If you know hockey, you know that players have roles. They’re not all expected to generate scoring chances.

Sorry, what?

That’s funny. Scoring chances are kind of fundamental to the game, every player’s one and only task in the game is to get more of them than their opposition.

You know, you go on about numbers in hockey but the fact is you understand neither. The numbers above, they paint a pretty clear picture. I.e. Bolland played horribly in his designated “checking” role, and Malhotra should not see good players.

And you know what? It’s obvious from the icetime totals that McLellan knows Malhotra blows at the checking role, because Pavelski ended up seeing more of Toews and the outchancing rates between Toews and Pavelski were similar to those between Thornton and Bolland.

And obviously there’s no fucking way that Quennenville expects long term winning by just squaring off in their top 6, which is why he tried to move Toews off to Malhotra whenever he could. Now if Bolland had actually been neutralizing Thornton then Quennenville could have kept on letting Toews outchance Pavelski and thereby completely neutralize every worthwhile Sharks player.

It was pretty clear from the get-go that you wouldn’t know these things, I just would have expected that you keep your ignorance on HF Boards or something.

There’s no doubt that Bolland’s line got out-chanced against HTML. But that number in a vacuum isn’t particularly useful. Everything’s relative. HTML are going to out-chance 99% of the lines they face. What would be useful — and I’m not sure if the data is available for this — is knowing how many chances a hypothetical league average line would do against HTML. That number is what should be the baseline (as opposed to a net breakeven of zero chances for/against). For example, from your numbers above, if Thornton vs Bolland is +1.9 chances/15 but an average line is going to give up +2.5 chances/15 against Thornton, Bolland is almost certainly doing a good job.

There’s no way that McLellan would choose to use Malhotra against Toews when he has Thornton and Pavelski at his disposal. THe H2H icetimes bear that reality out.

So it’s not so much that Malhtora didn’t do his job, so much as it is that McLellan really didn’t want to put Malhotra out there against the Hawks’ best, but was forced to due to the Sharks’ lesser forward depth and Quennenville’s tactics.

In any case, this is being painted as some sort of “Bolland vs. Malhotra” battle, that couldn’t be farther from the truth. The choice was Bolland and Pavelski facing the other team’s best, and neither did their jobs. But Quennenville had the luxury of better forward depth, so he switched off Toews to Malhotra when he could, and obviously won that battle.

But he wouldn’t have had to do that in the first place, if Bolland had actually done his job and neutralized Thornton. In a similar vein, if Pavelski had been neutralizing Toews, then likely all of these numbers would look different – Quennenville would have likely tried the PVP matchup with Toews-Thornton instead and, McLellan likely would have accepted that since he’s been oing it all season.

Thornton vs. Bolland: +1.9 chances/15
Toews vs. Malhotra: +4.1 chances/15

First line vs. checking line in both instances. Chicago’s first line is generating surplus chances at more than double the rate as San Jose’s. Ergo, Bolland is doing his job better than Malhotra, and/or Thornton is doing his job worse than Toews.

Why are you comparing first line vs. checking line in both instances? In the Shark’s case, it was Pavelski’s line that got the bulk of TOI against Toews, in fact, nearly double that of Malhotra. By playing Bolland against Thornton, he got Toews against Pavelski, and …they didn’t do as well against Pavelski as Thornton did against Bolland.

So
Bolland allowed more chances to Thornton than the rest of the NHL allowed to Thornton during the regular season.
Bolland allowed more chances to Thornton than Pavelski did to Toews.

No matter how you try and slice this, Bolland did not “shut down Thornton”, except on the scoreboard, which as we all know is luck if you’re getting outchanced. Any other mental acrobatics you’d like to use to try to defend this “shut down” narrative?

stick to the new narrative that Thornton actually destroyed Bolland in the series

You do have experience with narratives.

I think it’s up for debate whether or not the strategy was working. The chances suggest that SJ made better use of their OZ time, but the Corsi numbers suggest that Chicago was generating more time in the OZ. Eventually those measures will come together but over a short sample I think it’s pretty hard to prioritize one over the other and say “that’s the way it is.” Toews (16) / Kane (16) were killing by Corsi in the first two games and Hossa (14) / Sharp (5) were doing well too. Bolland (-10) / Ladd (-6) had a tough time against Joe. This looks to me like the kind of thing Quenneville was expecting and suggest that Quenneville wasn’t crazy for sticking to the game plan. But whether you think it was effective or not, my main point is that I think it’s overwhelmingly likely that sheltering the other two lines by running Bolland at Thornton was Quenneville’s plan even though he probably knew that Bolland probably wouldn’t outchance in that scenario.

I know that Pahlsson’s group did in fact outscore their opposition in that playoff run though I don’t recall whether or not they outplayed them. After all, Dave Bolland is +6 in these playoffs. But even if they did, I think you’re crazy if you think that’s what Carlyle was expecting them to do. Instead, he was expecting them to do something less than get totally killed while providing significant cover for Teemu’s line and Getzlaf’s line. And believe me, coaches tried to get away from it. It wasn’t the easiest thing to do. They all broke up their top lines in much the same way the Sharks did once they went to Chicago. In my opinion it was clear that by the time Games 3 and 4 rolled around, McLellan was not happy with Bolland’s line playing all of this top offensive guys – or maybe he wasn’t happy with Hossa and Toews consistently getting the softer part of the lineup – and tried to do something about it.

I agree with that last bit. In this case the advanced stat is IMO our scoring chances, which it seems are popular to the verge of being accepted holus bolus as the new Jerusalem. I like them a lot but not quite that much. In this case they are obscuring the very obvious context that Thornton wasn’t scoring against Bolland so Q had no reason to switch away from it. From the perspective of the scoreboard, the match-up worked.

Now with a line like HTML you don’t “give” the puck to them, they just take it. They’re bigger and/or have better hands than anybody who plays against them. They’re going to own the puck at times, sometimes shifts on end, they’re just that good. So trying to limit the damage is an obvious strategy to me, shit, it’s pretty much all you can do. Keep them to the outside as much as you can, and let your goalie see the shot. But on those occasions that they turn it over, go for it.

The other thing that scoring chances doesn’t yet cover is that if the group that goes +3/-5 includes a clearcut breakaway among their 3 and doesn’t allow anything of the kind among the 5 going the other way, that the metric hasn’t really nailed it. Yet. I have high hopes for the method, but there are rough edges that need to be identified.

My view is somewhat neutral: I don’t accept that Bolland is quite the hero he’s being played up in the media, and the scoring chance data supports that position. Neither do I accept the other extreme – what I jokingly referred to above as the “new narrative” – that he was “killed” by Thornton. He wasn’t.