Just a comment on the plus minus stat. Are you guys that are not taking the plus minus stat as a serious stat, supporting Corsi?

If so, how is Corsi calculated? How is plus minus calculated?
- Aussiepenguin

The two aren't mutually exclusive. One doesn't have to support plus/minus but distrust CORSI. Stats are a perspective on team and player performance. CORSI is a possession stat. It isn't perfect. Osprey cited Dustin Brown's decent CORSI numbers but his -10. If I tried to sell the fact that Brown was worth his contract based on his CORSI then that wouldn't confirm a subjective "eye test", his points/60, etc. relative to his cap hit.

With CORSI, the theory is the more you have the puck, the more shots you get, and the more shots you get (eventually) you will score more goals because the league average save percentage is around 91-92%. The last I-don't-know-how-many Cup winners were in the top 5 in CORSI. So that's added to its legitimacy as have the poor arguments of people that promote +/-.

Even strength, non-penalty shot goals while a player is on the ice - (even strength, non-penalty shot goals against while a player is on the ice + short handed, non-penalty shot goals against while a player is on the ice)=a player's plus minus

Shot attempts while a player is on the ice - shot attempts against while a player is on the ice=CORSI

The only completely reliable, unambiguous individual stat is games played. That tells the whole story: player x played at least 00:01 of y number of games. Not all that useful but it is indisputable.

+/- is not a good way to evaluate players.
- Jason_Lewis

i watch all the game and he was not so bad, i'll take a chance with a guy like that a low cost... i dont said that he's a superstar lol

Jason Lewis: Why Did Luke Schenn Not Fit?
- Jason_Lewis

More analysis than Luke Schenn is worth.

But... kudos, I guess.

Also, corsi is not a standalone. It should be complimented with other stats ala scoring chances, expected goals, zone starts, zone exit/entry, deployments, TOI, g/ga and various other rate statistics.

The world of advanced stats goes much deeper than just corsi.

Also, corsi is not a standalone. It should be complimented with other stats ala scoring chances, expected goals, zone starts, zone exit/entry, deployments, TOI, g/ga and various other rate statistics.

The world of advanced stats goes much deeper than just corsi.
- Jason_Lewis

Then why hasn't more data been introduced with plus minus stats?

They are generated in the exact same way. I find it amusing to see the people criticising plus minus then in the next breath support Corsi. Same formula for both stats.

Who says a team has 'possession' when a shot on goal occurs? The shooter may have slapped a puck off the stick of an opponent that goes on goal?

I realise Corsi is a product that creates further data, but if your core data is flawed doesn't that mean everything from that is also flawed? After all how many teams have "pass first" players skating around the offensive zone that give up break aways (as an example)??? Which team had better possession - the team that completed 10 passes to each other in the O zone without a SOG or the team that had the puck on their stick for 3 seconds but got a shot on goal?

Plus minus isn't a stat to bet your life on, far from it. But Corsi isn't either. Laugh at 1 laugh at both in my opinion!

The two aren't mutually exclusive. One doesn't have to support plus/minus but distrust CORSI. Stats are a perspective on team and player performance. CORSI is a possession stat. It isn't perfect. Osprey cited Dustin Brown's decent CORSI numbers but his -10. If I tried to sell the fact that Brown was worth his contract based on his CORSI then that wouldn't confirm a subjective "eye test", his points/60, etc. relative to his cap hit.

With CORSI, the theory is the more you have the puck, the more shots you get, and the more shots you get (eventually) you will score more goals because the league average save percentage is around 91-92%. The last I-don't-know-how-many Cup winners were in the top 5 in CORSI. So that's added to its legitimacy as have the poor arguments of people that promote +/-.

