20

Projections, Projections and Projections

Projections are fun. Anyone can make a prediction at the start of the season on who they think will make the playoffs and who won’t. For years there have been some obvious projections, like the New England Patriots making the playoffs in a brutal AFC east, or the Pittsburgh Penguins making the playoffs in the NHL. There are usually top teams who most expect to make it and of course there are always surprises, both positive and negative, in who makes the postseason or doesn’t.

Making predictions based on a gut feel is how most people do it. You research the off-season moves, look at the roster, and then take a stab at which 16 teams will make the NHL playoffs and which 15 won’t. Lately, more people are digging deeper and coming up with a formula and basing it on previous results, health and schedule.

Dom Luszczyszyn from the Athletic has created his own model. He’s clearly put a lot of work into it. He explains how it works during the season.

Advertisement - Continue Reading Below

The projections are based on 50,000 simulations of the remainder of the season, which factors in each team’s projected strength, current health, and their strength of schedule (which includes opponent strength, venue and rest). You can read more about the model here.

It is important to note this isn’t his opinion. He uses data only. He sent out this tweet yesterday and it caught my eye.

I glanced at it and was surprised by Vegas having a 21% higher chance of making the playoff than Edmonton, and a 26% better chance than Arizona, despite being seven and six points behind those teams. It sent me down the usual online rabbit hole, that I seemingly can’t avoid when it comes to stats and numbers, and I ended up looking at every team and their projections.

Advertisement - Continue Reading Below

ATLANTIC DIVISION…

Team CURRENT RECORD PROJECTED REST OF SEASON FINAL PROJECTION
Boston 17-3-5 35-17-5 52-20-10
Florida 12-8-5 27-23-7 39-31-12
Toronto 12-10-4 33-17-5 45-27-9 (missed a game in their projection)
Montreal 11-8-5 28-23-7 39-31-12
Tampa Bay 12-8-2 38-17-5 50-25-7
Buffalo 11-10-4 23-26-8 34-36-12
Ottawa 11-13-1 22-27-8 33-40-9
Detroit 7-17-3 20-27-8 27-44-11

Tampa Bay has been really good for years, and they were 45-9-3 from November 29th to the end of the season last year, but this is a different team. I still understand why they are ranked this high, as most of us expect them to be better than they have been. Last season the Maple Leafs had a better start and they finished 28-20-8 down the stretch. A coaching change makes it very difficult to predict what a team will do.

His model has Boston, Tampa Bay and Toronto making the playoffs.

METRO DIVISION…

TEAM CURRENT RECORD PROJECTED REST OF SEASON FINAL PROJECTION
Washington 17-4-5 31-20-5 48-24-9
New York Islanders 16-5-2 29-23-7 45-28-9
Pittsburgh 14-7-4 34-18-5 48-25-9
Carolina 15-9-1 31-20-6 46-29-7
Philadelphia 13-7-5 29-21-7 42-28-12
New York Rangers 12-9-2 27-25-8 39-34-10 (one extra game in their projection)
Columbus 10-10-4 29-22-7 39-32-11
New Jersey Devils 8-11-4 26-25-8 34-36-12

The model has the Penguins really heating up and the Islanders winning just under half of their remaining games. The Islanders have been the best team in the Metro division (in the regular season) from November 29th, 2018, up to today. Last year they went 36-18-5 and have continued that pace, going 16-5-2. They were 52-23-7 in their last 82 games, but this model has them going 29-23-7. It seems odd to me.

The Capitals, Islanders, Penguins, Hurricanes and Flyers will be in the playoffs. The Flyers will finish eighth with 92 points.

