NHL Referees Like Even things up (538 Sports Article)

[xxgt465xx]

http://fivethirtyeight.com/features...to-get-you-if-they-already-got-the-other-guy/

Pretty interesting stuff right here. Sort of statistically confirms a lot of fans' suspicions. People will still call it BS, but statistics are rather hard to argue with.


Also, if i posted this in some way that isn't acceptable here then let me know or just edit it so its good.

Click to read more...

 

[coolboarder]

Evening up the calls to make them appear impartial is one of my biggest pet peeve in the hockey. No matter the score, if a trailing team is owed a penalty, the trailing team can try break the rules without any call is unfair to the team that earned their way by playing the rules and would look out for one small infraction that they let it go earlier and made this same call while other team can make the same infraction that they called would let it go and let them play. This is not a fairest way to officiate a game.

Just like a football team, they earned their way through the field and a defending team made a turnover and would call a pass inference that is unearned just because he appeared to be interfered with but made a great play to swat the ball down first and made a catch but the referee saw differently and made a small call which is not legitimate infraction made a huge difference all because of the time and the quarter of the game and who they are trailing is unearned. This is the same in hockey, it must be earned, not given to even up the calls.

 

[FissionFire]

The whole evening up phenomena is a big reason why teams are able to compete better, especially in the playoffs. In fact, many teams have made it part of their strategy. The old Minnesota and Anaheim teams were great at it. The Devils in the 90s. More recently Boston and Detroit have mastered it. Once referees stop being afraid of having a team get 4-5x the PP chances in a game then things will change. Right now in trying to appear impartial the refs are just being unintentionally biased. I don't think it's any type of conscious choice, just human nature. It's hard to believe that a team having even a 3+ edge in PPs in a game is such a statistical rarity, and practically nonexistent in the playoffs.

 

[TheDevilMadeMe]

The whole evening up phenomena is a big reason why teams are able to compete better, especially in the playoffs. In fact, many teams have made it part of their strategy. The old Minnesota and Anaheim teams were great at it. The Devils in the 90s. More recently Boston and Detroit have mastered it. Once referees stop being afraid of having a team get 4-5x the PP chances in a game then things will change. Right now in trying to appear impartial the refs are just being unintentionally biased. I don't think it's any type of conscious choice, just human nature. It's hard to believe that a team having even a 3+ edge in PPs in a game is such a statistical rarity, and practically nonexistent in the playoffs.

The 90s Devils were regularly last place in both penalties for and against. If anything, they were victims of their own discipline and refs not wanting to call lopsided penalties on the other team.

 

[Lacaar]

For me it sounds wrong when the refs say they have to "manage" the game.

They don't/shouldn't manage anything. You call penalties when they happen. that's it.
The fact they feel they need to manage anything tells me they aren't impartial.

And that's what hurts the quality of the game.
As soon as a team has figured out the Ref's "Management style" they take advantage.

Don Cherry complains that there's so much stick work because there's not much fighting. I say it's because the refs insistence on "managing" the friggin hockey game.
If you got a penalty every time you slashed... you'd stop slashing. But if you know the ref won't call it.. slash away.

We're down by 2 goals.. slashing is permitted.
We've had the last 3 penalties. Slashing is permitted.
I broke a wrist slashing. Whoops.
I broke a stick slashing. Penalty!

Hell it's probably why the playoffs tend to get less entertaining. They've stickworked each other so much they're all sore.

 

[Canadiens1958]

Ppo/ppoa

Suggest checking the PPO/PPOA before going forward. 2014-15 below:

http://www.hockey-reference.com/leagues/NHL_2015.html

 

[xxgt465xx]

Suggest checking the PPO/PPOA before going forward. 2014-15 below:

http://www.hockey-reference.com/leagues/NHL_2015.html

Any explanation of your point?

 

[Doshell Propivo]

Hell it's probably why the playoffs tend to get less entertaining.

They do?

 

[Canadiens1958]

Explanation

Any explanation of your point?

Yes. PPOs range from 212 to 294 while PPOA range from 186 to 308. League average for the season was 251.

The two ranges suggest that various factors were not considered.

Specifically fatigue - are teams playing their 3rd game in 4 nights more prone to penalties like hooking, holding, interference as opposed to fresh teams playing their first game in 4 nights.

Mandatory calls - puckover the glass. Their impact.

Just a short list of questions about the study.

