Sorting the teams in the NHL standings

[Guest BCM]

In the NHL, a 13-12 team would be listed ahead of a 12-10 team. But in the NBA or MLB, the 12-10 team would be listed ahead. In the NHL, if a 15-10 team lost five games while a 14-11 team doesn't play (I know that's not likely - it's just an illustration), the formerly 15-10 team would still lead the 14-11 team. In the NBA or MLB, if that were to happen, the 15-10 (or now 15-15) team would fall behind the 14-11 team.

Why does the NHL do this differently? It can't be because they have ties (or now overtime losses), as a tie (or now OTL) has the net value of half a win and half a loss. Why not simply use games behind and winning percentage as is used in MLB and the NBA?

Does anyone have an explanation for this?

[Joey v3.4 45]

*Shrug* In the end they all play 82 games, so it doesn't matter come day zero.

I usually sort by point percentage, if I want to see more accurately who is doing overall better. Point Percentage is pretty accurate.

I don't know why the NHL does it differently than many other teams though.

[Buppy 1,720]

The NHL sorts mid-season rankings by points. Pretty simple really.

It's not like you get anything for mid-season rankings. So it doesn't really matter. Point% is certainly a better indicator of how well a team is doing, but there is that old saying about a bird in the hand. Points you actually have are better than points you could potentially get.

15-14-1 is better than 16-17-1, but there's always a chance the first team could lose all four games in hand and end up 15-18-1. So until they actually get the points they stay behind in the rankings. Like Joey said, we all end up playing 82 games, so the teams with games in hand will either make up the point difference and move ahead, or lose and stay behind.

[Echolalia 2,961]

NHL fans are better at math so we don't have to be spoon-fed win percentages to know what's up

[eva unit zero 271]

The NFL has only wins and losses.

The NBA has only wins and losses.

MLB has only wins and losses.

The NHL has wins, regulation losses, and overtime/shootout losses.

Unlike the other three, the NHL uses a 'point' system. This is due to the existence of ties for much of the history of the league. Games were worth two points; a team got both in a win and only one in a tie. Made sense.

As for your poll, the reason teams are sorted as such is this:

The NHL determines division championship and playoff seeding using the following criteria:

Points. The top eight teams in points in each conference are selected to play in the playoffs, seeded by points except for division champions. If there are ties in points which affect seeding, the following tiebreakers are used:

1) Number of games played. Two teams equal in points will be seeded based on the number of games played, from fewest to most.

2) Wins. A 42-40-0 team will be seeded above a 38-36-8 team.

3) Head-to-head record. Whichever team gained more points in head-to-head games is seeded ahead.

4) Goal differential.

Teams are sorted in the standings based on these rules. A team with a 0-0-11 record would be placed ahead of a team with a 5-0-0 record. At the end, it all washes out. If you want standings based on winning percentage, work them out yourself.

[Buppy 1,720]

...the following tiebreakers are used:

1) Number of games played. Two teams equal in points will be seeded based on the number of games played, from fewest to most.

...

Beside the point, but I've always thought it strange how the NHL has that as the first tiebreaker, when the only time a tiebreaker is really needed is at the end of the season when everyone will have played the same number of games...

[SouthernWingsFan 854]

I think this question can only be answered by Professor Frink.

[Theophany 110]

The NFL has only wins and losses.

Actually, the NFL has ties as well. Happened last year.

Edited December 27, 2010 by Theophany

[Guest BCM]

*Shrug* In the end they all play 82 games, so it doesn't matter come day zero.

My befuddlement doesn't have to do with the records at the end of the season, but during the course of the season - when inequities in numbers of games played is quite common.

I usually sort by point percentage, if I want to see more accurately who is doing overall better. Point Percentage is pretty accurate.

Point percentage is better than points earned, as using point percentage, a team at 15-14-1 will be listed ahead of a team at 16-17-1. Furthermore, point percentage causes you to fall back when losing, while points treats a loss the same as not playing.

[Guest BCM]

The NHL sorts mid-season rankings by points. Pretty simple really.

It's not like you get anything for mid-season rankings. So it doesn't really matter. Point% is certainly a better indicator of how well a team is doing, but there is that old saying about a bird in the hand. Points you actually have are better than points you could potentially get.

If it didn't really matter during the season, then why does virtually every major newspaper in the USA and Canada publish the NHL standings on a daily basis during the season - and use a method to sort the teams in which a loss is equivalent to not playing?

