What If the NHL had a March Madness-style playoff committee?

What factors should be considered?

Regulation wins, home wins, regulation plus overtime wins, five-on-five goal differential, Natural Stat Trick's five-on-five expected goals, team save percentage, power play, penalty kill, quality of victory and overall record should all be considered by our mythical selection committee. 

The Colorado Avalanche lead the NHL in regulation wins (45), five-on-five goal differential (plus-84), penalty kill (84.2 percent) and overall record. They are also second in Natural Stat Trick's five-on-five expected goals stat (56.07 percent). 

Colorado looks like a clear top seed and is No. 1 in five of our eight metrics. The Carolina Hurricanes are No. 1 in NST's expected goals metric, Edmonton is No. 1 on the power play and Minnesota is No. 1 in team save percentage.

Some factors may be weighted greater than others — this isn't the BCS, after all —- so the committee doesn't need to use the final score for each team. 

Division winners

Carolina and Colorado have already clinched their respective divisions. For this exercise, assume the current division leaders (Tampa in the Atlantic, Edmonton in the Pacific)  who haven't clinched will do so.

In the metrics laid out, Tampa emerges as one of the top teams — second in three categories and top five in many others. Carolina has multiple top-five finishes. Edmonton has only a top-five finish (power play). Buffalo has multiple top-five finishes. Dallas has multiple top-five finishes. Pittsburgh, Boston and Washington also have a top-five finish in one of our eight metrics.

By and large, this creates a pretty clear idea of the one-seeds.

Regions

Similar to the NCAA's Frozen Four or like the NHL's divisions now, the committee will split the teams into four regions with four teams each, seeded one through four. Unlike the current format, there is no need to remain tied to the division. Out-of-division opponents can be seeded, helping the top seeds with more favorable travel arrangements without sacrificing opponent ease or difficulty. The better seed will host the worse seed.

The winners of Region One and Region Four meet in the semifinals in Bracket One, while the winners of Region Two and Region Three meet in the semifinals in Bracket Two.

BRACKET ONE

Region One

No. 1 Colorado vs. No. 4 Vegas Golden Knights

No. 2 Boston Bruins vs. No. 3 Pittsburgh Penguins

Region Four

No. 1 Buffalo Sabres vs. No. 4 Washington Capitals

No. 2 Carolina Hurricanes vs. No. 3 Columbus Blue Jackets

BRACKET TWO

Region Two

No. 1 Tampa Bay Lightning vs. No. 15 New York Islanders

No. 2 Montreal Canadiens vs. No. 3 Ottawa Senators

Region Three

No. 1 Dallas Stars vs. No. 4 Utah Mammoth

No. 2 Minnesota Wild vs. No 3. Edmonton Oilers

The NHL playoff committee seeds all 16 

1. Colorado Avalanche
2. Tampa Bay Lightning
3. Dallas Stars
4. Buffalo Sabres
5. Carolina Hurricanes
6. Minnesota Wild
7. Montreal Canadiens
8. Boston Bruins
9. Pittsburgh Penguins
10. Ottawa Senators
11. Edmonton Oilers
12. Columbus Blue Jackets
13. Washington Capitals
14. Utah Mammoth
15. New York Islanders
16. Vegas Golden Knights

Which real-life team gets snubbed?

There are two pretty large snubs — one in each conference. Anaheim is a near-lock to make the playoffs in real life despite a late-season swoon. In this model, the Ducks get placed head-to-head with several similar Eastern Conference teams. The New York Islanders, Columbus Blue Jackets and Washington Capitals all bring significantly better advanced metrics, particularly in the goaltending department. 

Meanwhile, Philadelphia, which has the final Metropolitan Division playoff spot in real life, takes a pounding using this method because of weak metrics in regulation wins (T-25th), home wins (T-24th), power play (32nd), penalty kill (20th) and team save percentage (25th). 

Carolina is an automatic bid, but the committee would likely drop it from a No. 1 seed because of comically weak goaltending. Carolina's team save percentage is bottom five in the NHL, while Colorado, Tampa, Dallas and Buffalo are in the top eight in the league. 

No metric is perfect. The committee would likely reward Buffalo for a few different reasons, including a stretch since Thanksgiving in which it has won 38 of 55 games. Additionally, the Sabres are being held down by NST's expected goals metric, which is a useful but imperfect measurement. 

The lesson from this exercise is that there's likely no silver bullet to fixing the NHL's playoff format, but this version would have the potential to be a lot of fun. 

It would also never get adopted because it isn't objective, which would ultimately lead to allegations of bias for or against certain markets.

Alas, we can only dream.