NHL Standings under a 3-point system

First, a recap.

Back in March 2012, I solved the problem with the NHL standings point system.

In August 2013, Grantland touches on it but even when talking about a 3-point system neglects the key difference that OT wins should be worth 3 points and OT losses should be worth nothing.

In January 2014, the same guy at Grantland writes an entire article about it, repeating the same stuff.

- - -

Now then. I thought it would be interesting to compare the current 2-1-0 system to my 3-2-1-0 system.

Here's the current standings for above .500 teams, as always sorted by points percentage.

Here are the standings with 3 points for a ROW, 2 points for a SOwin, 1 point for a SOloss, 0 points for a ROL.

Ducks: 113 points / 153 possible = .739
Penguins: 98 points / 144 possible = .681
Blues: 94 points / 141 possible = .667
Avs: 93 points / 144 possible = .646
Bruins: 92 points / 144 possible = .639
Blackhawks: 97 points / 153 possible = .634
Sharks: 90 points / 147 possible = .612
Kings: 85 points / 147 possible = .578
Lightning: 85 points / 150 possible = .567
Canadiens: 81 points / 147 possible = .551
Canucks: 78 points / 150 possible = .520
Wild: 79 points / 153 possible = .516
Coyotes: 69 points / 144 possible = .479

In terms of league standings, the Penguins move up from 4 to 2, the Blackhawks fall from 2 to 6.

- - -

This system increases the value of a regulation or overtime win. It decreases the value of a shootout win, shootout loss, and overtime loss.

To repeat, it will reward teams that are winning outright in regulation and overtime play.
It will punish teams that are scraping by with shootout wins, shootout losses or overtime losses.

I think these are good, fair things.

I didn't expect to actually make much of a difference in the standings order. I think the real benefit comes with the standardization that every game is worth the same amount of points. And that there is little incentive to play for overtime or play for a shootout, resulting in better hockey. 

Updating NHL Playoff Picture

We're over halfway though the NHL season. With the new divisional playoff structure, I wondered where (if any) our playoff races will be.

Here's the current playoff picture:

Great. Here it is with league standings ranking next to each team.

This is my concern, dude.

Under the old system, the Avs would be the 4th best team in the West and draw the 5th best team in the West, the Sharks. But if they passed Chicago and became the 3rd best team in the West, they would get the 6th best, the Kings.

Currently, they are 3rd in the Central and draw the Blackhawks. If they pass Chicago and become the 2nd best team in the Central, they draw...the Blackhawks.

The bright side is that they have a comfortable lead on the Wild and Minnesota is still projected to make the playoffs anyways. So it would seem Colorado will be playing Chicago or St. Louis, and they've had better success against Chicago. Bring on the Blackhawks.

- - -

In the West, 7 teams have a playoff odds of over 87%. Minnesota, Phoenix and Dallas are fighting for the last spot.

In the East, only 4 teams have odds above 60%. 9 teams have a chance of claiming the 4 other spots.

Comparing Hockey Standings

I'm slightly obsessed with the math of the NHL.

Here are the current standings based on points percentage:

While fair, it still compares an uneven number of games and uses percentages/projection to balance it.

Here's what the standings look like for the top 11 teams, through 45 games played:

Anaheim 69
St. Louis 69
Chicago 67
Pittsburgh 65
San Jose 62
Colorado 61
Boston 60
LA 59
Tampa Bay 58
Montreal 55
Vancouver 55

So there you go. It's basically the same thing. These numbers give you a true way of saying, the Penguins are 4 points behind the Blues. But they also ignore the actual results since the cutoff date. For example, Anaheim went 3-0 since the 45 game mark, giving them the best points percentage in the league. At any rate, this only confirms the validity of the points percentage standings and the ridiculousness of the current standings that have shown the Blackhawks over the Blues all year.

