How Hockey ChangedStandings and results
Does Scoring Depth Win Stanley Cups?
Vegas won a Stanley Cup with nobody over 28 goals. Florida won the next one with a 57-goal scorer. Across 1,758 team-seasons and 966 playoff series since 1917, how a team spreads its goals adds almost nothing to what its goals for and against already tell you, in October or in June. The deeper team won 474 of those series; goal differential alone predicted 469. For more than a century, Stanley Cup champions have come from both ends of the depth table.
By Edgehalla ResearchPublished 11 min read
On June 13, 2023, the Vegas Golden Knights beat the Florida Panthers 9-3 and won the Stanley Cup without a 30-goal scorer. Jonathan Marchessault had led them with 28. Only two teams in the league had spread their goals more evenly.
A year and 11 days later, Florida won the Cup. Sam Reinhart had scored 57, and only one team leaned harder on its top two scorers.
Hockey says depth wins in the spring, and Vegas fit. Florida won with a star. The two champions did share one thing, and it was goal differential. Florida’s was the league’s best, Vegas’s the ninth-best. Once you know how many goals a team scores and allows, how it spreads them adds almost nothing.
On this page
Hockey has always believed depth wins in the spring
Ask a coach. In April 2017, Alain Vigneault’s New York Rangers had just beaten Montreal, and NHL.com’s Dan Rosen wrote that the coach had brought up “the need to have four interchangeable lines, all with speed, skill and scoring, in order to win in the NHL, especially in the Stanley Cup Playoffs.” One of his forwards, J.T. Miller, put it more simply: “We don’t really have a 1, 2, 3 and not even a 1, 2, 3, 4. That’s definitely a weapon for us.”
Ask the losers. Leon Draisaitl’s Oilers went out to Vegas in 2023. Asked by the Associated Press that fall what stood out about the champions, he said: “They’re deep. They have four lines that they can roll against any line out there.”
Or ask the papers. In March 2023, with the Bruins on their way to a 65-win season, The Boston Globe’s Julian Benbow asked how scoring depth carries from the regular season into the playoffs. “Depth generally pays off — both in the regular season and the playoffs,” he wrote. Brad Marchand agreed: “That’s typically when most teams will make really good Cup runs.”
We tested it on every NHL team since the league began: 1,758 team-seasons from 1917-18 to 2025-26, with goals split by team for every traded player, and all 966 playoff series between NHL teams. We counted depth five different ways. None of them adds much to what a team’s goals for and against already say. Not in the regular season, not in the playoffs, not in any era. And not when we checked our guesses on seasons we had kept out while making them.
Once goals are known, depth is worth under a point a season
Across 1,758 team-seasons, a typical gap in depth is worth at most 0.6 standings points a season once goals for and against are known; in 966 series the deeper team won 474 against 469.3 predicted.
The deepest team ever came out of Boston
On April 8, 1978, in Boston’s next-to-last game of the season, the Bruins were up 2-1 in Toronto with the Maple Leafs’ net empty. Bob Miller scored into it at 19:54 of the third period. It was his 20th goal, and it made him Boston’s eleventh 20-goal scorer of the season.
“Bob Miller sealed the record in the penultimate game of the season when he notched his 20th goal with an empty-netter in the B’s 3-1 win over Toronto,” the Bruins’ own history of the era says. No team has matched it before or since. Three teams have had ten, the last of them the 1980-81 St. Louis Blues. Don Cherry’s team got 283 of its 333 goals from those eleven men, from Peter McNab’s 41 down to Miller’s 20 and Don Marcotte’s 20.
Their defenceman Brad Park, who scored 22 of those goals, explained his part to The Boston Globe’s Kevin Paul Dupont when the team was honoured 40 years later. “No. 1, when I shot the puck from the point, I never looked at the net,” he said. “Because if I couldn’t miss the first [defender], the net was irrelevant.”
No other team has had more than ten 20-goal scorers
Only three teams have come within one of Boston.
The most 20-goal scorers on one team
Every NHL team with nine or more 20-goal scorers in a regular season
The 1977-78 Bruins had 11 20-goal scorers; three teams had ten and 9 more had nine.
They finished 51-18-11 and went to the Stanley Cup Final for the second spring in a row. If depth wins anything, this was the team it should have won it for.
The stars’ share of the goals hasn’t moved since the 1940s
Boston’s eleven 20-goal scorers are one way to size up depth. Another is to add up a team’s two best scorers. When Vegas won its Cup, Jonathan Marchessault and Jack Eichel scored 20.6% of its goals. Across the league, that number has barely changed in 80 years. In the early NHL, when teams used 11 to 17 skaters a season, the top two scored about half of a team’s goals (49.5% in 1920-21). By 1940-41 it was 28.1%. Since 1942 the league average has stayed between 24.9% and 31.5% every season. It was 26.0% in 2025-26.
