hey could you expand on the fact that wsh has a better defensive rating that Quinn Hughes? I’m very new to hockey and the most common things I’ve heard, esp across other platforms like tiktok, that Hughes is one of the best defence men in the league and will’s defensive game needs heaps of work??? Like verbatim I’ve heard ppl call Hughes the best defenseman on multiple occasions so how does this discrepancy happen? Is it just different people’s interpretations of stats and the game?
I would literally love to, Anon! Buckle up, because this can get a little technical at times but I’ll do my best to make everything easy to understand. Full analysis of Quinn Hughes & Will Smith's defensive metrics, plus my take on the discrepancies between the metrics and how people discuss both are under the cut!
The first thing I want to make abundantly clear is that when people say “best defensemen” they almost never exclusively mean “best defender” —confusing, I know. They mean best player in that position, which includes far more than just defensive impact. Like with forwards, truly elite defensemen are expected to have a pretty complete 200ft game. This means impacting in all three zones, and having high usage rates. That said, important to note that even though I’ll be almost exclusively using the numbers here, my bias does exist. I flat disagree with people who think Quinn Hughes is the best defenseman in the league. I do think he’s an elite, game changing player, but I don’t think he’s the best player in the league, at his position or in general. For Will Smith, I agree with the notion that his defensive game needs work, but I think it’s in some very specific areas (not necessarily all around or “heaps of work” as you’ve worded it here) and I think that he’s more disruptive than he’s often given credit for. But you asked about the numbers, so let’s dig into that.
For this, I’ll be looking at Net Rating, rather than the individual stats that go into the model, otherwise this would be far too long. Still, there’s nuance in these stats that must be acknowledged. That said, Net Rating (and defensive and offensive rating) has been developed with a lot of awareness of that nuance and very careful precision to try and account for it, so I feel pretty comfortable trusting the numbers.
Net Rating is essentially a measure of a player’s value in wins—an all encompassing stat when looking at players’ skill and impact over the whole game—that can be broken down into two sections: Defensive Rating and Offensive Rating. Here's how both measures break down:
Offensive Rating: A weighted combination of goals, primary assists, secondary assists, individual expected goals, faceoffs, penalties drawn, expected goals for impact at 5-on-5, goals for impact at 5-on-5, power-play goal impact and usage.
Defensive Rating: A weighted combination of blocked shots, faceoffs, penalties taken, expected goals against impact at 5-on-5, goals against impact at 5-on-5, penalty kill impact and usage.
Offensive and Defensive Rating are both centered around an average skater at their position where positive is always good. An average player by Net Rating is right at zero. Note, because goalies exist, offensive metrics still hold more weight because they can be more easily isolated to an individual player.
Okay now that we understand the numbers, let’s look at Quinn Hughes’ stats analytics from this season:
His O Rating of +21.5 puts in him the 99th percentile, is the 6th highest in the league, and the 2nd highest among defensemen. His Net Rating of +14.6 puts him in the 96th percentile, is the 35th highest in the league, and the 17th highest among defensemen. What stands out here is that his D Rating is -7, which is in the 1st percentile (meaning the bottom 1% of the entire league). Out of 716 total players who played in 20 or more regular season games, Quinn Hughes’ defensive rating ranks 709th, or 8th worst in the league.
Here’s his rating for every game this season by percentile:
As you can see, his O Rating started high and pretty much stayed there, really hitting his stride just before the trade to Minnesota. His Net Rating is a similar story, with less consistency. His D Rating is all over the place though, he had a lot of pretty good defensive games, and a lot of really terrible defensive games. What this tells me is that his defensive impact is extremely inconsistent and likely pretty susceptible to opponent match ups.
My personal ethos is that no matter how talented a skater is in one area, their game has to be more comprehensive to rightfully be considered “the best,” and especially with defensemen, I can’t justify giving the title of best defenseman in the league to a skater who is so completely inconsistent in their own zone, especially to such an extent that their overall defensive rating on the season ranks 8th worst in the entire league among all skaters.
For comparison, this is Cale Makar’s overall stats analytics and rating percentiles over the course of the season:
His +17.6 O Rating and +23.4 Net Rating both put him in the 99th percentile, and his +5.9 D Rating puts him in the 97th percentile. Among all skaters, his O Rating ranks 15th, D Rating ranks 27th, and Net Rating ranks 7th.
Looking at their Net Rating side by side, here’s how they compared throughout the season:
While Quinn Hughes had some spikes where he played a better overall game than Makar did, overall throughout the course of the season Makar played not only a better, more comprehensive game, but did so with far more consistency than Hughes did. Consistency matters, but also requires a bigger picture perspective, which sometimes gets missed.
