Why is +/- a bad stat?

[EdJovanovski]

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

Click to read more...

 

[LeafFever]

It was a good stat years ago. But the fact is it almost always depends on how good your team is. there are some severe examples of people leading the league in Worst +/-, and then having the best after a trade because they went from a bad team to a good one. (Chara is the best example).

 

[VinikToWinIt]

I think it's a bad stat when comparing players from different teams. It can be pretty telling when comparing players on the same team, although obviously it ignores situational usage and should be taken with a grain of salt.

 

[IamNotADancer]

You could literally lay unconscious on the ice and still get a + or a - . The stat doesn't really reflect the players direct contribution to success or failure.

 

[Shakey Rustie]

You could literally lay unconscious on the ice and still get a + or a - . The stat doesn't really reflect the players direct contribution to success or failure.

Yeah but you could do the same and get a secondary assist

 

[The Winter Soldier]

The stat is actually useful if you compare it to players on your own team and look in relation of how high or low they play in the line up. For instance a first pairing D man that is -10 tells you they are not really that effective on a first pairing, but a player that is similar playing on the 2nd pairing is +10.

Context is vital when using plus minus. When people say plus minus is a bad stat, it is when they do not use context. As with most stats, more layers of information is better. Another example, if your team is +50 in goal for/against ES. and you have a player that is a minus on the roster. Then that player is either miscast or not very effective.

 

[JoVel]

Yeah but you could do the same and get a secondary assist

Or a primary assist or a goal.

 

[HenrikW]

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

Edit

A dominant forward can be held back by poor goaltending and defense. A good defender can be held back by a lack of forward holding possession and the goaltender.

 

[JoVel]

It's in favor of players who plays no PK and tons of PP. Playing short-handed is bound to add to your minus', not only will the opposing team have better chance of scoring but it also takes away from 5v5/PP time because you'll need rest after playing shorthanded

It's also very dependant on the rest of the team. A dominant forward can be held back by poor goaltending and defense. A good defender can be held back by a lack of forward holding possession and the goaltender.

You don't get a minus of you get scored on on a power play. The goal scorers don't get a plus either.

 

[IamNotADancer]

Yeah but you could do the same and get a secondary assist

Once in a while sure, but with +/- you can be in a coma while they wheel your bed out there and at the end of the season you can still lead the league in +/-

 

[Laineux]

Because it lacks any meaningful predictive value.

It doesn't predict future success nearly as well as multitude of other stats.

 

[IamNotADancer]

It's in favor of players who plays no PK and tons of PP. Playing short-handed is bound to add to your minus', not only will the opposing team have better chance of scoring but it also takes away from 5v5/PP time because you'll need rest after playing shorthanded

It's also very dependant on the rest of the team. A dominant forward can be held back by poor goaltending and defense. A good defender can be held back by a lack of forward holding possession and the goaltender.

+/- doesn't exist during the PP

 

[DRW204]

imo it doesn't measure defensive prowess of an individual player across teams. i think it is a good stat to measure against players on their own team though.

 

[HenrikW]

You don't get a minus of you get scored on on a power play. The goal scorers don't get a plus either.

I stand corrected

 

[Quid Pro Clowe]

It's vague and often misleading.

 

[542365]

It's a reflection almost entirely of usage and team quality and people somehow twist it into meaning how good a player is defensively. It's incredibly stupid. It varies wildly from one season to another also, which is another sign of a poor statistic. It's not based on any skill a player does or does not possess. It's based on usage, team quality, and luck.

 

[pheasant]

You could literally lay unconscious on the ice and still get a + or a - . The stat doesn't really reflect the players direct contribution to success or failure.

Yes indeed!

I remember a story from Jeff O'Neil where he said that he got a -2 in a game because he jumped on the ice as the other team was on the rush. His skates barely hit the ice, both times, as the other team scored. He got a -2 for just coming off the bench for a change when he was supposed to.

+- isn't a great stat, because it can be influenced by so many factors, some of which have nothing to do with the players performance. Any player on a good team will have better +- than the same guy on a bad team, for example.

It can be useful when comparing players on the same team, like if one guy is a huge - while the rest are above 0. Or comparing one guys performance over a stretch with line 1 vs line 2.

But in most situations it is almost meaningless. You should never compare 2 players seasons using +- devoid of any other context. There has to be a reason and a proper frame of reference for it to matter.

 

[Nithoniniel]

Well, you're talking about a stat that is basically measuring what your whole unit does based on rare events (goals), where the parameters are all over the place (even strength, but also counts short handed goals, empty net goals etc), while not taking context into account at all.

Any time it's hard to tell what a stat actually measures, it's most likely a bad stat. Plus/Minus measures, roughly, what happens for the team when the player happens to be on the ice, in some but not all situations.

Another way to look at it is that stats are good because they have predictive or descriptive value. Some stats are good because you can use them to describe what happened in a game (descriptive value). Other stats are good because you can use them to predict what will happen going forward (predictive value).

A good example to look at would be Taylor Hall in Edmonton. His Plus/Minus there was really not good, despite him producing great and having good shot metrics, micro stats and so on. But when you look deeper, you notice that all of that could be pinned on two things: 1) Other teams scoring shorthanded goals, 2) His bad team playing a lot of games where they pull the goalie, and Hall being on the ice for those situations. If you take away the shorthanded goals and empty net goals, which says very little about his own ability, Taylor Hall was a plus player in Edmonton, despite how bad the team was.

Plus/Minus is derived from, in approximate order:
1) The strength of your team.
2) The relative ease of your minutes.
3) What your linemates does.
4) Your usage.
5) What your opposition does.
5) Your own ability.

A few years ago, the best two-way player in modern times was barely even. Talking about Patrice Bergeron, of course. He was terrific that season, just as always.

 

[OiledGun]

It also doesn't take into account that sometimes the best players on a good team will be out there when the goalie is pulled - which could easily end up a minus. Or in the last couple of minutes chasing a goal and pushing (i.e. making riskier plays to try and score a goal). Oliver Ekman-Larsson is an example of a very good player with a bad plus minus due to the team he plays on.

 

[CanadianPensFan1]

Its no better or worse than any other stat. Just needs to be taken in context.

The truth is, people will always say X stat is more important than Y but fail to take in to account variables and context.

 

[WingsMJN2965]

Jeff Schultz was +50 one season.

Any other questions?

 

[urho]

I think it's a bad stat when comparing players from different teams. It can be pretty telling when comparing players on the same team, although obviously it ignores situational usage and should be taken with a grain of salt.

But isn't this partly the case with advanced stats as well?

 

[qwerty]

It’s not a bad stat. It’s actually a really good stat if you don’t use it as some kind of indicator of how good defensively one individual is. But heck, who wouldn’t take players who are all +30’s? That’s 30 goals more than you gave up at even strength. Sounds like an effective line to me.

 

[TheDeuce]

You could literally lay unconscious on the ice and still get a + or a - . The stat doesn't really reflect the players direct contribution to success or failure.

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


m.

 

[Rants Mulliniks]

Like any stat, it isn't bad, it is how it is used.