Why doesn't the league track errors?

[Bending and Tending]

Then would you also argue what is the point of tracking assists? There must be over 100 passes in a single game.

The difference is that keeping track of passing is almost always clear-cut. Player A makes a pass to Player B who receives the puck or Player A makes a pass to Player B and the puck: misses, is deflected, intercepted, etc.

Errors are much more difficult to assess as they are very subjective.

If my teammate over commits and I have to cover for him, should I be given an error because I had to change my strategy to counteract a teammates mistake?

What if I'm trying to push a forward away from the front of the net and the screen causes a goal? My two options were to let the screen happen and leave a guy open, or try to move the guy in time so my goalie can see the puck.

 

[Michael Farkas]

Too subjective to be an objective statistic.



Beat me to it.

This mindset ******* (I guess that word is censored, "slows the growth of" would be a fitting replacement) statistics, it does not aid them.

The thought that the easily scraped metrics don't have bias (like Corsi) is patently false.

 

[Anglesmith]

Is the idea to track these just for goals against, or for scoring chances, too? If the former, I can't imagine it would be too difficult to do. If the league can "semi-officially" track giveaways and takeaways (about as subjective a stat as you can get), surely they can trust someone equally to assign blame on goals. I think there's probably too many scoring chances, which go too quickly and don't always lead to a stoppage, to track those.

I'd actually find this statistic interesting, especially if correlated with Corsi. I'm sure there are a lot of bad Corsi players who don't generally make the mistakes that lead to goals and vice versa.

 

[PensBandwagonerNo272*]

This mindset ******* (I guess that word is censored, "slows the growth of" would be a fitting replacement) statistics, it does not aid them.

The thought that the easily scraped metrics don't have bias (like Corsi) is patently false.

I think Corsi is **** too.

I mean there are no statistics that can be indicative of performance without context, but "errors" in a game like hockey would be way too subjective to be useful.

 

[Sojourn]

I can just imagine all the errors that would happen late in a period when the ice conditions degrade, and the puck is bouncing and rolling all over the place. That would be a difficult stat to take seriously.

 

[Anglesmith]

I can just imagine all the errors that would happen late in a period when the ice conditions degrade, and the puck is bouncing and rolling all over the place. That would be a difficult stat to take seriously.

Are you saying goals are only scored late in periods?

 

[Sojourn]

Are you saying goals are only scored late in periods?

Nope. Just that I'd expect the number of errors to increase late in a period, when the ice conditions can start to disrupt the play.

 

[Anglesmith]

Nope. Just that I'd expect the number of errors to increase late in a period, when the ice conditions can start to disrupt the play.

They do, but the type of errors being talked about here probably wouldn't include flubbing a pass or fanning on a shot (unless it leads to a goal). From what I understand, it is a matter of being responsible for a goal against (or maybe a chance against). This includes positioning and defensive plays, which aren't affected by the ice at all.

 

[Sojourn]

They do, but the type of errors being talked about here probably wouldn't include flubbing a pass or fanning on a shot (unless it leads to a goal). From what I understand, it is a matter of being responsible for a goal against (or maybe a chance against). This includes positioning and defensive plays, which aren't affected by the ice at all.

Giving the puck away is exactly the kind of thing impacted by ice conditions.

 

[Anglesmith]

Giving the puck away is exactly the kind of thing impacted by ice conditions.

True, but the ability of the opposition to capitalize on such an error is also affected. I'm just saying, if breakdowns only occur towards the end of a period, wouldn't goals also only occur at the end of the period?

 

[Butch 19]

Then would you also argue what is the point of tracking assists? There must be over 100 passes in a single game.

An assist leads to a goal. The ultimate point of the entire game!

An error may lead to a goal, but much more than likely it won't.

 

[Sojourn]

True, but the ability of the opposition to capitalize on such an error is also affected. I'm just saying, if breakdowns only occur towards the end of a period, wouldn't goals also only occur at the end of the period?

Whoa, wait a minute. You're arguing against something I never said. At no point did I say that breakdowns only occur towards the end of a period. You took a statement I made an extrapolated an argument I wasn't trying to make.

All I said was that deteriorating ice conditions would lead to more errors, and that I'd have a tough time taking the stat seriously.

 

[Anglesmith]

Whoa, wait a minute. You're arguing against something I never said. At no point did I say that breakdowns only occur towards the end of a period. You took a statement I made an extrapolated an argument I wasn't trying to make.

All I said was that deteriorating ice conditions would lead to more errors, and that I'd have a tough time taking the stat seriously.

I feel like you're using a different definition of error than the OP, then. In the context of this discussion, an error refers to a poor play by an individual (ie a breakdown) that leads directly to a goal against.

If the understanding, as I think Whiskey is proposing, is that errors are handed out every time a goal is scored, where applicable, then by no means should there be a bias towards the end of the period. If there is, it would be no different than the bias we see with goals.

 

[tsujimoto74]

But giveaways aren't the only form of an error. Missed coverage in front of the net etc.

True but that'd be darn near impossible to track, given how fluid hockey is. Say a defender goes deep in the O-zone and none of the forwards pick up on that and cover the point for him. A 2-on-1 break starts the other way with the lone defender back, he misplays it, and a goal is scored. Who do you assign an 'error' to? All 5 guys? The guy who made the most recent mistake? There's no good way to track that.