15-14-1 is better than 16-17-1, but there's always a chance the first team could lose all four games in hand and end up 15-18-1. So until they actually get the points they stay behind in the rankings. Like Joey said, we all end up playing 82 games, so the teams with games in hand will either make up the point difference and move ahead, or lose and stay behind.

Doesn't the old saying of the bird in the hand apply to the lost point opportunities as well? Just as the 16-17-1 team has 16 wins (or 32 points) which cannot be taken away, they also have 17 losses (or 34 games in which they failed to earn a point) which cannot be taken away.

What if we were comparing a 17-16-1 team to a 14-15-1 team? You could just as easily say that since the 14-15-1 team hasn't lost 16, there is no guarantee they won't win their next four games in hand.

The NFL has only wins and losses.

The NBA has only wins and losses.

MLB has only wins and losses.

The NHL has wins, regulation losses, and overtime/shootout losses.

Unlike the other three, the NHL uses a 'point' system. This is due to the existence of ties for much of the history of the league. Games were worth two points; a team got both in a win and only one in a tie. Made sense.

I already touched on this. A tie (or an OTL) has the net value of half a win and half a loss - so the reason cannot be because of the presence of ties.

As for your poll, the reason teams are sorted as such is this:

The NHL determines division championship and playoff seeding using the following criteria:

Points. The top eight teams in points in each conference are selected to play in the playoffs, seeded by points except for division champions. If there are ties in points which affect seeding, the following tiebreakers are used:

We know it's points. The question is why points when more viable options such as games behind or points percentage are available?

1) Number of games played. Two teams equal in points will be seeded based on the number of games played, from fewest to most.

2) Wins. A 42-40-0 team will be seeded above a 38-36-8 team.

3) Head-to-head record. Whichever team gained more points in head-to-head games is seeded ahead.

4) Goal differential.

Teams are sorted in the standings based on these rules. A team with a 0-0-11 record would be placed ahead of a team with a 5-0-0 record. At the end, it all washes out. If you want standings based on winning percentage, work them out yourself.

Would you contend that 10-70-2 is a better W-L-OTL record than 9-0-1?

Beside the point, but I've always thought it strange how the NHL has that as the first tiebreaker, when the only time a tiebreaker is really needed is at the end of the season when everyone will have played the same number of games...

Maybe the NHL has a contingency plan in place in the event the season ends with not all teams playing the same number of games.

Let's say the season ends. The Blues & Stars are fighting for the final playoff spot and the season ends prematurely. The Blues are 40-41-1 (81 pts) while the Stars are 40-39-0 (80 pts, with three games remaining). You're the commissioner. Who do you give that final playoff spot to? The 40-41-1 team or the 40-39-0 team?

[Buppy 1,720]

If it didn't really matter during the season, then why does virtually every major newspaper in the USA and Canada publish the NHL standings on a daily basis during the season - and use a method to sort the teams in which a loss is equivalent to not playing?

Doesn't the old saying of the bird in the hand apply to the lost point opportunities as well? Just as the 16-17-1 team has 16 wins (or 32 points) which cannot be taken away, they also have 17 losses (or 34 games in which they failed to earn a point) which cannot be taken away.

What if we were comparing a 17-16-1 team to a 14-15-1 team? You could just as easily say that since the 14-15-1 team hasn't lost 16, there is no guarantee they won't win their next four games in hand.

We know it's points. The question is why points when more viable options such as games behind or points percentage are available?

Would you contend that 10-70-2 is a better W-L-OTL record than 9-0-1?

Maybe the NHL has a contingency plan in place in the event the season ends with not all teams playing the same number of games.

Let's say the season ends. The Blues & Stars are fighting for the final playoff spot and the season ends prematurely. The Blues are 40-41-1 (81 pts) while the Stars are 40-39-0 (80 pts, with three games remaining). You're the commissioner. Who do you give that final playoff spot to? The 40-41-1 team or the 40-39-0 team?

I said it doesn't matter. Never said people didn't care. Sports sites print standings because people want to see them, and they print them using the NHL's standard;which is points.

Games in hand would in the analogy be the birds in the bush. They are already factored in by virtue of only a two point lead, instead of more. But you still can't count the points until they're actually earned. If it really makes you feel better, go ahead and sort teams by point %. Like I said it means nothing until the end of the season anyway, when the games even out.

