Ranking Soccer Players By Following the Bouncing Ball 142
sciencehabit excerpts from an interesting report on statistics for soccer, in the stats-obsessed world of sports: "Only a handful of soccer ranking systems exist, most of which rely on limited information: the number of goals scored in a match, the number of goals assisted, and some indices of a match's difficulty and importance. ... So researchers turned to an unlikely source: social networks. Applying the kinds of mathematical techniques used to map Facebook friends and other networks, the team created software that can trace the ball's flow from player to player. As the program follows the ball, it assigns points for precise passing and for passes that ultimately lead to a shot at the goal. Whether the shot succeeds doesn't matter. Only the ball's flow toward the goal and each player's role in getting it there factors into the program's point system, which then calculates a skill index for each team and player."
incomplete metrics (Score:5, Insightful)
How about off-ball activity that contributes? Moving across a zone or defender to clear space for someone who actually handles the ball? What about the guy who makes a brilliant cut but doesn't get served well by a teammate, so never handles the ball?
What about defense?
Never mind the fact that this metric would be biased against Italian league players, where ball control and quality opportunities is more important than number of shots. You could game this system very easily by cranking shots from 30 yards.
Soccer doesn't lend itself well to statistical analysis of players. That's one of the things that makes it a beautiful sport and fun to discuss, IMO.
offensive, isn't it? (Score:5, Insightful)
assigns points for precise passing and for passes that ultimately lead to a shot at the goal
calculates a skill index for each team and player.
Wow, that's really going to tell you about a players defensive skills, isn't it.
Not that those could possibly important in a game where usually only one or two balls make it to the net the whole game. I mean, it's not like defense would play much of a role there.
Bouncing ball? (Score:3, Insightful)
Re:Um ... (Score:4, Insightful)
But that could be because both teams played AMAZINGLY well.
Or it could be because one player messed up a lot on each team (the one that actually shot).
Basically, this is just software that analyzes individual players performance leading up to shots - assists. Their JOB is to get the ball up to the striker. Their job is not necessarily to actually score. The scorers, though, get all the glory. Perhaps this software will help that?
Re:Flawed metric (Score:4, Insightful)
It is less flawed than the current methods mentioned in the summary. In fact, it does MORE to measure the team effort than a metric like goals scored. This is what we might call an incremental improvement. /facepalm
Re:Um ... (Score:3, Insightful)
But you get 1st downs, which has nothing comparable in soccer. It also can put you in field goal range, which has no parallel in soccer. In US football, the team with the most offensive yards almost always wins. Does soccer have a similar outcome? And in US football, yards are only counted FORWARD. If the fullback runs 30 yards left, then 30 yards right, then is tackled on the line of scrimmage, he has gained exactly zero yards. Comparing to US football isn't a fruitful exercise, they are just too different.
NFL football is more like a blend of chess and raw violence. It is a series of calculated moves, not a continuous flow of play. Not better or worse, just not really comparable.
Re:Um ... (Score:4, Insightful)
NFL football is more like a blend of chess and raw violence.
If NFL football is chess, soccer is go. The difference? It actually requires talent to be good at goh, whereas a supercomputer can beat anyone at chess. Skilled athletes excel at soccer, overweight drug addicts who should have failed out of high school win football games.
Re:Bouncing ball? (Score:3, Insightful)
Re:incomplete metrics (Score:3, Insightful)
The problem is, the best defense is often one where a player is so well marked no one even tries to pass it to him.
Also, how do you "interfere with a goal"? It's either a goal or it isn't and if it is then the interference sure didn't do much.
Re:Fantasy Leagues? (Score:2, Insightful)
Actually football (okay, soccer...) fantasy leagues follow quite simple metrics, it's just about goals, assists, yellow and red cards.
Some system takes into account only individual stats, others also team stats (eg, a bonus for a defender if his real team doesn't concede goals).
There may be some more or less convoluted bonus and combo rules (say, all your forwards score goals and all your defenders' teams don't concede...) but that's pretty much it.
