Do SHL teams that keep the same players win more?

Modestly. Across 52 team-seasons (2021/22–2024/25), returning-core share correlates 0.32 with points percentage. Cores that returned ≥70% of regulars averaged 0.361 pts%; 55–70% returners averaged 0.322.

Does an older roster win more?

No. Mean core age versus points share is essentially zero (−0.05). SHL cores in this window sit around 26 years old either way.

Is the head coach or the sporting director the key?

The lake has coaches, not GMs. After 19 in-sample coaching changes, points share barely moves (−0.011) versus keeping the coach (−0.005). That is a club-persistence story, not a 20-year director ranking.

Who does the model pick for SHL 2026/27 top four?

Frölunda, Växjö, Skellefteå, Rögle — last year’s table plus coach continuity. Prior, not a lock. Björklöven is promoted and unscored.

SHL clubs beat coaches: returning cores win, age does not, and 2026/27 still looks like last year’s top four

The hiring myth is that the name behind the bench decides the table. The fan myth is that a veteran core that has been together forever is either a dynasty machine or a tired group that needs a tear-down.

We can measure both — in the SHL window the lake actually owns.

The short version

52 team-seasons, SHL regular 2021/22–2024/25. Core = skaters with ≥10 games, keyed by name + birthdate.
Returning-core share vs points%: r = 0.32. Cores that kept ≥70% of regulars sat at 0.361 pts%; the 55–70% band sat at 0.322.
Mean age vs points%: r = −0.05. Age is not the lever.
19 coaching changes vs 29 keeps: delta −0.011 vs −0.005. The club’s level mostly stays. That is the honest coach-vs-organisation test we can run without a GM table.
2026/27 prior (not a lock): regular-season top four Frölunda, Växjö, Skellefteå, Rögle. Same four as the playoff final-four prior. Björklöven unscored.

Returning cores

Returning share	Team-seasons	Mean pts%
≥70%	18	0.361
55–70%	31	0.322
<55%	3	0.331

That is roughly six extra points over a 52-game season for the sticky cores versus the middle band — not a dynasty law, not noise either. Frölunda’s 2024/25 core returned 78.6% and posted 0.503 pts%. HV71 returned 57.9% and posted 0.210.

Age is a wash

Mean core age in the sample is 26.4. Under-26 cores (0.339 pts%) and 26–28 cores (0.336) are the same team in different jackets. We do not have a “too old” SHL in this window. Almost nobody’s core mean sits over 28.

Coaches: tenure is weak, replacement is weaker

Head-coach tenure vs points%: r = 0.16. First-year coaches (0.335) are not a disaster; 3+ years (0.354, n=15) are a little better; one five-year run (Thomas Berglund, Luleå 2024/25) is not a sample.

The cleaner cut is keep vs change:

	n	Mean Δ pts%	Mean pts%
Same head coach	29	−0.005	0.341
New head coach	19	−0.011	0.337

Replacing the coach is not a reset button. It is also not a poison pill. The organisation’s prior level is the thing that shows up again. That is as close as this lake gets to “is it the coach or the sporting leadership?” — and the answer is mostly the club.

Last year, from the standings table

shl_analytics 2025/26 grundserie (52 games). This is the SoT for last season’s table — not a web guide that puts Skellefteå first in the regular season.

Club	Pts	Pts%
Frölunda HC	89	0.571
Växjö Lakers HC	81	0.519
Skellefteå AIK	77	0.494
Rögle BK	75	0.481
Luleå HF	68	0.436
Färjestad BK	68	0.436

Playoff games in the same table: Rögle 17, Skellefteå 15, Luleå 14, Växjö 10. HockeyNews records Skellefteå as SM gold and Rögle as the finalist — journalistic, not a series-winner field in the lake.

2026/27 prior

Score = 2025/26 pts% + a small continuity bump from 2024/25 returning share − 0.015 if the club hired a new head coach (HockeyNews, 9 June 2026). Leksand is relegated; Björklöven is promoted and has no SHL prior here.

Regular-season top four: Frölunda (Robert Ohlsson) · Växjö (Björn Hellkvist) · Skellefteå (Daniel Hermansson) · Rögle (Dan Tangnes).

All four keep last year’s bench. The five new coaches sit at clubs that were 7th–13th: Luleå (Dennis Hall), Örebro (Martin Filander), Linköping (Ante Karlsson), Timrå (Mikael Gath), HV71 (Niklas Eriksson).

Playoff final four prior: the same four. Last spring Luleå outplayed that ranking in May; this model haircuts them for a new coach and does not pretend Omark’s retirement is in the spreadsheet.

This is a prior, not a lock. Summer roster moves after 2024/25 are only partly in the returning-share term. 2026/27 games have not been played.