Even strength, non-penalty shot goals while a player is on the ice - (even strength, non-penalty shot goals against while a player is on the ice + short handed, non-penalty shot goals against while a player is on the ice)=a player's plus minus

Shot attempts while a player is on the ice - shot attempts against while a player is on the ice=CORSI

The only completely reliable, unambiguous individual stat is games played. That tells the whole story: player x played at least 00:01 of y number of games. Not all that useful but it is indisputable.
- Only_A_Ladd

Corsi can be referred to as 'the law of averages' stat then? Also backing that little gem up is the fact that analytics guys love the 'it all washes out in the end' if someone suggests that a player having no part in the play still gets a positive Corsi stat for that play.

Both are very raw data. It's just more data has been introduced with Corsi than plus minus. But as I replied in the other comment, if your core data is flawed everything from that is also compromised.

Then why hasn't more data been introduced with plus minus stats?

They are generated in the exact same way. I find it amusing to see the people criticising plus minus then in the next breath support Corsi. Same formula for both stats.

Who says a team has 'possession' when a shot on goal occurs? The shooter may have slapped a puck off the stick of an opponent that goes on goal?

I realise Corsi is a product that creates further data, but if your core data is flawed doesn't that mean everything from that is also flawed? After all how many teams have "pass first" players skating around the offensive zone that give up break aways (as an example)??? Which team had better possession - the team that completed 10 passes to each other in the O zone without a SOG or the team that had the puck on their stick for 3 seconds but got a shot on goal?

Plus minus isn't a stat to bet your life on, far from it. But Corsi isn't either. Laugh at 1 laugh at both in my opinion!
- Aussiepenguin

There are kind of some fundamental misunderstandings of how corsi is calculated in your response, so I'll just post this. It's a good read
http://www.arcticicehocke...ly-asked-questions-3-what


But as far as them being sort of the same. No, they are not.

Plus/Minus attempts to put a black and white stat on a game that is played 95% in gray area. It assumes or tries to establish a result but does nothing to represent the process. Shot attempts, AKA corsi or fenwick (If you are using the unblocked variety) analyze more of the process. As do scoring chances and expected goals.

Let's say the Kopitar line is on the ice. They play a near two minute shift in their own zone, get hemmed in, and allow a whopping 10 shot attempts through on net. Jonathan Quick stands on his head and none go in. The puck clears the zone, the line changes, and the Jeff Carter line comes on. The opposing team enters the zone, takes one shot attempt it goes in.

On the score sheet the Jeff Carter line is a minus, the Anze Kopitar line is an even.

In corsi or Fenwick however, we would see the Carter line at a -1, while the Kopitar line at a -10.

It is more representative of the flow of play than plus minus. You cannot look at a stat sheet and gauge how well a player did based on plus/minus on most nights. Yes, nights where players are a -4 and -5 or something extreme are generally bad corsi nights as well. HOWEVER, a simple -1, -2, or a +1, +2 are not at all indicative of quality of shifts, or play. If you looked at the boxscore without seeing the game you'd just say..oh Jeff Carter's line was a -1, they must not have played very well.

If you look at a +10 corsi player on a night and a -10 corsi player on a night, that generally means that when they were on the ice there were either way more shot attempts against them or for them. THAT is more indicative of the quality of the game.

Even that, however is grey. This is why we incorporate Scoring Chance data, Expected Goals, or matchups/usage into the mix. Were the shot attempts being taken good shot attempts? Or just low scoring attempts? Where did the start of play happen? Was it a Dzone start or Ozone. Who were they playing against? Etc.

Plus/minus literally gives you no information other than this guy was on the ice when his team scored or got scored against. It could have been after 20 shot attempts were taken against his line or after 1. There is no differentiation to a +1. No plus/minus stat is created equal, and given the tools we have it should be considered an antiquated, semi-useless stat at this point.

Hopefully that kind of clarifies just why we take so little credence as analytics people in plus/minus.

No, they are not.

Plus/Minus attempts to put a black and white stat on a game that is played 95% in gray area. It assumes or tries to establish a result but does nothing to represent the process. Shot attempts, AKA corsi or fenwick (If you are using the unblocked variety) analyze more of the process. As do scoring chances and expected goals.

