Jaffka's gossip thread on the ins and outs at the 'Bok

[Chrisfu]

He isn't Cruyff. But neither is he Nicky Brookman.

This is Nicky Brookman.

[giraffesarecool]

Stats are sometimes bollocks though.

What does that even mean?

Does it, perhaps, mean that you have a pre-conceived idea of what Muamba does, and when that is borne out by the statistics, you can say "Stats are sometimes bollocks", and relate how you once saw him cock-up badly, ignoring all the times he didn't?

Or does it mean something else? If so, what?

I'm always entertained by the arguments people make along the lines of "You can prove anything with statistics" - no you can't. And they're certainly more useful than taking atypical occurrences and suggesting they're the norm...

Stat's themselves aren't bollocks, but 'you can prove anything with statistics' whilst an obvious exaggeration it's hardly entirely untrue. You could argue the best passer in a team is the one with the higher completed passes % against the number of attempts, whilst the flaw with the stats here is that it doesn't take into account the style of pass, the range, the intent, or anything about it. You can only really understand a very limited amount of information from using stats, and you actually need to watch and understand football to interpret everything else.

Stats alone mean very little.

[boltonboris]

Smashing the ball at somebody's chest 4 yards away then watching it bounce off the player in a general direction, would still class as a completed pass.

Doesn't mean it's actually successful

[Puskas]

Smashing the ball at somebody's chest 4 yards away then watching it bounce off the player in a general direction, would still class as a completed pass.

Doesn't mean it's actually successful

Does it?

In which case, Opta need to change their definition of a completed pass.

But I'd still maintain that such a "pass" was not Muamba's usual - he was more likely to do a simple ball of a few yards, and do it successfully.

[Puskas]

Stat's themselves aren't bollocks, but 'you can prove anything with statistics' whilst an obvious exaggeration it's hardly entirely untrue. You could argue the best passer in a team is the one with the higher completed passes % against the number of attempts, whilst the flaw with the stats here is that it doesn't take into account the style of pass, the range, the intent, or anything about it. You can only really understand a very limited amount of information from using stats, and you actually need to watch and understand football to interpret everything else.

Stats alone mean very little.

So how would you determine the best passer in the team without using statistics?

Surely the basis would be how many successful passes they made, compared to how many they attempted?

Rather than just making stuff up?

[thebish]

Stats are sometimes bollocks though.

What does that even mean?

Does it, perhaps, mean that you have a pre-conceived idea of what Muamba does, and when that is borne out by the statistics, you can say "Stats are sometimes bollocks", and relate how you once saw him cock-up badly, ignoring all the times he didn't?

Or does it mean something else? If so, what?

I'm always entertained by the arguments people make along the lines of "You can prove anything with statistics" - no you can't. And they're certainly more useful than taking atypical occurrences and suggesting they're the norm...

Stat's themselves aren't bollocks, but 'you can prove anything with statistics' whilst an obvious exaggeration it's hardly entirely untrue. You could argue the best passer in a team is the one with the higher completed passes % against the number of attempts, whilst the flaw with the stats here is that it doesn't take into account the style of pass, the range, the intent, or anything about it. You can only really understand a very limited amount of information from using stats, and you actually need to watch and understand football to interpret everything else.

Stats alone mean very little.

hmmm.... your argument seems to be that you can use statistics in an argument (either well or badly) - which isn't saying very much.

you certainly haven't backed up the notion that "you can prove anything with statistics" - which (as puskas points out) is manifestly bollox, and hairy, sweaty ones at that!

[thebish]

BTW...

THIS giraffe is cool!!

[Tombwfc]

Do they calculate passing % like that? In any case... http://www.whoscored.com/Teams/92/Show/England-Bolton" onclick="window.open(this.href);return false;

[ohjimmyjimmy]

Look at how shit that Sordell is.
Stats don't lie !

[Dave Sutton's barnet]

Do they calculate passing % like that? In any case... http://www.whoscored.com/Teams/92/Show/England-Bolton" onclick="window.open(this.href);return false;

Nice to see the central midfielders ticking it over. Irksome to see Bogdan turning it over.

[giraffesarecool]

Stats are sometimes bollocks though.

What does that even mean?

Does it, perhaps, mean that you have a pre-conceived idea of what Muamba does, and when that is borne out by the statistics, you can say "Stats are sometimes bollocks", and relate how you once saw him cock-up badly, ignoring all the times he didn't?

Or does it mean something else? If so, what?

I'm always entertained by the arguments people make along the lines of "You can prove anything with statistics" - no you can't. And they're certainly more useful than taking atypical occurrences and suggesting they're the norm...

