Why is +/- a bad stat?

[Duke Silver]

Ignores WAY too much context to be useful.

People use it incorrectly as a measure of an individual's defensive ability. It's not.

If Plus/Minus was a true indication of defensive ability, Jeff Schultz would have a Norris trophy.

 

[lomiller1]

It has its benefits. If you're -30 and the rest of your team is +20 better...we know you suck.

Are you sure? To get to -30 you need to be receiving a lot of ice time, probably a good amount of PP time and probably you are being relied on when your team is down by a goal in the final min of a game. Would a player who sucks really be used like this?

The truth is that the legitimately bad players tend to end up with middle of the pack +/- relative to their teams.

 

[Menzinger]

Ignores too much context and goals are too rare of an event to draw inferences from

 

[tarheelhockey]

It has its benefits. If you're -30 and the rest of your team is +20 better...we know you suck.

Or you could be the one good player who skates with 2 ECHL level linemates.

 

[Kahvi]

But can't the same thing be said of Corsi and Fenwick?

I wondered this myself lately as well. Shots on goal happen more often than goals (of course), so sample size favours Corsi and Fenwick

 

[zeke]

Why is +/- considered such a bad stat? and what do you consider to be the best stats?

1. it's a team stat.
2. it ignores quality of competition, meaning that the best defensive players on a team will often have the worst plus minus thanks to logging all the tough matchups, while the sheltered bad players will have the best.
3. team goals for and against are a very crude way of trying to measure any individual player's offensive or defensive play.

you throw these 3 things together and the stat becomes useless in every possible way.

it's not just a bad stat - it's so awful that much of the time it tells you the exact opposite of what you think it's telling you. it not only has no value as a stat - it actually has negative value.

 

[Maukkis]

Looking at Haula's +/-, he was a -4 in that game. Heck, if you want to include special teams in this equation, he even scored twice - and received no +'s for those goals.

Everything else, most of which was very positive, is disregarded by only looking at +/-. The stat ignores way too much to be considered credible.

Everything that is necessary has already been said, but I'll add this: +/- is good for identifying players on non-playoff teams, particularly defensemen and forwards playing (a lot of) defensive minutes. Out of the 50 worst players in terms of +/-, just five are on playoff teams.

 

[mouser]

+/- isn't inherently a bad stat. It has usefulness when used in an appropriate context.

The issue is too often posters will reference +/- in a bad context.

 

[KMart27]

It is viewed as a defensive stat but it isn't really a defensive stat.

 

[jetsforever]

I stand corrected

Only shorthanded goals count for +/- during a PP, so playing shorthanded a lot will actually help your +/- over time.

 

[Throw More Waffles]

So you have the best shutdown line in the league on the highest scoring team in the league, the line is not expected to score and will occasionally get scored on, but much less than others would playing against top players. They are -20, the team is +30. What has +/- told you in this situation?

Couldn't such situations be given to any stat? Including assists and goals? If a player who usually has a lot of assists all of a sudden gets very few assists one year, we just say he had a lousy year. We don't make a whole bunch of weird excuses.

In your example above, I'd say that line is still pretty lousy. They get scored on far too often compared to how much they score. They're a liability.

 

[lomiller1]

+/- isn't inherently a bad stat. It has usefulness when used in an appropriate context.

The issue is too often posters will reference +/- in a bad context.

I hate to have to keep repeating this, but what you say is may be true of GF%, it’s not the case for +/-. The formula for +/- is broken so it’s not a stat that should be used in any context.

 

[LeafsNation75]

It's a team stat, not individual player. Kessel didn't magically become some defensive specialist overnight when he got traded to Pittsburgh. Roman Polak is not better than Morgan Rielly.

Maybe it had to do with Kessel playing in Toronto but when he was always bashed by fans of other teams one stat they would always bring up was how bad his +/- was.