 

[hatterson]

Yes. PPOs range from 212 to 294 while PPOA range from 186 to 308. League average for the season was 251.

The two ranges suggest that various factors were not considered.

Specifically fatigue - are teams playing their 3rd game in 4 nights more prone to penalties like hooking, holding, interference as opposed to fresh teams playing their first game in 4 nights.

Mandatory calls - puckover the glass. Their impact.

Just a short list of questions about the study.

I don't really see how any of these would affect the study in a meaningful way.

Is there a reason to believe that PPO vs PPOA ranges would affect the order in which teams take penalties?

Is there a reason to believe that mandatory penalties are more likely committed by the team that is "owed" a penalty?

Or that tired teams are more likely to take (or receive) penalties more when they're owed?

 

[GoldenBearHockey]

I don't really see how any of these would affect the study in a meaningful way.

Is there a reason to believe that PPO vs PPOA ranges would affect the order in which teams take penalties?

Is there a reason to believe that mandatory penalties are more likely committed by the team that is "owed" a penalty?

Or that tired teams are more likely to take (or receive) penalties more when they're owed?

Yes, there are reasons to believe that tired teams will take more penalties,

There are also reasons why teams trailing tend to create more penalties or put another way, teams leading, tend to take more penalties,

When teams are leading, they will tend to sit back in the last 10 min of a game, and protect that lead, aka, over the red line, dump, go from a hard forecheck to a soft trap etc, which means the trailing team has the puck more. It's much easier to draw a call when you have the puck and you are attacking the net which you have to do when you are trailing, than it is when you don't have the puck and don't have to attack.

 

[hatterson]

Yes, there are reasons to believe that tired teams will take more penalties,

There are also reasons why teams trailing tend to create more penalties or put another way, teams leading, tend to take more penalties,

When teams are leading, they will tend to sit back in the last 10 min of a game, and protect that lead, aka, over the red line, dump, go from a hard forecheck to a soft trap etc, which means the trailing team has the puck more. It's much easier to draw a call when you have the puck and you are attacking the net which you have to do when you are trailing, than it is when you don't have the puck and don't have to attack.

You must have missed the meaning of "when they're owed"

The study strictly looks at the effects of penalties previously called in the game and the probability of the next penalty, not how score or tiredness affects overall rates.

I don't really see a reason to believe that a tired team that has been called for 2 fewer penalties than their opposition thus far in the game is more likely to commit another one as opposed to a tired team that has been called for 2 more.

No, this isn't an exhaustive review, and there's certainly things that can be added, but I don't see how any of the points raised would weaken the underlying result it seems to be pointing at.

 

[GoldenBearHockey]

You must have missed the meaning of "when they're owed"

The study strictly looks at the effects of penalties previously called in the game and the probability of the next penalty, not how score or tiredness affects overall rates.

I don't really see a reason to believe that a tired team that has been called for 2 fewer penalties than their opposition thus far in the game is more likely to commit another one as opposed to a tired team that has been called for 2 more.

No, this isn't an exhaustive review, and there's certainly things that can be added, but I don't see how any of the points raised would weaken the underlying result it seems to be pointing at.

The bolded is why I don't lend credence to the study, it doesn't take into factors that it should, how many of those so called "owed" penalties are like someone said due to mandatory penalties, puck over glass etc..

 

[xxgt465xx]

The bolded is why I don't lend credence to the study, it doesn't take into factors that it should, how many of those so called "owed" penalties are like someone said due to mandatory penalties, puck over glass etc..

Long-run statistics generally negate those factors.

 

[Canadiens1958]

Ranges

I don't really see how any of these would affect the study in a meaningful way.

Is there a reason to believe that PPO vs PPOA ranges would affect the order in which teams take penalties?

Is there a reason to believe that mandatory penalties are more likely committed by the team that is "owed" a penalty?

Or that tired teams are more likely to take (or receive) penalties more when they're owed?

No one is claiming that the ranges affect the order in which teams take penalties.PPO range for 2014-15 was 212 to 294, a swing of 82 penalties. This shows that when compiled over a season the results reflect(not affect) that certain teams may be prone to forcing the key first penalty of the game.

Likewise PPOA with a range of 186 to 308, a swing of 122 penalties reflects that certain teams may be more likely to take the key first penalty.

The issue then becomes why are certain teams prone to take first penalty, which if we give even the slightest credibility to the concept of a team being "owed a penalty becomes the triggering event - until you have the first penalty none of the two participating teams may be viewed as being "owed" a penalty.