Let's say the Kopitar line is on the ice. They play a near two minute shift in their own zone, get hemmed in, and allow a whopping 10 shot attempts through on net. Jonathan Quick stands on his head and none go in. The puck clears the zone, the line changes, and the Jeff Carter line comes on. The opposing team enters the zone, takes one shot attempt it goes in.

On the score sheet the Jeff Carter line is a minus, the Anze Kopitar line is an even.

In corsi or Fenwick however, we would see the Carter line at a -1, while the Kopitar line at a -10.

It is more representative of the flow of play than plus minus. You cannot look at a stat sheet and gauge how well a player did based on plus/minus on most nights. Yes, nights where players are a -4 and -5 or something extreme are generally bad corsi nights as well. HOWEVER, a simple -1, -2, or a +1, +2 are not at all indicative of quality of shifts, or play. If you looked at the boxscore without seeing the game you'd just say..oh Jeff Carter's line was a -1, they must not have played very well.

If you look at a +10 corsi player on a night and a -10 corsi player on a night, that generally means that when they were on the ice there were either way more shot attempts against them or for them. THAT is more indicative of the quality of the game.

Even that, however is grey. This is why we incorporate Scoring Chance data, Expected Goals, or matchups/usage into the mix. Were the shot attempts being taken good shot attempts? Or just low scoring attempts? Where did the start of play happen? Was it a Dzone start or Ozone. Who were they playing against? Etc.

Plus/minus literally gives you no information other than this guy was on the ice when his team scored or got scored against. It could have been after 20 shot attempts were taken against his line or after 1. There is no differentiation to a +1. No plus/minus stat is created equal, and given the tools we have it should be considered and antiquated, semi-useless stat at this point.

Hopefully that kind of clarifies just why we take so little credence as analytics people in plus/minus.
- Jason_Lewis

I'll respond very quickly as I have to go out. But as I said above, Corsi has created other forms of data & is also used with that. It has/is evolving as a stat where as plus minus hasn't & isn't. They are both generated the same way though so why has 1 grown & the other not? If further data were to be put with both stats equally they would both be useful. As I said earlier I'm comparing Corsi to plus minus as both are generated the same way.

There may be part 2 to this later today/tonite as I do look only at the box score for a lot of games!

I'll respond very quickly as I have to go out. But as I said above, Corsi has created other forms of data & is also used with that. It has/is evolving as a stat where as plus minus hasn't & isn't. They are both generated the same way though so why has 1 grown & the other not? If further data were to be put with both stats equally they would both be useful. As I said earlier I'm comparing Corsi to plus minus as both are generated the same way.

There may be part 2 to this later today/tonite as I do look only at the box score for a lot of games!
- Aussiepenguin

Plus/minus hasn't evolved because you literally can't evolve it. It is the end product. One goal for. One goal against. There is no where else you can go with it.

Corsi IS the evolution of such a stat, as it tries to explain the flow of play versus putting a single, finite value on it like one goal against, one for.

Well put.
- Jason_Lewis

Also, corsi is not a standalone. It should be complimented with other stats ala scoring chances, expected goals, zone starts, zone exit/entry, deployments, TOI, g/ga and various other rate statistics.
- Jason_Lewis

You complimented the strawman that +/- is flawed when used as a standalone stat and then suggested that Corsi isn't standalone, itself. You then went on and double-downed on that by treating +/- all by itself and re-iterating that Corsi should be complemented with other data.

Plus/minus literally gives you no information other than this guy was on the ice when his team scored or got scored against.
- Jason Lewis

Corsi For/Against literally gives you no information other than this guy was on the ice when his team shot or was shot against. It's easy to reduce a stat to seeming irrelevance like that because every stat gives you very little information and not nearly enough to make a judgment on its own. If you're forcing one stat to stand all on its own, but granting another the complement of additional data, you're holding a double standard.

Then why hasn't more data been introduced with plus minus stats?

They are generated in the exact same way. I find it amusing to see the people criticising plus minus then in the next breath support Corsi. Same formula for both stats.

