General Transfer Rumour Discussion Thread

[milo]

not always right though are they

Of course not.it is about trying to edge the odds in your favour

 

[JerusalemMan]

Yes you can - algorithms. That's what they do - use big data to predict future trends

Algorithms. You use that word as though it is some magic bullet, a solve-all miracle.

 

[milo]

Algorithms. You use that word as though it is some magic bullet, a solve-all miracle.

It's obviously not but you can certainly learn about players by studying them and their strengths and weaknesses. Given the financial gap between us and the clubs above us, it seems like a sensible option for us.

 

[Yiddo]

People really think that it works?

You can buy 10 right backs from the championship without looking at them and the chances are at least 1 will work out. What's the difference?

 

[Mumorn]

Real Madrid are close to signing Juventus Striker Fernando Llorente (Tuttosport)

 

[billyiddo]

People really think that it works?

You can buy 10 right backs from the championship without looking at them and the chances are at least 1 will work out. What's the difference?

well one is using the available data to help make an informed decision and one is rolling the dice

 

[jimmyb]

People really think that it works?

You can buy 10 right backs from the championship without looking at them and the chances are at least 1 will work out. What's the difference?

I won't pretend to know the first thing about algorithms or how they could help scouts and coaches to identify the right players, predict how they will perform or indicate how best they should be managed. But I'm fairly sure that clubs wouldn't bother with them if they were of no use whatsoever.

 

[K.D.D.D.D.Soc]

Real Madrid are close to signing Juventus Striker Fernando Llorente (Tuttosport)

That's alright, he's passed it, I've read it on here

 

[indianspur]

It's obviously not but you can certainly learn about players by studying them and their strengths and weaknesses. Given the financial gap between us and the clubs above us, it seems like a sensible option for us.

i think the top clubs also use algorithms/data/stats

 

[Baleforce]

it's only as intelligent as the person or people who define it

 

[DublinSpur]

Many moons ago when I was working in industrial systems design and optimisation we used a whole bunch of algorithms and analytic tools to predict potential future requirements and developments. It helped us allocate limited resources to the correct areas in the production-distribution network.

However it's not a magic bullet, and nobody who works with future trends / behaviour / developments believes it is (nobody sane anyway). I've never met Mr. Mitchell, but I guarantee he does not claim his system is 100% accurate... infallible. If he does then he's a fricking idiot. And anyone who blithely assumes his predictions will be 100% accurate and - as a result - fails to build contingency and back-ups into our team and preparation is just as idiotic.

And personally I don't think Pochettino is an idiot.

 

[parklane1]

Algorithms. You use that word as though it is some magic bullet, a solve-all miracle.

It is the new buzz word that's for sure

 

[Baleforce]

somewhere in Birmingham Tim is on the phone to his head scout...

"I don't know if it's Alfred or Albert, Spurs and Southampton keep banging on about him"

"no, I don't know how to spell the facking surname, try it like it saands init, G O R I T T E M"

 

[milo]

i think the top clubs also use algorithms/data/stats

Absolutely. Emirates Marketing Project have a massive team doing match analysis and made a lot of their data publicly available a few years ago. This is one of the reasons that I struggle to understand how anyone could be against using it.

 

[jimmyb]

Many moons ago when I was working in industrial systems design and optimisation we used a whole bunch of algorithms and analytic tools to predict potential future requirements and developments. It helped us allocate limited resources to the correct areas in the production-distribution network.

However it's not a magic bullet, and nobody who works with future trends / behaviour / developments believes it is (nobody sane anyway). I've never met Mr. Mitchell, but I guarantee he does not claim his system is 100% accurate... infallible. If he does then he's a fricking idiot. And anyone who blithely assumes his predictions will be 100% accurate and - as a result - fails to build contingency and back-ups into our team and preparation is just as idiotic.

And personally I don't think Pochettino is an idiot.

But you would agree that anyone who blithely dismisses the usefulness of algorithms as nonexistent is every bit as much an idiot?

 

[DublinSpur]

But you would agree that anyone who dismisses the usefulness of algorithms as nonexistent is every bit as much an idiot?

Absolutely. Statistical analysis and bespoke algorithms can be useful tools in predicting likely outcomes.

Thing is though, I can't say I've carefully read every post on this thread, but I don't recall anyone at all claiming that "the usefulness of algorithms as nonexistent"; merely people cautioning against relying on them too much. However I have, on the other hand, read posts insisting that because Alli and Pritchard have (apparently) been recommended by Mitchell's system, we don't need to worry about them failing and can dispense with back-ups and contingency.

I have nothing against using Mitchell's system to shape our squad, but we absolutely need to build in a contingency for when it gets things wrong (which like all analytical systems, it will do).

 

[milo]

People really think that it works?

You can buy 10 right backs from the championship without looking at them and the chances are at least 1 will work out. What's the difference?

Do I think that looking into something before investing millions is better than acting on a hunch? Yes.

 

[Yiddo]

Like I said earlier, there is no way anyone can predict for sure if a player is going to be a success in a team. It's all guess work.

 

[jimmyb]

Absolutely. Statistical analysis and bespoke algorithms can be useful tools in predicting likely outcomes.

Thing is though, I can't say I've carefully read every post on this thread, but I don't recall anyone at all claiming that "the usefulness of algorithms as nonexistent"; merely people cautioning against relying on them too much. However I have, on the other hand, read posts insisting that because Alli and Pritchard have (apparently) been recommended by Mitchell's system, we don't need to worry about them failing and can dispense with back-ups and contingency.

I have nothing against using Mitchell's system to shape our squad, but we absolutely need to build in a contingency for when it gets things wrong (which like all analytical systems, it will do).

That's either an exaggeration of GB's post or you're just reading too much into it. I don't think that he was suggesting anything so extreme. On the flip side, there was a post dismissing the notion of using algorithms to help scouts and coaches in the ways suggested as "a load of cobblers".

 

[DublinSpur]

That's either an exaggeration of GB's post or you're just reading too much into it. I don't think that he was suggesting anything so extreme. On the flip side, there was a post dismissing the notion of using algorithms to help scouts and coaches in the ways suggested as "a load of cobblers".

Obviously I may have picked it up wrong (and I can't be ars'd to go back over multiple threads), but I'm pretty sure GB has said that he's happy to see Dembele leave and does not see a new CM as a necessity because Mason and Bentaleb, Pritchard and Alli are sufficient for that position this season. Even excluding the "lack of experience" issue, that leaves us with sod-all injury cover which is plain reckless.