 

[WhiskeyYerTheDevils]

I'll reiterate that I would only propose that the league tracks errors on scoring plays. Seems like there was some confusion.

As to the subjectiveness, I'd argue it's no more subjective than takeaways and penalties. They seem able to do it in baseball, so I don't see the issue. If I were a player, I'd prefer the stat to be tracked. It's better than +/- IMO.

 

[Mitts]

Way too complex and is subjective.

 

[CanadienShark]

That is so subjective, and could seriously hurt certain players financially. If stat trackers just "have it out" for a certain guy and we all know how important stats are in negotiations, isn't that kind of problematic? Not to mention the million other issues with it. I mean, we already track things like giveaways. That's enough.

With the number of events that occur in the game, it's simply not feasible to analyze every single one and judge whether or not it was an error. That would require exponential amounts of analysis for every second of gameplay. That's just silly.

 

[Anglesmith]

That is so subjective, and could seriously hurt certain players financially. If stat trackers just "have it out" for a certain guy and we all know how important stats are in negotiations, isn't that kind of problematic? Not to mention the million other issues with it. I mean, we already track things like giveaways. That's enough.

With the number of events that occur in the game, it's simply not feasible to analyze every single one and judge whether or not it was an error. That would require exponential amounts of analysis for every second of gameplay. That's just silly.

I mean, giveaways are equally subjective and no one really cares about that stat anyway. I don't see the harm in it. I don't know how hard it would be to assign errors without actually analyzing a good sample of goals and trying to determine how or if I would assign errors on the play. Maybe it's harder than I'm thinking, or maybe not.

And again, this isn't about every second of gameplay. It's just about scoring plays.

 

[Brainiac]

Because the league itself would end up at the top of the list every year...

 

[CanadienShark]

I mean, giveaways are equally subjective and no one really cares about that stat anyway. I don't see the harm in it. I don't know how hard it would be to assign errors without actually analyzing a good sample of goals and trying to determine how or if I would assign errors on the play. Maybe it's harder than I'm thinking, or maybe not.

And again, this isn't about every second of gameplay. It's just about scoring plays.

I don't think giveaways are as subjective as "losing your man." That's just me though. A giveaway has some ambiguity to it, whereas "losing your man" is 100% ambiguous. And I'm not even a fan of giveaways being tracked. I don't see significant harm if a stat (or stats) gained traction off a site like HFBoards. Then perhaps we could look at the use of it in relative isolation from the NHL and NHLPA. If it gained traction, I'm sure the NHL would take notice and track it themselves.

I know I exaggerated with the "every second of gameplay," but how do you define a scoring play? Is a scoring play when Karlsson starts with the puck behind his own net, dangles through multiple defenders and puts it glove-side top corner? Or is it scoring play when Glass skates it into the zone, fumbles the puck on a weak shot and banks it off a defender and into the net? I'm not trying to be rude, I'm just trying to say it's really tough when there is literally no black and white that we can easily observe. If someone can come up with a valid instrumental variable however, I will be impressed and interested to hear their argument.

 

[Stuzchuk]

Like moving the Flames to Calgary?

Biggest error possible.

Atlanta Flames were a legend.

yes they were... werent they the only city that got two tries ?

 

[WhiskeyYerTheDevils]

I don't think giveaways are as subjective as "losing your man." That's just me though. A giveaway has some ambiguity to it, whereas "losing your man" is 100% ambiguous. And I'm not even a fan of giveaways being tracked. I don't see significant harm if a stat (or stats) gained traction off a site like HFBoards. Then perhaps we could look at the use of it in relative isolation from the NHL and NHLPA. If it gained traction, I'm sure the NHL would take notice and track it themselves.

I know I exaggerated with the "every second of gameplay," but how do you define a scoring play? Is a scoring play when Karlsson starts with the puck behind his own net, dangles through multiple defenders and puts it glove-side top corner? Or is it scoring play when Glass skates it into the zone, fumbles the puck on a weak shot and banks it off a defender and into the net? I'm not trying to be rude, I'm just trying to say it's really tough when there is literally no black and white that we can easily observe. If someone can come up with a valid instrumental variable however, I will be impressed and interested to hear their argument.

Like I said, not every goal has an error associated with it. A lot of goals would have no errors.

But if Karlsson skates through three guys who all missed on a poke check and let him through, you could plausibly give all 3 guys errors.

 

[CanadienShark]

Like I said, not every goal has an error associated with it. A lot of goals would have no errors.

But if Karlsson skates through three guys who all missed on a poke check and let him through, you could plausibly give all 3 guys errors.

But how do you define a missed pokecheck?

If you're skating by at full speed and realistically I have no chance at getting it but make a hail-mary dive for it, is that a "missed" pokecheck?

Or what if you're skating by at slow speed and I'm not even looking, so you skate right by. Is that a "missed" pokecheck?

The issue I have is not necessarily with your idea, rather that I don't see how it could reasonably be implemented. However my favourite subject is econometrics, so I'd be really excited and interested if someone could find a set of instrumental variables to accurately measure a "missed" pokecheck or "losing your man."

Basically, I'm not trying to attack your idea because I think it's silly. I like it, in fact. I'm criticizing it because I don't see it as plausible right now.

 

[josra33]

yes they were... werent they the only city that got two tries ?

I guess the Winnipeg Jets dont currently exist....

 

[thammias]

Does the NFL track errors when a cornerback blows his coverage or offensive linebacker misses a block?