...

Plus minus isn't a stat to bet your life on, far from it. But Corsi isn't either. Laugh at 1 laugh at both in my opinion!
- Aussiepenguin

You seem to get it. Stats are just facts. They are not good or bad. They can only be used or misused. People will misuse +/-, a stat that everyone understands, so that they can promote fancier stats which they believe that they understand better than others. Everyone knows that +/- has issues (like every stat), but proclaiming it nearly worthless is just a symbolic show to make yourself look progressive, knowledgeable and sure of yourself. It's apparently not good enough to simply use the new stats alongside the old. You have to actually discredit the old ones, even if there's still some merit in them and what you're promoting has some issues of its own. People tend to give more credibility to those who speak absolutes in a convincing manner (i.e. "+/- bad, Corsi good") than to those who try not to.

Plus/minus hasn't evolved because you literally can't evolve it. It is the end product. One goal for. One goal against. There is no where else you can go with it.

Corsi IS the evolution of such a stat, as it tries to explain the flow of play versus putting a single, finite value on it like one goal against, one for.
- Jason_Lewis

There are a few things to be said about your response. The first would be you are correct when you say there is 1 goal for (an actual data event) & 1 against (again an actual event). So you have an actual event to detail. What details come from that is up to which argument you are for or what data you want to use that is available. Plus minus is a lazy stat as it simply snapshots who is on the ice when a goal is scored. There is so much more data to be used if you want to dissect the play that leads to the goal scored - but it appears there is no formula or 'goal information' that does that due to its time consumption of reviewing each goal, but it can be done. So plus minus stagnates as the snapshot when a goal is scored.

Corsi however (even though formulated the 'exact' same way only there is no actual event - it's an opinion based on human judgement that a shot towards goal is a SOG), is evolving using math calcs as they find a fit with how a game 'can' flow. Corsi is also a stat generated over different games (the more the better right?), so the data used to produce a result is flawed due to it having so many different environments it's being produced from formulated to provide that single result - that's not even taking into conditions of the actual player (fit, injured, sick, playing out of position, playing in a team that's strong/weak etc etc). Single game results are considered a question mark because Corsi statistics like to have high volumes of data to get a better 'prospective', so single game Corsi stats aren't taken too seriously by the analytics world (or so I'm told please correct me if I'm wrong).

So when a player is given a positive Corsi point because he was on the ice at the time a shot was judged to be on goal even though he had zero input into that shot, what other stats can be formulated based on that positive Corsi event? What growth can be seen in his personal statistics due to that single 'perceived positive event'?

You seem to get it. Stats are just facts. They are not good or bad. They can only be used or misused. People misuse +/-, a stat that everyone understands, so that they can promote fancier stats which they believe few but them fully understand. Everyone knows that +/- has issues (like every stat), but proclaiming it nearly worthless is just a symbolic show to make yourself look progressive, knowledgeable and sure of yourself. It's apparently not good enough to simply use the new stats alongside the old. You have to actually discredit the old ones, even if there's still some merit in them and what you're promoting has some issues of its own. People tend to give more credibility to those who speak absolutes in a convincing manner (ex. "+/- bad, Corsi good") than those who try not to.
- Osprey

Most stats are facts - not all. An event that gives a positive/negative point when a player does zero to cause the result, I wouldn't call that stat a fact.

I love stats & fully support the efforts to dissect games to produce stats that are correct & not needing a math formulae to find a result. But in hockey there is so much kaos, finding actual events to formulate detailed statistics is hard when there is so much change that never repeats or has no pattern. That's why hockey is so great due to the complete unpredictability of the games. Corsi is a false economy, & going back to my original comment, people criticising plus minus then supporting Corsi is insane due to them both being formulated the same way.

There are those that cherry pick to provide 'support' for their opinion. Anyone can find something that will support their individual belief with these 'new wave' analytics which in itself should be a huge warning that the analytics isn't a proven truth. If it's not a proven truth then it's not a fact. They're getting there & the data is there it just all needs to be utilised.