At this point the various factors - fatigue, talent, mandatory,obstruction penalties, physical penalties, etc enter the picture.

Look at the penalties for the October 28,2014 game Montréal at Calgary, Canadiens were playing their 3rd game in 4 nights, having played the Rangers in Montréal on the 25th, in Edmonton on the 27th and in Calgary on the 28th. Calgary had a home game on the 25th against Washington so they were sitting at home waiting to play a tired team that had traveled across two time zones. Canadiens received the last 2nd period penalty plus three late(second half) third period penalties. Was Calgary somehow owed these penalties or was it fatigue? Six possibly seven of the Montreal penalties were fatigue type penalties - kneeing could go either way.

http://www.hockey-reference.com/boxscores/201410280CGY.html

 

[hatterson]

No one is claiming that the ranges affect the order in which teams take penalties.PPO range for 2014-15 was 212 to 294, a swing of 82 penalties. This shows that when compiled over a season the results reflect(not affect) that certain teams may be prone to forcing the key first penalty of the game.

Likewise PPOA with a range of 186 to 308, a swing of 122 penalties reflects that certain teams may be more likely to take the key first penalty.

I don't understand how PPO or PPOA ranges show anything about which teams are more or less likely to take the first penalty in a game.

The issue then becomes why are certain teams prone to take first penalty, which if we give even the slightest credibility to the concept of a team being "owed a penalty becomes the triggering event - until you have the first penalty none of the two participating teams may be viewed as being "owed" a penalty.

At this point the various factors - fatigue, talent, mandatory,obstruction penalties, physical penalties, etc enter the picture.

The thing is, it isn't about one penalty, and it isn't just about the first penalty, it's about any time a team has been called for fewer penalties than the opposition. Even if you showed that those factors make a given team more prone to take the first penalty, I fail to see why they would suddenly be less likely to take the second penalty.

Also the idea of first penalty of the game doesn't really match the overtime results.

Overtime produces the most dramatic results, with the second penalty call of the extra period about three times as likely to be whistled on the team that received the first power play.

Look at the penalties for the October 28,2014 game Montréal at Calgary, Canadiens were playing their 3rd game in 4 nights, having played the Rangers in Montréal on the 25th, in Edmonton on the 27th and in Calgary on the 28th. Calgary had a home game on the 25th against Washington so they were sitting at home waiting to play a tired team that had traveled across two time zones. Canadiens received the last 2nd period penalty plus three late(second half) third period penalties. Was Calgary somehow owed these penalties or was it fatigue? Six possibly seven of the Montreal penalties were fatigue type penalties - kneeing could go either way.

http://www.hockey-reference.com/boxscores/201410280CGY.html

And I could find you a ton of games where a tired team took early penalties and then the non-tired team got the "even up" call. It's not a 1 or 2 percent that's being shown, it's a massive variance between likelihood to take a penalty when "owed" versus when not being owed.

And either way, this game actually has the *opposite* thing happening where Calgary was owed several penalties at the end of the game put didn't get any. Presumably your explanation is that Montreal was tired and defending a lead and thus more likely to be in a position to take penalties, however if that holds in any kind of statistically significant number of games it would actually strengthen the relationship in the rest of the games.


Anyone who has watched a decent amount of hockey has, at some point, seen a team take a few penalties in a row and mentally (or verbally to someone watching with you) "well we all know who's going to get the next penalty" and it feels like often we're right.

The numbers show that, on a very statistically significant level, that's not just confirmation bias that we feel we're right more often than not.

 

[GoldenBearHockey]

I don't understand how PPO or PPOA ranges show anything about which teams are more or less likely to take the first penalty in a game.



The thing is, it isn't about one penalty, and it isn't just about the first penalty, it's about any time a team has been called for fewer penalties than the opposition. Even if you showed that those factors make a given team more prone to take the first penalty, I fail to see why they would suddenly be less likely to take the second penalty.

Also the idea of first penalty of the game doesn't really match the overtime results.





And I could find you a ton of games where a tired team took early penalties and then the non-tired team got the "even up" call. It's not a 1 or 2 percent that's being shown, it's a massive variance between likelihood to take a penalty when "owed" versus when not being owed.

