Eye Test: How can Shea Weber's stats be this bad?

Shea Weber of the Nashville Predators is to many the consensus best NHL defenseman, but his advanced statistics are plain not good and here's why.

Typically, these Eye Test articles entail me reconciling my head with my gut—what I knowstatistically to be true about a player with the raw impression I get from watching him. For this installment, in the interest of full disclosure and at serious risk of being corny, I have to involve a third organ-as-metaphor: my fickle heart. 

Ever since Dad Page turned to me, a high school freshman in Nashville, and said, “Number 48 ... the Weber kid ... watch him,” the Predators' franchise player has been of special interest to me. More than any other single player, Shea Weber is responsible for my enthusiasm for hockey. Given that background, it’s very tough—if not downright impossible—for me to perform an unbiased, clinical accounting of his game.

SI.com's 2014-15 NHL midseason award winners

. He is, to many, the consensus choice as the best defenseman in hockey. In a bigger market, he would likely already be working on his third or fourth Norris Trophy. At various moments in Sochi last year, he seemed to be single-handedly willing Canada to its latest gold medal.

But Weber’s advanced statistics are plain not good. His Corsi percentages during the three seasons since longtime partner Ryan Suter left—45.9% in 2012–13, 48.1% in ’13–14, 49.1% this year—are pedestrian, though they are steadily improving. And any mention of mitigating factors, such as his heavy workload and his inexperienced teammates, only further challenge his perceived elite status, since comparable players, including Zdeno Chara, have always demonstrated an ability to elevate the stats of weaker partners, and in the toughest minutes. 

Surprisingly, even proponents of advanced stats have been slow to condemn the Preds’ captain. The opportunity to tear down a previously unimpeachable player—as sabermetrics did with Derek Jeter—should galvanize the movement. But Canadian nationalism, combined with the inconvenient fact that the Weber/Suter partnership had previously validated the numbers, has made the Weber case a divisive one. 

I have an answer: Weber’s play doesn’t facilitate efficient zone exits, leading to prolonged defensive zone shifts that sink his Corsi rating.

This thesis, of course, raises some more interesting meta-questions that are central to this column: Does this mean Shea Weber isn’t actually good? What does it say about the Predators' strategy? What does it say about Corsi’s ability to evaluate different phases of the game? These will all be touched on to varying extents in the course of breaking down Weber’s game.

The Salad Years

Weber, unlike past Eye Test subjects, has the advantage of two disparate bodies of work to compare. With Suter as his partner, Weber logged four seasons of very good possession stats (Fenwick For, or FF) while often playing on poor possession teams:

year	Weber FF%	Team FF%
2008–09	53.0%	45.7%
2009–10	53.2%	50.8%
2010–11	50.1%	48.6%
2011–12	50.5%	44.6%

It’s worth noting that this season, Weber has a Fenwick of 50.3%, on par with what were widely considered his best years with Suter. Now, 50.3% on a team that is at 53.3% collectively is a far cry from 50.5% on a team that's a miserable 44.6% (see chart below). The gulf in Weber’s Fenwick percentages partly reflects the difference between Nashville’s second pairing in 2011–12, which was led by possession-stat sinkhole Kevin Klein, and this season’s second pairing, which is led by Ryan Ellis, a great puck-mover. But the bottom line remains: Elite defensemen typically play well above the level of their team. With Suter, Weber did just that. Without Suter, Weber has mostly just broken even:

year	Weber FF%	Team FF%
2012–13	47.8%	47.1%
2013–14	50.0%	50.0%
2014–15	50.3%	53.3%

What made Weber and Suter great partners was their ability to effectively keep forwards to the outside, then break up possession before the offensive cycle could get somewhere dangerous. Watch this shift from the 2011 playoffs against the Canucks. The Predators’ top pairing bottled up Daniel and Henrik Sedin all series thanks to play like this. 

The sequence starts off a Weber turnover to Daniel Sedin behind the net. Weber and Suter are both fast and each grab a man before taking turns playing off the other. First, Suter intercepts the puck on a turnover forced by Weber, and then Weber knocks the puck loose from a player pinned by Suter:

Ultimately, Suter proffers the final whooping and initiates the zone exit:

Both Suter and Weber can be physically punishing and are extremely athletic. They played off each other extremely well, anticipating the exact moment to provide support. But Suter, crucially, is the one who got the puck going in the other direction.

