+/- Improvement Idea

[neutral zone trap13]

I feel that +/- is not an accurate stat to determine how good a player is at defence (or offence) So here is my idea: A players +/- compared to the team's +/-. So for example, right now, Claude Giroux is -2 on the Philadelphia Flyers who are so his new +/- would be -14. Jason Garrison is +8 on the -16 Florida Panthers so he would be +24. +/- isn't a super big deal, but I believe that this would be an improvement and would make it so that just being on a good team won't give you a very high +/-.

Click to read more...

 

[Wedontneedroads]

Team A: +50
Player A: +25
Player A New +/-: -25?

Team B: -50
Player B: 0
Player B New +/-: +50?

Am I missing something? If not then that doesn't make a whole lot of sense to me.

 

[EVBetting Site]

I feel that +/- is not an accurate stat to determine how good a player is at defence (or offence) So here is my idea: A players +/- compared to the team's +/-. So for example, right now, Claude Giroux is -2 on the Philadelphia Flyers who are so his new +/- would be -14. Jason Garrison is +8 on the -16 Florida Panthers so he would be +24. +/- isn't a super big deal, but I believe that this would be an improvement and would make it so that just being on a good team won't give you a very high +/-.

+/- is not a defensive stat. For whatever reason it has become synonymous with measuring a player's defensive ability. It's an overall stat that literally measures how many more goals you score at ES (generally) compared to your opponents while that player is on the ice. Good players who have high +/- ratings help their teams be better teams, teams are not born as a "good team", causing their players all to have ridiculous +/- ratings. Punishing a player because he makes his team better is counter productive to say the least.

What it does not include however, are the matchups; who you are playing against, who is helping you do that, what situations/role your coach is using you in (offensive role with high zone start, etc), and minutes (both quantity and quality).

 

[Paranoid Android]

I think a simple improvement would be divide the +/- column into two different columns, a "+" and a "-"

 

[Hockey Team]

I feel that +/- is not an accurate stat to determine how good a player is at defence (or offence) So here is my idea: A players +/- compared to the team's +/-. So for example, right now, Claude Giroux is -2 on the Philadelphia Flyers who are so his new +/- would be -14. Jason Garrison is +8 on the -16 Florida Panthers so he would be +24. +/- isn't a super big deal, but I believe that this would be an improvement and would make it so that just being on a good team won't give you a very high +/-.

That would make it so every player's a minus on a great team with a huge + goal differential, and every player a plus on a horrendous team with like a -60 goal differential.

Sorry, but this is a horrible idea to make an already hard to use stat even more useless.

 

[Drunkspleen]

+/- is always going to be a flawed stat, honestly, they shouldn't be combined.

You could infer a lot more about a player's value by looking a their + and their - entirely separately, this is why whenever I'm looking at players in a way that +/- might be relevant, I tend to opt instead for Goals For/60 minutes and Goals Against/60 minutes on and off the ice.

 

[Korpse]

Team A: +50
Player A: +25
Player A New +/-: -25?

Team B: -50
Player B: 0
Player B New +/-: +50?

Am I missing something? If not then that doesn't make a whole lot of sense to me.

Player A new: 0

It really doesn't make sense. The best way to use +/- is to compare to teammates. Which isn't a perfect solution but the closest thing to it.

 

[number72]

also divide by TOI so it is normalized for how much a player spends on the ice

 

[McIlrathsWrath]

Same kind of flaw exists with PP opportunities. Both should be looked at and reformed

 

[Tom Polakis]

Team A: +50
Player A: +25
Player A New +/-: -25?

Team B: -50
Player B: 0
Player B New +/-: +50?

Am I missing something? If not then that doesn't make a whole lot of sense to me.

It makes sense in showing if a player has under- or overachieved relative to his team, but the OP's overly simple approach will exaggerate those differences, as you have shown.

A relevant +/- stat would somehow normalize a player's +/- against the team's number; maybe we need something involving a third-order equation.

 

[HanSolo]

+/- is stupid. Especially since power plays aren't counted. I don't get it. You're still on the ice when a goal is scored.

 

[GuineaPig]

Also, adjust +/- to account for goaltender save percentage at even strength.

 

[JQR]

It's just such a horribly misleading stat. People will naturally look at a + rating as a good thing and a - rating as a bad thing, even though there's a lot more to it. I hate it.

 

[MrGold]

+/- is stupid. Especially since power plays aren't counted. I don't get it. You're still on the ice when a goal is scored.

+/- is a measure of offensive/defensive domination.

The "+" has to do with offensive domination

The "-" has to do with defensive domination

You are not awarded a "+" on the PP because it's a given that you're in a domination position.

You are not awarded a "-" on the PK because it's a given that you're being dominated.

A player with a better +/- than his team average +/- can be said to fulfill his role in a better way than the average player on his team.

 

[nik jr]

thread with adjusted +/- (excludes SH scoring) since NHL started officially recording it in '68.

http://hfboards.mandatory.com/showthread.php?t=591548

I think a simple improvement would be divide the +/- column into two different columns, a "+" and a "-"

that is how it is done in sweden, and it is a much better method.


for example, gretzky was +71 in '86. only a few players on terrible teams had worse than gretzky's -120, but no one had better than his +191.

gretzky was +518 over his career, but had more ESGA and almost certainly more minuses than any player in history.

 

[a pony]

It's called Corsi, forget that +/- ever existed