There are kind of some fundamental misunderstandings of how corsi is calculated in your response, so I'll just post this. It's a good read
http://www.arcticicehocke...ly-asked-questions-3-what


But as far as them being sort of the same. No, they are not.

Plus/Minus attempts to put a black and white stat on a game that is played 95% in gray area. It assumes or tries to establish a result but does nothing to represent the process. Shot attempts, AKA corsi or fenwick (If you are using the unblocked variety) analyze more of the process. As do scoring chances and expected goals.

Let's say the Kopitar line is on the ice. They play a near two minute shift in their own zone, get hemmed in, and allow a whopping 10 shot attempts through on net. Jonathan Quick stands on his head and none go in. The puck clears the zone, the line changes, and the Jeff Carter line comes on. The opposing team enters the zone, takes one shot attempt it goes in.

On the score sheet the Jeff Carter line is a minus, the Anze Kopitar line is an even.

In corsi or Fenwick however, we would see the Carter line at a -1, while the Kopitar line at a -10.

It is more representative of the flow of play than plus minus. You cannot look at a stat sheet and gauge how well a player did based on plus/minus on most nights. Yes, nights where players are a -4 and -5 or something extreme are generally bad corsi nights as well. HOWEVER, a simple -1, -2, or a +1, +2 are not at all indicative of quality of shifts, or play. If you looked at the boxscore without seeing the game you'd just say..oh Jeff Carter's line was a -1, they must not have played very well.

If you look at a +10 corsi player on a night and a -10 corsi player on a night, that generally means that when they were on the ice there were either way more shot attempts against them or for them. THAT is more indicative of the quality of the game.

Even that, however is grey. This is why we incorporate Scoring Chance data, Expected Goals, or matchups/usage into the mix. Were the shot attempts being taken good shot attempts? Or just low scoring attempts? Where did the start of play happen? Was it a Dzone start or Ozone. Who were they playing against? Etc.

Plus/minus literally gives you no information other than this guy was on the ice when his team scored or got scored against. It could have been after 20 shot attempts were taken against his line or after 1. There is no differentiation to a +1. No plus/minus stat is created equal, and given the tools we have it should be considered an antiquated, semi-useless stat at this point.

Hopefully that kind of clarifies just why we take so little credence as analytics people in plus/minus.
- Jason_Lewis

Just to clarify your first couple of sentences. If a Pens player shoots at the net & scores, each Kings player gets a -1 & Pens player a +1. If that same shot is saved by Quick, each Pens player gets a Corsi +1 & Kings Corsi -1. So where is the difference you are saying?

I'm just feeling the general analytical spiel here that everything you guys want to say is relevant & anything anyone says that opposes that are not understanding the 'data' or are misunderstanding the results. Your example above can be interpreted a few different ways, & what you are saying about anyone's understanding of the data you have provided is so naive. Anyone that looks at any data associated with hockey & knows how hockey can change, will also understand what isn't in the totals shown - apparently you suggest for plus minus that's a definitive stat that no one can accept any other way other than its most basic form, but Corsi should be taken differently understanding that is only 1 form of data that has to be utilised with other data? Isn't Corsi actual shots on goal?? Your answer here would be appreciated.

My example above with players that have a pass first mentality that sit in the O zone but dont shoot gets what credit on your possession flow through Corsi? Alternatively the breakaway or counter attack players that shoot every chance they get shows what type of possession? Your immediate reply may be to look at other stats, but on face value which is what we are talking about, your Corsi stat does not give you a true indication of the game. That's what you are talking about isn't it - face value statistics?? I can go further & break down the individual Corsi stats which are worse (team Corsi I accept as a legitimate stat as it's the 'team' shots on goal although as mentioned above, don't count it as actual 'possession flow' as it's not a true event for possession it's more 'probability' of possession). Legitimate shots on goal can be counted as they are true events (based on human perception).