And either way, this game actually has the *opposite* thing happening where Calgary was owed several penalties at the end of the game put didn't get any. Presumably your explanation is that Montreal was tired and defending a lead and thus more likely to be in a position to take penalties, however if that holds in any kind of statistically significant number of games it would actually strengthen the relationship in the rest of the games.


Anyone who has watched a decent amount of hockey has, at some point, seen a team take a few penalties in a row and mentally (or verbally to someone watching with you) "well we all know who's going to get the next penalty" and it feels like often we're right.

The numbers show that, on a very statistically significant level, that's not just confirmation bias that we feel we're right more often than not.

Agreed to a point, it would be a wonderful thing to know WHAT those penalties are, meaning, are they borderline calls, are they solid penalty calls etc,

From officiating, I know that if I have two teams that are similar, and out to just play hockey, if one team isn't being aggressive but gets 3-4 calls in a row, I will be sensitive to a borderline call on the other team, meaning, yes it's a penalty, but 90% of the time, it's borderline let go, in this case, it's a call, is that game management, ABSOLUTELY. Nothing wrong with game management, it prevents much more dust ups than it causes.

 

[Canadiens1958]

Clarifications

I don't understand how PPO or PPOA ranges show anything about which teams are more or less likely to take the first penalty in a game.



The thing is, it isn't about one penalty, and it isn't just about the first penalty, it's about any time a team has been called for fewer penalties than the opposition. Even if you showed that those factors make a given team more prone to take the first penalty, I fail to see why they would suddenly be less likely to take the second penalty.

Also the idea of first penalty of the game doesn't really match the overtime results.





And I could find you a ton of games where a tired team took early penalties and then the non-tired team got the "even up" call. It's not a 1 or 2 percent that's being shown, it's a massive variance between likelihood to take a penalty when "owed" versus when not being owed.

And either way, this game actually has the *opposite* thing happening where Calgary was owed several penalties at the end of the game put didn't get any. Presumably your explanation is that Montreal was tired and defending a lead and thus more likely to be in a position to take penalties, however if that holds in any kind of statistically significant number of games it would actually strengthen the relationship in the rest of the games.


Anyone who has watched a decent amount of hockey has, at some point, seen a team take a few penalties in a row and mentally (or verbally to someone watching with you) "well we all know who's going to get the next penalty" and it feels like often we're right.

The numbers show that, on a very statistically significant level, that's not just confirmation bias that we feel we're right more often than not.

A team with 308 PPOA was more than likely to have the first penalty called against them than the team with 186 PPOA. In fact they are more than likely to have any penalty called against them.

Please do not presume anything about my explanations. The link clearly shows that Montreal never had the lead in the game until winning in the shootout.

Watching hockey since the mid fifties at the NHL level.

 

[hatterson]

A team with 308 PPOA was more than likely to have the first penalty called against them than the team with 186 PPOA. In fact they are more than likely to have any penalty called against them.

I still fail to see how this would affect the likelihood that, if they receive the first penalty, their opponent is more likely to receive the next one.

Please do not presume anything about my explanations. The link clearly shows that Montreal never had the lead in the game until winning in the shootout.

Watching hockey since the mid fifties at the NHL level.

Regardless of what the score was (I didn't pay too much attention to the box score because it's meaningless on a statistical level), the overall point I was making still stands.

And I was forced to try and presume something about your explanation because it had minimal substance to it. It was an anecdotal reference of a single game, asking questions that no one except the referee on that night could know, in a conversation about a statistical review of more than 10,000 games.

I don't know if the referee that night thought Montreal was owed a powerplay or not. Maybe Calgary played extremely clean in the third, maybe Montreal's calls were extremely blatant and the ref couldn't ignore them, maybe the Habs had been chirping him all night and he didn't want to give them a break. Either way, it's irrelevant unless you can show some kind of statistical trend to it.

Agreed to a point, it would be a wonderful thing to know WHAT those penalties are, meaning, are they borderline calls, are they solid penalty calls etc,

Completely agree. You could go into a lot more detail for something like this and try and separate out automatic calls (or at least ones with minimal wiggle room like too many men) from judgement calls. Or see if anything regarding significant penalties like match penalties or other majors affects this.

 

[Ralph Spoilsport]

What does the referee stand to gain from even-up calls? If he tries to appease one team by giving them an "owed" power play he draws the disapproval of the other, assuming that coach sees it for what it is. Doesn't make sense. The ref has to please both parties. If the call is weak he degrades his reputation and loses the respect of both coaches.