Since 1942, a team’s top two have scored 25% to 32% of its goals
Average share of a team’s regular-season goals scored by its two leading goal scorers, every season since 1917-18
Show data
| Season | Top-two share of team goals |
|---|---|
| 1917-18 | 50.9% |
| 1918-19 | 45.5% |
| 1919-20 | 45.5% |
| 1920-21 | 49.5% |
| 1921-22 | 48.5% |
| 1922-23 | 50.0% |
| 1923-24 | 48.4% |
| 1924-25 | 52.4% |
| 1925-26 | 48.8% |
| 1926-27 | 39.0% |
| 1927-28 | 42.7% |
| 1928-29 | 43.2% |
| 1929-30 | 38.8% |
| 1930-31 | 38.7% |
| 1931-32 | 33.4% |
| 1932-33 | 32.8% |
| 1933-34 | 33.8% |
| 1934-35 | 32.5% |
| 1935-36 | 32.9% |
| 1936-37 | 31.0% |
| 1937-38 | 29.8% |
| 1938-39 | 29.1% |
| 1939-40 | 27.1% |
| 1940-41 | 28.1% |
| 1941-42 | 26.8% |
| 1942-43 | 29.1% |
| 1943-44 | 30.1% |
| 1944-45 | 30.3% |
| 1945-46 | 28.6% |
| 1946-47 | 29.6% |
| 1947-48 | 29.5% |
| 1948-49 | 27.4% |
| 1949-50 | 28.0% |
| 1950-51 | 27.9% |
| 1951-52 | 29.5% |
| 1952-53 | 30.3% |
| 1953-54 | 28.9% |
| 1954-55 | 29.0% |
| 1955-56 | 30.8% |
| 1956-57 | 29.7% |
| 1957-58 | 29.0% |
| 1958-59 | 31.5% |
| 1959-60 | 28.2% |
| 1960-61 | 28.5% |
| 1961-62 | 28.1% |
| 1962-63 | 29.6% |
| 1963-64 | 28.4% |
| 1964-65 | 27.0% |
| 1965-66 | 28.6% |
| 1966-67 | 25.1% |
| 1967-68 | 26.9% |
| 1968-69 | 29.2% |
| 1969-70 | 27.5% |
| 1970-71 | 25.8% |
| 1971-72 | 28.4% |
| 1972-73 | 28.0% |
| 1973-74 | 27.2% |
| 1974-75 | 26.2% |
| 1975-76 | 26.5% |
| 1976-77 | 25.5% |
| 1977-78 | 26.4% |
| 1978-79 | 25.8% |
| 1979-80 | 27.3% |
| 1980-81 | 26.0% |
| 1981-82 | 25.9% |
| 1982-83 | 26.1% |
| 1983-84 | 25.7% |
| 1984-85 | 25.5% |
| 1985-86 | 24.9% |
| 1986-87 | 25.7% |
| 1987-88 | 27.6% |
| 1988-89 | 27.1% |
| 1989-90 | 26.9% |
| 1990-91 | 26.1% |
| 1991-92 | 25.9% |
| 1992-93 | 27.6% |
| 1993-94 | 27.2% |
| 1994-95 | 27.4% |
| 1995-96 | 27.2% |
| 1996-97 | 27.8% |
| 1997-98 | 26.9% |
| 1998-99 | 27.8% |
| 1999-00 | 26.3% |
| 2000-01 | 27.8% |
| 2001-02 | 27.4% |
| 2002-03 | 27.3% |
| 2003-04 | 25.3% |
| 2004-05 | – |
| 2005-06 | 27.1% |
| 2006-07 | 27.3% |
| 2007-08 | 28.1% |
| 2008-09 | 26.4% |
| 2009-10 | 25.6% |
| 2010-11 | 25.5% |
| 2011-12 | 27.4% |
| 2012-13 | 26.8% |
| 2013-14 | 25.5% |
| 2014-15 | 25.1% |
| 2015-16 | 26.2% |
| 2016-17 | 25.3% |
| 2017-18 | 25.4% |
| 2018-19 | 26.1% |
| 2019-20 | 26.0% |
| 2020-21 | 25.6% |
| 2021-22 | 26.6% |
| 2022-23 | 26.3% |
| 2023-24 | 26.5% |
| 2024-25 | 25.9% |
| 2025-26 | 26.0% |
- 1942-43: Original Six
- 1967-68: Expansion
The line is flat. The 20-goal count, the number people quote, swings. Teams averaged 6.1 20-goal scorers in 1980-81 and 3.4 in 2010-11, and 4.1 in 2025-26. Rosters didn’t get deeper and then shallower. Scoring went from 3.8 goals per team per game in 1980-81 to 2.7 in 2010-11, and the count of 20-goal scorers followed it. Boston’s eleven came in the high-scoring 1970s, too.