I don't know for sure, but I would assume the NHL would have some contigency in the event that the season couldn't be completed. Probably something like throwing out any 'extra' games for teams that have played more than the team playing the fewest.

But here's a scenario to illustrate the value of points:

Team 1: 45-27-10 82 games 100 pts .610 point%

Team 2: 40-22-16 78 games 96 pts .615 point%

Team 2 is better, right? Not so fast. Because of the wins, team 2 actually needs 5 points to pass team 1. But Team 2 can not maintain that point%. You can't earn fractions of points.

4 points = .500 point%

5 points = .625 point%

So even though Team 2 was technically better, they still need to improve their play over those last 4 games.

[eva unit zero 271]

I already touched on this. A tie (or an OTL) has the net value of half a win and half a loss - so the reason cannot be because of the presence of ties.

We know it's points. The question is why points when more viable options such as games behind or points percentage are available?

Would you contend that 10-70-2 is a better W-L-OTL record than 9-0-1?

Maybe the NHL has a contingency plan in place in the event the season ends with not all teams playing the same number of games.

Let's say the season ends. The Blues & Stars are fighting for the final playoff spot and the season ends prematurely. The Blues are 40-41-1 (81 pts) while the Stars are 40-39-0 (80 pts, with three games remaining). You're the commissioner. Who do you give that final playoff spot to? The 40-41-1 team or the 40-39-0 team?

The commissioner wouldn't "decide" to give the playoff spot to one or the other. The rules are already in place for what happens if not all teams finish the season with the same number of games played.

Example: Detroit goes 49-26-7 while Chicago for some reason does not end up playing/finishing a game, and finishes with a 48-24-9 record. Both teams have the same number of points, and Chicago couldn't have had a better record than Detroit based on pure win/loss record if they had played that last game and won. But in this instance, Chicago gets the division crown because the first tiebreaker is, as I posted earlier, games played. Now throw in a makeup game, which the Hawks lose. Detroit wins the division on tiebreaker #2, which is games won.

Sure, a team with a winning record deserves to be ranked ahead of a team with a losing record. But the standings, as I said before, are based on the rules for playoff seeding. Which means that teams which have played more games are more likely to be ranked higher as the season progresses,

[Guest BCM]

Sure, a team with a winning record deserves to be ranked ahead of a team with a losing record. But the standings, as I said before, are based on the rules for playoff seeding. Which means that teams which have played more games are more likely to be ranked higher as the season progresses,

If you use games behind or points percentage, you will end up with precisely the same playoff seedings as you would if you use points. That being said, why not simply use games behind or points percentage during the season. One of the best examples of how points is flawed is that when it comes to the primary means of sorting the teams in the standings, a regulation loss is precisely the same as not playing.

Let's say the Wings are 15-5-0 (30 pts) and the Blackhawks are 15-5-1 (31 pts). The Blackhawks lose two games while the Red Wings don't play. The Wings are 15-5-0 (30 pts) and the Blackhawks are 15-7-1 (31 pts). Had the Blackhawks not even played those two games, they would be 15-5-1, which is the same amount of points (31) as 15-7-1.

Is there a particular reason the NHL doesn't want teams to fall back in the standings as a result of a loss?

[Wing-in-avs-town 123]

At 15-14-1, you are still 2pts down, but you have 4 games in hand. I'd take my chances with that!

[Guest BCM]

At 15-14-1, you are still 2pts down, but you have 4 games in hand. I'd take my chances with that!

The point is not to project what 15-14-1 will turn into after another four games, but to make a direct comparison to 16-17-1.

To take it to another extreme, let's say the Wings go 18-0-0 in games in which Zetterberg scores a goal. In games in which Zetterberg doesn't score a goal, the Wings go 19-17-5. Would you say they did better with Zetterberg scoring or with Zetterberg not scoring?

[eva unit zero 271]

If you use games behind or points percentage, you will end up with precisely the same playoff seedings as you would if you use points. That being said, why not simply use games behind or points percentage during the season. One of the best examples of how points is flawed is that when it comes to the primary means of sorting the teams in the standings, a regulation loss is precisely the same as not playing.

Let's say the Wings are 15-5-0 (30 pts) and the Blackhawks are 15-5-1 (31 pts). The Blackhawks lose two games while the Red Wings don't play. The Wings are 15-5-0 (30 pts) and the Blackhawks are 15-7-1 (31 pts). Had the Blackhawks not even played those two games, they would be 15-5-1, which is the same amount of points (31) as 15-7-1.

