A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back

Minnesota Wild v St Louis Blues - Game Three

The following is a guest post courtesy of Geoff Upchurch. If you enjoyed it, give him some feedback on Twitter @GeoffUpchurch!

It’s been a strange year so far for me as a Blues fan. I had the general “Hooray, here comes the season!” vibe when Prospect Camp rolled around, tempered by the thought that the Blues had made good moves, if not particularly ground-breaking ones. And then, the injuries all came pretty much at once, and I had the same expectations that a lot of fans had: “Well, hopefully they’re not out of it around the turn of the year when they get most of these guys back.”

Of course, as we know now, 21 games in, a funny thing happened on the way to that expected middling start while waiting for a third of the top-nine forwards and a top-four defenseman to return. The Blues ignored the expectations and the predictions, and have won a hair under 75% of their games so far. Steen came back six games in and helped solidify the top six. Now, all signs point to Bouwmeester playing and Berglund eyeing a return in the near future, and if the D pairs as tweeted by Lou Korac and others are accurate, Gunnarsson is shaping up to be the odd man out.

Is this really the move the team should be making?

Plenty of people have been pleased with Gunnarsson’s play so far this year. Like the rest of the D corps, he’s chipping in on the offensive side more than under the previous regime. Let’s have a look at the numbers, shall we? I’ll be pulling them all from hockey-reference.com, so feel free to check any math I wind up doing on them. Yes, there will be advanced, or fancy, stats. Wiki has a very basic explanation here.

First, from a historical perspective, what are the Blues expecting to draw into the lineup with a healthy Bouwmeester?

He’s played four full seasons here, plus the last third of the lockout-shortened 2013-13 campaign, comprising 321 regular season games. In that time, he’s posted 91 total points, ranging from 37 in his first full season here, to 13 the following year. Not coincidentally, that 13-point season was also the season in which his 737-game iron man streak ended due to the dreaded “lower body injury”.

He hasn’t particularly recovered in the point production department since then, with seasons of 19 and 15 points the last two years. Some of this can be accounted for by a reduction in power play time, as the 37-point year included a goal and 10 assists with the man advantage, but that doesn’t close the whole gap by any means. Bouwmeester’s even strength play in 2013-14, by itself, out-points any of the following three seasons by at least seven points, whether you leave the other seasons power play production in the mix or not.

Also, while plus/minus isn’t accepted as a particularly meaningful stat in the modern age, he followed up a +26 performance in the 13-14 season with seasons of +7, (-4), and +6. Not really killing the team, and not really showing heavily on the positive side, either.

A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back
(Photo by Dilip Vishwanat/Getty Images)

To round out our look at Bouwmeester, I’ll add in some fancy stats. His even strength Corsi and Fenwick (pick your poison, depending on how you feel about blocked shots) has been slowly declining from his first full season here, 53.3% CF and 53.0% FF in 2013-14, to 50.0% CF and 51.2% FF in 2014-15, 49.0% CF and 48.8% FF in 2015-16, and finally 47.8% CF and 48.1% FF last season.

Some of this should be expected because of how the Blues have chosen to use Bouwmeester over the years. In 2013-14, he actually received slightly more offensive zone starts than defensive zone starts (51.2%/48.8%), which rose to 52.2% defensive starts the following year, then 53.8%, and finally 58.7% in 2016-17. Without trying to pile on his entire advanced stat line, the simple truth is what you would expect: The possession percentages track up and down fairly straightforwardly with how much he starts in the D zone.

In other words, if the Blues, and fans, are expecting negligible offense and solid, but not great, fancy stats out of Bouwmeester in a heavy-lifting defensive role, they will likely get their wish. For his career, he hasn’t fallen under 45% CF or 44% FF, even under heavy defensive usage conditions.

Now that we’ve established what we would be expecting if we’re getting a healthy Bouwmeester back, where does it make sense to put him?

Other than for the sake of arguing, I doubt anyone will come along and say that he ought to be replacing Pietrangelo or Edmundson, so I’m going to leave them out of the following segment. Personally, I doubt there’s a reason to make a case for Parayko, either, even though plenty of fans’ eye tests say he’s had one of the rougher starts to the season amongst the D corps, given his history plus the handedness issue. But, what the heck, let’s leave him in to make the comparisons more fun, eh?

