Story Archive

Alex Ovechkin starts his final season on 929 goals. He was still a top-25 five-on-five scorer last season, but Washington has cut him to 13 minutes a night, and the minutes decide whether he reaches 950.

Dylan Larkin was booed, then scored in his season debut. Five straight 30-goal seasons make the case for Detroit's price. His five-on-five numbers explain why nobody has paid it.

Barkov is out six to eight weeks and Marchand was helped off in Los Angeles. Florida already played a season without its captain, and it says Sam Reinhart belongs on the wing.

The Rangers are 4-1 and Alexis Lafreniere has a hat trick. Their goaltending leads the league, but at five-on-five they are almost the same team that finished last in the East.

Vancouver has scored 19 goals in four games, the most in the NHL, on 10 expected. Paul Cotter has four on seven shots. The chances still rank near the bottom of the league.

Knies and Marchenko finished one point apart last season. At 5-on-5 one sat in the 90th percentile for shot attempts and the other in the 10th. Here is what each club bought.

We rebuilt the standings on 27 November across nine seasons. Teams in a spot made it 78 per cent of the time, but the gap from eighth tells you far more than the line.

Florida spoiled Carolina's banner night 1-0 in overtime. The regular-season rivalry is even. The playoff one is 8-1, and Carolina's Cup came the spring Florida missed.

Buffalo lost its opener 6-3 and the market expects a 12-point drop from 109. The opener means nothing. Alex Tuch's departure with his finishing means a lot.

Utah's plus-28 goal differential was worth about 100 points last season. Dylan Guenther's 40 goals came on 27.6 expected. One claim holds up and the other is a stretch.

Pucky weighted every rostered player's age by the minutes he played, then checked it against the table. The correlation with points is 0.11. The useful number turns out to be a different one entirely.

Charlie McAvoy is serving six games in October for a slash delivered on 1 May. The carry-over rule is defensible. What it does to the punishment, and to Boston's 24th-ranked penalty kill, is a separate question.

Brock Faber had surgery on a fractured skull and is out indefinitely. Minnesota's fourth-ranked defence was built on two men playing 24-plus minutes. The honest accounting is stranger than the headlines.

Bouchard opened with a hat trick and five points and Edmonton still lost 6-5 after leading by three. The Oilers have been seventh or better in scoring in three of the last four years and were 25th in goals against last season. That is the whole story.

Los Angeles opened with an 8-4 loss in Denver and the 2028 cap-space theory took off. The cap maths is real. The plan was never reported. The warning was in last season table all along.

The Canucks finished 32nd for the first time since 1971-72, and they deserved to: 44 percent of the chances at 5-on-5, 31st in the league. But they were also the second-unluckiest team in hockey, 38.9 goals below expected, worth twelve to fifteen standings points. The summer's moves were right. The bounce back is the problem.

Detroit missed the playoffs for a tenth straight season, four points short, and the only real flaw was finishing: 19 goals below expected at 5-on-5, 29th in the NHL, concentrated in Marco Kasper, Andrew Copp and Emmitt Finnie. That corrects itself. Instead the summer brought a new GM, a captain's trade request and an unsigned Simon Edvinsson.

Ilya Sorokin saved 25 goals above expected, fourth in the NHL, and the Islanders still missed the playoffs. At 5-on-5 they finished 31st of 32 in finishing, and most of that gap belonged to one player: captain Anders Lee, nine goals on 21 expected. He signed in Utah. Every replacement is on a one-year deal.

Dallas finished third overall with 112 points and created chances at even strength like a wild-card team: 2.36 expected goals per 60, 22nd in the NHL and last among the eight 100-point teams. Then the cap forced it to sell Mavrik Bourque, one of its few even-strength drivers, for picks.

Tampa was the fifth-best team in hockey by chance share and lost a Game 7 to Montreal, a 48 percent team with the best finishing luck in the NHL. Then it let Darren Raddysh sign an eight-year deal in Toronto and replaced him with 36-year-old John Carlson at the same cap hit. Kucherov has one year left.

Pittsburgh's goaltending was average last season only because Stuart Skinner's plus-8.3 cancelled out Arturs Silovs' minus-11.9. Skinner signed in Winnipeg on July 1. The Penguins go into 2026-27 with Silovs and a 22-year-old, an old core, and 98 points that rested on 20 goals of finishing luck. The numbers on what that costs.

Columbus was a 52 percent team with almost no luck either way, 92 points and out of the playoffs. It let its captain Boone Jenner and Mason Marchment walk and paid a second, a third and a fifth for Valeri Nichushkin, a 59 percent play-driver whose price was set by a 10.6 percent shooting year. The trade is right. The summer is not finished until Jet Greaves and Adam Fantilli are signed.