Two ways to count depth give opposite answers
Whether deep teams finish higher depends entirely on which of those two measures you pick.
Start with the top two. In 81 of 106 seasons, the teams leaning hardest on their top two had the better records. In a typical season, teams sit about 4 percentage points either side of the average: 30% of the goals from the top two, say, against an average of 26%. A team that much more top-heavy than average finished about two standings points higher over 82 games. (That is 1.5 points of points %, the share of available standings points a team takes, for a gap of one standard deviation.)
Now count 20-goal scorers, the way coverage usually does, and the answer reverses. A team one typical gap above average on that count finished about ten standings points higher (6 points of points %). Deep teams win, by a lot.
Goals sit behind both results. Twenty-goal scorers come from teams that score. Within a season, the teams with the most of them are mostly the teams with the most goals: a correlation of 0.68 with goals for, where 1 would mean the two lists line up exactly and 0 that they have nothing to do with each other. The top-two share tilts the other way for a similar reason. Stars are how teams get goals. Teams whose top two carry more of the load score slightly more than the rest.
Give two teams the same goals and depth stops mattering
The fair test, then, is two teams with the same goals. The 1999-2000 season has one. Ottawa and Florida each scored 244 goals. Ottawa allowed 210, Florida 209. Florida rode Pavel Bure, who scored 58, and Ray Whitney, who scored 29. Together they had 35.7% of the team’s goals, the most top-heavy split of the 28 teams that season. Ottawa’s top two, Marian Hossa and Shawn McEachern, had 29 each, 23.8%, the eighth-deepest split in the league. Florida finished with 98 points and Ottawa with 95. Both went out in the first round.
That is the test we ran on every team at once. Each team was set against teams that scored and allowed about as many goals as it did, with every season measured against its own league average. Then we asked what depth added. Under every way of counting depth, its value collapses. The most it is worth is about 0.6 standings points a season for a typical gap in depth (0.37 points of points %). A typical gap in goals scored, about 13% of a team’s total, is worth about nine standings points (6 points of points %). The crumbs left over still disagree. The share measures lean a hair toward top-heavy teams, the 20-goal count a hair toward deep ones.
Every measure of depth shrinks to almost nothing once goals are counted
Points of points % for a typical gap in depth (one standard deviation deeper), all 1,758 team-seasons 1917-18 to 2025-26, compared within each season. One point of points % is about 1.6 standings points over 82 games.
Comparing all teams, top-heavy teams win more by share (−1.48) and deep teams by 20-goal count (+6.05); comparing teams with the same goals for and against, all five are between −0.31 and +0.37.
- All teams compared
- Teams with the same goals for and against
The first bar in each pair compares all teams. The second compares teams with the same goals, and it barely leaves zero.
No era breaks the pattern. Split the century into six eras and, with goals held level, depth is never worth more than about two standings points a season either way (from −1.19 to +0.79 points of points %). The biggest numbers come from the smallest eras, and they switch between favouring deep teams and favouring top-heavy ones from one era to the next.
We also checked the goals-only guess on seasons it had never seen. We built it on the odd-numbered seasons and tried it on the even ones. Then we built it before the 1967 expansion and tried it after, and before the 2005 salary cap and after, and each the other way round. Going on goals alone, the guess misses a team’s final total by about five to eight standings points. Adding depth never shrinks that miss by more than 0.05 standings points (0.03 points of points %).
The most top-heavy teams got what their goals earned
If leaning on two stars were a weakness, it would show at the far end of the list. The most top-heavy team since expansion was the 1993-94 St. Louis Blues. Brett Hull scored 57, Brendan Shanahan 52, and together they had 40.4% of the team’s 270 goals. The Blues went 40-33-11. In the first round Dallas swept them, 5-3, 4-2, 5-4, 2-1, the last on April 24, 1994, in St. Louis.
That looks like the case against stars until you check the Blues’ goals. They finished about ten standings points better than their goals for and against predicted (6.4 points of points %). The trouble was at the other end of the ice. St. Louis allowed 283 goals.
The 1998-99 Ducks (Teemu Selanne 47, Paul Kariya 39, 40.0%) were second on the list and finished within about two standings points of their prediction (1.3 points of points %). Third were the 2018-19 Oilers: Leon Draisaitl 50, Connor McDavid 41, 39.7% of the goals, the most top-heavy team relative to its own season in NHL history. They missed the playoffs at .482, about one standings point from what 229 goals for and 271 against predicted.
The ten most top-heavy teams reached no finals, and their goals predicted fewer than one
Line up all ten next to what their goals predicted.
The ten most top-heavy teams since 1967-68 did, on average, a little better than their goals predicted
Points % above or below the league average: actual, and what goals for and against predicted
The ten most top-heavy teams finished +1.36 points of points % from what their goals predicted, on average; none reached a Stanley Cup Final, where goal differential predicted 0.6.