I already fixed this (nhl standings points)

In today's Grantland How To Fix the NHL column there is this paragraph about the standings points:

Many fans have suggested making all games worth three points. Teams would get three points for a regulation win, two for an OT/shootout win, one for an OT/shootout loss, and nothing for a regulation loss. You'd have all games worth the same amount of points, and you'd encourage teams to go for the win in regulation.

Of course this is better than what we have, but it's still not perfect. As I wrote here over a year ago, the best solution is to have three points for a regulation or OT win, two for a shootout win and one for a shootout loss.

The Blackhawks' Streak

I decided I wasn't going to comment on the streak until it was over.

- - -

Prior to last Friday, the Chicago Blackhawks had played 24 games.

14 of those games were decided in regulation. Chicago won all 14.
4 of those games were decided in the 5 minute overtime period. Chicago won all 4.
6 of those games were decided in a shootout. Chicago won 3 of those 6.

So it was a 24 game streak in which they earned a point in every game. Which is very impressive.
But it's even more impressive when you realize they didn't even lose a game in overtime, just shootouts.

- - -

To review, here are two ways of awarding standings points.

Current NHL 2-point plan
Reg Win/OT Win/SO Win = 2 points
OT Loss/SO Loss = 1 point
Reg Loss = 0 points

Hoagie Central's NHL 3-point plan
Reg Win/OT Win = 3 points
SO Win = 2 points
SO Loss = 1 point
Reg Loss/OT Loss = 0 points



I'm going to look at some top teams and see how they fair under both plans. 

- - - 

Currently, Montreal has earned 38 points in 26 games. That's a points percentage of .731
Under the 3-point plan they would have earned 52 points in 26 games. That's a points percentage of .666

Currently, Anaheim has earned 39 points in 24 games. That's a points percentage of .813

Under the 3-point plan they would have earned 53 points in 24 games. That's a points percentage of .736

Clearly, the NHL's current plan inflates records so that almost every team seems like they're above .500. 

What about the Blackhawks through 24?

They had earned 45 in 24 games. For a crazy percentage of .938
Under the 3-point plan they would have earned 63 points in 24 games. That's a points percentage of .875. 

That's crazy good and doesn't have quite the same drop off as other top teams when you take away loser points. 

NHL Realignment: Take Two

In October 2011, I solved the Winnipeg problem with a plan to keep the six divisions. The NHL did not choose this plan.

Then in December 2011, the NHL unveiled their "four-conference" plan. I wasn't so much concerned with the team placement as I was with the playoff plan, or lack of one.

For the Avalanche it was keeping their division, but subtract Minnesota and add San Jose, LA, Anaheim and Phoenix. Not a big gain or loss there.

And then this week the NHL announces a new "four-division" plan and playoff plan. The playoffs would start as divisional playoffs, but with wild-cards per conference. That sentence is confusing. Let's try again. Each division gets the top 3 teams. And then there are two wild-card spots for the West. And two for the east. Here's what it looks like:

So the wild cards fix the imbalance problem with some divisions having 7 teams, and some 8, right?

Nope.

The West has 14 teams, the East 16.

The West will have 8 teams in, 6 out.
The East will have 8 teams in, 8 out.

In other words, in the West once you takeaway the top three spots in the two divisions, you have 8 teams fighting for 2 wild-card spots. But in the East, you have 10 teams fighting for 2 spots.

If the West and East both had 15, it wouldn't matter if each conference had uneven divisions. How does nobody in the NHL league office realize this? I guess they probably did but think that timezones are more important than fairness. They have 16 teams in the Eastern timezone so they put 16 teams in the Eastern conference. It prevents Detroit or Columbus from having to play at inconvenient times, but is that really worth it?

And because of the wild-cards, you're not going to have true divisional playoffs, which I thought was one of the selling points. So the playoff plan is kind of a disaster.

As for the Avalanche, I love, love love the new division. I never gave a shit about Vancouver/Calgary/Edmonton. But gaining rivalries with St. Louis and Chicago I think if fantastic. And I can get more excited about playing Minnesota, Dallas and Nashville than Phoenix and Anaheim.