First, a quick rundown of on-ice PDO for the D corps, with the on-ice shooting percentage and on-ice save percentage in parentheses:

Gunnarsson – 107.1 (10.8% oiSH% + 96.3% oiSV%)
Pietrangelo – 103.9 (10.5% oiSH% + 93.4% oiSV%)
Parayko – 102.6 (10.0% oiSH% + 92.6% oiSV%)
Edmundson – 102.0 (11.2% oiSH% + 90.8% oiSV%)
Dunn – 101.7 (8.5% oiSH% + 93.2% oiSV%)
Bortuzzo – 99.7 (8.7% oiSH% + 91.0% oiSV%)

While not great for making the comparison, because these are for all situations rather than even strength only like the PDO numbers above, I will note that the team average for shooting% is 9.7, and the team average for save% is 91.7, for an all-situations PDO of 101.4.

Now, on to the individuals.

A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back
(Photo by Rich Gagnon/Getty Images)

The Case for Colton Parayko

In fact, let’s jump right in with ol’ Colt55. His history is pretty short, being that he’s a third-year pro this season. He tallied up nine goals and 33 total points in his rookie campaign, and followed that up with four goals and 35 total points last year. This year, he has two goals and eight points already, pacing to break the 30-point mark again. Yes, he’s had his lapses in the early going this season, but his counting stats are still solid, and we can expect that he will easily out-produce Bouwmeester over the remainder of the campaign.

His fancy stats actually back this up in spades. If anything, last season was actually his down year, regardless of what our lying eyes might be trying to tell us. Parayko’s rookie campaign featured a 54.8% CF and a 55.4% FF with a slight bias to starting in the offensive zone (51.2% of the time). The so-called sophomore slump hit his second season, a 50.8% CF and 50.9% FF showing, while his usage narrowed even closer to 50/50, with a slight edge to defensive zone starts (50.4%). So far this season, you ask? Yeo has thrown Parayko right into the heavy defensive lifting, with 60.1% of his starts occurring in the defensive zone. Regardless, his Corsi and Fenwick numbers have rebounded, to 53.3% CF and 55.2% FF respectively.

Verdict: The kid’s doing all right. If anything, his underlying numbers are actually substantially better than the eye test might indicate. Add in his right-handed shot, and, well… we did just put him in the list for fun, after all.

A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back
(Photo by Justin K. Aller/Getty Images)

The Case for Robert Bortuzzo

Let me just go ahead and throw this out there before the stats: I like Bortuzzo as the third pairing RHD more than a lot of Blues fans, whether the numbers back me up or not. He’s not great by any stretch of the imagination, but let me put it to you another way. Would you rather play Bortuzzo on the third D pairing for some toughness, or Chris Thorburn at 4RW?

Now, the numbers. Bortuzzo is not going to bring this team much in the way of offense-oriented counting stats. For his career, not just with the Blues, he averages a whopping four-and-a-half points a year. His Blues-only numbers are pretty much identical to that. Maybe a hair lower, certainly nothing worth quibbling about. He does have one goal and four total points on the year, which is tenths of a percent better as a pace than his career year with 10 assists for Pittsburgh in 2013-14. Pardon me if I’m not reaching for the party favors and making obnoxious noises with a kazoo over it. Maybe the fancy stats can help his case.

Generally, Bortuzzo simply gets different usage than Bouwmeester has in the past. Excluding his first “full” year with the Blues, in which he played 40 games and took 58% of his starts in the D zone, he’s been starting a point or three over 50% of his time in the O zone, both with the Blues, and his career in the league. Possibly surprisingly, he’s been a positive driver of possession, even in that 58% D zone start outlier year in his time with the Blues, averaging 53% CF and 55% FF (ranging from 50.9% CF on the low end in 2016-17 to 57.3% CF in 2014-15 after being traded here from Pittsburgh, and 53.9% FF last year as a low to 60.4% FF in that same 2014-15 partial year). He’s sitting at 51.2% CF and 51.7% FF this year, though I will note that essentially the entire team other than Dmitrij Jaskin was wildly underwater in possession fancy stats for the first ten games or so of the year, and have mostly been seeing substantial improvement lately.

Finally, I will point out that Bortuzzo’s career high in games played for a regular season is 54, with Pittsburgh in the 2013-14 season. This is probably the biggest mark against him, and I don’t personally think that the stats justify it.

Verdict: He wouldn’t be my choice, but he’s been the coach’s choice often enough over his career that it wouldn’t surprise me to see him draw out.