- Actual
- Predicted from goals
As a group, the ten most top-heavy teams since 1967-68 reached no finals and won no Cups. Their goal differentials predicted 0.6 finals and 0.3 Cups. Stretch it to 25 teams and it is still no finals against 1.7 predicted. Twenty-five teams that good would all miss the final about one time in five on luck alone. As a group they finished right around the league average, and most of them were average teams that happened to have two great scorers. The ten deepest teams by the same measure averaged about five standings points below the league average (3.0 points of points %); the deepest of all, the 2008-09 Islanders, had no player above 18 goals and finished last in the league. One of those ten deepest, the 1985-86 Calgary Flames (Hakan Loob 31, Dan Quinn 30), reached the final.
And the most famous top-heavy team of all won. The 1984-85 Oilers had Wayne Gretzky’s 73 goals and Jari Kurri’s 71, 35.9% of the team’s 401. They clinched the Cup on May 30, 1985, 8-3 over Philadelphia. In those playoffs the two scored 36 of the Oilers’ 98 goals, 36.7%, a bigger share than in the regular season.
The playoffs don’t reward depth either
The spring is where depth is supposed to pay, so we went series by series. In each of the 966 series, we asked which team was deeper and what each team’s regular-season goal differential alone predicted. The deeper team, by top-two share, won 474. Goal differential predicted 469.3. Other yardsticks agree. One asks how many equal scorers a team’s goals add up to: that 1999-2000 Ottawa team came to about 14, Bure’s Florida about 10. By that count the deeper team won 457 series against 460.4 predicted. By 20-goal scorers, it won 445 against 455.4.
Nick Zararis, a hockey writer who doubts the depth argument, put the opposite case on Substack in 2025: “the later in the playoffs a team goes, the more star power matters.” Round by round, the series don’t show that either. In the first round the deeper team won 249 of 506, where goal differential predicted 242.8. In the rounds between, it won 167 of 353 against 172.7. In the 99 all-NHL finals, it won 53 against 50.9. Stars came out about six series ahead of the prediction in the middle rounds. Depth came out about two ahead in the finals. Luck alone easily produces gaps that size, in all three rounds.
Stanley Cups by depth, against what goal differential predicted
Every NHL team since 1917-18, split into fifths within each season by the share of goals its top two scored
The most top-heavy fifth won 26 Cups against 28.2 predicted, the deepest fifth 13 against 16.4.
- Cups won
- Predicted from goal differential
One more version of the claim holds that stars get checked out of it in the playoffs, and the third line has to win the series. The top two regular-season scorers did score a smaller share of their team’s goals in the playoffs, 24.6% against 27.4% across 1,069 playoff teams. But most of that drop would happen even if nobody checked anyone. A player who leads his team in goals has usually had some luck to get there: a few pucks off a defenceman’s skate, a goalie fanning on a long shot. That luck doesn’t follow him into April. So we replayed every season a hundred times. Each player scored at his true rate, with the bounces drawn fresh each time, and we picked the top two from each replay. Nobody checked anybody in the replay. Even so, the top two’s share drops by 2.6 percentage points from the regular season to the playoffs. The real playoffs drop only 0.2 points more than that.
The Globe’s numbers were right, and goal differential predicted its Cups
The Globe’s piece offered the strongest evidence the depth case has. “Since 2000, 39 teams have had at least six 20-goal scorers; six reached the Stanley Cup Finals and four won it,” Benbow wrote. “But 15 lost in the first round and 25 didn’t make it past the second round. Three missed the playoffs.”
Counting from the 2000-01 season, our data reproduces every number: 39 teams, six finals, four Cups, 15 first-round exits, three misses. And four Cups from 39 teams looks like a lot. The average team in those seasons won a Cup 3.3% of the time; these teams won 10.3% of the time.
But six 20-goal scorers is a lot of goals, and these were very good teams. On average they outscored their opponents by about 20% of a typical team’s goal total. In a season like 2025-26, that is a plus-50 team. Give each of them the Cup chance its goal differential alone implies and the expected totals are 6.9 finals and 4.1 Cups. They got six and four. Bring it up to 2025-26 and there are 60 such teams, with eight finals and five Cups against 8.9 and 5.2 predicted.
Teams with six or more 20-goal scorers won the Cups their goal differentials predicted
Stanley Cup Finals reached and Cups won, against what each team’s regular-season goal differential predicted
2000-01 to 2021-22: 6 finals and 4 Cups against 6.9 and 4.1 predicted; 1999-00 to 2025-26: 8 and 5 against 8.9 and 5.2.
- Actual
- Predicted from goal differential
Benbow counted right. His deep teams were also teams that piled up goals and kept them out. Teams with six 20-goal scorers win Cups at the rate any team with their goal differential does.
