But but but show me another one... This is beginning to look like the Monty Python "but what else have the Romans done for us!?" scene...

Anyways, I do believe that we are running bad with this, however... Any manager worth his salt would cry a river about this and hand out A4 printouts of the red card stats to innocent bystanders at press conferences.

That would subconsciously influence the refs and it would turn into unintentional positive treatment... Make it happen Lego Pep!

Lmao @ reds and penalty differential should converge to zero. One of the dumbest things ever said itt.

QFT. Especially as the short period of time analysed has obviously been cherry-picked to prove otherwise.

Yes, that is dumb.

Let us not be distracted by that and focus on my objectively funny post instead

the arguments that you would make regarding the null hypothesis of the differential of redcards/penalties should converge to zero would likely be the following:

teams with more touches in the box would get more penalties, teams that concede less possession in their box give up less penalties and this can be adjusted for.

here's 18/19 data that shows an almost non existent correlation to touches in the box and winning penalties:

id be pretty confident there isnt a strong enough correlation in a larger dataset to be statistically significant that more touches in the box means more penalties.

same thing with fouls, teams that draw more fouls you would say should have their opponents receiving more red cards, and teams that concede more fouls would draw more red cards.

this also shows very little correlation from the data ive briefly looked at (i can make a graph later when i want to put more effort into it).

I wonder how much reputational feedback loops affect this. After the Kane Vs Robertson incident the refs said openly that they judge fouls based on player reputation/past red card history, which obviously means that since Kane wasn't red carded in the past he's less likely to be red carded when he deserves it next week and even less likely than that in the future. Vice versa for someone like Xhaka.

If you get a few players with bad negative feedback loops you could end up with a bunch of marginal decisions going against you.

Exactly!

This is why you need to cry about it on all possible channels. Get on it LP!

i was actually about to post something regarding this. its a bit harder to quantify, but the way you would go about it would probably be:

what rate did the player get red cards/yellowcards/penalties concede before joining or after leaving a team, and then when they were on the team. if the rates are similar enough then its an issue of you getting a player with known issues should you have a problem with the frequency of these results. if the rate with a specific team is higher than his historic rates with other teams, then your argument is better that the team is being targeted. ultimately i think on a per player level theres likely not enough of a sample size to prove statistical significance either way.

id probably be inclined to agree a players perception/reputation effects how calls are made against them, but its not significant enough to explain all of the variance in the numbers.

How can you say the correlation is almost non-existent when there's literally a trendline in the graph going up to the right lol. r2 (how the **** do you do squared on a keyboard) is really low I suppose, so my bad.

Sample size of less than one season is obviously an issue, I can't find data from opposition touches in box outside of that season. Red card vs fouls do seem more correlated but still low

Thank god the international break ends tonight

fyp

Don't just show us a graph, run it as a regression and tell us if it's significant/what the standard error/odds ratio is etc.

I hope there is a Friday night game to put an end to all this bullshit?

There is, but reading the bullshit again will probably be preferable

This is undoubtedly high quality work. However, the limiting factors are rarely considered. For example, the ability of the players/coaches to understand wft the stats nerds are going on about and the intelligence of your average footballer to implement it at the appropriate in-game situation.

Fortunately baseline without best player in the world and Mane still high enough to comfortably dispatch with league dross such as Arsenal.

Lol ducksauce, I'm sure if prime Barca played Aldershot FC in a 1000 game series, the penalty differential would tend to zero.

comparing penalty stats over a season or even several is just so useless when you think that a team averages a penalty every what, 5 games at best? same for red cards. it's the equivalent of saying my team runs bad because they had more fouls whistled against them in the last 25 minutes

then while penalties have the same probability of being scored, they're not created equally and behind the penalty calls, there are all kinds of flavours : i.e. ones that prevent an almost definite goal and others because a non-dangerous cross hit the ball of the defenders 20 yards away from goals, the stonewalls ones, the in between ones, the bad ones, etc

then we have fans using those cherry picked stats to try to prove their point about unlucky their team is (because obviously my team is always unlucky) and it just get aids

reminds me the time where man united fans were saying that them getting 20 penalties a season wasn't an outlier

This conspiracy theory would explain all those red cards under Wenger

There I was thinking they were just dirty ****s

Arsefan collectively going full joediego is just the spice the thread needed.

I love that we are now applying lol xg methods to fat bloke our team is the only team that's hard done by whining

I had forgotten about the Partey red card and just got reminded. It has really brightened up my day.

My favorite part is that the data does show a correlation. It isn’t extremely strong, probably because one season is not much data as mentioned above. There will obviously be outliers over one season. There are also more reasons the statement was dumb, including non-conspiracy driven ones.

All this talk now of bias due to reputation. What about the fact that the perceptions of who is a dirty or foul-prone player are often correct? David Luiz didn’t just get a lot of red cards just because he was a big-haired foreigner - he was also a brash idiot and in physical decline during his Arse years.

There are also a few players that go years with hardly any fouls or cards - surely there is more in that than luck or ref bias. Having a player like that on your team will obviously impact the results.

Anyway - I’m ready for some relegation-ball later. I’m predicting one penalty apiece in this one.

Didn't think he'd play after just coming back from Africa but his enthusiasm on the sideline must have convinced Arteta to bring him on