Stat's themselves aren't bollocks, but 'you can prove anything with statistics' whilst an obvious exaggeration it's hardly entirely untrue. You could argue the best passer in a team is the one with the higher completed passes % against the number of attempts, whilst the flaw with the stats here is that it doesn't take into account the style of pass, the range, the intent, or anything about it. You can only really understand a very limited amount of information from using stats, and you actually need to watch and understand football to interpret everything else.

Stats alone mean very little.

hmmm.... your argument seems to be that you can use statistics in an argument (either well or badly) - which isn't saying very much.

you certainly haven't backed up the notion that "you can prove anything with statistics" - which (as puskas points out) is manifestly bollox, and hairy, sweaty ones at that!

I wasn't trying to back it up. It's an obvious exaggeration (as I said) but many things can still be 'proven' by statistics, which was exactly what my post was an example of.

The meaning of 'you can prove anything with statistics' is not that things are actually proved, it's that you can use statistics to form the basis of your argument, whilst conveniently neglecting to provide contradictory or counter-argument stats/information. My argument was such to display that whilst 'you can prove anything with statistics' it's very possible that nothing at all is proven, and instead the readers of said stat are being misinformed.

[giraffesarecool]

Stat's themselves aren't bollocks, but 'you can prove anything with statistics' whilst an obvious exaggeration it's hardly entirely untrue. You could argue the best passer in a team is the one with the higher completed passes % against the number of attempts, whilst the flaw with the stats here is that it doesn't take into account the style of pass, the range, the intent, or anything about it. You can only really understand a very limited amount of information from using stats, and you actually need to watch and understand football to interpret everything else.

Stats alone mean very little.

So how would you determine the best passer in the team without using statistics?

Surely the basis would be how many successful passes they made, compared to how many they attempted?

Rather than just making stuff up?

You would use statistics, no one suggested otherwise. You've totally misread what I've said I think.

[thebish]

The meaning of 'you can prove anything with statistics' is not that things are actually proved, it's that you can use statistics to form the basis of your argument,

only if you radically redefine the word "prove" to mean what it doesn't mean!!

[Lord Kangana]

Define "best".

If, for example, your centre half spent his entire game giving the ball to the goalie to give it the big heave-ho (for he himself was incapable of a decent forward pass), then he would probably be considered the "best" passer in the league by stats alone, whilst the keeper would be castigated for losing possession 50% of the time. Then the discussion would be "well I don't know why they're so shite, as he's clearly an excellent passer". Which would be bollox. And it would be. You know it really, in your heart of hearts, deep inside.

[giraffesarecool]

The meaning of 'you can prove anything with statistics' is not that things are actually proved, it's that you can use statistics to form the basis of your argument,

only if you radically redefine the word "prove" to mean what it doesn't mean!!

That's the whole point of the saying...

[Puskas]

Define "best".

If, for example, your centre half spent his entire game giving the ball to the goalie to give it the big heave-ho (for he himself was incapable of a decent forward pass), then he would probably be considered the "best" passer in the league by stats alone, whilst the keeper would be castigated for losing possession 50% of the time. Then the discussion would be "well I don't know why they're so shite, as he's clearly an excellent passer". Which would be bollox. And it would be. You know it really, in your heart of hearts, deep inside.

Well, yes - if your ridiculously contrived situation ever happened in the real world, then the simple statistics of pass completion may not provide a full picture. However, I'm not sure it has. And further statistics - number of forwards passes, number of passes to the goalkeeper, number of players within a certain distance - would also help give a clearer picture - indeed, they would, anyway. I don't know if these statistics are kept by Opta, or whoever, though.

My point, such as it was, was really just reacting to the over-the-top and silly remark that Muamba was the worst player in the Premier League at passing, dribbling and shooting. Since he didn't spend all his time passing back to the goalkeeper, and made a reasonable number of different sorts of passes, I think we can, more or less, take the 70-odd percent pass completion as being fairly accurate.

My other point was reacting to the "all stats are bollocks" remark. Which is also silly. They are useful - more so than picking out an isolated error, and trying to say that it backs up the claim that Muamba is the worst passer in the league. It doesn't - it's an isolated error. Even Taddy Nowak's passes went astray sometimes...

[Lord Kangana]

It does.

Exhibit A, for the prosecution: Michael Carrick.

The prosecution rests your honour.

[Lord Kangana]

Happen in the real world that is.

[bobo the clown]

He isn't Cruyff. But neither is he Nicky Brookman.

This is Nicky Brookman.

Are you sure ? I thought that was Steve Taylor.

I'm not sure Nicky B could get into that position.

[Bruce Rioja]

He isn't Cruyff. But neither is he Nicky Brookman.

This is Nicky Brookman.

Are you sure ? I thought that was Steve Taylor.

I'm not sure Nicky B could get into that position.

Why Steve Taylor, clown?