Champions come from both ends of the depth table
Look at who actually wins. Of 106 NHL champions, 48 spread their goals more evenly than the middle team of their season, slightly fewer than half. In the cap era it is 10 of 21. The champions of 1979 to 1993, the years of the Islanders, the Oilers and Mario Lemieux’s Penguins, leaned top-heavy (5 of 14 deeper than the middle team). Those champions also ranked, on average, 3.4 in goal differential, and 14 Cups are too few to make a pattern of.
The last few springs show the range. Vegas won in 2023 as the league’s third-deepest team, with Jonathan Marchessault (28) and Jack Eichel (27) its top scorers. Florida won in 2024 with Sam Reinhart (57) and Carter Verhaeghe (34), and won again in 2025 as a middle-of-the-pack team, 18th of 32 in depth. The 2008-09 Penguins, with Evgeni Malkin and Sidney Crosby, sat almost exactly at the league’s middle (top two 26.4%, middle team 26.9%) and won too.
Goal differential is what the champions have in common. Era by era, their average rank in regular-season goal differential ran from 1.8 to 5.7 (1 = the league’s best), in leagues of 3 to 32 teams.
Cup winners land on both sides of average depth, about half each
How much deeper (above zero) or more top-heavy (below) each NHL champion was than its season’s average team, by top-two share of goals
Depth vs season average (in typical gaps, or standard deviations)
Season
Show data
| Season | Depth vs season average (in typical gaps, or standard deviations) | |
|---|---|---|
| Toronto Arenas 1917-18 | 1,917 | 0.83 |
| Ottawa Senators 1919-20 | 1,919 | 0.67 |
| Ottawa Senators 1920-21 | 1,920 | −1.16 |
| Toronto St. Patricks 1921-22 | 1,921 | −0.42 |
| Ottawa Senators 1922-23 | 1,922 | 0.42 |
| Montréal Canadiens 1923-24 | 1,923 | −0.97 |
| Montreal Maroons 1925-26 | 1,925 | −0.72 |
| Ottawa Senators 1926-27 | 1,926 | 0.35 |
| New York Rangers 1927-28 | 1,927 | −0.13 |
| Boston Bruins 1928-29 | 1,928 | 1.32 |
| Montréal Canadiens 1929-30 | 1,929 | −0.94 |
| Montréal Canadiens 1930-31 | 1,930 | 0.72 |
| Toronto Maple Leafs 1931-32 | 1,931 | −1.31 |
| New York Rangers 1932-33 | 1,932 | −1.66 |
| Chicago Blackhawks 1933-34 | 1,933 | −1.14 |
| Montreal Maroons 1934-35 | 1,934 | 0.93 |
| Detroit Red Wings 1935-36 | 1,935 | 0.99 |
| Detroit Red Wings 1936-37 | 1,936 | −0.07 |
| Chicago Blackhawks 1937-38 | 1,937 | −1.26 |
| Boston Bruins 1938-39 | 1,938 | 1.44 |
| New York Rangers 1939-40 | 1,939 | −1.13 |
| Boston Bruins 1940-41 | 1,940 | 0.90 |
| Toronto Maple Leafs 1941-42 | 1,941 | −0.11 |
| Detroit Red Wings 1942-43 | 1,942 | 1.00 |
| Montréal Canadiens 1943-44 | 1,943 | 0.80 |
| Toronto Maple Leafs 1944-45 | 1,944 | 0.48 |
| Montréal Canadiens 1945-46 | 1,945 | −0.88 |
| Toronto Maple Leafs 1946-47 | 1,946 | 1.05 |
| Toronto Maple Leafs 1947-48 | 1,947 | 0.27 |
| Toronto Maple Leafs 1948-49 | 1,948 | −1.26 |
| Detroit Red Wings 1949-50 | 1,949 | −0.49 |
| Toronto Maple Leafs 1950-51 | 1,950 | −0.19 |
| Detroit Red Wings 1951-52 | 1,951 | −1.52 |
| Montréal Canadiens 1952-53 | 1,952 | −0.44 |
| Detroit Red Wings 1953-54 | 1,953 | −0.49 |
| Detroit Red Wings 1954-55 | 1,954 | 0.55 |
| Montréal Canadiens 1955-56 | 1,955 | −1.14 |
| Montréal Canadiens 1956-57 | 1,956 | −0.40 |
| Montréal Canadiens 1957-58 | 1,957 | 0.98 |
| Montréal Canadiens 1958-59 | 1,958 | −0.58 |
| Montréal Canadiens 1959-60 | 1,959 | 0.97 |
| Chicago Blackhawks 1960-61 | 1,960 | 0.79 |
| Toronto Maple Leafs 1961-62 | 1,961 | 0.62 |
