Jecross does have a point. If you look at Pardew's last 3 seasons, there are long runs of games (within and between seasons) where the team are either picking up a lot of points, or not many.

Whether there are better explanations than he's a streaky manager (transfers, favourable fixtures, momentum, external factors, etc), or whether this type of occurrence is normal, or if it dates back over his whole career, is a question for people better at stats and with more time on their hands.

However, the pattern does suggest there are probably better explanations than "random variance".



(The column with 1 and 2 refers to first 19 and last 19 games respectively).

There were some (very believable) rumours doing the rounds about the circumstances that gave rise to Pardew getting that job.

Are you referring to the story about Ashley doing competetive boozing against a Polish analyst and then puking into the fireplace immediately after he'd won?

No, that was far better than this one.

That also was a little more than a rumour, given it was told under oath in a courtroom.

cleverleyshareprice.jpg

Yeah going to need more than this?

i believe the pards rumour is that that got pissed up and borrowed/blew through a load of mike ashley's money one night in a casino and couldn't pay it back, so they came to an arrangement which involved pards signing an 8 year contract for 600 quid a week or something

Yeah but that was like part a) of his argument, which despite being difficult to prove, is far easier than part b) that the manner in which he fosters his teams' spirit has a causal relationship with the streakiness.

That table means nothing. I could arbitrarily pick a week, say week 24 of a season and show you stats from the previous 19 games and the next 19 games which may show that Pardew is a perfectly consistent manager.

Yeah, I agree the "team spirit" argument is bollocks, as it's just not that easy to build team spirit if you are the one responsible for arsing it up.

e.g.

Go on then. Using week 24 as a mid point (with week 5 as the other midpoint), construct the same table for the same 3 seasons.

Over more than 1 season? Go on then. Also he didn't select a week to make a point, he picked the midpoint.

The team spirit argument was a suggestion as I can't think of a better one, and it was reminiscent of Keegan who was dominant at Newcastle until it all fell apart- happy for people to put forward others, but the phenomenon is pretty clear looking at his career.

Maybe he's just a great manager but a despicable **** and it takes him a while to get to know his player's wives for the latter to show through.

Another suggestion could be that he comes up with the odd tactical masterpiece but is too slow to change it when he gets found out.

Are you daft? I am not claiming that he is not streaky. I am arguing that his streakiness is not statistically significant. Now you come in with some logic-defying numbers that state he is streaky, so I couldn't resist calling you out on it. For instance - if a team went W10 L9 W9 L10, your week 19 stat would say that team is super consistent.

See above. I am not trying to say that he is not streaky. Even if he is streaky, the midpoint reasoning is stupid.

Or variance?

Fair enough. How did you calculate statistical significance?

I didn't and don't want to. I am just saying that it should be the default unless proven otherwise. He claimed that Pardew was streaky due to team bonding and Pardew not having a beer with the fans or Pardew dancing. I just pointed out that the most likely answer is variance and that it is so easy to write a narrative like most of the ****ty pundits do.

fyp.

Why are you so desperate to find an excuse that isn't either variance, or one that doesn't fit a nice linear narrative (like a combination of smaller issues relating to the club/players/external factors)?

If we track 100 managers over 10 years, and we say that Pards is average (which is pretty generous but whatever), and the average is 33% win, 33% draw, 33% loss, some managers will get quite quickly to those %s and mostly stay there whereas some take very streaky routes. There doesn't have to be a causal reason for this, sometimes it's just how the chips fall.

tbf if the gambling rumour is true then pards clearly did his bollocks in a major way that night, which lends credence to the idea that he's an abnormally high variance individual

Making dreams come true.

Gosh, comprehension skills lacking ITT. I even gave it to you that he is streaky. All I said was that your method of showing it was garbage. If you want me to actually show you that he is or is not streaky within a certain confidence interval with a certain probability, I could do that. It is not worth my time though, unless you bet that I cannot do it.

This isn't even my argument. You just keep nit-picking at one little thing, without defending your original statement, which was wrong.

No, I didn't. I claimed he was streaky (in a statistically significant way), and put forward his ability to motivate players (i.e. harbour team spirit) as an explanation for that, in absence of me being able to think of another one - given he seems to me to be a complete tool.

since 2012/13 season pards has a record of just 4 wins out of 18 league games during april. this 22.2% win record is approximately half of his average win record over the same period in all remaining months

i offer the theory that pards is the easter bunny and that his egg-hiding duties negatively impact his ability to deliver motivational team talks. prove me wrong