And this way, Colorado doesn't have to get out their passports as much.

Scoopy Puck Goals

As far as I'm concerned, the original is Michigan's Mike Legg in 1996:


Crosby did it in 2003:


Last year Mikael Granlund did it:


And then yesterday, Miks Indrasis did it too:


I'm sure that there are others. God bless the internet.

Fixing the NHL Standings with a 3-Point System

Current NHL Standing System
Reg Win/OT Win/SO Win = 2 points
OT Loss/SO Loss = 1 point
Reg Loss = 0 points

The first flaw in this system is that not all games have equal points. If St. Louis beats Nashville in regulation, two points are awarded to St. Louis. If St. Louis beats Nashville in overtime or the shootout, two points are still awarded to St. Louis, but Nashville has earned a point as well. This just became a three point game. It doesn't really make sense that some games are more valuable than others.

The second problem is even bigger. For games tied late in the 3rd, both teams have an incentive to play cautiously. Just getting to overtime, assures both teams half credit. With the good possibility of earning full credit.

This system doesn't make sense.

Let's look at two options.

Proposal A: Conventional Wisdom's 3-point plan
Reg Win = 3 points
OT Win/SO Win = 2 points
OT Loss/SO Loss = 1 point
Reg Loss = 0 points


This solves the first problem easily. All games are worth 3 points.

This also solves the second problems. Teams have an incentive to win in regulation (thought teams still could play cautiously to make it to overtime, at least insure earning a single point).

In a recent ESPN post about this, someone criticied this plan as such:

"You could have a tie game with five minutes to go in regulation and suddenly a team pulls their goalie because they desperately need those three points to move up the standings late in the season," the GM said. "And then they get an empty-net goal scored on them and the other team gets handed three bonus points instead, which could have a huge impact on whatever club was chasing that team. Get it?"

Yes that would be a shame if a team desperate for three points pulled their goal in a tied game. But if you look at the numbers, I doubt it would happen unless it was the last game of the year. Because the chance to get two points in a overtime win, is much better than the risk of getting zero in a reg loss.

This plan solves a lot of problems and makes a lot of sense. But here's a question...should winning in the five minute overtime be less valuable than winning in regulation? Imagine if winning in extra innings or overtime in the NBA or NFL was worth less than a regulation win. Also imagine, that just getting to extra innings helped you in the standings.

It seems contrary to just about everything else we know about sports.

So let's take a step back. Why don't we just count wins?

The answer, clearly, is that now with the addition of the shootout, it would be too arbitrary to award a full win based on winning a three-shot shootout.

Right? Isn't that the answer? So why should we penalizing overtime winners and rewarding overtime losers in hockey, but not in any other sport? Even currently, if you win in overtime, it's a win.

So let's take a look at an alternate idea, one that I haven't seen talked about at all.

Proposal B: Hoagie Central's 3-point plan
Reg Win/OT Win = 3 points
SO Win = 2 points
SO Loss = 1 point
Reg Loss/OT Loss = 0 points

Let's look at the problems with the current system.

Yep, all games are worth the same amount of points.

Now what about teams tied in the 3rd playing cautiously? I think this would work even better than Proposal A. There is no incentive to get to overtime. If you make it to overtime, you still haven't earned a point yet.

So teams would play hard in the 3rd. And I think it would make overtime even better as well. For a team that is excellent in the shootout, like Colorado (7-1) or Detroit (7-2), currently they have an incentive to play cautiously in overtime. But with an overtime win being worth 3 and a shootout win being worth 2, I think all teams would play to win in OT.

And it makes sense, according to the reason why we don't just count wins in the first place. Because the shootout is sort of arbitrary and winning the shootout isn't really the same as winning the game in regulation or overtime. So the fans still get a winner or loser. And if your team loses in the shootout, at least you still got a point out of it.

