• Moderators, please send me a PM if you are unable to access mod permissions. Thanks, Habsy.

The James Mirtle Mea Culpa Thread

leafman101 said:
01-OVERALL-web.jpg


http://www.sportsnet.ca/hockey/nhl/introducing-the-shot-quality-project/
Click to expand...

aaaaaaand here it is.
 
The current flaw with this is the imperfection in the data. The technology they're using to track shot and goal locations still isn't accurate enough imo. I'd like to see tracking capabilities built right into the puck....which would also have goal line applications. If the technology was demonstrably accurate enough, there would be no more pucks underneath goalies, but over the goal line, not being called goals.
 
leafman101 and ME, I get the idea that using the "hotzone" is what we have and works to some degree, but hockey is such a fluid thing that I don't believe shot position is close to enough.

You would need shot position, who the shooter is,was the shooter in their "hotzone"?...just a couple things I can think up right away. Kessel with a wrist shot at one position is worth more than someone else in a "better" position.
I'm jumping in and out here, kind of busy, but I think I git the gist of it across.
 
Bleedsblue&white said:
leafman101 and ME, I get the idea that using the "hotzone" is what we have and works to some degree, but hockey is such a fluid thing that I don't believe shot position is close to enough.

You would need shot position, who the shooter is,was the shooter in their "hotzone"?...just a couple things I can think up right away. Kessel with a wrist shot at one position is worth more than someone else in a "better" position.
I'm jumping in and out here, kind of busy, but I think I git the gist of it across.
Click to expand...

yep, these are things we've discussed a bit before. But the Corsi crowd loves their simple counting stats instead of asking for better.
 
Bleedsblue&white said:
leafman101 and ME, I get the idea that using the "hotzone" is what we have and works to some degree, but hockey is such a fluid thing that I don't believe shot position is close to enough.

You would need shot position, who the shooter is,was the shooter in their "hotzone"?...just a couple things I can think up right away. Kessel with a wrist shot at one position is worth more than someone else in a "better" position.
I'm jumping in and out here, kind of busy, but I think I git the gist of it across.
Click to expand...

True. A PK slapper from the end boards is a lot more dangerous than a floating wrist shot from the blueline too. But there's only so much to handle. I mean, if you want to go to that length, a Kessel wrist shot from one location might be nothing, but take the same location and park a guy in front of the goalie and suddenly it's a big scoring chance.

Probably the easiest first cut will be to take the heatmap of positions, split that into a few zones, and maybe also break them down into 2-3 classes of shots - hard shots and soft shots (or maybe slap shot, wrist shot, and deflections?), then you should be able to get a good idea of the overall quality of shots.
 
UWHabs said:
True. A PK slapper from the end boards is a lot more dangerous than a floating wrist shot from the blueline too. But there's only so much to handle. I mean, if you want to go to that length, a Kessel wrist shot from one location might be nothing, but take the same location and park a guy in front of the goalie and suddenly it's a big scoring chance.

Probably the easiest first cut will be to take the heatmap of positions, split that into a few zones, and maybe also break them down into 2-3 classes of shots - hard shots and soft shots (or maybe slap shot, wrist shot, and deflections?), then you should be able to get a good idea of the overall quality of shots.
Click to expand...

Thing is, the technology already exists to track the speed, height, etc of a shot and classify it based on speed (baseball's pitchtrax system can differentiate pitches in baseball off the the spin and angle of movement ffs...surely we can track the speed of a puck). Then all you need is real time, accurate player location data (which exists in european football) tracking where a player is on the ice at all times, what speed they're moving, the approx amount of force of hits etc.

This is why I groan at the word "advanced" being used by the metrics crowd. They have **** all for data, and that's not their fault, but it's still embarrassing that they'll take something as simple as shots for/against, play with it a bit, make it a few different types of statistics (rates, %'s, etc) and then try to pawn it off on us as advanced and worthy of discussion along side Baseball's sabremetrics, the NBA's stats gathering (Synergy >>>>>>>>> what hockey metrics people are doing. You want to quantify how good a defender a NBA player is, synergy can tell you his individual shooting percentage against in isolation situation, zone defence situations, help defence situations, etc), and even what is tracked in european footie. A statistic can only be as good as the data used to create it, and theses guys are childs play compared to what is out there in other sports.
 
UWHabs,
it's because hockey is so fluid that you need to find ways to be more accurate. We don't have the nice stops and starts that baseball does for instance. IMO, if this stat crowd wants legitimacy in the hockey circles, they have to account for more than they are.
Much more.
 
UWHabs,
it's because hockey is so fluid that you need to find ways to be more accurate. We don't have the nice stops and starts that baseball does for instance. IMO, if this stat crowd wants legitimacy in the hockey circles, they have to account for more than they are.
Much more.
 
For sure. It's a big step up from the old system where they only used goals, assists, PM, and +/- to judge players, but it's still stone age compared to other sports.
 
You must log in or register to reply here.
Back
Top