We all practice biased impartiality to some extant so I'm sure it exists in the NHL too, I'm just not sure how widespread it is. Most likely found where inexperienced refs cross paths with persuasive veteran coaches.

Otherwise, don't penalties have a way of evening themselves out naturally, without referee assistance? If Team A gets three in a row, why wouldn't Team B get the fourth? Are they supposed to have immunity by that point?

 

[hatterson]

What does the referee stand to gain from even-up calls? If he tries to appease one team by giving them an "owed" power play he draws the disapproval of the other, assuming that coach sees it for what it is. Doesn't make sense. The ref has to please both parties. If the call is weak he degrades his reputation and loses the respect of both coaches.

We all practice biased impartiality to some extant...

You basically just answered your own question. No matter how one-sided a game is, if the penalties end 7-1 in favor of one team, people are going to say that the reffing was one-sided.

Otherwise, don't penalties have a way of evening themselves out naturally, without referee assistance? If Team A gets three in a row, why wouldn't Team B get the fourth? Are they supposed to have immunity by that point?

If each team is equally likely to take a penalty, then the next expected penalty should be 50% to each team regardless of current gap or streak.

If one team is more likely to take a penalty overall, I'd expect the team with more penalties to be the team more likely to take the next one as opposed to the team that has less.

 

[Ralph Spoilsport]

People can say what they want but referees are not accountable to fans.

But they are accountable. Unless I'm mistaken, after every game the coaching staffs of the teams involved complete assessments of the officials' performance, which are submitted to the league and form the basis of their job evaluations. That's kind of important, no? Whatever the refs are doing, they are doing with the approval of the league and its teams. And I can't see a ref who persistently hands out even-up calls having a very long NHL career.

If one team is more likely to take a penalty overall, I'd expect the team with more penalties to be the team more likely to take the next one as opposed to the team that has less.

And I'd expect both teams to get their share, depending on their style of play.

The stats indicate an irregular pattern to the distribution of calls. The reason for that pattern is still up in the air. Where some would say that teams were "owed" a call, I'd just say they were probably due anyway. Either way, we are both projecting our own biases, but I'd give the refs the benefit of the doubt since there is a review process in place.

 

[Doctor No]

People can say what they want but referees are not accountable to fans.

In theory, I agree. In practice? Eh.

 

[SillyRabbit]

Because of this very well known fact, wouldn't every team benefit from playing a very physical, obstruction-heavy type of game?

Basically, a team should play "on the edge" and try to get away with whatever they can. When they get away with it, they will benefit from a competitive advantage. When they don't, they will get a penalty.

But who cares? Because even if the other team plays a much more disciplined game, the referees are just going to try to even up the penalties by the end of the game.

So the team gets the advantage of physically intimidating their opponents, shutting down potential opportunities with obstruction, pissing the other team off with excessive roughing and potential cheap-shots, and in the end we will still end up with the same amount of penalties.

Even if the penalties end up 5-4 in the other teams favor, the extra advantage gained by playing "on the edge" will more than make up for that.

Now if there was no bias towards evening up the calls, you might see the "on the edge" team finish the game with 1 PP, while the disciplined team got 9 PP's. That would discourage the heavy obstruction and "on the edge" play, because any advantage they gained would be offset by the disparity between the PP chances.

 

[GoldenBearHockey]

Because of this very well known fact, wouldn't every team benefit from playing a very physical, obstruction-heavy type of game?

Basically, a team should play "on the edge" and try to get away with whatever they can. When they get away with it, they will benefit from a competitive advantage. When they don't, they will get a penalty.

But who cares? Because even if the other team plays a much more disciplined game, the referees are just going to try to even up the penalties by the end of the game.

So the team gets the advantage of physically intimidating their opponents, shutting down potential opportunities with obstruction, pissing the other team off with excessive roughing and potential cheap-shots, and in the end we will still end up with the same amount of penalties.

Even if the penalties end up 5-4 in the other teams favor, the extra advantage gained by playing "on the edge" will more than make up for that.

Now if there was no bias towards evening up the calls, you might see the "on the edge" team finish the game with 1 PP, while the disciplined team got 9 PP's. That would discourage the heavy obstruction and "on the edge" play, because any advantage they gained would be offset by the disparity between the PP chances.

So then,

Why don't we?

Maybe is it because the players, coaches, and GM's realize this?

Maybe they realize there is no ref bias, and it's only created by fans?