I really don't see a problem with this idea.

nhl math

First off, the way ESPN and even the NHL show standings is flawed. Here's ESPN, with teams sorted by total points.

And here's how it should be, via nhltruestandings.com, sorted by points earned divided by points possible.

All season, San Jose has been lower in the ESPN standings that they really should be, and Colorado has been higher. And you can have silly things like this. Now that we've got that straight, I want to look at the true standings to make sense of the playoff race.

Looking at the race in the West for the 8th spot, we see:

Dallas .556
LA .548
Calgary .540
Colorado .540

Which is all fine and dandy. But hard to draw conclusions from. Instead, I'm more interested in the right hand column, the pace.


Dallas 91.11
LA 89.94
Calgary 88.61
Colorado 88.51

For an idea of how these change, here's how these paces looked at the all-star break a month ago:

LA 95.12
Minnesota 92.04
Dallas 88.83
Colorado 86.82

Calgary 85.28

Colorado and Calgary have upped their pace over the last month, while LA, Minnesota, and Dallas have all gone down. So even with 63 games played, these numbers are flexible.

Heck, Phoenix went from 52 points at 50 games, to 73 points at 62 games. That upped their pace to 96.55 from 85.28 in 12 games.

- - -

But I want to focus on the 8th spot. So here are the pressing numbers.


Dallas 91.11
LA 89.94
Calgary 88.61
Colorado 88.51

The real question is what will it take to get Colorado's pace up to a 91 or 92.

The Avs have played 63, 19 to go.

They have 68 points right now.

Since their pace right now is 88.51, and you can't earn half a point, I'll figure both 88 and 89 as Current.
And I'll figure 92 as what it takes to make the playoffs.

Current Pace A: 88 = 20 new points in 19 games
You could expect the Avs to go 10-9-0 or 9-8-2 at this rate.

Current Pace B: 89 = 21 new points in 19 games
You could also expect the Avs to go 10-8-1 or 9-7-3 at this rate.

Playoff Pace: 92 = 24 new points in 19 games
To make the playoffs, they would need to go 12-7-0 or 11-6-2.

In other words, to make the playoffs, Colorado would need to win two games that they would normally lose.

Which is a tricky concept. Let's say they win tonight against Anaheim. It's hard to say that win is one of those two. They already have a winning record, so it's not enough to win games. They need to win two extra.

Or, another way of looking at is, to get 24 new points in 19 games, would be 63.2% of the remaining points possible. Over 63 games, they've earned 54% of the points possible. Which makes it seem like quite a challenge, to suddenly play like a much better team down the stretch. Certainly it's more daunting than the ESPN standings seem, just two points behind Dallas, just one game.

i could've been a contender

So one thing that I thought about with the Chiefs over the last couple years, particularly in 2005 when they missed the playoffs at 10-6, and in 2006 and 2010 when they lost in the wild-card round...is that it's not really enough to just make the playoffs--you need a Super Bowl contender.

Take the Raiders this year. They're only one game out of the division. But even if they won the division and hosted a playoff game...they're not winning the Super Bowl this year. So what's the point of getting all excited about making in to the playoffs? But then you have teams like the Packers who win the Super Bowl as the 6th seed and the Cardinals that win the World Series as a wild-card. So you have evidence that all you need to do is make into the playoffs and anything can happen.

So which is it?

- - -

What made me think of this? The Avs are currently 12th in the conference. They've played 30 games. If they had won 3 more, they'd be 8th in the conference and in the playoffs. So I've been eyeing that line and hoping they could win enough to get to that 8th spot. But should I be? Is getting in the playoffs with a mediocre team enough? Or to win a championship, do you have to have one of the best teams?