| Toronto Maple Leafs 1962-63 | 1,962 | 0.12 |
| Toronto Maple Leafs 1963-64 | 1,963 | 0.61 |
| Montréal Canadiens 1964-65 | 1,964 | 0.70 |
| Montréal Canadiens 1965-66 | 1,965 | 1.15 |
| Toronto Maple Leafs 1966-67 | 1,966 | 0.87 |
| Montréal Canadiens 1967-68 | 1,967 | 0.38 |
| Montréal Canadiens 1968-69 | 1,968 | 0.29 |
| Boston Bruins 1969-70 | 1,969 | 0.02 |
| Montréal Canadiens 1970-71 | 1,970 | 0.22 |
| Boston Bruins 1971-72 | 1,971 | −0.79 |
| Montréal Canadiens 1972-73 | 1,972 | 0.86 |
| Philadelphia Flyers 1973-74 | 1,973 | 0.47 |
| Philadelphia Flyers 1974-75 | 1,974 | −0.56 |
| Montréal Canadiens 1975-76 | 1,975 | −1.02 |
| Montréal Canadiens 1976-77 | 1,976 | −1.30 |
| Montréal Canadiens 1977-78 | 1,977 | −1.05 |
| Montréal Canadiens 1978-79 | 1,978 | −0.14 |
| New York Islanders 1979-80 | 1,979 | −1.27 |
| New York Islanders 1980-81 | 1,980 | −0.85 |
| New York Islanders 1981-82 | 1,981 | −0.89 |
| New York Islanders 1982-83 | 1,982 | −1.29 |
| Edmonton Oilers 1983-84 | 1,983 | −1.81 |
| Edmonton Oilers 1984-85 | 1,984 | −2.23 |
| Montréal Canadiens 1985-86 | 1,985 | 0.62 |
| Edmonton Oilers 1986-87 | 1,986 | −1.83 |
| Edmonton Oilers 1987-88 | 1,987 | 0.92 |
| Calgary Flames 1988-89 | 1,988 | −0.39 |
| Edmonton Oilers 1989-90 | 1,989 | 0.46 |
| Pittsburgh Penguins 1990-91 | 1,990 | 0.65 |
| Pittsburgh Penguins 1991-92 | 1,991 | −0.63 |
| Montréal Canadiens 1992-93 | 1,992 | 0.74 |
| New York Rangers 1993-94 | 1,993 | 0.09 |
| New Jersey Devils 1994-95 | 1,994 | −0.41 |
| Colorado Avalanche 1995-96 | 1,995 | −0.09 |
| Detroit Red Wings 1996-97 | 1,996 | −0.53 |
| Detroit Red Wings 1997-98 | 1,997 | 1.28 |
| Dallas Stars 1998-99 | 1,998 | −0.04 |
| New Jersey Devils 1999-00 | 1,999 | 0.63 |
| Colorado Avalanche 2000-01 | 2,000 | −1.80 |
| Detroit Red Wings 2001-02 | 2,001 | 0.07 |
| New Jersey Devils 2002-03 | 2,002 | 0.94 |
| Tampa Bay Lightning 2003-04 | 2,003 | −0.87 |
| Carolina Hurricanes 2005-06 | 2,005 | 0.12 |
| Anaheim Ducks 2006-07 | 2,006 | −0.71 |
| Detroit Red Wings 2007-08 | 2,007 | −0.27 |
| Pittsburgh Penguins 2008-09 | 2,008 | 0.00 |
| Chicago Blackhawks 2009-10 | 2,009 | 1.20 |
| Boston Bruins 2010-11 | 2,010 | 0.62 |
| Los Angeles Kings 2011-12 | 2,011 | 0.58 |
| Chicago Blackhawks 2012-13 | 2,012 | −0.88 |
| Los Angeles Kings 2013-14 | 2,013 | −0.78 |
| Chicago Blackhawks 2014-15 | 2,014 | 0.05 |
| Pittsburgh Penguins 2015-16 | 2,015 | 0.02 |
| Pittsburgh Penguins 2016-17 | 2,016 | −0.88 |
| Washington Capitals 2017-18 | 2,017 | −1.41 |
| St. Louis Blues 2018-19 | 2,018 | 0.26 |
| Tampa Bay Lightning 2019-20 | 2,019 | 0.10 |
| Tampa Bay Lightning 2020-21 | 2,020 | 0.88 |
| Colorado Avalanche 2021-22 | 2,021 | 1.06 |
| Vegas Golden Knights 2022-23 | 2,022 | 1.44 |
| Florida Panthers 2023-24 | 2,023 | −1.91 |
| Florida Panthers 2024-25 | 2,024 | −0.03 |
| Carolina Hurricanes 2025-26 | 2,025 | 1.06 |
Dots above the line are champions deeper than their season’s average team, dots below more top-heavy. Every era has winners on both sides.
Count the goals, not the scorers
- If you want to know how good a team is, its goals for and against tell you nearly everything the split of its scoring could. Depth adds less than one standings point a season under any definition we tried.
- Building around two stars won’t sink a team in April. Four lines that all score won’t lift one either. Both kinds win series at the rate their goal differentials say they should.