Is there a particular reason the NHL doesn't want teams to fall back in the standings as a result of a loss?

Take it up with the NHL Board of Governors.

The NHL playoff seeding is based on points, therefore the standings are based on points. That's pretty much it. Sure, you would get the same result from point percentage if all teams have played the same number of games, but that's not what the rule is.

As far as games behind, how exactly do you calculate that? With wins and losses, you're fine. A 5-4 team is two games behind a 7-2 team, and 1.5 games behind a 7-3 team. But when you add in the overtime loss, it throws it all off. So you would have to throw out the overtime loss and make all losses worth nothing - a straight W-L record. At which point you could calculate games behind, but how does that work for the conference? Seems like winning percentage would be better than games behind, especially given the conference format the NHL uses.

[Guest BCM]

Take it up with the NHL Board of Governors.

I'd guess if no one in Hockeytown knows, then perhaps the NHL Board of Governors knows.

The NHL playoff seeding is based on points, therefore the standings are based on points.

That does nothing to answer my question. Why does the NHL use points when there are more viable options available such as points percentage and games behind?

When using points percentage and games behind, a loss causes your team to fall back. But when using points, a loss has the same value as not playing. Why does the NHL want a loss to have the same value - when it comes to sorting teams in the standings - as not playing?

That's pretty much it. Sure, you would get the same result from point percentage if all teams have played the same number of games, but that's not what the rule is.

I'm not questioning what the rule is. I'm questioning why they have that rule.

As far as games behind, how exactly do you calculate that? With wins and losses, you're fine. A 5-4 team is two games behind a 7-2 team, and 1.5 games behind a 7-3 team. But when you add in the overtime loss, it throws it all off. So you would have to throw out the overtime loss and make all losses worth nothing - a straight W-L record. At which point you could calculate games behind, but how does that work for the conference? Seems like winning percentage would be better than games behind, especially given the conference format the NHL uses.

Since an OTL (or any game in which you earn just one point) has the net value of half a win and half a regulation loss, it would be just as easy to use games behind as it is in the NBA or MLB.

Example:

Detroit 22-9-4 (48 pts);

Take the four OTLs and make half wins and half losses.

Now you have 24-11-0 (also 48 pts) or 13 games above .500

Chicago 20-14-3 (43 pts)

Take the three OTLs and make them half wins and half losses

Now you have 21.5-15.5-0 (also 43 pts) or 6 games above .500

Detroit is 13 games above .500, while Chicago is 6 games above .500. Using the games behind system of the NBA or MLB, Chicago would be 3.5 games behind Detroit. Very simple.

[Buppy 1,720]

I'd guess if no one in Hockeytown knows, then perhaps the NHL Board of Governors knows.

That does nothing to answer my question. Why does the NHL use points when there are more viable options available such as points percentage and games behind?

When using points percentage and games behind, a loss causes your team to fall back. But when using points, a loss has the same value as not playing. Why does the NHL want a loss to have the same value - when it comes to sorting teams in the standings - as not playing?

I'm not questioning what the rule is. I'm questioning why they have that rule.

Since an OTL (or any game in which you earn just one point) has the net value of half a win and half a regulation loss, it would be just as easy to use games behind as it is in the NBA or MLB.

Example:

Detroit 22-9-4 (48 pts);

Take the four OTLs and make half wins and half losses.

Now you have 24-11-0 (also 48 pts) or 13 games above .500

Chicago 20-14-3 (43 pts)

Take the three OTLs and make them half wins and half losses

Now you have 21.5-15.5-0 (also 43 pts) or 6 games above .500

Detroit is 13 games above .500, while Chicago is 6 games above .500. Using the games behind system of the NBA or MLB, Chicago would be 3.5 games behind Detroit. Very simple.

You're taking this way too seriously.

The point system is pretty easy to understand. At the end of the season when everyone has played an equal number of games, the only time it makes any difference, it produces the exact same results as point% or games back.

If you really need to see how teams rank by point%, go to NHL.com and click on P%. There. Problem solved.

If you really want to ***** about the NHL point system, ***** about 3 point games/loser points. Whining about meaningless mid-season rankings is just pointless.

[dropkickshanahans 463]

Actually, the NFL has ties as well. Happened last year.

It did?

[Guest BCM]

You're taking this way too seriously.

The point system is pretty easy to understand. At the end of the season when everyone has played an equal number of games, the only time it makes any difference, it produces the exact same results as point% or games back.