- - -

Going back for the last ten years, here are the regular season records for the Cup winners:

2011: Bruins (7th best record in NHL)
2010: Blackhawks (3rd)
2009: Penguins (8th)
2008: Red Wings (1st)
2007: Ducks (4th)
2006: Hurricanes (4th)
2004: Lightning (2nd)
2003: Devils (4th)
2002: Red Wings (1st)
2001: Avalanche (1st)

I think this is pretty revealing. For seven seasons, the Cup winner had a top-4 record in the league. Over the last three years, it's expanded, but not by much. The Bruins were a 3-seed and the Penguins were a 4-seed.

So in modern hockey, in a league where SIXTEEN teams get in the playoffs, you still have to be a top team to win it all. I suppose the 7-game series nature helps that quite a bit.

So what about other sports?

NFL 

2010: Packers (8th best) [They were a 6-seed but 10-6, while the 4-seed Seahawks were 7-9.]
2009: Saints (2nd)
2008: Steelers (5th)
2007: Giants (8th)
2006: Colts (4th)
2005: Steelers (9th)
2004: Patriots (2nd)
2003: Patriots (1st)
2002: Bucs (3rd)
2001: Patriots (7th)

Considering only 12 teams make it in, the fact that 4 of the last 10 NFL Champions were 7th best or lower does suggest that the NFL playoffs are more wide open than the NHL playoffs. Though it should be noted when we've had 6-seed winners, they were still in the top 9 in the league.

MLB

2011: Cardinals (8th best)
2010: Giants (5th)
2009: Yankees (1st)
2008: Phillies (5th)
2007: Red Sox (2nd)
2006: Cardinals (13th) [Won a 6-team division with a 83-78 record. Only played 161 games due to a rainout that was never made up because it didn't matter.]
2005: White Sox (2nd)
2004: Red Sox (3rd)
2003: Marlins (7th)
2002: Angels (4th)

There's just as many winners from outside the top-4 teams as from the top-4. Considering only 8 teams get in, I'd say it's reasonable for a fan to just hope their team makes the postseason.


NBA

2011: Mavericks (5th)
2010: Lakers (2nd)
2009: Lakers (2nd)
2008: Celtics (1st)
2007: Spurs (5th)
2006: Heat (4th)
2005: Spurs (3rd)
2004: Pistons (6th)
2003: Spurs (1st)
2002: Lakers (3rd)

Much like the NHL, it's not enough to sneak into the playoffs. You need to be a top-5 team in the league during the regular season. So come the 2012 NBA playoffs, we should look at the top 5 teams and be able to cross off the other contenders.

Of note: The Cavs with and without LeBron. I didn't realize how good their regular season records were since they never won a title.

2009: 1st in the NBA
2010: 1st in the NBA
2011: 29th in the NBA

Also of note: From 1980-2010, there were 31 NBA Champions. They were distributed among 8 franchises: Lakers, Celtics, 76ers, Pistons, Bulls, Rockets, Spurs, Heat. The 2011 Mavericks make it 9 franchises out of 32 titles.

In other words, of all the NBA fans born since 1976, only one-third of them has seen their team win a championship. 

- - -

So to get back to the original question. The Avs currently have the 25th best record in the NHL. Right now the 8th seed in the West is the Sharks with the 14th best record. I guess I should stop hoping they sneak up to the 14th best record, and hope they rebuild enough to be a top-5 team.

NHL realignment: Playoff Problems

30 teams.

Currently in 6 even divisions. With wild card spots.

Next year in 4 uneven conferences. With possibly no wild-card spots?

The plan is for the top 4 teams in each conference to make the playoffs.
So half the teams will have a 50% chance of making the playoffs.
And the other half will have a 57% chance.

Okay, so I understand that it sucks to have teams like Columbus and Detroit playing a bunch of games on the West coast throughout the regular season. I don't have a beef with the schedule, though it does take rivalries like Avs-Red Wings down to 2 games a year (ie, Detroit plays only once in Denver.) Applies to Blackhawks-Canucks, Bruins-Flyers, etc.