Just to clarify your first couple of sentences. If a Pens player shoots at the net & scores, each Kings player gets a -1 & Pens player a +1. If that same shot is saved by Quick, each Pens player gets a Corsi +1 & Kings Corsi -1. So where is the difference you are saying?

I'm just feeling the general analytical spiel here that everything you guys want to say is relevant & anything anyone says that opposes that are not understanding the 'data' or are misunderstanding the results. Your example above can be interpreted a few different ways, & what you are saying about anyone's understanding of the data you have provided is so naive. Anyone that looks at any data associated with hockey & knows how hockey can change, will also understand what isn't in the totals shown - apparently you suggest for plus minus that's a definitive stat that no one can accept any other way other than its most basic form, but Corsi should be taken differently understanding that is only 1 form of data that has to be utilised with other data? Isn't Corsi actual shots on goal?? Your answer here would be appreciated.

My example above with players that have a pass first mentality that sit in the O zone but dont shoot gets what credit on your possession flow through Corsi? Alternatively the breakaway or counter attack players that shoot every chance they get shows what type of possession? Your immediate reply may be to look at other stats, but on face value which is what we are talking about, your Corsi stat does not give you a true indication of the game. That's what you are talking about isn't it - face value statistics?? I can go further & break down the individual Corsi stats which are worse (team Corsi I accept as a legitimate stat as it's the 'team' shots on goal although as mentioned above, don't count it as actual 'possession flow' as it's not a true event for possession it's more 'probability' of possession). Legitimate shots on goal can be counted as they are true events (based on human perception).
- Aussiepenguin

Thank you for this response. I think it gets at the core disconnect.

CORSI versus +/- is not a battle over which one shows the truth. Each are a measure of performance. Neither perfectly tells the story. Each are flawed. One is more flawed than the other, and the other tells a better story. Advanced statistics are a means to get closer to measuring the true value of player and team performance, but it is a futile pursuit. It should be reframed as a pursuit to get closer to measuring the true value of a player but not the actual truth. No statistic well ever truly and accurately measure performance and get at this point of capturing what you describe as a "true event". Ever. It won't happen.

[double post]

[triple post]

[quadruple post!]

[quadruple post!]
- Only_A_Ladd

(frank) deeznuts!!!

(frank) deeznuts!!!
- Aussiepenguin

sick kid+3 a.m. in the morning=nothing is functioning right

Most stats are facts - not all. An event that gives a positive/negative point when a player does zero to cause the result, I wouldn't call that stat a fact.
- Aussiepenguin

That's still a fact. It says that the result occurred while the player was on the ice. That's all. Anything more that you read into is how you choose to use that fact to draw a conclusion or support a point. A stat doesn't tell you what contributed to it. It just is. It's up to you to guess at what contributed to it.

Corsi is a false economy, & going back to my original comment, people criticising plus minus then supporting Corsi is insane due to them both being formulated the same way.
- Aussiepenguin

Yes, I've said pretty much the same thing. They're calculated the same flawed way (by whether the player was simply on the ice when the event occurred) and they both tell you very little on their own, but some people insist on choosing sides and tossing one out to embrace the other, which is nonsensical.

CORSI versus +/- is not a battle over which one shows the truth. Each are a measure of performance. Neither perfectly tells the story. Each are flawed. One is more flawed than the other, and the other tells a better story. Advanced statistics are a means to get closer to measuring the true value of player and team performance, but it is a futile pursuit. It should be reframed as a pursuit to get closer to measuring the true value of a player but not the actual truth. No statistic well ever truly and accurately measure performance and get at this point of capturing what you describe as a "true event". Ever. It won't happen.
- Only_A_Ladd

The thing is that Corsi and +/- don't tell the same story. Corsi gives you a better sense of the "flow of play" (as Jason put it) than +/-, but it tells you absolutely nothing about the results. Did any of those shot attempts turn into goals for or against? Who knows. +/-, on the other hand, tells you the opposite: how many goals were scored for versus against, but tells you nothing about the flow of play. Because they tell virtually opposite stories, a wise analyst would use both to help construct a total picture, not discard the one that's been around a while and isn't as sexy.

