What happens when a team disinterested in playing defence faces one that has mastered the art?
That’s what we appear to have out west in one of the more interesting first-round series between the San Jose Sharks and Vegas Golden Knights. The series has been perplexing, largely because San Jose appears to be playing a different sport than their Pacific Division rival.
After Game 2 I re-watched every shot sequence against San Jose and there were two common themes: (a) complete indifference to disciplined, off-the-puck play; and (b) fatally flawed goaltending that is incapable of bailing the team out of trouble.
The goaltending issue is a lost cause at this point. Martin Jones is a complete mess and Aaron Dell isn’t any better. Whether or not San Jose wins this series, you have to imagine that this off-season will be spent trying to repair the position. It’s not tenable even in the short term, and Jones in particular has had enough bad sequences to last a lifetime. (San Jose won’t be able to bury his contract, but they should be able to find someone to shoulder some or most of the burden going forward.)
Fixing the rest of the roster is another challenge altogether. This is a very talented and very capable San Jose Sharks lineup. This regular season they finished with 101 points (6th best in the NHL), thanks in large part to the league’s second best offence. But the issue that the Sharks had in the regular season has manifested in a big way this postseason. Everything falls apart when they don’t own the puck in the offensive zone and are forced to play stretches of hockey off the puck.
That’s what’s happening in this series. Vegas is one of the few teams that can skate with San Jose, and when the Golden Knights have had opportunities, they’ve made San Jose look silly.
Compounding the issue is the fact that the Golden Knights have been fantastic defensively and are backstopped by a much more capable goaltender in Marc-Andre Fleury.
The crazy part of this series is that the run of play on the surface seems relatively fair. San Jose and Vegas have traded shots most of the series – the Sharks are carrying 51.6 per cent of the shot share, but have also been trailing in the majority of these minutes. Adjust for score effects and the Sharks are carrying about 49 per cent of play – essentially dead even.
A 50 per cent team in shot share will generally carry 50 per cent of the goals over longer periods of hockey. The two components that can change the translation between shot share and goal share are big variances in shooting talent or big variances in goaltending talent. A team with an elite power play (like Washington) should convert on marginally more opportunities than their competitors. A team with an elite goaltender (like Dallas and Ben Bishop) should be able to thwart marginally more scoring opportunities against. But at the end of the day, the relationship between shot share and goal share remains steadfast.
The plot below emphasizes that point. Even across tiny four-game intervals, there is a fairly tight relationship between shot shares and expected goals (shot shares adjusted for the quality of those shots, including angles, types, distances, et al.). In other words: winning the shot advantage means winning the scoring chance advantage, which usually means outscoring your opponent:
That brings us back to San Jose. They have traded shots with Vegas and if we only knew that data point, we might infer that their record (1-3) and goal differential (-7) through four games was poor luck mixed with possibly shoddy goaltending. But that’s not the case. When you start adjusting for shot quality, the narrative changes considerably:
Again, this is a fairly rare occurrence – shot shares and expected goals tend to line up quickly over small samples and over longer samples the goals do as well. In San Jose’s case one of two things must be going on: they are taking a lot of lower-quality shots which is polluting the sample, or they are conceding a lot of quality shots against to pollute the sample, or both.
So let’s look at that. We know how many shots each team has taken and faced, and we know how to apply the appropriate quality adjustments on those shots. To that end, what has each shot been worth in terms of marginal expected goals for each team in the playoffs?
San Jose has been bad defensively – they are leaking more shots against than most teams, and those shots tend to be of a higher quality – but it’s not as if they are alone. Boston and Winnipeg are conceding a ton of shots against. Carolina is giving up very few shots, but when they do they tend to come from extremely dangerous scoring opportunities. All in, we would say San Jose is about the fourth-worst defensive team in the first round – and again, this being completely blind to the goaltending talent behind the skaters.
Differential is a different story. Consider Winnipeg for a moment: The Jets would expect to concede about 3.2 goals against per 60 minutes based on the way their playing. Their opponent in St. Louis is at 2.8 goals against per 60 minutes. That’s pretty close, and one of the big reasons that series is knotted up at two apiece.
San Jose and Vegas couldn’t be more different. Since we know the shots each team has faced and the expected number of goals each team should have given up, we can calculate expected save percentages – the point being that it is blind to goaltender performance. Here is how that looks, with Vegas and San Jose highlighted:
Let’s try and summarize the issues here. San Jose is bleeding tons of quality chances against. On top of that, they have a goaltending tandem that’s incapable of stopping those shots. Their opponent in Vegas has suppressed shot quality like no other team has in the postseason, at least on a per-shot basis. When things get dire, Fleury routinely bails the Golden Knights out.
It’s hard to explain what’s more perplexing – San Jose’s inability (outside of game one, I suppose) to pierce Vegas’ interior, or San Jose’s indifference when it comes to defensive zone play. Either way, it’s created quite an unexpected mismatch.
The Sharks no longer have room for error. It’s win or go home the rest of the way. Based on this analysis, I don’t like their chances.