But what is unacceptable is having unequal chances to make the playoffs. How to fix this?

Plans

Plan 0: The old system
Plan A: As proposed, top four from each conference
Plan B: Top three from each conference, with four wild cards - divisional playoffs
Plan C: Put a 7-team conference and an 8-team conference on one side of the bracket, top two from each conference, four wild-cards
Plan D: Contract two teams, get down to 7 teams in each conference and do the top four just fine.

Personally, I like wild-cards as it keeps interest in playoff races. Anyways, let's compare plans 0-C using last years records. I know that Plan D would work, but contracting two teams is it's own ball of wax.

Plan 0

Notes: The Rangers had the 18th best record but made the playoffs over Dallas and Calgary because the West was stronger this year.

Plan A

Conference A
1 Vancouver
4 Phoenix

2 San Jose
3 Anaheim

Conference B
1 Detroit
4 Dallas

2 Nashville
3 Chicago

Conference C
1 Boston
4 Buffalo

2 Tampa
3 Montreal

Conference D
1 Washington
4 NY Rangers

2 Philadelphia
3 Pittsburgh

Notes: Changes from Plan 0--LA with 98 points misses the playoffs, Dallas with 95 points makes them. All divisional playoffs for the first two rounds, every year.

Plan B

Conference A
1 Vancouver
4 Dallas (wild-card: Vancouver was #1 division winner, so they get Dallas the #4 wild-card)

2 San Jose
3 Anaheim

Conference B
1 Detroit
4 Los Angeles (wild-card: Detroit was #3 division winner, so they get LA the #2 wild-card)

2 Nashville
3 Chicago

Conference C
1 Boston
4 Phoenix (wild-card: Boston was #4 division winner, so they get Phoenix the #1 wild-card)

2 Tampa
3 Montreal

Conference D
1 Washington
4 Buffalo (wild-card: Washington was #2 division winner, so they get Buffalo the #3 wild-card)

2 Philadelphia
3 Pittsburgh

Notes: Top 16 teams in NHL made playoffs, even better than Plan 0. While fair, this creates playoffs that aren't close in travel.

Plan C

This is a pain in the ass only because they put the two 7-team conferences on the east coast. If Columbus was in Conference D (green on the map) it would be pretty simple.

I'll do it both ways to see what I mean.

Plan C.1 

Conference A +  D Bracket

1 Vancouver
8 Los Angeles

2 Washington
7 Phoenix

3 Philadelphia
6 Anaheim

4 Pittsburgh
5 San Jose

Conference B + C Bracket

1 Detroit
8 Dallas

2 Boston
7 Buffalo

3 Tampa
6 Montreal

4 Nashville
5 Chicago

As you can see the B/C bracket is easy. The A/D bracket sucks, east-west travel all over.
But if we swap Columbus to conference D...

Plan C.2

Conference A + B Bracket (West)

1 Vancouver
8 Chicago

2 San Jose
7 LA

3 Detroit
6 Phoenix

4 Anaheim
5 Nashville

Conference C + D Bracket

1 Washington

8 NY Rangers

2 Philly

7 Buffalo

3 Pitt

6 Montreal

4 Boston

5 Tampa

Would you look at that, it's virtually identical to Plan 0. We still don't get the top 16 teams in the league, but the travel is fine. It's the easy way to fix the regular season travel without affecting the playoffs. 

Conclusion

Plan 0: Has worked well for the playoffs. But it's going away.

Plan A: Unfair percentage wise. Negatively impacted 1 team in 2011. Could be fixed with a tweak that says a 5th place team in an 8-team C could take the spot of a 4th place team in a 7-team C. In this example, LA would have taken the Rangers spot to play Washington. But then that makes Washington one of the only teams travelling far and they had the 2nd best record.

Plan B: I prefer it to Plan A. I think the wild-card races create more interesting games in the last two months of the regular season, which is always important. And I don't mind forcing teams to travel in the playoffs, if it's been helped in the regular season. At least we get the top 16 teams in the NHL.