If you really need to see how teams rank by point%, go to NHL.com and click on P%. There. Problem solved.

It is an issue when teams haven't played the same number of games. Let's say the Wings go 18-0-0 in games in which Zetterberg scores a goal. In games in which Zetterberg doesn't score a goal, the Wings go 19-17-5. Would you say they did better with Zetterberg scoring or with Zetterberg not scoring?

If you really want to ***** about the NHL point system, ***** about 3 point games/loser points. Whining about meaningless mid-season rankings is just pointless.

Giving a point to the team that loses in overtime is like giving a half a win to the baseball team that loses a game in extra innings.

[Buppy 1,720]

It is an issue when teams haven't played the same number of games. Let's say the Wings go 18-0-0 in games in which Zetterberg scores a goal. In games in which Zetterberg doesn't score a goal, the Wings go 19-17-5. Would you say they did better with Zetterberg scoring or with Zetterberg not scoring?

Giving a point to the team that loses in overtime is like giving a half a win to the baseball team that loses a game in extra innings.

It's not an issue because no one is saying anything is 'better' than anything else.

It says one thing and one thing only: Who has collected the most points so far.

Teams don't get trophies for being in 1st after 37 games, aren't eliminated from the playoffs because they were last after 59 games, players don't get bonus checks for being within 2.5 games of the division lead...

Mid-season rankings don't mean anything. If you don't like the way the standings are sorted on websites and in newspapers, no one is forcing you to pay any attention to them. I just gave you a link to a list sorted by point% if you're so interested in it. No one is trying to stop you from going by that.

[Guest BCM]

It's not an issue because no one is saying anything is 'better' than anything else.

It says one thing and one thing only: Who has collected the most points so far.

Teams don't get trophies for being in 1st after 37 games, aren't eliminated from the playoffs because they were last after 59 games, players don't get bonus checks for being within 2.5 games of the division lead...

Mid-season rankings don't mean anything. If you don't like the way the standings are sorted on websites and in newspapers, no one is forcing you to pay any attention to them. I just gave you a link to a list sorted by point% if you're so interested in it. No one is trying to stop you from going by that.

Let's say you're the head coach. Your job is to determine who to put on the power play. Your research tells you that when Zetterberg and Datsyuk are on the power play, the Wings are 18-0-0 (36 pts) and when Zetterberg and Datsyuk aren't on the power play, the Wings are 19-17-3 (41 pts). This is the only data you have. Do you go with Zetterberg and Datsyuk on the power play? Or not?

[Buppy 1,720]

Let's say you're the head coach. Your job is to determine who to put on the power play. Your research tells you that when Zetterberg and Datsyuk are on the power play, the Wings are 18-0-0 (36 pts) and when Zetterberg and Datsyuk aren't on the power play, the Wings are 19-17-3 (41 pts). This is the only data you have. Do you go with Zetterberg and Datsyuk on the power play? Or not?

You're not getting the point.

NHL standings are sorted by who has the most points. Nothing really to do with who is 'better'. Look back at my first post, I said 15-14-1 is a better record, it just doesn't matter.

[Guest BCM]

You're not getting the point.

NHL standings are sorted by who has the most points. Nothing really to do with who is 'better'. Look back at my first post, I said 15-14-1 is a better record, it just doesn't matter.

What if you need to make an assessment?

Let's say the Wings go 9-7-1 (19 pts) in December and 9-4-0 (18 pts) in January. You're assigned the task of reporting to the public if the Wings had a better W-L-OTL record in January than they did in December. Since you voted for 15-14-1 over 16-17-1, I'll take that to mean you'd say the Wings did better in January (9-4-0) than they did in December (9-7-1). Am I right?

[paulwoodsfan 52]

Because there is always a discrepancy in games played, the points column is not completely meaningful. I always look at how many games over .500 (in the win-loss column) a team is to figure our relative strength. For instance, both Pit and Det have 50 points but Det is 14 games over .500 (23-9) while Pit is 13 games over (24-11). Pit has played one more game. Even though they are tied atop the standings, Det has the league's best points-attained percentage, which projected over a full season would have them at No. 1. The NHL uses points actually attained rather than percentage of possible points attained, which would provide a more accurate picture of teams' relative strength. This was especially noticeable early in the season, when Det had three or four teams ahead of them in points. They had earned a higher percentage of possible points than the other teams but had played fewer games so were ranked lower even though their actual success rate was higher.