Keeper stats

[JG]

Top five seasons: adjusted save percentage
1. 2001 Metros, Tim Howard +7.2%


So Howard saved about 13.5 goals compared to his expected save percentage...that's almost four wins due to his goalkeeping.

[numerista]

So Howard saved about 13.5 goals compared to his expected save percentage...that's almost four wins due to his goalkeeping.

To clarify, JG:
The Metros had 42 points in 2001. Replacing Howard with an average keeper, we would've expected them to get only about 30 points?

If so, wow ...

[JG]

To clarify, JG:
The Metros had 42 points in 2001. Replacing Howard with an average keeper, we would've expected them to get only about 30 points?


I hadn't actually run the numbers before--it turns out that the extra 13 goals would only be a 9 point difference. But the Metros overachieved a bit in 2001 compared to their goal differential...a team with a GD of 38-48 (the metros GD with an "average" keeper) would be expected to get 30 points from a 26-game schedule.

Presumably we could get "point values" for every goalie this way.

[mpruitt]

I hadn't actually run the numbers before--it turns out that the extra 13 goals would only be a 9 point difference. But the Metros overachieved a bit in 2001 compared to their goal differential...a team with a GD of 38-48 (the metros GD with an "average" keeper) would be expected to get 30 points from a 26-game schedule.

Presumably we could get "point values" for every goalie this way.
JG, would you mind expanding on this. I don't understand how you're calculating expected goals and how you're relating that to points? Please forgive my ignorance with some of this stuff.

[ChrisE]

JG, would you mind expanding on this. I don't understand how you're calculating expected goals and how you're relating that to points? Please forgive my ignorance with some of this stuff.

Exactly why we need a website (;)). I had the same reaction when I read this a month ago, it's based on a thread JG posted here:

http://www.bigsoccer.com/forum/showthread.php?t=80294

[Elninho]

I have another theory for the offside calls correlation.

More offside calls could correlate to, not only more aggressive offside traps, but also more cherry-picking attackers. (Roy Lassiter and Mamadou Diallo come to mind.) These players are going to try to beat the offside trap every time, and even if they're offside most of the time, they will beat the offside trap sometimes. This leads to more breakaway opportunities.

In the current MLS, I see a lot more goals coming from combination plays and less from breakaways. I used to sort MLS goals by MLSnet match report descriptions. The "breakaway" column in recent seasons is a lot shorter than in previous years. I don't know where my old data went - I suspect a lot of it got destroyed by a hard disk glitch a couple years ago. But I know that, in one of the earlier seasons, as many as 8% of all MLS goals were scored on breakaways, and in the current season (with my analysis now aided by video) it's been only 2.8% so far. Incidentally, the percentage of goals scored by headers has gone up dramatically - near 20% this season, compared to 14.8% in 2003 and around 12% in the first few seasons.

[AussieVamp2]

Exactly why we need a website (;)). I had the same reaction when I read this a month ago, it's based on a thread JG posted here:

http://www.bigsoccer.com/forum/showthread.php?t=80294

thanks,

That is definitely interesting