2006–2020 NHL Re-Draft Team Grades

The NHL Draft: A time for unbridled optimism, where a downtrodden fanbase can be revitalized in the hope that their collection of teenage magic beans will someday blossom into a Stanley Cup trophy. But as exciting as draft day is, it’s what comes afterwards that really captivates that fanbase. Fans seem to take personal pride when a 4th round flier starts looking like he’s going to become an NHL regular; and they are just as quick to call their scouts donkeys when the 1st round pick’s trajectory looks to be losing some steam. And when those prospects enter the NHL, the fans don’t forget those feelings of pride and frustration; on the contrary, they only compound and lead to fiercer arguments as those aforementioned magic beans grow into their end potential.

For that reason, I was not surprised to see how much traction that JFreshHockey was getting for his recent fan-voted NHL re-drafts. The final graphic included a comparison of the re-draft position to the player’s actual draft position; you could quickly see which players fans considered steals, busts, etc. Now, these results were adjudicated by the court of public opinion. I have to give credit for the genesis of my article to Twitter user Cody Magnussen, who took fan subjectivity out of the equation by utilizing an objective measure to perform a re-draft. He utilized a statistic from Evolving Hockey called Goals Above Replacement (GAR), which is an all-encompassing statistic that boils down the total amount of value that a player contributes to their team into a single number.

Seeing that JFresh and Cody had individual draft classes handled, I decided that I wanted to do some analysis on how teams had performed in aggregate throughout the years. Taking a page from Cody’s book, I decided that it would be easiest to do so via an objective statistic, for which I chose another Evolving Hockey statistic, Standing Points Above Replacement (SPAR), which is an offshoot of GAR, and is an extension to the more commonly known WAR statistic.

For each draft class from 2006 through 2020, I ranked each player in their respective draft class by career cumulative SPAR. I then calculated 2 statistics:

• The 1st statistic, the Ranking Delta, is simple algebra. It is the difference between the player’s SPAR rank in the draft class vs. where they were drafted.
• The 2nd statistic, SPAR>BPA, is slightly more complicated and is most easily explained via an example. David Pastrnak was drafted #25 overall in the 2014 class, and is ranked 1st in his draft class today with a career Standing Points Above Replacement (SPAR) of 46.7. For the SPAR>BPA statistic, I calculated the difference between Pastrnak’s actual SPAR (46.7) and the SPAR of the 25th ranked player in that class, Michael Amadio (10.2 career SPAR). It’s a simplified way of comparing the value that the team got from their draft pick vs. the value of the theoretical Best Player Available (BPA) at the team’s draft position (in this case, #25) if all players were drafted in the “correct” order (ranked by SPAR).

Obviously this SPAR>BPA statistic is a hypothetical exercise, because players are not drafted with perfect hindsight in “correct” order. The Delta statistic does not require these hypotheticals. The SPAR>BPA statistic’s advantage, though, is that it not only details which players have justified a higher position in a re-draft, but also by how much that draft pick has paid off for a team. With this statistic, you can see how much value a team left on the table with their pick; or from a more optimistic perspective, you can better quantify just how much of a steal a certain draft pick was.

I aggregated each team’s Delta and SPAR>BPA for each player drafted from 2006 to 2020 (Vegas is not included in this analysis), which resulted in the rankings below. For both of these statistics, I isolated for 1st round selections, selections from the 1st 3 rounds, and for a team’s entire draft. Note that optimal outcomes are a lower Delta and higher SPAR>BPA.

You can see that these results are correlated, but are not identical. San Jose, for example, has done an excellent job finding NHL-caliber players in the late rounds of the draft throughout the years, but that value has not necessarily translated to star players. We’re talking the Justin Brauns and Tommy Wingels’ of the world; guys taken in the 6th and 7th round that may never have made an all-star game, but became real NHL regulars. Then you compare that to the Capitals, who have a lower quantity of late-round finds, but a higher quality — like Braden Holtby, taken #93 overall in 2008 (which translates to an expected/BPA SPAR of -7.8) but now ranked #2 overall in the class with 51.2 SPAR over the course of his career, for a SPAR>BPA of 59.

Both of these statistics were focused around how well a team drafted relative to the draft positions they were assigned to. But I also wanted to consider the amount of value (SPAR) from the draft that a team got on an absolute basis. Because in the end, you don’t win Stanley Cups based upon “value drafted greater than expected”; you win them based on the actual overall value on the ice.