Weber provided the physicality, Suter the skating. Weber shot hard, Suter passed well. These common summations of their chemistry were oversimplifications—Weber can pass fine, Suter is a physical force—but rang fundamentally true. Weber, the more imposing of the two, may have forced pucks free from opponents at a higher rate than his partner, but he also almost always relied on Suter to move the puck, regardless of which of them had done the job of getting it back. 

We can quantify this tendency now, thanks to the meticulous zone entry/exit data compiled by Corey Sznajder last year. Among the defenseman who are generally considered elite, notice the only one who converted defensive zone touches to controlled exits at a rate of less than 20% last season:

Player	Carry %	Pass %	Possession %
Erik Karlsson	13%	19%	32%
Alex Pietrangelo	6%	22%	28%
Ryan Suter	4%	23%	27%
Duncan Keith	2%	24%	26%
P.K. Subban	7%	19%	25%
Drew Doughty	7%	16%	24%
Mark Giordano	3%	18%	21%
Zdeno Chara	2%	19%	21%
Shea Weber	2%	16%	17%

Funny enough, Roman Josi, Weber’s current partner, ranked above Suter, with 28% of touches converted to exits with possession. So to simply say that Weber requires a puck-moving partner and no longer has one would be wrong. The problems go deeper than that. 

The Post-Suter Era

To test my suspicion that the Weber/Josi partnership fails in initiating exits, not executing them, I watched a handful of Nashville games, both from this season and from Weber’s years with Suter.

Weber logs a ton of TOI for the Predators, so he plays with all the team’s forwards. But on a disproportionate number of the shifts he plays behind the Paul Gaustad-anchored fourth line, since Gaustad and Weber are integral to coach Peter Laviolette’s extreme zone matching tendencies. (Gaustad leads the league with 66% of shifts started in the defensive zone. His 7% started in the offensive zone is, unsurprisingly, dead last.) 

Based on my own observations, these defensive-zone clean-up assignments seemed to be where Weber’s possession stats were getting hammered the most. So I got out the stopwatch and compared how long it took Weber/Suter to get the puck out of the defensive zone versus Weber/Josi. I watched about 20 defensive zone starts for each pairing, trying to focus on games against teams such as the Canucks and the Blackhawks (i.e., teams that have had the same top lines for several years), in which Weber was drawing roughly the same assignments. 

Plays when the other team had established possession stood out. When Nashville won a face-off outright, or quickly gained possession, the Preds had no trouble exiting with either pairing. But when the shift dragged on for Weber/Josi, it really dragged on, whereas Weber and Suter were usually able to stop the bleeding. The cut-off point seemed to be 10 seconds. If a shift exceeded 10 seconds, Weber/Suter always seemed to be one move from a solution. But after 10 seconds for Weber and Josi, goalie Pekka Rinne needed to be on red alert:

Pairing	Avg. Zone Time <10 sec.	Avg. Zone Time >10 sec.
Weber/Suter	5.38 seconds	18.0 seconds
Weber/Josi	5.96 seconds	28.4 seconds

Since we looked at Weber and Suter against Vancouver, here’s Weber and Josi in the defensive zone last season against the Canucks (albeit not the Sedins):

While Weber bodies his man behind the goal line, Josi tentatively moves to engage Radim Vrbata. I got the sense watching Josi that he was more concerned about staying in between his man and the net than about puck retrieval. But in the eternal balance of not making a mistake versus making a play, fancy stats always seem to favor the player who takes a risk.

Vrbata sends the puck low before Josi can get to him, but Weber makes an expert read, cutting the play off on the boards. He then seems to panic a bit and sends the puck right back up the boards, where Josi is no longer standing, having already retreated back to defend the slot: 

We can glean from Sznjader's data that Weber is not a turnover machine (pretty much every guy on the above “elite defensemen” chart had a turnover rate that hovered around 6%). But he does love to ream the puck up the boards. And while he may not be technically creating turnovers on plays where he defers moving the puck to Josi (or whoever), he is increasing the number of touches until the zone exit, creating a greater chance that someone else will screw up along the way. 

By the way, in case you’re wondering how this shift ended: Weber was penalized for brutalizing a Vancouver forward that Josi was unable to muscle away from the goal mouth. 

