Columbus Blue Jackets

What the Quant Says

262 goals, 14th in the league, off 2.484 expected goals per 60 at even strength — 20th in hockey.

At five-on-five it scored 177 against an expectation of 158.5. Eighteen and a half goals of finishing gained, 5th in the league. A club that generated chances at a below-average rate converted them at a well above-average one, and the 14th-place goal total is the product.

Patrik Laine had 26 goals in 56 games at 15.6 per cent, off a 101.5 mph shot — the 2nd-hardest of the 365 forwards the tracking covers — taken from a median of 36 feet with 22 per cent of his attempts from the slot, the lowest share of any qualified forward. Boone Jenner had 23 goals in 59 games from a median of 22 feet at 95.9 mph.

Jack Roslovic had 22 goals at 16.8 per cent, the 90th percentile, from a median of 31 feet at the 91st. Jakub Voracek had 56 assists and six goals, at 4.1 per cent shooting — the 11th-lowest of the 393 qualified forwards.

Cole Sillinger scored 16 goals in 79 games at eighteen years old. Egor Chinakhov scored 7 in 62 at twenty-one, off a 97.2 mph shot taken from a median of 35 feet.

Johnny Gaudreau arrives with 40 goals, 75 assists and a plus-64. His 4.23 even-strength points per 60 is the highest figure of any forward in this book, and he is joining the club that finished 31st in chance share.

What the Scout Says

Gaudreau is the arrival and he is genuinely the real thing: 5-9, 163 pounds, and the highest even-strength scoring rate in the league. He holds the puck until something opens — the giveaway rate is at the 97th percentile and it is the price of the vision, not a flaw beside it. He does not hit anybody and does not need to.

Laine is the contradiction. The shot is the second-hardest in the league among forwards and he takes it from further out than almost anyone, which works because the release is instant and does not work because his ice is being outchanced sixty-forty. He is a large man who is difficult to find in his own end.

Jenner is the counterweight: hard from close, 95.9 mph from twenty-two feet, blocks shots at the 89th percentile among forwards, and neither fast nor a heavy skater at twenty and a half minutes a night. Voracek passes and no longer shoots — six goals in 79 games at 4.1 per cent, and 56 assists.

Roslovic is speed at the 90th percentile who is never fouled, which is what going around the outside looks like in the data. Sillinger works and is not quick. Chinakhov has the shot and started every shift in his own end. Danforth is 5-8, hits at the 85th percentile and never turns the puck over.

What the Quant Says

300 goals against, 28th in the league — off 3.080 expected goals against per 60 at even strength, which is more than any other club gave up.

At five-on-five it conceded 214 against an expectation of 195.3. Nineteen goals worse than chances that were already the worst in the league, which is the goaltending, and it is the compounding problem on this page.

Zach Werenski took 45.6 per cent of the expected goals across 1,344 minutes at 25:40 a night, the 97th percentile in ice time. Vladislav Gavrikov took 45.4 per cent across 1,495 — more minutes than anyone here — with 3:09 a night shorthanded at the 97th percentile. Andrew Peeke took 43.7 per cent, the 7th percentile, and played all 82 games.

Not one Columbus defenceman finished above the 20th percentile in chance share. Werenski at the 15th is the best of them, and he is the club's best player.

Jake Bean took 42.7 per cent across 1,207 minutes with 1.79 hits per 60 — the 10th percentile — and an 87.1 mph shot at the 4th. Adam Boqvist scored 11 goals on 64 shots, a 17.2 per cent conversion that is the highest of the 202 qualified defencemen, and took 45.8 per cent of the chances.

Erik Gudbranson arrives from Calgary with a plus-15 and 53.8 per cent of the chances. Paired with Nikita Zadorov for 1,708 attempts that pair took 56.5 per cent; without Zadorov, Gudbranson was at 44.1 across 481. The 6-5 frame and the 708 career penalty minutes are what the club is actually buying.

What the Scout Says

This is the problem and every number says so. Not one Columbus defenceman finished above the 20th percentile in chance share, on the club that gave up more expected goals per 60 at even strength than anybody in the league.

Werenski is the exception in kind rather than degree. He is a genuine number one — 25:40, shots at the 197th of 202, points at the 90th percentile — and he defends entirely by leaving with the puck. 1.13 hits per 60 is the 2nd percentile. At twenty-five minutes a night against the other side's best, that is a great deal of escaping to have to do.

Gavrikov is the one man here who takes hard minutes and does not leak: 22:17 a night, the lowest offensive-zone share of any defenceman on the roster, and giveaways in the bottom fifth of the position. He fouls when beaten and 68 penalty minutes is the bill.

Peeke is a 6-3 shot-blocker with no offence at all — 2.45 shots per 60, the 8th-lowest among qualified defencemen — and he played the whole 82. Bean has a shot at the 4th percentile in speed and does not hit. Boqvist scores and cannot stay in the lineup.

Gudbranson is the addition, 6-5 and 222 with 708 career penalty minutes, and the with-and-without-you data says the good Calgary season belonged to the man beside him.

What the Quant Says

Columbus goaltending came in 18.7 goals below the chances it faced, 26th in the league.

That figure sits behind the highest rate of expected goals against in hockey, which is the caveat and does not dissolve the number: the arithmetic already accounts for the chances, and the crease was still worse than them.