Plan C:
C.1 sucks.
C.2 is pretty darn good. Requires flipping a team like Columbus (though they are in the Eastern time zone and are relatively new, so I don't see the big deal) to an Eastern division. Fixes regular season travel and keeps the playoffs as they have been. Edit: It would also work to move Phoenix to Quebec and put them in Conference C.

Plan D:
Would be sweet. But isn't on the table right now.

Post Script

Plan A vs Plan 0 in modern history...

How unfair is Plan A? Since this is the proposed plan, I think it warrants a larger investigation.

Plan A in 2010

Colorado, with the 12th best record gets bumped for St. Louis at 15. Philly and Montreal still get in at 18 and 19.

Plan A in 2009

Carolina and the Rangers get bumped with the 11th and 12th best records for Florida and Buffalo at 14 and 18. Interestingly, Carolina and Rangers were teams 5 and 6 in a 7-team conference. So this isn't a product of 7 teams vs 8, this is just what happens because conferences are never perfectly balanced.

Plan A in 2008

Another strong year for 7-team conference D. Washington at #12 gets bumped for #20 Chicago in an 8-team conference.

Plan A in 2007

Here the Thrashers with the 12th best record get bumped for Toronto at 18. This was the case of a 5th team in an 8-team conf vs a 4th team.

Verdict on Plan A


In the last 5 seasons, 6 teams that made the playoffs would not have under the proposed new rules. All 6 would have been replaced with teams with worse records, as would always be the case.

Conferences with teams negatively affected
D: 3 times
A: 2 times
B: 1 time

Conferences with teams positively affected
C: 3 times
B: 3 times

In case you forgot, A & B have 8 teams. C & D have 7 teams. At least for the last five years, it's more about who's in your conference than how many teams. If you're in conference B or C, you should be pretty happy that 4 of your teams are getting in the playoffs. D or A, not so much.

Every year the 12th best team in the NHL doesn't make the playoffs. But it's plausible for Bettman to argue that the increased rivalries and decreased travel is worth that tradeoff.

solving nhl realignment

For the 2012-2013 NHL season something has to give. Because the Atlanta Thrashers moved to Winnipeg and became the new Jets, we've got a geography problem.

We need to get Winnipeg to the Western Conference, which means we need to get a different team to the  East.

Move #1: Nashville Predators goes to the East, Southeast division. 

Solves the vacant SE division spot, creates a border rivalry (Tennessee/Carolina) and Tennessee feels more like a southeast team then a midwest team anyways. (Think SEC football not Big Ten)

Now we could just move Winnipeg to the Central, taking the Predators spot and call it a day. But...

Move #2: Minnesota Wild goes to the Central division. 

In the 80's the North Stars played in the same division as Chicago and Detroit, and just feels more like a midwest division. Minneapolis and Chicago and Detroit are already football, baseball rivals, etc.

Move #3: Winnipeg Jets goes to the Northwest division. 

Now you've got the Jets playing in a division with the three other Western Canadian teams.

Simple, elegant.

- - -

The only team this makes for long travel is...the Colorado Avalanche. There's not a great solution for this. But if you wanted to do one it would be: swap the Sharks and Avalanche.

Reasons for: You could rename the Pacific division, the Southwest. Which fits better for Dallas, Phoenix, Colorado.

Reasons against: It makes sense to keep the California teams together. Colorado is a legacy Canadian team so it makes sense for them to be in the division with 4 Canadian teams.

OR

If you wanted to keep thinking...why not put both Minnesota and Winnipeg in the Northwest? Minnesota feels more Canadian than Colorado. Which means Colorado would slide into the Central and get divisional games against Chicago, St. Louis and Detroit! To me, Denver feels more like a midwestern city--they already play Kansas City in football. I think adding rivalries with the Red Wings and Blackhawks would be awesome.