Washington remains near the top of the list, but that does not necessarily indicate a pattern. Nashville, for example, has generally drafted quite well relative to expectations and is 4th in Total SPAR>BPA, but that has not translated to excellent results on an absolute basis. A lot of that has to do with the fact that since 2006, the Preds have only picked higher than 11th on 2 occasions, and those selections were Seth Jones at #4 and Colin Wilson at #7.

Chicago is a fascinating case study; the team got a ton of value in the 1st round (a couple players named Kane and Toews that you may have heard of), but the results outside of that were disappointing overall. That may have played a role in why the wheels ended up falling off at the end of the dynasty , as there were fewer entry-level deals to replace the cap casualties that emerged. It’s a somewhat similar story in Edmonton, which made the slam-dunk McDavid and Draisaitl picks but did not have a whole lot of high-value late-rounders to speak of.

That observation about the Oilers becomes even clearer when you provide the additional context of the “expected” value that a team could have theoretically picked with the Best Player Available at their draft position. The below chart is essentially the SPAR>BPA statistic in expanded form:

You can see in the above table how Nashville had the 5th lowest value theoretically available for their selections in the 1st round. Detroit also stands out — they did not pick in the top 10 until 2017, so they simply haven’t had an opportunity to acquire very highly-touted prospects in the draft until recently. And folks wonder why the Yzerplan is going to take some time to take shape.

I will also concede that Pittsburgh got a bit of a raw deal here — if this exercise extended back to 2005, Sidney Crosby’s SPAR would be included, and that single selection would bring their 1st Round drafted SPAR to a respectable level. Unfortunately, there is not SPAR data available from the 2006–07 season, which affects players like Crosby, Anze Kopitar, and Marc-Edouard Vlasic. The Kings also lose out on 158 SPAR and 119.2 SPAR>BPA from Kopitar and Jonathan Quick in the 2015 draft, so I promise I’m not singling out the Pens.

Anyway, aggregated numbers are boring. Fans love slapping labels on players. So let’s do that!

Superstars clearly matter. When you look at the teams that have won the Stanley Cup the past decade, it’s generally been teams that have drafted Superstars — Tampa, Colorado, Chicago, Washington, Boston, and also Pittsburgh and L.A. who had drafted superstars in the years just prior to this data set. It seems to correlate with championships better than any of the statistics that I’ve discussed in this article. For that reason, Stars fans should feel really good about the direction that their organization is headed in with their young core.

But then you have the Blues. They may not have had a roster filled with superstars, but they generally hit on their high-end prospects (with 5 “Stars” emerging from the 1st round) and had a couple of steals later in the draft (namely Parayko and Binnington). Mix in a little ‘Gloria’ juju, and you have a Stanley Cup in St. Louis. This should provide some hope to teams with similar profiles like the Hurricanes and Islanders.

I’m happy to dig into the methodology behind what makes players “Superstars,” “Stars,” “Steals,” and “Busts,” but I assume at this point in the article very few people are interested. Rest assured that these labels have come about via a consistent formula, not through my own subjective opinion.

As I wrap this up, I’m going to make the obvious note that there is obviously some nuance lacking in this analysis. Some of these players were traded and never played a game for the team that drafted them. Hell, Frederik Andersen was drafted twice and ended up leading both the 2010 and 2012 draft classes in SPAR, so Carolina gets credit for a player that the Ducks drafted again 2 years later. But I do not have the patience to comb through the over 3,000 draft picks from 2006 through 2020 to properly assign value to teams that acquired players via methods other than the draft. That’s also not what I was after when I began this analysis. I simply wanted to see which draft rooms were proven down the line to have made the best calls on draft day (and, admittedly, I was curious to know which ones made the worst calls, too).

A closing note — there is clearly a great deal of luck (both good and bad) embedded within these statistical conclusions. There are any number of reasons outside anyone’s control that a player may not develop into who the organization thought they were selecting — injury, personal reasons, or otherwise. It’s also quite apparent based on this analysis that teams can compete for championships even if they have a few “Busts.” With that said, if your organization has selected 11 Busts within a 15-year window, maybe that indicates that there is a not-so-great pattern at play (sorry, Yotes fans).