As for Will Smith, here’s his overall stats analytics and rating percentiles over the course of the season:
His +7.2 O Rating puts him in the 89th percentile, his -1.3 D Rating puts him in the 25th percentile, and his +5.8 Net Rating puts him in the 84th percentile. For his first year at a new position and only his second year in the league, that’s incredibly good impact. In fact, offensively and overall I’d argue that’s damn good (at least top 16% of the entire league good according to the numbers). His defensive game does need work (remember the average for position would be 0, so his -1.3 rating does mean that defensively, his impact is below average for a forward with his usage metrics) but the numbers aren’t as concerning as people often make it appear with how they talk about it.
Here’s how his and Quinn Hughes’ D Ratings compared over the course of the season:
Like noted above, Quinn Hughes was all over the place, but Will’s rating was on an upward trend at the end of the season as he worked on his defensive game, and he was trending back towards league average for forwards, which is impressive given his already above average offensive and net impact ratings. Will’s got work to do, no doubt, but his defense is no where near as terrible as people claim (besides, defense at the NHL level is usually where skaters require the most development anyway) and he’s got the IQ, talent, and work ethic to continue to improve, especially as he continues to get more comfortable on the wing and playing at the NHL level in general.
So to answer your question about discrepancies between the way people talk about Quinn Hughes and Will Smith and the actual statistics, I think it’s a combination of people’s interpretations of the game and also Quinn Hughes’ impressive offensive impact that really explains it. Quinn Hughes is a high profile, impactful, highly intelligent and skilled skater. His Net Rating isn’t misleading, and it does mean that overall he’s a player in the top 5% of the league (ranks 35th among all skaters), but so much of that is entirely offensive impact. I just don’t think that’s enough to call him the best defenseman in the game. I wouldn’t even place him top 3.
As for Will, I honestly can’t say why people act like he’s not as good as his is. Part of it could be the Mack comparisons just because they’re playing every game together, and Mack’s insane—I honestly don’t know.
What I do know is that he’s an insanely talented player with a high ceiling that he’s nowhere near hitting, and largely I have zero doubt that he’s going to prove everyone who’s ever under-sold him or doubted him completely wrong. That includes everyone who likes to doom-post about Will being moved to the fourth line or possibly traded. All those people are doubting and underselling Will as well. He’s a damn good hockey player, one of the biggest assets the Shark’s have, and is coming off of an impressive sophomore season. He’s not going anywhere.
This is so long but it was a really fun deep dive, and I hope that answered your question!
I was both laughing and in awe to your response re: disastrous Blackhawks trade just now - it's always a great time reading your hockey analysis :). Stats! Real numbers in excels! Good insights!
Please share more on advanced stats and defensemen analysis. Curious minds love to learn.
Omg thank you!!! That post is so long fr, but I always try to be thorough with any hard analysis and I think that defensemen by and large get screwed over a lot by stats in hockey. Dom Luszczyszyn (responsible for the score cards and actually developing the metric of Game Score for players and teams) is like seriously my GOAT for fixing a hole in advanced metrics that largely didn't actually account for defensive impact, and really focusing on balancing the playing field in that way.
The issue with the way that the NHL does summary statistics (or one of them, anyway) is really that the only stat that measures defense at all on that list is +/-, which is a famously problematic stat for a lot of reasons but most notably because it's not independent of teammates, so it's not really a great read on the individual player themselves. While it can tell us if a player was on the ice when a goal is scored against them (thus implying poor defense) it's not actually a good measure of defensive impact. When the league (or anyone really) is ranking skaters using summary statistics then, it's almost entirely based on offense. I despise ranking defensemen using offensive metrics because that's not their primary job on the ice (not that a high scoring defenseman isn't good, but it should, in my opinion, be secondary to defensive metrics). The Norris, given to the best defenseman in the league, always ends up going to a defensemen at the top of the statistical rankings, which is problematic because those statistical rankings don't measure defensive impact at all.
You can find some basic defensive statistics under the leagues miscellaneous stat listings. For the more basic defensive stats, I like to look at Hits, Blocks, and Takeaways per 60 (raw numbers are influenced by total games played and TOI, the average of per 60 for the most part fixes this problem), because I think these give a good idea of defensive skill breakdown. However, even these are tough to grade for an actual measure of true defensive impact. Game Score is, at it's core, a measure of impact, which is neither groundbreaking or that different from something like a production stat. But under the surface it's pretty ingenious, it accounts for the differences in defensive and offensive play driving and counter-balances to provide weight for strictly defensive impact. You can read the whole methodology behind the statistical model here. I also am a huge fan of offensive, defensive, and net rating—the first two provide a really solid isolated stat, and the last one is a very good yet simple overall impact statistic. If you love numbers, you should really dive into game score cards because that statistical analysis is really changing the game.
To illustrate all of this, let's look at some defensemen and their metrics.