Seattle generated chances at the worst rate in the NHL last season. In one summer it paid a late first and a second for Mackie Samoskevich, a 59 percent play-driver who shot 7.5 percent, and gave six years at $5.75 million to Bobby McMann, a 45 percent winger who shot 15.8. The numbers say it got one of those exactly right.

Anaheim was the third-best chance-creating team in the NHL last season and one of the three unluckiest. Then it matched an $18 million offer sheet for Leo Carlsson, paid Mintyukov $7.2 million, and moved four of its top six defencemen, plus McTavish, to make the cap work. The chances were real. The blue line that produced them is gone.

Boston was a 47 percent team at 5-on-5, 20th in chances created and 28th in chances allowed, and won 100 points on 51 goals of finishing and goaltending luck, the most in the NHL by twelve. Its summer answer was two first-round picks for JJ Peterka, a 25-goal winger who replaces the 25-goal winger it let walk.

The Oilers were a 52 percent team at 5-on-5, fourth in the NHL in chances, and were outscored by 13 because their goaltending was 18 goals below expected. They traded the goalie who finished plus-8, let the one who finished plus-2 walk, and go into 2026-27 with Tristan Jarry (minus-4, four straight years of decline), a 37-year-old Frederik Andersen off an .874, and Devon Levi.

Washington missed the playoffs by three points with Logan Thompson saving 29 goals above expected, the best in the league, in front of a defence that allowed chances at the 25th-best rate. It spent the summer on wingers: $84 million through 2034 for Alex Tuch, whose plus-25 was plus-3 by chances, and a cheaper trade for Jordan Kyrou, who actually fixes the problem.

The Devils scored 136 goals on 170 expected at 5-on-5, minus-34, the worst finishing luck in the NHL by 13 goals. Then they signed Anthony Mantha, 33 goals on 18.3 expected, and paid a first and a second for Luke Evangelista, 12 goals on 19.4. Both will regress. Only one of them was priced right.

Pavel Dorofeyev scored 37 goals for Vegas. Nineteen came on the power play, tied for the highest share of any 33-goal scorer, and at 5-on-5 he scored at the same rate as Mika Zibanejad, who the Rangers already pay $8.5 million. New York's problem was even strength. It spent two firsts and $11 million a year on the one thing it was good at.

The Blues created 222 expected goals last season, 31st of 32. Their summer was one trade: Jordan Kyrou and a late first out, Mason McTavish and Connor McMichael in. McTavish generated more chances per minute than any Blue and gave up more than any Duck. Which number travels decides the deal.

At 5-on-5 last season Brady Tkachuk and William Eklund each scored 11 goals and 31 points. Tkachuk did it in 373 fewer minutes while Ottawa owned 62 percent of the ice. Eklund was a 49 percent player on a 48 percent team. The Senators sold their identity and bought a fifth-round goalie.

Florida's bounce-back consensus rests on Barkov, both Tkachuks and a crease it says it fixed. The offence will come back. But the Panthers ranked 27th in slot shots against with the same blue line, and they replaced the 58th-ranked goaltender in the league with the 57th.

Toronto allowed 295 goals and answered with $21 million for a 38-year-old coming off .877 and 12.2 goals below expected, while trading away its only goaltender above water. The 295 were built on 285 expected, second-worst in the league. The fix is the coach, not the crease.

Hellebuyck went public with his trade request and Buffalo is the name attached. But he saved 5.5 goals above expected last season while Luukkonen saved 10.5 and Lyon 12.6. The Sabres would be paying a fortune to downgrade.

Ten teams will open camp without a captain, believed to be the most in league history. The vacancies map almost exactly onto franchises between eras, and the candidates, Carlsson, Celebrini, Werenski, show who the next decade belongs to.

Carolina is advancing trade talks on Alexander Nikishin, a 24-year-old who just played 81 games for a champion. The reason it can: the Hurricanes controlled 57 percent with him on the ice and 56 without him.

Victor Hedman played 33 games last season after an extended absence on personal leave. Tampa still posted a 54 percent share and plus-57 differential. The unresolved question is what the blue line looks like across 84 games.

Every preview names goaltending as Colorado's weakness. Scott Wedgewood saved 23.1 goals above expected, fifth best in the league, with Blackwood adding 8.3. On a 57 percent team with a plus-101 differential, the narrative is simply wrong.

Calgary has built its rebuild on the premise that goaltending is settled. Devin Cooley saved 17.3 goals above expected in 31 games; Dustin Wolf was 1.8 below in 57. On a team that scored 208 goals, that gap is the whole margin.

Ottawa controlled 55 percent of the expected goals and outscored opponents by 30, with four forwards driving play at elite rates. Drake Batherson's 33 goals and 62 percent share show the Senators are not a sleeper anymore. The crease is still the ceiling.