I mean not easy to know how reliable those stats are as the author admits he felt Liverpool were hard done by that's why he did the analysis. Not too mention it success Liverpool have been significantly disadvantaged more then Arsenal so it counters the other position that we are getting hard done by.
I don't fully understand that red card probability calculation above. I tried to see if there was anything else and canon stats did some analysis on January regarding the 8 arsenal reds which he stated
"Using this and 8 red cards for 55 matches now looks like a massive outlier 2.37 mean absolute deviations from the expected mean. This is still just short of the statistical significance threshold at the 0.05 p-value level but is very very close (0.08 here).
This isn’t exactly a satisfying answer to say this is an outlier but still not such an outlier to not be random chance at this point but that is where the data points right now."
So the Arsenal ones seem to be in line with normal variation although he did mention the city ones look less likely