Zach Werenski, this season's Norris winner, ranked 2nd in points for all defensemen (behind Bouchard, who routinely has his impact written off because he plays with McDavid but we'll come back to that). Points say absolutely nothing about his defensive skill or impact. Looking at his average game score, he ranks 4th among defensemen. His net rating ranks him at 5th. Pretty good, but overall he's not the top defenseman using overall impact metrics. If you isolate his defensive net rating he actually ranks 70th among all defensemen, it's his 5th place offensive net rating that accounts for a lot of his overall impact. So, best defenseman?
Circling back to Evan Bouchard, these metrics aren't 100% independent of teammates, but they are formulated specifically to at least account for that as much as possible, so largely I'm tossing the "It's because he plays with McDavid" argument out a window. That might be valid when looking at his leading points stat, but let's look at the other stuff. Among defensemen, Bouchard ranks: 1st in game score, 1st in net rating, 1st in offensive rating, and 100th in defensive rating. Again, offensive dmen are incredible, but let's not sacrifice the defensive parts of the game here.
The top 5 defensemen in game score and in net rating are the same (just switched around a bit) so lets look at the other 3 from the top 5:
Cale Makar: 2nd in game score, 4th in net rating, 21st in defensive rating and 5th in offensive rating.
Lane Hutson: 3rd in game score, 2nd in net rating, 32nd in defensive rating, and 3rd in offensive rating. (this is fucking insane as a sophomore dman by the way)
Moritz Seider: 5th in game score, 3rd in net rating, 2nd in defensive rating, and 14th in offensive rating
The 3 finalists for the Norris this year were Werenski, Makar, and Rasmus Dahlin, who ranked 7th in game score, 6th in net rating, 49th in defensive rating, and 6th in offensive rating.
With all this in mind (noting that I have not looked at TOI or usage metrics, which would impact my thoughts here but I want to keep it to defensive and overall metrics) my top 3 finalists for the Norris, with a focus on defense and overall impact but still weighting offensive impact, would probably actually be: Cale Makar, Lane Hutson, and Moritz Seider, and because he's top 15 in everything Seider might take it for me. I'd also have to take a serious look at Bouchard because ranking 1st in both comprehensive categories despite ranking 100th in defensive net rating is no joke, but at the end of the day this is an award for excelling defensively, and I hate that it tends to go to defensemen who excel primarily at offense.
This has been an incredibly long rant about defense and statistical analysis, thank you for coming to my ted talk. I'll leave you with Lane Hutson's overall score card for the season because he's a lunatic and freakishly skilled:
Note: All of the above rankings are among the top 125 (out of 239) defensemen who played a minimum of 20 regular season games in overall impact game score, so it's possible that D and O ratings could actually rank a little lower, especially for those who have notably low rankings on one or the other.
just thought you should know you're getting some shit on twitter for the qh/wsh post 😭😭
https://x.com/calesbians/status/2071586239963160743
CRYINGGGGGGG LMAO. Okay for anyone #confused I did not make those numbers up. Those are the actual numbers from the Net Rating Model, which was developed years ago by Dom Luszczyszyn, an extraordinarily well-respected and established statistical analyst that currently handles a lot of The Athletic's hockey analysis. For the most recent season, Quinn Hughes' defense rating is -7 and Will Smith's defense rating is -1.3. Now, the rating is based on a scale where 0 is the average for the player's position. I was very specific in my wording for the OG post and no where in it do I say that Will Smith is a better defender than Quinn Hughes, because that's not what these numbers tell us. What they do tell us is that Quinn Hughes is a significantly worse defender compared to every other defenseman in the league than Will Smith is compared to every other forward in the league.
Anyway, to the person who tweeted that, idk what you want me to say. Will Smith does have a better defense rating that Quinn Hughes, so go argue with the statistical model, but I doubt that will change it.
I looked at the same website as you did and Quinn’s defense rating is +3 and WSH is -1.3
So this is a good example of what actually misrepresenting the numbers looks like. Quinn Hughes' defense rating for the postseason was in fact +3. I didn't include either Hughes' or Makar's playoff numbers because the point of the post was to look at regular season stats. Obviously, Will Smith does not have a postseason performance to look at, and it's not statistically relevant to compare one player's postseason performance to another player's regular season performance. I'm going to assume this was an honest mistake on your part and just let you know that you have to toggle to "Regular Season" under "Game Type" on the website. Right now, it auto-populates for the playoffs because those are the most recent stats. Hughes's playoff defense rating was +3, his regular season defense rating was -7. I included links in my original post for this very reason, so that people could see the numbers themselves. I have also linked both Hughes' postseason and regular season stat cards in this post for your convenience.
Disclaimer, this is the last time I'm going to answer an ask specifically about this. The numbers are what they are, I did not come up with them, and I certainly didn't misrepresent them. I'm only answering this ask because I want to address the different ratings for anyone who may have accidentally looked at playoff numbers and been confused.