Here's an example:
http://fantasy.premierleague.com/M/help.mc?category=scoring
Stats and metrics really don't belong to soccer, it's a game of subjective evaluations.
In fact, here in Italy (but I guess it's the same in other countries) sports newspapers evaluate players with grades after every match, to give a sense of how any of them performed.
You get a table with grades given by the journalist, in a scale from 1 to 10 just like in our school system, and our "fantacalcio" games are based on those numbers (plus all the various bonus and malus).
Pretty subjective, but that's part of the fun of soccer, the discussions about performances and matches are neverending as there aren't objective stats to call in to support your views. =)
Sounds rather slanted (Score:5, Insightful)
Under those circumstances, Spain played an amazing game against Switzerland this week: Hundreds of accurate passes that ended in shots. More passes in one half than most teams make in an entire game. And yet, they didn't score, and lost the game against a team that had 25% ball position, but actually managed to score.
It would also mean that every Italian national team from the last 30 years happens to be terrible, despite their world championship titles.
Re:incomplete metrics (Score:1, Insightful)
I agree, basing the judgement on how the player handles the ball is missing most of what really goes on.
Its not all about the ball.
Things are a bit more obvious in American Football, its obvious there that many players are never intended to interact with the ball at all. Instead their job is to block the other teams players and keep them out of the action, or to create a distraction, or to keep the other sides best players out of an area of the field due to threat of injury etc.
A player may never touch the ball all season, but they may also be keeping the other sides best player out of the game which is just as valuable as being the scorer.
Re:Um ... (Score:3, Insightful)
Points for not scoring? Isn't that the same as a woman telling you that she just wants to be friends because your friendship means more than a relationship would?
No, it's like giving a guy points for how many numbers he gets and how many hot chicks actually flirt back with him. Then you can see who's better with the ladies even if you're comparing two Slashdotters and the scoreline would typically be a nil-all draw.
Re:incomplete metrics (Score:2, Insightful)
Never mind the fact that this metric would be biased against Italian league players, where falling on the ground and begging for a foul when another player is within arm's reach is more important than number of shots.
FTFY.
Re:Um ... (Score:4, Insightful)
Yes, in general the team with the longest possession, most corner shots and most penalties (in an offensive, goal-able position) for them usually wins. There are already a few comparable statistics in place to gauge whole teams, but estimating the "game value" of a certain player is often rather hard. A player may be "valuable" just by being on the field without a single ball contact. There are players who have to be covered tightly so they CANNOT touch the ball and cannot be a sensible place to pass to, because there is ALWAYS an opponent with him. His value lies in the ability to tear apart the defense of the opponent because he has to have a watchdog, often two. He will not be counted as "valuable" in this new scoring system, even though he is probably one of the most valuable players in the team.
It's like using yards carried for football and considering the offense line useless because they don't really carry the ball anywhere.
Re:Call it right (Score:3, Insightful)
I call that other game "American Rugby," since that's what it is.
Re:Um ... (Score:4, Insightful)
Any football fan will tell you that when two teams play AMAZINGLY well the result will be more like 5-5 rather than 0-0.
I strongly disagree. There is much more than offense to consider in a good game of football. If the game results 5-5 it is rather clear that both defenses have failed at their job. Even a game ending 0-0 can be extremely interesting to watch for a football connoisseur. Football is not just about making goals, it's also about not conceding them.
Re:Call it right (Score:3, Insightful)
Re:Sounds rather slanted (Score:5, Insightful)
Spain *are* a much better team than Switzerland and this system would show that. Have them play a thousand times, and Spain would win the vast majority. So I'm not sure I see your point.
You do make a good point about Italy. However I'd be interested to see what the system actually says about Italy before condemning it.
Re:I dunno... (Score:1, Insightful)
Re:Um ... (Score:3, Insightful)
I get the impression that it's more like Rugby, though, with different types of "score" giving a different point value. Oh, and with body armour, totally non-fluid gameplay, and game times which rival Test cricket.
Quantitative Analysis not Possible (Score:3, Insightful)
Trying to measure a qualitative activity with quantitative tools is meaningless.