The thing is that Corsi and +/- don't tell the same story. Corsi gives you a better sense of the "flow of play" (as Jason put it) than +/-, but it tells you absolutely nothing about the results. Did any of those shot attempts turn into goals for or against? Who knows. +/-, on the other hand, tells you the opposite: how many goals were scored for versus against, but tells you nothing about the flow of play. Because they tell virtually opposite stories, a wise analyst would use both to help construct a total picture, not discard the one that's been around a while and isn't as sexy.
- Osprey

Which is definitely why rate statistics of goals for/scoring chances/shooting percentages are great compliments to corsi! Again, corsi is not a standalone but it tells you way more than plus minus, thus making it severely antiquated. Especially when you consider goals for/against per 60 along with it. (More accurately represents production in respects to ice time.). So, to follow up, it's not really "tossing one to embrace the other" it's the simple fact that +/- offers very little analytical value considering what we have aside from it, that actually does a better job with the same concept.

Overall, good discussion on the different mindsets I'd say.

I always like Luke Schenn, it is a shame that a guy with so much upside was not able to live up to his potential. The guy just doesnt look confident in his ability. This guy IMO had the tools to be a stud Dman. Physically he was built like Pronger (Not saying he could be anywhere near Pronger's skill level), he even had a nice blast from the point. Problem like Lewis mentioned was there was no instinct. His shot as powerful as it was was not very accurate and he struggles against faster players. He had the skills to be a good defenceman but his decision making is what kills him.

Which is definitely why rate statistics of goals for/scoring chances/shooting percentages are great compliments to corsi! Again, corsi is not a standalone but it tells you way more than plus minus, thus making it severely antiquated. Especially when you consider goals for/against per 60 along with it. (More accurately represents production in respects to ice time.). So, to follow up, it's not really "tossing one to embrace the other" it's the simple fact that +/- offers very little analytical value considering what we have aside from it, that actually does a better job with the same concept.
- Jason_Lewis

The fact that you can go to a lot of trouble (analyzing Corsi and "rate statistics for goals for/scoring chances/shooting percentages") to get a more accurate picture doesn't make +/- "antiquated." +/- is a very useful boiling down of the data, an approximation, if you will. Naturally, the full collection of data tells you a lot more. If a scientific journal publishs a 20-page report on climate temperatures, those 20 pages will naturally tell you a lot more than the abstract on the first page, which summarizes and boils down the findings, giving you a general idea of the data, but that doesn't take away from the usefulness of the abstract.

As a bean counter, you may scoff at +/- and prefer to jump into the sea of data, and that's your right, but that doesn't take away the value of such a stat. It's valuable to those who don't want to pore over the various data and it can even be valuable to folks like you, though you may not care to admit it. For example, if you see a high-Corsi player with a high +/- (like Toffoli), you can assume that there isn't a problem there, but, if you see a high-Corsi player with a poor +/- (like Brown), that might be a red flag that encourages you to do more investigation, and that investigation would most likely reveal areas where the player has struggled (like in shooting percentage and shot quality). You're not likely to see someone with a poor +/- whose enhanced stats are all sterling. You're going to find some stats which explain why more goals were scored against him at even strength than for.

+/- is very easy to understand, is very easy to calculate and tells you a reasonable amount, considering. That's rather impressive and shouldn't be discounted. Yes, you can tell more by digging into stats that are harder to understand and calculate, but poo-pooing the simple computation to suggest that the complex, time-consuming one is the only proper way can come across as elitist. Not everyone has the luxury of making this his job and spending the better part of his afternoon poring over and making sense of the data. For those without time to kill, +/- gives a quick and general idea of whether a player is struggling or not. There may be more accurate ways to judge that, but they're not necessarily "better" just because they're more accurate. There's beauty in a simple solution, even if it's not the most accurate or efficient one.