- A pile of 20-goal scorers mostly means a pile of goals. When the count rises, it is usually because the team, or the league, scored more.
Boston still has the record. Montreal still has the Cup.
The newest champion had depth and goals both. Carolina won in 2026 as one of the league’s deepest teams, seventh of 32, with Seth Jarvis its top scorer at 32. Its goal differential was third-best.
Forty-eight years earlier, the 1977-78 Bruins and their eleven 20-goal scorers met Montreal in the Final. The Canadiens had five 20-goal scorers. They also had 359 goals for and 183 against, +176, the best goal differential in the league. Boston won Games 3 and 4 at home. Montreal won the series 4-2, the last game 4-1 in Boston on May 25, 1978.
Nobody has had eleven 20-goal scorers since. By that count, the 1977-78 Bruins are still the deepest team in NHL history. The Cup went to the team with five.
Share this article
Line tools
Live Edgehalla tools on the same subject.
- NHL line combinations today Forward lines, defence pairs and PP units for all 32 teams.
- Power play tonight PP1 units and power-play matchups for tonight’s games.
- Fantasy hockey projections Tonight’s projected stats for every skater and goalie.
- NHL team hubs Lines, goalies, injuries and playoff odds for all 32 teams.
Show the work
- Tests written down
- Not in advance (exploratory)
- Results computed
Every NHL regular-season team-season from 1917-18 to 2025-26 (1,758; the 1917-18 Wanderers, six games, dropped). The NHL’s season totals give a traded skater one combined line, so each team with a traded skater was re-read with the totals filtered by franchise, which splits his goals by team; skater goals then equal team goals for in every team-season but 18, each one short by a goalie’s goal. Depth five ways (top-two share of goals, Gini of the top 12 scorers, effective number of scorers, schedule-scaled 20-goal scorers, top-two share of points), each a z-score within its season. Model: points % on goals for and against relative to the league, all demeaned within season, plus one depth measure; 90% ranges from 1,000 resamples of whole seasons. Holdouts: odd vs even seasons, before vs from 1967-68, before vs from 2005-06, each both ways. Playoffs: 966 NHL-vs-NHL series (2020 round robin and series against PCHA/WCHL teams excluded), a symmetric logistic on the difference in goal differential, with and without the difference in depth; finals and Cups predicted from goal differential and league size, fitted on all 106 champions. Star share: top-two regular-season scorers’ share of playoff goals against a benchmark that redraws each season from luck-adjusted (shrunk) scoring rates. Not pre-registered.
| Test | Result | Verdict |
|---|---|---|
| Depth (top-two share) predicts points %, raw | −1.48 points per SD deeper (top-heavy teams win more); top-heavy teams better in 81 of 106 seasons | Real, but it is goals |
More tests below
| Test | Result | Verdict |
|---|---|---|
| Depth (20-goal scorers) predicts points %, raw | +6.05 per SD; correlation with goals for 0.68 | Real, but it is goals |
| Depth adds to goals for and against: Top-two share of goals (lower = deeper) | −0.14 per SD (90% −0.28 to −0.02), n 1,758; residual SD 3.23 to 3.22 | Tiny |
| Depth adds to goals for and against: Gini of the top 12 scorers (lower = deeper) | −0.10 per SD (90% −0.22 to +0.02), n 1,758; residual SD 3.23 to 3.23 | Tiny |
| Depth adds to goals for and against: Effective number of scorers (higher = deeper) | −0.31 per SD (90% −0.46 to −0.18), n 1,758; residual SD 3.23 to 3.21 | Tiny |
Continues below
| Test | Result | Verdict |
|---|---|---|
| Depth adds to goals for and against: 20-goal scorers, schedule-scaled (more = deeper) | +0.37 per SD (90% +0.19 to +0.58), n 1,758; residual SD 3.23 to 3.22 | Tiny |
| Depth adds to goals for and against: Top-two share of points (lower = deeper) | −0.13 per SD (90% −0.26 to 0.00), n 1,758; residual SD 3.23 to 3.22 | Tiny |
| The depth effect differs by era | Six eras × five definitions: −1.19 to +0.79, signs flip between eras | Noise |
| Depth improves out-of-sample forecasts | Six folds × five definitions: RMSE change −0.033 to +0.030 points (base 2.9-4.8) | No |
| Effective scorers: stable sign | Controlled −0.31 (90% −0.46 to −0.18); negative in all six holdout folds (−0.19 to −0.52). Points toward top-heavy teams; about 0.5 standings points a season. One untested guess: stars on teams already winning collect empty-net goals | Stable but tiny; not a finding |
Continues below
| Test | Result | Verdict |
|---|---|---|
| Deeper team wins more playoff series | 474 of 966 vs 469.3 predicted; depth coefficient +0.009 log-odds per SD (90% −0.073 to +0.094) | No |
| Stars matter more in later rounds (finals) | Finals: deeper team won 53 of 99 vs 50.9; coefficient +0.202 (90% −0.129 to +0.527) | No; leans the other way, noise |
| Cups by depth fifth | 26/28.2, 19/21.0, 22/19.1, 26/21.2, 13/16.4 (actual/predicted, most top-heavy to deepest) | As predicted |
| Stars fade in the playoffs | Top two: 27.4% regular, 24.6% playoffs (−2.9); luck alone −2.6; beyond luck −0.2 (eras −1.4 to +1.5). Benchmark check: its regular-season share is 26.4% vs 27.4% actual | No beyond luck |
Continues below
| Test | Result | Verdict |
|---|---|---|
| Boston Globe 2023: six 20-goal scorers, six finals, four Cups since 2000 | Reproduced exactly from 2000-01 (39 teams); predicted from goal differential 6.9 finals, 4.1 Cups | True and fully explained by goal differential |
| Teams with nine or more 20-goal scorers | 13 teams, 2 finals, 0 Cups against 2.9 predicted. Picked after the fact from about ten groups we looked at; the large version of the same idea (the deepest fifth by schedule-scaled 20-goal scorers, 342 team-seasons) won 38 Cups against 39.6 predicted | Noise; not a finding |
| Champions are deeper than average | 48 of 106 deeper than their season’s median team; cap era 10 of 21 | No |
Frequently asked questions
Does scoring depth win the Stanley Cup?