Advertisement - Continue Reading Below

CENTRAL DIVISION…

TEAM CURRENT RECORD PROJECTED REST OF SEASON FINAL PROJECTION
St. Louis 15-5-6 30-20-6 45-25-12
Dallas 15-9-2 31-19-6 46-28-8
Winnipeg 15-9-1 27-23-7 42-32-8
Colorado 14-8-2 32-20-6 46-28-8
Nashville 11-9-4 31-20-6 42-29-10 (missing one game in their projection)
Chicago 10-9-5 28-23-7 38-32-12
Minnesota 10-11-4 29-21-7 39-32-11

Based on these projections the Blues win the division with 102 points followed closely by Colorado and Dallas with 100. The Avs have played quite well despite major injuries and I can see why they are projected to win the most games from here on in the West. Speaking of the Avs, last year they limped home from November 29th, winning only 23 of their final 57 games (23-24-10), but they still made the playoffs and then absolutely dominated the Calgary Flames in round one. They are a great example of how a team can ride a good start all the way to the post season.

However, Nashville winning 31 (possibly 32 when the projection is fixed) seems high to me. I recognize past success has something to do with it, but their goaltending has been brutal. Pekka Rinne has a .889sv% in 14 starts and Juuse Saros has a .891 in 10 starts. That would be a major turnaround for the Preds to win 31 (or 32 when projection is fixed).

The model has the Blues, Stars, Avalanche, Predators and Jets making the playoffs, with Winnipeg getting the second wildcard at 92 points.

PACIFIC DIVISION

TEAM CURRENT RECORD PROJECTED REST OF SEASON FINAL PROJECTION
Edmonton 16-8-3 25-23-7 41-31-10
Arizona 15-8-3 25-23-7 40-31-10 (missing one game in their projection)
Vancouver 12-10-4 28-22-7 40-32-11 (one extra game in their projection)
Vegas 12-11-4 31-18-6 43-29-10
Calgary 12-12-4 28-20-6 40-32-10
San Jose 13-12-1 26-23-7 39-35-8
Anaheim 11-11-4 26-23-7 37-34-11
Los Angeles 10-13-4 22-27-6 32-40-10

According to the model only Vegas and Calgary are going to win at least half, or more, of their remaining games. Past seasons clearly hurt the Oilers and Coyotes in this model, and I have no doubt that the acquisition of Dave Tippett and Ken Holland doesn’t factor in much to this type of equation.

Advertisement - Continue Reading Below

Currently, teams in the Western Conference are on pace to have the second wildcard spot be 85 points, which is quite low. I agree with the model that has a team needing 92 points to get in. It is interesting that this model has the Oilers, despite having a below average second half, still finishing second in the Pacific Division.

It is hard to say who will get the third spot, as Arizona is currently missing a game in their projection and Vancouver has one too many. They will need to fix that. But it has the Flames and Sharks both missing the postseason.

Advertisement - Continue Reading Below

WRAP UP…

Like any predictions, there will be some we disagree and agree with, even if it is based solely on emotion and stats, or a computer based equation running 50,000 simulations. I understand why some think there is a bias when the Maple Leafs are projected to lose the same amount of games as the Bruins and Lightning from now on, and that the Eastern Conference is home to the top four teams with the most victories — Tampa, Boston, Pittsburgh and Toronto — the rest of the way. That rarely happens where one conference has the top-four teams for that long, so it is unlikely.

On November 20th I wrote why I felt the Oilers would be a playoff teamI based my prediction solely on what I’ve seen this year, combined with their remaining schedule and how the rest of the Pacific Division was playing.

I only focused on the Pacific Division, and while Dom and my prediction methods are different, we came to the same conclusion on the Oilers. Oilersnation, plan your April accordingly.

PARTING SHOT…

Our 9th annual Month of Giving begins on Monday. This year I have outlined most of the packages we will be auctioning off. I have a few more to add. I’m just waiting on confirmation, but you can click here to see the different packages for each day. We have some unique coaching and dinner experiences, some fun evenings out, golf packages, a few necessities for your house and more.

Thank you so much for all the sponsors who graciously donated their package and to the individuals who donated their time. It is awesome. And thanks in advance to all of you who will make bids and donate money to help others so they can have some more happiness and enjoyment in their life.

Recently by Jason Gregor: