dumb with small samples
As already hinted, the sample sizes that ruho6000 is comparing is an obvious problem. We are dealing with 2 x 30 matches each. And quite noteworthy, almost 1/3 of those matches ended as a rage quit, meaning that this isn’t a comparison of 2 x 30 full matches.
The problem with insufficient samples – and these clearly are – is that they produce random results. In a nutshell, the observed variations may be the product of natural variation. You simply can’t rule that out with two samples of <30 matches each.
A valid tactic
A third problem is ruho6000’s idea of “going dumb”. The question that pops up is how dumb it really is. Giving away possession isn’t a bad idea per se: “Whoever has the ball is more likely to make a mistake”, as Jose Mourinho famously puts it. Additionally, bringing on fresh strikers from the beginning of the second half perhaps isn’t a bad idea either. Maybe this does give you a competitive edge.
Ruho6000 states in a comment that “pretty much all of my games followed the same pattern: I had 60-70% possession, 10-15 shots on target, 2-3 goals and the opponent had 2-3 shots on target and 1-2 goals.” Perhaps what really changed here is that he learned to play a less naïve game and apply the same cynicism as his opponents.
Additionally, we don’t really see any reason for EA to consider a defensive approach a sign of weakness or consider it “going dumb” for that matter. So, why in the first place would EA want to help someone out who clearly is doing fine albeit taking a defensive approach?
To cut it short, Ruho6000’s experiment doesn’t suggest that you can improve your performance by going dumb in the first half and it contains no information suggesting that scripting, momentum, handicapping or DDA are part of online FIFA.
We of course offered Ruho6000 the chance to comment on this post and respond to some of our questions. He denied to do so.