Elvis Merzlikins went 27-23-7 with a .9073 and a 3.22 in 59 games, on 1,920 shots faced — the heaviest workload of his career — with 2 shutouts. His career line across 120 games is a .9132, and the .9227 he posted as a rookie remains the best of the three seasons by 15 points.

Joonas Korpisalo went 7-11-0 with a .877 and a 4.15 in 22 games. That is 25 points below his own career figure of .9016 across 182 games and the worst of the seven seasons he has played.

Jon Gillies arrives having gone 3-10-2 with a .8873 in 20 games for St. Louis and New Jersey — the most he has played in a season, nine years after being drafted. 6-6, and a career .8928 across 32 games.

Columbus spent its summer on offence and this is the number it did not address. A club that is 32nd at generating chances against it needs a crease that steals games, and the measured record says it has one that loses them.

What the Scout Says

Merzlikins is a reactive, athletic goaltender who plays deep and beats shots with his hands rather than by taking away the angle — the method that produced a .9227 as a rookie and a .9073 last season behind a much worse team. Nothing measured separates a goaltender being overwhelmed from a goaltender declining, and 1,920 shots is a large sample of the first and a poor test of the second.

Korpisalo is the opposite build and the opposite method: 6-4, upright, positional, covering the net by standing in it rather than by moving to it. That style survives a poor defence badly, because lateral puck movement asks the feet for work the reach then has to do. A .877 is what that looks like when it goes wrong.

Gillies is 6-6 and blocks. Thirty-two NHL games in nine years and a career .8928.

Behind the worst even-strength defensive record in the league, all three of these save percentages are partly a club figure wearing a name. The 18.7 already accounts for that, and it still says minus.

What the Quant Says

26th in the league at 5.754 expected goals per 60, over 4:33 a night.

Jakub Voracek ran it: 3:13 a night at the 90th percentile among forwards, with 40 per cent of his points coming on it — the 96th. Zach Werenski took 3:13 from the blue line, the 95th percentile among defencemen, with 29 per cent of his points on it.

Patrik Laine took 3:41 a night, the 94th percentile, and scored 5 power-play goals in 56 games. On a unit ranked 26th, a 101.5 mph one-timer from the circle is the single asset worth building around, and the unit was not built around it.

Boone Jenner scored 8 power-play goals in 59 games from the net front — more than any other Blue Jacket — and took 2:58 a night at the 85th percentile.

Johnny Gaudreau arrives with 2:57 a night and 6 power-play goals, and 22 per cent of his production came on the man advantage, the 63rd percentile: he is a five-on-five player who happens to be excellent on the power play, not the other way round.

Adam Boqvist took 1:56 a night, the 78th percentile among defencemen, and scored 3. The pieces for a top-ten unit are on the roster. They have not yet been on the ice together.

What the Scout Says

Voracek at the top, Laine in the circle, Werenski walking the line, Jenner in the paint — and 26th in the league.

The parts are better than the whole. Voracek's distribution is the 96th percentile in power-play share of points; Laine's one-timer is the second-hardest shot in the league; Werenski takes 3:13 a night from a defenceman who shoots from 43 feet rather than 55. That should be a top-ten unit and it was not.

Gaudreau arrives and does not obviously solve it, because he is not principally a power-play player — 22 per cent of his points came on the man advantage, which is ordinary. What he does solve is the entry, and a unit that ranked 26th at generating chances usually has a zone-entry problem before it has a personnel one.

Jenner is the net front and scored eight of them in 59 games. Boqvist is the second-unit quarterback. There is no obvious reason this should not be much better.

What the Quant Says

18th in the league at 7.012 expected goals against per 60, over 4:53 a night.

On a club that finished 32nd at even strength, a mid-table penalty kill is the one defensive number that is not embarrassing, and it is not an accident: Columbus put its two best defensive players on it and left them there.

Andrew Peeke killed 3:13 a night, the 98th percentile among defencemen, and Vladislav Gavrikov 3:09, the 97th. Between them that is more than six minutes a night of shorthanded defence from a pair that also took the hardest even-strength assignment.

Sean Kuraly killed 2:22, the 98th percentile among forwards, and Eric Robinson 2:19, the 97th. Kuraly recorded 11.79 hits per 60, the 93rd percentile, and Robinson posted the 93rd-percentile top speed among forwards at 23.4 mph — a heavy centre and a fast winger, which is the conventional pairing and it worked.

Gustav Nyquist killed 1:46 at the 84th percentile and scored 4 shorthanded points.

Erik Gudbranson arrives having killed 2:55 a night in Calgary, the 94th percentile among defencemen. He is the first genuine addition this unit has had, and it did not need one as badly as anything else here did.

What the Scout Says

The one thing on this club that works, and it works because two defencemen never come off it.

Peeke and Gavrikov take six minutes a night between them, both above the 97th percentile at the position. Peeke blocks everything and holds a gap; Gavrikov closes and finishes the check and occasionally takes the penalty. Neither of them is a puck-mover and on the kill that does not matter.

Kuraly and Robinson are the forward pair and they are opposites — a 6-2 centre hitting at the 93rd percentile and a 6-2 winger with a top speed at the 93rd. Robinson's speed is the more valuable of the two shorthanded, where the difference between a clear and an icing is one stride.

Nyquist takes the second unit and scored four shorthanded points doing it. Gudbranson arrives as a third defenceman who killed 2:55 a night in Calgary.

None of this is why the club conceded 300 goals. The kill was 18th; the even-strength defence was 32nd.