Examining how Morgan Rielly's partners affected him

This past season, Morgan Rielly completed his ELC and signed a 6-year, $30M deal with the Leafs. The Leafs front office has made it clear that the future of the Leafs defense lies with him. From Mike Babcock's usage of him, to everything the team and front office has said publically, Rielly has had a lot expected of him in his young career and for the most part, he has been a beacon of hope on an otherwise uninspiring blueline.

His offensive talents are prodigious. One of the best skaters in the league, Rielly uses that skill well to generate offense at a rate consistent with top pairing defenders. Unfortunately, he hasn't shown the same level of ability on the other side of the puck. Rielly hoards shots against like they're going out of style. DragLikePull had an excellent article about this last season, and came to the conclusion that Rielly is more likely to be a glass cannon than a do-everything defenseman.

Now of course, there are caveats. Rielly is 22 - a lot of great players weren't even NHLers at that age. Rielly also was entrusted to take top pair minutes by Mike Babcock, and do so without a great deal of sheltering, either by zone or competition. Perhaps most importantly, Rielly was saddled with a fairly mediocre partner in a clearly-over-his-head Matt Hunwick, which had the knock on effect of forcing Rielly to play the right side on defense (his off-side). Domenic Galamini has written about the importance of handedness among defense pairings, and found that on average, a defenseman playing their off-side has a negative affect on their pairing of about 2 Corsi events per 60 minutes (compared to a player with the same partner and possession ability playing their natural side).

I wanted to take a look at these last two factors in particular, as they go hand in hand. That is, how did Rielly's pairing (and relatedly, the side that pairing forced him to play) affect his results? All numbers are 5v5 from HockeyAnalysis. These numbers are not score-adjusted - as a result, the shot results are rosier than reality, since the Leafs were often behind and thus, their shot numbers are likely aided by score-effects juice that is not really accounted for here (unfortunately, score-adjusted WOWY data is a hard find).

Rielly's Partners

The first thing we should understand is who Rielly played with, and how often. Below is a table depicting Rielly's most common defence partners, and his results with said partner.

There's a lot of data there, so lets simplify this and only look at TOI.

From this, we see that the only sizeable samples are Hunwick - Rielly and Marincin - Rielly, who were both primary defensive partners for Rielly at different points in the year. Polak's time with Rielly tended to come in assorted fragments during games where Rielly was paired with someone else (usually Hunwick). Rielly was (somewhat randomly, to my memory) chained to Scott Harrington for three games early in the season. The rest of the players were people he spent sporadic shifts with in games where he was primarily partnered with someone else. For the rest of this analysis, I'm going to drop everyone who spent less than 10 minutes with Rielly this season. This is somewhat arbitrary, and I wouldn't quibble if you wanted to filter more heavily.

Rielly's Performance

Rielly's performance (here, measured by shot results) varied greatly depending on his partner.

I'm largely going to ignore guys like Carrick and Harrington, based on the extremely small sample sizes. Really, the same could be said about every pairing except Hunwick, Marincin, and maybe Polak.

Like most charts about the Leafs, you can identify the section where you want to be by finding Jake Gardiner. I'm quite curious about the contexts in which the Gardiner-Rielly pairing was deployed. Their results were really stellar together. Perhaps it would make sense long-term to use them in high leverage minutes, even if they don’t primarily play together.

If we look at Rielly's two most common partners, we see they're adjacent to one another, though Marincin-Rielly has a modest advantage over Hunwick-Rielly in both shot suppression and shot generation. This is as close to a true comparison between pairings as we'll get. Both of these pairings saw heavy minutes when used, and Rielly played the right side on both. I think we can take this as a positive sign. I'm the President of the Martin Marincin fan club, but even I'll admit that he's probably a little out of his depth on a top pairing. If Rielly can tread water with Marincin in that role, it's a positive development, especially given how abject his results were with Hunwick.

What’s more is that there isn’t any evidence that the pairings were deployed significantly differently from one another. While I couldn't find how competition varied by pairing, we can look at zone starts as an alternative. These have limited explanatory power in terms of shot results - I feel like I link this in every piece, but Micah McCurdy had an excellent discussion of this here. That said, I think it carries some information about how the coaching staff views the pairings.

I don't think there's an appreciable difference in the zone starts of Hunwick - Rielly and Marincin - Rielly, which to me, indicates that they were probably seen similarly by Babcock in terms of how he wanted to deploy them. I do think there's an argument for Marincin - Rielly's results being understated here, as they played a large portion of their games together down the home stretch of the season, when the Leafs were playing an AHL set of forwards.

Going back to the bubble chart, there's something else I found worthy of note. The three defensemen that Rielly played the most on his natural side with were Polak, Phaneuf, and Corrado. All three pairings were more offensively potent than any of Rielly's pairings where he played RD (with the exception of Gardiner, who is a total outlier on the Leafs defense in the best possible way). There's not enough data here to say this with certainty (I want to stress that - aside from Hunwick and Marincin, we're looking at pretty small TOI figures), but perhaps for Rielly, the handedness effect manifests itself mostly on the offensive side of the game. I’d like to explore this more in a future piece, because there’s a lot to unpack here. For now, it’s interesting to note how Rielly seemed to be disproportionately affected on offense by playing his off-side. If true, this would reduce the strength of the argument that Rielly’s poor defensive numbers are a result of him playing an unfamiliar side of the ice.

What Did We Learn

I think there's a few things to take away from this. The first is that the Hunwick-Rielly pairing really didn't work. I'm sure that's no surprise to anyone who watched the Leafs this year, but the numbers corroborate it. The second is that Marincin-Rielly held their own, though they were certainly not ideal for a 'top pairing' either. That might sound negative, but I actually think it's fairly impressive that Rielly can tread water in tough minutes playing with someone like Marincin. Marincin, much as I love him, is probably not a top pairing defensemen on a team attempting to win. Last year, he was asked to be for a significant portion of the year

So Rielly is no Jake Gardiner, who seemingly carries everyone to impressive shot results, but you know what? Most players aren't. Most players - even "number 1 defensemen" - play with partners who are talented and well above average in their own right. Rielly really hasn't had that opportunity. When Rielly has been played with a good player (Gardiner), he's succeeded. To get a better idea of whether he can play top pairing minutes going forward, they need to pair him with one for the long term.

All stats from HockeyAnalysis.