A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back
(Photo by Derek Leung/Getty Images)

The Case for Vince Dunn

Plenty of concern out there in questioning who would come out of the lineup when Bouwmeester returned that it would wind up being the rookie leftie Dunn. He certainly has some offensive flair, if not necessarily on pace to match, say, Parayko’s rookie numbers. Still, two goals and five points, all at even strength, through a quarter of the season is nothing to sneeze at. He’s also earned enough trust from the coaching staff to be paired with Pietrangelo in situations where the Blues need more even strength offense. That alone probably seals his spot in the lineup, unless the wheels come off.

Dunn has seen by far the highest offensive zone starts by percentage of any of the Blues D, at 61.9%, and he’s parlayed it into a 54.9% CF and a 54.6% FF. Not the amazing showing that Parayko’s positive possession stats with essentially reversed zone starts is, certainly, but Dunn’s working reasonably well with what he’s being given.

A note not applicable to the others, Dunn is still waiver-exempt, so if he is the odd man out, the Blues can send him down to play minutes in the AHL rather than letting him rot in the pressbox, if they so desire.

Verdict: The team clearly likes his offensive upside, and his usage is almost the exact opposite of Bouwmeester’s. Should be safe unless somebody forgot to tighten the lugnuts on the new car.

A statistical analysis of the St. Louis Blues defense now that Jay Bouwmeester is back
(Photo by Harry How/Getty Images)

The Case for Carl Gunnarsson

Gunnarsson is the other left-handed shot that we might expect to come out of the lineup in favor of Bouwmeester. Historically, his counting stats are better than Bortuzzo’s by about three times. Though, let’s be honest, 13 points per year versus four isn’t hugely meaningful in the grand scheme of things. It’s all a far cry from solid-to-good offensive defensemen posting 30-40 point years, and the handful of elite offensive defensemen surpassing even that. He is showing up in the early going, with three goals and four total points.

I haven’t brought it up before now, because the other three defensemen mentioned are all shooting at fairly normal or career levels (all in the 3-5% range), but Gunnarsson is scoring on 15.8% of his shots so far, compared to his career average shooting percentage of 5.3%. Funnily enough, it’s actually the two “safe” defensemen who are posting likely-unsustainable shooting rates aside from Gunny (Petro at 10.1% compared to his career 5.7%, and Eddy at 15.8% versus his career 4.8%). That said, in terms of putting points on the board, Gunnarsson is benefiting from some luck to be tied with Bortuzzo in points.

With Bouwmeester out, Gunnarsson has been leaned on this season to take a heaping serving of defensive zone starts, at 61.7%. Unlike Parayko, who has borne up under this load to post positive possession stats, Gunny is currently sitting at 45.4% CF and 49.3% FF, both his worst showings during his four seasons in St. Louis. Interestingly, Toronto, during some of their mediocre-to-bad years in the early 2010s, attempted 60+% defensive zone starts for Gunnarsson in the 2012-13 and 2013-14 seasons, and he got shelled both times (44.5% CF and 45.0FF in 2012-13, and 39.7%CF and 40.8%FF the following year). More specifically, in 2013-14, Gunny’s team was out-shot 989-1500 while he was on the ice. 511 more shots against than for, or over six shots per game. The only player on the current Blues roster with a worse CF%? Chris Thorburn, at 45.1%. The next lowest defenseman is five points better than Gunnarsson, that being Edmundson at 50.6%. Gunnarsson leapfrogs a few forwards with his 49.3%FF, though that number is still nearly 2.5 points in arrears of the next D on the list in Bortuzzo (51.7%).

Verdict: His point total is being driven by unsustainable good luck, while his team is being out-shot substantially while he’s on the ice. Only an unsustainably high .963 save percentage is keeping him from being questioned as a boat anchor through 20 games. The heavy defensive usage is working out poorly, just as it has in the past. The best that I can say is that his possession numbers would likely improve to some degree if he were to slide down to more sheltered usage in the third pairing in place of Bortuzzo.

Overall, given how much the modern NHL favors handedness in defensemen, I think Gunnarsson is the choice as the defenseman to draw out of the lineup over the longer term. While his possession numbers could benefit from playing on the third pairing, at the very least, directly replacing him in the top four with Bouwmeester, and having Dunn stashed on the third pairing to up the offense in the top four in a pinch makes a lot more sense than expecting the team to roll out a lefty/lefty D pairing when they don’t have to. At the very least, a left-handed defenseman not named Carl Gunnarsson needs to be tasked with the high defensive zone starts job, be that Bouwmeester or Edmundson.

Arrow to top