 

[Alberta_OReilly_Fan]

Why is +/- considered such a bad stat? and what do you consider to be the best stats?

takes too much effort for casual fans to understand it. like many stats, its only as good as the understanding of what it means. plus/minus cant be universally applied. it wont mean the same thing to every player. its best measure is as a comparable. you try to measure apples to apples and oranges to oranges. you need to filter out influencers that are not balanced

if a player is a plus year after year after year... in whatever situation he finds himself... like a zdeno chara... then the stat has some meaning. but if a player is constantly stuck on a horrible team... the players he plays with have the same lousy number he has... the stat has less significance

and you have to factor in that hockey is a team game... a lot of players are given an assignment to be in a certain place to take away a certain part of the ice. if they are doing their job, and someone else breaks down and a goal is scored... obviously the plus/minus isn't as meaningful.

I guess in order to try to make this stat more meaningful... we need a plus/minus that only measures when you are directly involved in the goal. where you blew your assignment... or you were right there involved in the scoring play

but I wouldn't call it a bad stat... just one that requires a lot of work to get any true value out of in most cases and one that does get used too often by poorly informed fans to create a misunderstanding

 

[Atoyot]

Couldn't such situations be given to any stat? Including assists and goals? If a player who usually has a lot of assists all of a sudden gets very few assists one year, we just say he had a lousy year. We don't make a whole bunch of weird excuses.

In your example above, I'd say that line is still pretty lousy. They get scored on far too often compared to how much they score. They're a liability.

That's the point, there are too many confounding variables to the point where the stat literally tells you nothing as evidenced by the fact that it has zero predictive value.

 

[Throw More Waffles]

Ignores too much context and goals are too rare of an event to draw inferences from

So the amount of goals a player actual scored is significant. The amount of assists a player actually gets is significant. But how many goals go in while they're on the ice isn't significant? I fail to see why...

 

[Throw More Waffles]

That's the point, there are too many confounding variables to the point where the stat literally tells you nothing as evidenced by the fact that it has zero predictive value.

The best shut down line in the league wouldn't be -20 on the highest scoring team in the league. So your comparison has a faulty premise.

 

[ImNeverWrong]

Are you sure? To get to -30 you need to be receiving a lot of ice time, probably a good amount of PP time and probably you are being relied on when your team is down by a goal in the final min of a game. Would a player who sucks really be used like this?

The truth is that the legitimately bad players tend to end up with middle of the pack +/- relative to their teams.

Well David desharnais is -23 playing 13 mins a night. Safe to say he's not goood.

 

[PatrikBerglund]

It's only an important stat if my favorite player has a high plus-rating.

 

[Tom ServoMST3K]

I would probably take the league low +/- player on my team.

So I guess in that sense it's a useful stat

But Seriously, why do EN and SH goals count, but PP goals dont?

 

[Damisoph]

Too many variables, like lots of stats in hockey.

 

[LeafsNation75]

But Seriously, why do EN and SH goals count, but PP goals dont?

That's something I have also wanted to know.

 

[Baywulf]

VKG's EV GD is ~0 when WK is out of ice and it's ~+40 when he is playing. Does that tell you something?

 

[Cup or Bust]

That sounds like a misunderstanding of how +/- works as those teams with better goal differentials have PP goals counted in those totals. As an example, Tampa has 30 more PP goals than Edmonton. That's 30 goals towards their differential, none of which count toward player +/-.

Edit - that's just 5 on 4 goals for.

I was making a point of how hard it is to maintain a high +/- on a weak team in general. There is always exceptions to the rule but for the most part teams with positive goal differentials are better then teams with negative goal differentials. Tampa has an even strength goal differential of +48 compared to Edmonton's -1. McDavid is +22 and Kucherov is +23 and Stamkos is +18. Pittsburgh is +1 in even strength goal differential and Crosby is -2 and Malkin is +17. Colorado has a +9 even strength goal differential and MacKinnon is +15. It is a reasonable stat to use in that context.