It's just such an awful, awful post. Nate Silver is probably the most mainstream stats guy out there and he posted/got started on 2p2.

It has been dreadful, hasn't it? Watching Cardiff toiling makes me even more thankful that I'm not Welsh.

i finally remembered what this stats based spudsboom reminded me of - 2013/14 aka the suarez season aka the pool +5 points season

Yea some dreadful stuff. Seen Bristol miss about 6 free headers

I've watched Athletic tear up Gijon instead. No regrets

Yep same here.

Just play out the season on Excel

Totally forgot Scolari signed a multi year deal as well before getting the sack. £16m for eight months. lol Chelsea.

what is a worcester city and why are we playing them

They're a **** lower league team. You're playing them because you're also a **** lower league team.

Think someones made

Our analytics guys passed on Mané 15 months ago. They have a use, but way too much influence at elite clubs for such a young industry.

Me? I'm not mad, I think it's funny!

Pipe down fellow Spurs fans.

With several bigger clubs in transition/disarray, and Spurs team settled enough that we have gone 9 unbeaten including humping the title favourites, it's obviously RAWK-style boom talk for us to consider the possibility of improving one place on last season's finish.

I agree there is far too much of this thinking ITT, and perhaps in football generally. If you're taking a decent critical approach, you consider both inductive and deductive approaches, so sometimes the question is data driven, and sometimes theory driven.



So a good example of using data to drive a theory was Comoli purportedly signing Henderson and Downing because they had good recovery stats in the final third. That information is useful, but it doesn't make them good players.

If you're judging players it's far better to use the the deductive approach, use the eye test first up, and then when you think you've identified a good player, see if the stats agree. If you take the inductive approach and check the stats first, then you will have already formed an opinion about that player and the conclusions of your eye-test will probably be biased - as well as possibly ignoring other decent players on show.

It's such a lazy attack that cinarocket has done over and over again that he seems to think people who like to look at advance stats just completely rely on them.

xG is a great indicator to future success. Is it perfect? No, but it's also being worked on and made better every single day. Caley posts how he does it. He updates how he does it. Other people have slightly different models. I understand how some people don't like it, but read this article by him: http://cartilagefreecaptain.sbnation...051.1444936826

Has something better been written by anyone this season? People who rely on stats can be wrong, but the ones who are worth anything are seeing where they are wrong and continually improving their methodology.

Yeah I use Caley.

ElRazor, I'm finishing up Fergie's latest book, he has a section on stats and he clearly used the deductive approach.

From his method:

So this is exactly what I was talking about. This method is clearly driven by theory and watching football. Seems obvious, but as he says, it's really easy to just throw a lot of numbers into a regression and find things that are statistically significant. Ben Goldacre does a good job of explaining why that's bad here.

Is it good (for a non homer)?

Elrazor, my point is that cinarocket and you as well are railing against things nobody are doing.
As you note, the people behind these stats watch as much football as anyone on earth

Nobody in football, or nobody itt?

Either way, I think you're wrong. People consistently use stats to drive arguments, rather than create arguments supplemented by stats.

Yea, but they don't have masters.

---

Looking at the training pictures and reports, seems like the core apart from Mert, Cech, Walcott will get a complete rest, which I'm delighted about.

Starting team will most likely feature most of the following players:

Joel Campbell, Chambers, Debuchy, Gibbs, Iwobi, Bielik, Flamini and a couple more youngsters

Can't watch the game live, but I hope I can get a spoiler free place to download the game from. Prolly rojadirecta

If you lose I'll happily spoiler you via Twitter. Or carrier pigeon if you have no internet/phone access.