Which is only one of the five factors they use in the second step to estimate the transfer value. Overall the approach is quite valid, with correlative significance, albeit not regressive significance. But the inclusion of the book value adds what is essentially a meaningless accounting convention to the valuation of a unique asset. So it skews the data in the case of players with very high or zero book values, like Rudiger and Pickford.
he hardly played in the 2nd half of the season but i take your point. Sahko played 8 games last season so he didn't miss a year of football either and was a key player in Palace avoiding the drop
Almost every player was "overpaid" for, which is impossible by definition of market value, although they do admit this a sign of recent price inflation they didn't account for
Don't forget when we bought Sakho for £18 million from PSG he'd been benched, stripped of the captaincy and was getting criticism for his fitness and weight Sounds familiar, right Making £8 million on him is a result really
Adrien Silva: Fifa rejects Leicester City's application for midfielder http://www.bbc.co.uk/sport/football/41118927
The new PL TV deal started first - we were talking about £60m+ for Van Dijk before Neymar was even a story
True, it's how commodities work, it doesn't mean that all prices suddenly shoot up into the hundreds of millions but it incrementally raises and inflates the price of every other player beneath that top layer, artificially
TBF this is BS Deadlines are supposed to stop people arranging deals after the window closes, not having an #official form stamped 14 seconds late
35 factors actually which means the book value doesn't have the significance you're assigning it. Nobody could claim to put an exact valuation on a player but the science these people use has meant they have 81% correct prediction rate which makes it quite valid as you say and fairly interesting to talk about when there's little to no football going on Coincidentally last year they had Wijnaldum value down as £24.8m [£25m] and Mane as £35.5m [£34m] - I'd say that was fairly spot on.
do you know how the factors are weighed? are you assuming that book value is of the same import as the others?
This is true, however that also means a 19% inaccurate prediction rate. Given that the table posted above shows the biggest discrepancies between their predictions and actual values, it's fair to say these transfers are part of the 19%. Also it's very easy to create a model that has a high prediction rate when you keep adding in more and more variables, particularly ones like book value which will obviously have some colinearity with transfer fees. The inclusion of book value does actually explain most of their worst predictions as well. Not, imo, a coincidence that the vast majority of the most 'overvalued' players were those with little or no book value, nor that the three 'undervalued' players were the three with their book values still high.
Well if PSG peed in their pool, then they definitely peed in our pool, or got stage fright http://www.bbc.co.uk/sport/football/41174403
Virgil van Dijk available for Southampton selection after returning to first-team training http://www.skysports.com/football/n...ection-after-returning-to-first-team-training
