Media thread 2022/23

[halfcenturyup]

No, the Spurs value would be the total they spent (as a zero factor would be multiplied by 1 to find true value in the benchmark).

Not sure I follow this either. Zero multiplied by one is still zero. Or are you suggesting clubs who haven't won a trophy get allocated an unearned trophy for the analysis? Because this would go completely against the premise that only trophies won actually count?

 

[SebastianBlue]

I think we're agreeing in part - the post I replied to was talking about net transfer spend. I was that IF we had to compare, wages is a better metric - as it's by far the biggest outlay.

Not sure about Spurs benchmark being 1 - how could you then separate a team with one trophy from those with zero?

And surely Southampton are so far off financially that they can't really be expected to win anything, so are hardly a club which represents financial waste.

I imagine we do generally agree.

Regarding Spurs, their benchmark value (and any clubs with no trophies) would simply be excluded from the top line ranking because they have no trophies for the assessment period (they would be in a separate control set).

Which answers your second question: they represent lack of achievement and waste at *this level* (something I have qualified several times in this discussion), meaning they are not unpar with other clubs that have won silverware.

Southampton’s structure is very different to ours but a net spend or wages per median points or per median table position over multiple seasons would make them seem much closer in achievement.

Now, if we want to create a benchmark for being a well run club below the top level (the level at which silverware is the measure of success), median points would make sense as factor.

Again, my point is that using anything other than per trophy makes United, Spurs, and especially Liverpool look like they are on par with our achievement, because it skews to measures of simply playing at this level, not truly succeeding at this level. And I think we can agree that is not accurate.

 

[SebastianBlue]

How many seasons in your view so that it becomes statistically relevant?
Not trying to be tetchy, just interested even though I think your rationale is fundamentally flawed.

Well, given the level of assessment we are attempting to do (that is benchmark on success not merely playing at the top level), and the fact that there are 4 trophies available each season (there are actually more, but including those would skew things as only the champions of previous season tournaments can win them), and several of those are dependent on previous season achievement (creating a sort of weighting system inherently in the benchmark), I would use a sliding assessment timeframe of 10 seasons.

And I would ask you this: do you think using net spend or wages per median points or median table position would accurately reflective relative performance between City, Liverpool, United, Spurs, Chelsea, and Arsenal?

 

[SebastianBlue]

Not sure I follow this either. Zero multiplied by one is still zero. Or are you suggesting clubs who haven't won a trophy get allocated an unearned trophy for the analysis? Because this would go completely against the premise that only trophies won actually count?

No, the spending amount would be multiplied by one and be presented as a control set comprising clubs that have no trophy won during the assessment timeframe. I speak to this is a follow response to @bluenova.

This is pretty common method of normalising data for benchmarking (especially, again, when you have pretty desperate cohorts).

 

[halfcenturyup]

You’ve just answered your own question. ;-)

Again, relative performance (points or table position) would be the benchmark the likes of Levy would use, as it makes Spurs look like a well run, high achieving club.

So you are talking about amortised player cost, not net spend?

And we are talking principles here, not how good or bad the results make Spurs look .... I can understand why a City fan would like to base this performance analysis on trophies won, but I really can't understand it from a statistical point of view.

You really think a team that comes second in the league by, say, one point for five years at a fraction of the cost of the winners has performed financially much worse than the team that wins 5 times in a row having spent many times more?

 

[SebastianBlue]

So you are talking about amortised player cost, not net spend?

And we are talking principles here, not how good or bad the results make Spurs look .... I can understand why a City fan would like to base this performance analysis on trophies won, but I really can't understand it from a statistical point of view.

You really think a team that comes second in the league by, say, one point for five years at a fraction of the cost of the winners has performed financially much worse than the team that wins 5 times in a row having spent many times more?

No, I am talking total net spend, which means everything on the books, including wages off and on, transfers in and out, capital investment, etc. the net spend used for many of the debates amongst pundits and fans is far to simplistic.

And you are misunderstanding the benchmark we are (or at least I am) discussing: we aren’t determining which club are financially better or worse than each other, we are using (true) net spend to determine performance on the pitch relative to peers in the context of finances.

I actually talk about your version in another post and explain that if we want to do that, median points or median table position over a sliding assessment timeframe would make sense.

 

[halfcenturyup]

No, the spending amount would be multiplied by one and be presented as a control set comprising clubs that have no trophy won during the assessment timeframe. I speak to this is a follow response to @bluenova.

This is pretty common method of normalising data for benchmarking (especially, again, when you have pretty desperate cohorts).

I think we probably need to disagree on this, but how many clubs have won one of the four trophies in the last ten years? Would be a pretty short list. With a very long list for those that haven't won any. Again, I don't see the statistical relevancy when most of the population are exceptions.

 

[SebastianBlue]

I think we probably need to disagree on this, but how many clubs have won one of the four trophies in the last ten years? Would be a pretty short list. With a very long list for those that haven't won any. Again, I don't see the statistical relevancy when most of the population are exceptions.

I think it goes back to your last post which I think indicates we are actually talking about two very different benchmarks of performance.

 

[halfcenturyup]

I think it goes back to your last post which I think indicates we are actually talking about two very different benchmarks of performance.

My problem isn't what you put in the numerator, it is the relevance of the denominator. We have had, what, 40 clubs in the PL in the last ten years? How many of them have actually won a trophy? 8? 10 maybe? So you will be comparing the performance of the remaining, say, 30 purely on the basis of the amount spent. So Hull, say, would have been more successful than, say, Spurs. Or do you only compare clubs who have been in the PL for each of the last ten years, which weakens the analysis even further?

Interested in your reply but I won't be replying again, it's too boring for everyone, I imagine.

 

[Mazzarelli's Swiss Cheese]

Statistical cunts.