When Vegas Golden Knights GM George McPhee began the expansion draft process on June 18, he might have been overwhelmed by the possibilities. There were so many different directions he could have gone in to craft an identity for the NHL's newest team, each one with its own unique set of strengths and weaknesses.
In the end, McPhee went all-in for a model that builds toward the future. He accumulated a lot of picks and prospects in exchange for passing on some quality young players, and accepting and/or selecting several older players on pricey contracts.
It might be hard to argue against a strategy that allows for so many future opportunities, but a closer look at the numbers highlights the opportunity he had to craft one of the league's better blue lines, making the Golden Knights a potential playoff team in their inaugural season.
As it stands, the numbers confirm the Golden Knights might not be very competitive out of the gate.
Where exactly are they projected to land in the standings, and what could McPhee & Co. have done differently to optimize this singular opportunity?
Projecting the actual roster
I fed the Golden Knights' roster into the statistical model described in the opening chapter of Stat Shot, which measures each player's performance relative to his cap hit using a three-year weighted average of a catch-all statistic like GVT, regresses it to the league average to remove random variation, and makes adjustments for age.
The end result estimates that the Golden Knights rank 24th out of the NHL's 31 teams, which is pretty consistent with the traditional hockey perspective. While they might stay out of the league's basement, the Golden Knights are likely to remain outside the race for a berth in next season's playoffs.
The team's player usage chart helps to highlight the chosen roster's strengths and weaknesses. It places all of the team's players into context based on the percentage of shifts they start in the offensive zone relative to the defensive zone on the horizontal axis, and based on whether they take on top-line opponents or the depth lines on the vertical axis.
Finally, the size and shaded circles provide an indication of how well each player's team performed with that player on the ice, from a shot-based perspective. The shaded circles indicate positive numbers, while white circles are negative, and the size of the circles reflect the size of that advantage, or disadvantage.
In particular, the chart highlights the interesting choice that McPhee made for the team's defensemen, who are denoted in italics.
Their top defensemen, in terms of cap hit, ice time, and quality of competition are Marc Methot, Alexei Emelin, Jason Garrison and Luca Sbisa. In each case, these veterans have the big, white circles that denote poor shot-based metrics. Even Methot, who often played with the incredible Erik Karlsson with the Ottawa Senators, somehow managed a white circle.
Yes, the secondary defensemen are young, plentiful and exceptional -- even after Trevor van Riemsdyk and David Schlemko were traded to the Carolina Hurricanes and the Montreal Canadiens, respectively, the next day. But, they missed a real opportunity to craft a solid top four.
Instead of Methot, Emelin, Garrison and Sbisa, McPhee could have chosen from any number of the following, superior options:
Sami Vatanen or Josh Manson of the Anaheim Ducks
Jason Demers of the Florida Panthers
Matt Dumba of the Minnesota Wild
Calvin de Haan of the New York Islanders
Dan Hamhuis of the Dallas Stars
In each case, these defensemen have better numbers, and not just in terms of shot-based metrics or catch-all statistics. Even using the eye test, these are all defensemen who take on top opponents in both zones, who can be effective in all manpower situations, and can also contribute scoring. Furthermore, no four-player combination matches or exceeds the combined $17.2 million cap hit of the top four that McPhee actually chose.
In fairness, McPhee did make some pretty good side deals in lieu of selecting most of those players. He extracted Shea Theodore from the Ducks in return for selecting Clayton Stoner instead of Vatanen or Manson, and made a deal to get both Reilly Smith and Jonathan Marchessault from the Panthers, both of whom will be key, top-six forwards.
From Minnesota, McPhee landed Alex Tuch in exchange for signing Wild free agent Erik Haula, and received a windfall from the Islanders that included Jake Bischoff, a first-round pick in 2017 and a second-round pick in 2019, in exchange for selecting backup goalie Jean-Francois Berube and agreeing to take on Mikhail Grabovski's contract. No deal has been announced for the Dallas Stars, but there must have been some conversation before McPhee agreed to take Cody Eakin's contract off their hands.
Projecting the alternative roster
It's hard to disagree with those moves because they did land some pretty decent players, picks and prospects. But what if McPhee had just built the best possible blue line? Then, the team's player usage chart could have looked like this.
In this alternate reality, McPhee selected Manson, Demers, Dumba, de Haan and Hamhuis, which means losing Theodore, Stoner, Marchessault, Tuch, Berube and Eakin. It assumes that he still could have (and would have) acquired Smith.
Would McPhee have had to sacrifice some of the picks he acquired? No, not necessarily. By laying off Tampa Bay's other unprotected players, and agreeing to take on some contracts like those of Grabovski and Garrison, it's possible McPhee could still have negotiated most of those picks.
However, this alternative version of the Vegas Knights does assume that McPhee kept both van Riemsdyk and Schlemko for depth purposes, which means forfeiting the second- and fifth-round picks received for them, respectively.
Plus, this configuration still leaves an expansion draft pick from Montreal, Ottawa, Vancouver and possibly Tampa Bay, since the selections of Emelin, Methot, Sbisa and Garrison weren't necessary.
It also frees up $15.3 million in cap space because this new orientation uses up $11.6 million instead of the $26.9 million for the actual picks. That means McPhee would have the cap space to pick some higher-priced forwards such as Montreal's Tomas Plekanec or Ottawa's Bobby Ryan. Since Jon Merrill isn't required in this configuration, McPhee could have considered changing his New Jersey pick to Michael Cammalleri.
Or, in keeping with his overall philosophy of building to the future, McPhee could have picked some promising young players from those teams and saved the cap space, allowing him to either take on more bad contracts in exchange for picks and prospects, or focus on free agency.
What could have been
Even without considering those additional possibilities, adding this simple blue-line change to the statistical model lifts the Vegas Golden Knights from 24th in the NHL into the playoff mix at 14th. That means they could have been competitive right out of the gate, been in a superior position cap-wise, and without undue sacrifice for the future.
It's easy to play armchair GM, and it is hard to argue against an all-in strategy for the future, especially when almost all of the big contracts expire after the team's first season anyway (specifically, 19 of 31 players chosen will become unrestricted or restricted free agents next summer).
However, the numbers clearly suggest there was a real possibility of building an excellent blue line that could have thrust the Vegas Golden Knights into a surprising playoff position in their first season, without sacrificing many of the futures they wound up with. Only time will tell if the very difficult decisions McPhee made were the right ones.