The Parts Vs. The Whole

Weber passes the eye test for so many fans because he is the prototype for the modern defenseman. Unusually athletic for someone of his size and strength, he rarely gets beat in his own zone, and when he lands a hit, he never gets cheated. Those traits alone would make for a great blueliner, but add in the 105-mph slap shot and you've basically got a less-gangly Chara with a bigger mean streak.

Yet Weber’s penchant for crushing forwards doesn’t lead to the puck going up ice for a few reasons:

1. Deference: He hands the puck off to Josi, even when he might be in better position to move it. 
2. Refusal to carry: He often prefers to shoot the puck along the boards in both the neutral and defensive zones.

And Josi, despite being maybe an even better skater with the puck than Suter, exacerbates Weber's shortcoming because:

1. Lack of anticipation: Weber and Suter had an innate awareness of where the other was on the ice. This was particularly important for Weber since he could us his size to pin a player while waiting for Suter to skate by and take the puck up ice. Josi tends to hang back, which would work better if Weber was the kind of player who more readily handled his own exits.
2. Fewer overall touches: Skating ability doesn’t mean much if you don't have the puck on your stick. Suter didn’t need Weber to get it back—he could body check opponents on his own, leading to plays like in the second Sedin GIF above.

So what to make of a player like Weber? Predators fans—myself included—have never felt anything but confident with him on the ice. He doesn’t commit a mental miscues or get out-muscled by enemy forwards. Is it better to have a player more likely to make a defensive mistake but also more likely to move the puck up ice? The answer from the Corsi and Fenwick stats would seem to be yes. Does that mean that Nashville would be better off putting Ellis and Mattias Ekholm out to protect a one-goal lead with the opponent’s net empty? I suspect this suggestion would make every single NHL head coach and general manager roll their eyes. 

Should we count it against Weber that he chooses not to do something (skate with the puck) that he has the tools to do? Or, put another way, should we count it against Weber that he developed in concert with a complementary partner and the habits that he formed then are failing him now? 

I’ll stop asking questions now and tell you some things that I believe about Shea Weber, with the caveat that I am a longtime fan:

1. Given the choice between Weber and Suter, a team would be better off choosing Suter. The Wild defenseman is a low-maintenance puck-mover who drives possession and is more capable of creating chemistry with other partners (although Suter’s stats also seriously suffered after he left Nashville). I think I said as much when they were both impending free agents in 2011–12 (though I probably changed my mind the next week).
2. When Weber and Suter were paired together, Weber was the more valuable player. That may seem to contradict belief No. 1, but Weber simply didn’t make the same mistakes when he had Suter to play off of, which is different than Suter cleaning up after Weber’s errors or carrying Weber.
3. Weber is a better, more valuable player than Josi and a bunch of other players in their Corsi bracket. Replace Josi with Ekholm or Seth Jones and Weber would get on the same, maybe a little bit better since both players are bigger than Josi. Replace Weber with anyone else on the Preds’ roster and Josi would constantly get run around the defensive zone. Just look at Josi’s abysmal possession stats without Weber. 
4. Weber is uniquely more valuable to a good team than a bad team. Every defenseman’s job, late in a game in which their team is leading, is to prevent goals, not to drive possession. Teams retreat into defensive shells, especially good teams who trust their goalies. Nashville is both, and Weber is exactly the guy you want in those situations. 
5. Pietrangelo, Doughty, et al., are better, but it’s closer than you might think. Moving the puck is a huge thing, especially these days with the widespread emphasis on puck possession. But given Weber’s value on special teams and in late game situations, and considering the mostly pedestrian forwards he’s had to play with through the years, I don’t rate an elite defenseman like Duncan Keith much better.

Eye Test: Cody Franson's game is ugly, but his pretty stats don't lie

At worst, Weber is the best defenseman at playing away from the puck in a league that increasingly values players who play with the puck. He can hurtle the little disk 105-mph on a one-timer, knock it loose with a bone-crushing hit and block it with his rugged frame. But ask him to carry it onto his stick for more than five seconds and his attitude becomes, No ... how about you take this.

But crucially, that’s not a failure of talent, effort or hockey IQ. He simply doesn’t do something—something everyone knows he can do—often enough. So I’m left with what has become a familiar refrain of the Eye Test: The stats aren’t wrong, and yet everything they say about the player is true. Whether you think a defenseman is not elite because his stats can vary so wildly based on who he is partnered with, or because he’s a victim of his own habits, is academic. As someone emotionally invested in the the Predators’ success, Weber is still the guy I want on the ice in the last minute.