Theo Walcott


#1087

I didn’t really, but then I don’t think you were really reading me (actually, our interaction began, I remember now, with you saying no offense to me about not agreeing with comparing Pires to Walcott, something a4tt did and I made clear to you I never did nor would ever consider doing nor in my craziest fever dreams, lol). And even if I did, without having said, “I agree with a4tt re: his complete stance on Theo”, I don’t have such a low view of people’s intelligence that I feel I must repeat, “Ah, for the record, I don’t agree with Theo’s general stance on Theo, even if I happen to agree with him in this very miniscule facet of this argument, just in case there should be any misunderstandings!”, every time I mention a4tt in a sentence.


#1088

lol give me a break, man… you are an insane amount of work… feel for your gf.


#1089

@ljungbergkamp my mentions of a4tt.

This is indeed the pot calling the kettle black, when you were the one who has managed to persist with that awful analogy of yours and defend it throughout, despite, I’m sure, realising by now how little sense it makes.


#1090

LOL you win… I give up. If you choose to interpret my totally clear arguments the way you do, I have no hope!

Good luck with your future in statistical learnings!


#1091

If the parameters are little bit shit and little bit decent on his day I think that tells you all you need to know.


#1092

You are the only person here that has a leg to stand on when it comes to criticising Theo. When he was up for extension you are the only one who was against it. Everyone else was sweating bullets thinking we’d lose him for free, and then there was a big celebration once he ultimately signed on. All this talk about him losing a step or regressing is revisionist history by his detractors trying to spare their own blushes.


#1093

What does wanting him to resign have to do with anything? Maybe some of us were wrong about him back then? Maybe some of us thought he was still a useful option and feared we would be left with nothing instead? It really has nothing to do with our current analysis of Theo… we now have two full years of data since his last contract… and if we wanted to sign him back then and now think differently, how does that have anything to do with the validity of our opinion on him?

Genuinely curious - it seems reasonable to change one’s position with more data if you think it is warranted, no?

For the record, I don’t even remember my position on him back then. I assume I wanted him to resign because I still held onto hope that he might come good and I did fear we would be left with worse alternative, but I genuinely don’t remember what I said back then.


#1094

By the way the whole baseball vs football stats comparison is terrible. Baseball is all about 1 v 1 duels. It’s not even truly a team sport.


#1095

lol now this I agree with. A4tt you are truly a loose cannon haha, I cracked up when I read that post. Your trolling remains quite shit but your loose cannon act is growing on me.


#1096

Yeah and the whole argument went way over your head… the larger point is that a sport like football with 22 players and constant movement and dynamism is severely problematic to measure individual performance with stats comparatively… my point is that Football is NOTHING like baseball and hence why stats should be looked at with massive, massive grain of salt.

Stats in baseball are by definition measuring performance of discrete events with direct involvement… football is a sport of mostly non or indirect involvement…


#1097

This is true but tbh I think the use of stats in football is also going over your head too. Basketball or american tackleball are also not sports like baseball with discrete events and direct involvement, but that doesn’t mean stats can’t be used and used well. Non-direct involvements can be measured and controlled for as well, though this is where more sophisticated models in football are needed most, it’s true. But to take such a radical anti-stats stance with football is going to leave you behind the game in your understanding of it, much like those same baseball fans who still cling to measures like Wins and Saves.


#1098

I do think they are useful, but they handle outliers TERRIBLY… this is the major problem… in baseball, there is no such thing as a bad outlier in terms of good offensive stats… I am sure this is something you can relate to… I’ll leave it at that.

And just to be clear, baseball stats represent total direct involvement, whereas football stats are trying to show the quality of a player (or not) and can only adequately represent 3-5% of a player’s total involvement… it is a pretty steep gap and handicap.

Having studied black sheep events in business extensively, I can’t stress how big a deal this is enough.


#1099

All I got out of this was something about WINS and SAVES.


#1100

Not the good models. Read some more statsbomb and the philosophicable pontificasition thread :wink:


#1101

I have and they try, but it is supremely hard to measure the key things like weight of pass, direction, selection of optimal option under every situation, movement off the ball when is possession, movement off the ball when NOT in possession, etc. Not to mention things like risk taking when appropriate, etc. It is one of the reasons that certain players will never be properly represented by stats.


#1102

In order to prove your point you took two sports which aren’t comparable whatsoever and then took two statistical categories that arent even remotely similar and then tried to make an argument based upon those parameters that football stats are useless. Lol k :joy:


#1103

Like I said, you missed my point… if you read, you are trying to do so. I think we have well established the troll you are.


#1104

I don’t know about never, but certainly at the moment there are some that aren’t adequately measured. That said, in general I think the models that are meant to measure more general skillsets are giving us logical results without missing out on big players who should be there, which is promising. The things you mention can’t be measured directly but they probably can be, and already are, indirectly.


#1105

I think it is hard because you need some form of judgment for momentary “where should i move/be now” that computer stats will struggle with for a long time. And football is hugely about this. Fabregas successful pass was 10x better than Denilson successful pass (irrespective of “key” passes) and that is really hard to measure.

Agree with the larger point the we WILL improve - the computational power and manpower associated with this analysis increases daily - but we are still a ways off.


#1106

This is something that I was getting to. Most of the models will have Messi at or near the top along with the other top players.