Not on its own. Across 966 NHL playoff series since 1917, the deeper team (by the share of goals its top two scorers carried) won 474; regular-season goal differential alone predicted 469.3. The most top-heavy fifth of teams won 26 Cups against 28.2 predicted, the deepest fifth 13 against 16.4.
Which NHL team had the most 20-goal scorers?
The 1977-78 Boston Bruins, with 11: Peter McNab 41; Terry O'Reilly 29; Bobby Schmautz 27; Stan Jonathan 27; Jean Ratelle 25; Rick Middleton 25; Wayne Cashman 24; Gregg Sheppard 23; Brad Park 22; Don Marcotte 20; Bob Miller 20. Three teams have had ten. The Bruins lost the 1978 Final to Montreal in six games.
What is the most top-heavy team in NHL history?
Since expansion in 1967-68, the 1993-94 St. Louis Blues: Brett Hull (57) and Brendan Shanahan (52) scored 40.4% of the team’s goals. Relative to its own season, the most top-heavy team ever was the 2018-19 Edmonton Oilers (Leon Draisaitl 50, Connor McDavid 41, 39.7%).
Do star players score less in the playoffs?
A little, but no more than luck predicts. A team’s top two regular-season goal scorers scored 27.4% of its goals in the regular season and 24.6% in the playoffs, across 1,069 playoff teams. Removing regular-season luck alone predicts a drop of 2.6 points.
Did the Boston Globe’s depth numbers hold up?
Yes, exactly: from 2000-01 to 2021-22, 39 teams had six or more 20-goal scorers, and six reached the final and four won the Cup. Their goal differentials alone predicted 6.9 finals and 4.1 Cups.
More in Standings and results
- Verdict: Data story
Same Goals, Half the Blowouts
- Verdict: Data story
Vegas Had the Best First Season of Any NHL Expansion Team
- Verdict: Data story
Are the NHL’s Stars Pulling Away From Everyone Else?
- Verdict: Data story
The Schedule Got Kinder. The Second Night Didn’t.
- Verdict: Data story
Regular-Season Records Predict NHL Playoff Upsets Almost Exactly
- Verdict: Data story
Home Ice Lost Almost Half Its Edge in 1992. The Power Plays Didn’t Budge.
Sources and method
- NHL stats (NHL.com), every season since 1917-18
- NHL stats API: skater season totals, filtered by franchise for traded players, every season since 1917-18
- NHL stats API: team season totals (goals for and against, points %)
- NHL Records: playoff series (formats, rounds and series results)
- The Boston Globe, Julian Benbow: In the NHL, how does scoring depth translate from the regular season to the playoffs? (March 10, 2023)
- NHL.com, Dan Rosen: Rangers say scoring depth key against Senators (April 25, 2017)
- Associated Press, Stephen Whyno: Vegas Golden Knights winning the Stanley Cup shows the value of depth at every position (Oct. 8, 2023)
- NHL.com (Boston Bruins): The Lunch Pail A.C., top 10 moments from 1977-85
- The Boston Globe, Kevin Paul Dupont: Bruins’ goal-scorers of 1977-78 had 20-20 vision (Feb. 14, 2018)
- Nick Zararis: Does depth truly win championships? Rethinking NHL playoff success (May 14, 2025)
- NHL API reference (community documentation of the stats endpoints)
Research and analysis, not betting advice. Past results do not guarantee future returns; bet only what you can afford to lose.