Celebrini scored 45 goals on 31.8 expected, and San Jose is betting on the gap

Take Macklin Celebrini's shot chart from last season. Sort every unblocked attempt by where it left his stick. Now count the ones the model files as high danger.

Twenty-four, out of 411. That is 5.8 percent, and it is the smallest share of attempts taken from high-danger ice of any San Jose forward who played 750 minutes at five-on-five. Alexander Wennberg, at the other end of the same eight, took 26 of 132.

Celebrini scored 45 goals anyway, and 115 points, the highest single-season total by any player in franchise history, ahead of Joe Thornton's 114 in 2006-07. MoneyPuck's model, fed the same 411 attempts, expected 31.8 goals. The 13.2 between those figures was the fourth-largest surplus of goals over expected goals among the 411 NHL forwards who played 40 games or more, behind Cole Caufield, Morgan Geekie and Anthony Mantha.

The record is real. The question underneath it is open, and San Jose gets one season to answer it at a discount.

The cheap year#

Celebrini plays 2026-27 on the final year of his entry-level contract at a cap hit of $975,000. The extension he signed on July 29 begins the season after, at $18,800,000. PuckPedia and CapWages carry the same figures, as does the club's own release.

That is more than nineteen times the price, one year apart, for the same player. Whatever San Jose gets out of a roster built around a cheap superstar, it gets this year or not at all.

One qualifier belongs beside it. The entry-level deal carries $3,500,000 in performance bonuses on top of the $975,000, per PuckPedia, and bonuses a player earns are charged to the cap. The cheap year is cheap up front and can be billed later.

Even with him, and sinking without him#

At five-on-five last season, 73 goals were scored with Celebrini on the ice and 56 against, a share of 56.6 percent. The chances underneath those same shifts came to 57.99 expected goals for and 59.02 against, which is 49.6 percent.

Now the same measure with him on the bench. MoneyPuck's file rounds both to whole numbers and carries 50 percent with Celebrini on the ice against 46 percent without him.

That is the club in one line. Its best player drags it to even. Nobody else drags it anywhere.

The three arguments against me, and the first one is real#

One. Elite finishers repeat. Caufield is on that leaderboard most years, and so is Nathan MacKinnon, two places below Celebrini on it. The top of a finishing list is not drawn fresh every October. Some of those men are better at scoring than the model can see.

Two. He is twenty. His rookie season was 25 goals on 236 shots across 70 games. Last season was 45 on 287 across 82. A player improving that fast is not obviously a player who got lucky once.

Three. The low-danger scoring is itself a skill. Celebrini scored 20 goals on 290 low-danger unblocked attempts, 6.9 percent. That was ninth-highest by low-danger conversion rate among the 123 NHL forwards who took 150 or more such attempts, against a median of 4.3 percent. Beating goaltenders from outside the slot is a scarce thing to be good at.

All three hold, and none of them makes the bet free. Talent or variance, the size of the surplus still has to repeat.

Best case, for San Jose#

This is the entry where the club's interest and the player's do not come apart. A twenty-year-old ascending toward the top of the league is good for him and good for them, and nothing in San Jose's season wants less of him.

So the club's best case is not another point record. It is the 46 moving.

That means Michael Misa's line and Will Smith's shifts giving back less than the first line wins, and the club no longer needing a fourth-in-the-league finishing season to break even at five-on-five. A team seven seasons removed from a playoff berth does not get there by asking its best player for more. It gets there by needing him for less.

Worst case, for San Jose#

The surplus was percentage, and it goes.

If Celebrini finishes at the rate his shot locations support, the same 411 attempts are worth about 32 goals rather than 45. The on-ice goal share falls from 56.6 toward the 49.6 the chances actually supported. San Jose is then a club that was level at five-on-five with a franchise-record scorer on the ice, and at 46 percent the rest of the time.

The damage is not to him. He is twenty, signed through 2032, and a player who creates that many attempts will score plenty at any finishing rate. The damage is that the one season San Jose has him at $975,000 gets spent finding out.

What to watch in October#

The faceoff dot, and the zone it happens in. Celebrini won 625 of 1,285 draws last season, 48.6 percent, after 507 of 1,051 the year before, 48.2 percent. Two seasons, both under half. On last season's volume, eighteen more wins would have made it 50.0 percent, which is one extra draw won every four or five games.

The interesting half is where the deficit sits. In his own end he won 177 of 347, or 51.0 percent. In the offensive zone, 253 of 536, or 47.2. He is already above water where losing a draw costs the most, and under it where losing one costs the least. Box scores carry faceoff totals every night, and the league's game summaries split them by zone.

The shot mix. Twenty-four high-danger attempts out of 411 is what made last season a question. If that share climbs in October, the argument gets quieter, because the chances get better and the model stops disagreeing with the goal column.

The stakes#

The Athletic published a contention-cycle exercise on September 16 that rates every club on the present and the future, each on a scale to ten. San Jose came out at 1.7 present and 9.8 future, the highest future rating on its board and one of the two lowest presents, and Celebrini is the only Shark in its top tier on both lists. One player is carrying the whole distance between what this club is now and what it might become. Training camp opened today, September 17, per the club's own announcement, and the season count starts October 1 against Florida. It starts with a franchise-record scorer at a cap hit of $975,000, which is a thing that happens to this club once.

Previously in this series: No. 2, Will Smith.

Sources: MoneyPuck's 2025-26 season summary file for skaters, read 2026-09-17, for every shot-danger, expected-goal and on-ice figure here, including the eight-forward San Jose group, the 411-forward league group and the 123-forward low-danger group, each built from that file rather than inherited. The NHL's own statistics and player APIs, read 2026-09-17, for Celebrini's counting lines in both seasons and for his faceoff totals by zone, which agree with MoneyPuck's figure of 625 of 1,285 to the decimal. The San Jose Sharks' own site, read 2026-09-17, for the training camp announcement of September 16. PuckPedia and CapWages, in agreement, for the entry-level cap hit, the performance bonuses and the terms of the extension effective 2027-28. The Athletic, September 16, 2026, for the contention-cycle ratings and tiers.