It is simply insane to think that any meaningful player evaluation system will be, can be, derived from statistical data. As previously pointed out, even what information should be collected is an unkown, but so is how to collect it and how to evaluate it. And then to massage it into a player rating is simply insane.
Baseball pretty much exists for stats. But I have yet to see a stat that measures the difficulty of catching a fly ball. Or of how many double plays a player has broken up or the difficulty in beaking them up.
Let's look at a simple player-possession sequence. A player recieves the ball, moves it ten yard by dribbling, and then passes it to a team mate. Objective data can be found in - success or failure in trapping the ball, distance dribbled, maintaining possession or losing possession in the dribble, success or failure in the pass, and distance of the pass. The key components of the difficulty in recieving the pass successfully, in evading opposition during the dribble, the quality of the opposition as well as their numbers (and distance from the player in possession does alone not determine if the opponent is actively opposing), and then the pass from the player in possession - Right choice? best choice available? Wrong choice? difficulty of the pass, speed, indentification of the teammate's tactical plan upon reception (to lead or not to lead?), Pass to feet or other part of the body? Opposition to the pass?
And we are not even getting into the more difficult to quantitize off the ball data.
The last time I had a game analysed by a service, the info was not nearly that complete, and that was $150 per game, from a single camera angle. To get even half the stats required for any sort of half assed system would require multiple cameras, myriad subjective evaluations by different individuals, and many, many man hours. Perhaps it could be done for under $700 per match, but I doubt it. And that is per team per match. Even at only 40 games a season, that is $28,000. Anyone feel like paying an extra $1,500 a year per kid to have personalized stats of dubious quality?
Of course, coaches only work afternoons or evenings then a few hours a weekend. Perhaps Big Clubs could set up an inhouse video review of games for stat purposes. Another 20 hours of morning work a week for coaches who otherwise might have to get day jobs, say at $40 per hour. That would give them the time to properly review maybe two matches a week, for $800. There would also be the costs of video taping the matches, from (just as an example) 3 different perspectives. The equipment to synchronize the showing of the film, so that the 3 gamefilms are running at the same time. So that is 6 hours of videographer time - again, per match.
Would all this data be useful? Certainly, all data is useful. Would the utility be worth the costs? Not even within an order of magnitude.
The idea of meaningul personal stats is ludicrous. I once had a GK who had 320 saves in a HS season, all from legit shots on goals. But our team was weak, most of the girls had never played soccer before, and our defensive gameplan was to pack the penalty area and encourage the opponents to shoot from outside it. On another team, one with experience and quality, a championship caliber team, she would have had an average of 3-5 saves a game. And a lower GAA.
Unless the idea of personal stats is just another ploy to employ ever more "youth soccer professionals" this is an absolute joke.
Author
Topic: CAP - second stage of development (Read 1620 times)