How Hockey ChangedBodies
Major Junior Lost Half Its Share of the NHL Draft
In the drafts of 1969 to 1973, about three picks in four came from Canadian major junior. In 2022-26 it was just under two in five, though 2026 alone ran higher, at 45.5%, and European clubs and American junior teams took most of the rest. But compare players picked at the same spot, and the league they came from has barely changed their odds of making the NHL. American high schoolers were the exception. They made it about half as often as their draft slots predicted.
By Edgehalla ResearchPublished 13 min read
On June 26, the Toronto Maple Leafs took Gavin McKenna first overall. He had played 133 games for the Medicine Hat Tigers of the WHL and scored 244 points. Until August 2025, those games would have made him ineligible for NCAA hockey. He spent his draft season at Penn State instead.
In June 1969, St. Louis used the 66th pick on Tommi Salmelainen of HIFK Helsinki. He was the only player that year taken from a European club. He did not know he had been drafted.
Of the 84 players picked in 1969, 62 came from Canadian major junior.
On this page
Major junior went from three picks in four to just under two in five
Salmelainen was a lone European in a draft that belonged to Canadian junior hockey. In the five drafts from 1969 to 1973, the OHL, WHL and QMJHL supplied about three picks in four. Those three leagues, known then by older names, now form the Canadian Hockey League (CHL), the top tier of Canadian junior hockey. In the five drafts from 2022 to 2026 they supplied just under two in five. European clubs and American junior teams took most of the difference.
So is major junior still the best road to the NHL? It sends the most players, and it gets a big share of the highest picks. Take two players chosen at the same spot, one from the OHL and one from a Swedish club or the USHL, the top American junior league. Across five decades of drafts, players like those two made it about equally often. The one road that kept falling short ran through American high schools.
The bar for making it in this story is 200 NHL regular-season games in a player's first seven seasons, about two and a half full seasons. John LeClair, who turns up later, cleared it with 215.
Major junior lost half its share, and Europe took most of it
In the 2022-26 drafts, 39% of picks came from major junior, 33% from Europe, 17% from US junior and 4% from the NCAA.
Major junior dropped by about half. Europe went from almost nothing to about a third. Colleges shrank to a sliver.
Each pick counts under the league he played in during his draft season, as the NHL's draft record lists it. Patrick Kane came up through USA Hockey's National Team Development Program (NTDP), which runs the country's under-17 and under-18 teams, then spent his draft year with the London Knights, so he counts as major junior. Nico Hischier, a Swiss teenager, was drafted out of Halifax in the QMJHL, so he counts as major junior too. Auston Matthews, an American, spent his draft year with Zurich's club and counts as Europe. The record spells leagues 288 different ways over six decades ("OHA", "WCHL", "Minn.", "HIGH-MA", "SWEDEN-JR."). We sorted every spelling into seven roads: major junior, Europe, US junior, US college, US high school, other Canadian amateur hockey and everything else.
Europe and American junior teams took the space
The picks major junior gave up went somewhere. Most went to European clubs and, much later, to American junior teams. European clubs supplied 0.2% of picks in 1969-73. By 1999-2003 they supplied 36.2%, and their biggest single draft was 2001, at 41.2%. In 2022-26 it was 33.2%, about a third.
Year by year, Europe climbed to meet major junior, and US junior teams rose from zero
Share of all picks in each draft, by the league the player was drafted out of, 1969-2026
Show data
| Year | Major junior (CHL) | Europe | US junior | NCAA | US high school | Everything else |
|---|---|---|---|---|---|---|
| 1969 | 74% | 1% | 0% | 8% | 0% | 17% |
| 1970 | 75% | 0% | 0% | 14% | 0% | 11% |
| 1971 | 72% | 0% | 0% | 19% | 0% | 9% |
| 1972 | 80% | 0% | 0% | 14% | 0% | 7% |
| 1973 | 78% | 0% | 1% | 15% | 1% | 6% |
| 1974 | 69% | 2% | 4% | 17% | 0% | 8% |
| 1975 | 63% | 2% | 3% | 27% | 0% | 6% |
| 1976 | 72% | 6% | 2% | 19% | 0% | 2% |
| 1977 | 67% | 3% | 2% | 27% | 0% | 3% |
| 1978 | 55% | 7% | 1% | 31% | 0% | 6% |
| 1979 | 79% | 5% | 0% | 12% | 0% | 5% |
| 1980 | 66% | 5% | 0% | 20% | 3% | 6% |
| 1981 | 59% | 15% | 1% | 10% | 8% | 6% |
| 1982 | 51% | 14% | 3% | 8% | 15% | 10% |
| 1983 | 50% | 12% | 2% | 6% | 19% | 12% |
| 1984 | 43% | 15% | 2% | 9% | 18% | 13% |
| 1985 | 48% | 12% | 2% | 8% | 18% | 11% |
| 1986 | 48% | 11% | 1% | 9% | 16% | 15% |
| 1987 | 34% | 15% | 2% | 16% | 27% | 6% |
| 1988 | 34% | 16% | 2% | 19% | 23% | 7% |
| 1989 | 39% | 15% | 4% | 19% | 19% | 4% |
| 1990 | 35% | 21% | 2% | 15% | 22% | 5% |
| 1991 | 41% | 21% | 6% | 16% | 14% | 2% |
| 1992 | 47% | 31% | 3% | 3% | 10% | 5% |
| 1993 | 44% | 27% | 4% | 6% | 12% | 8% |
| 1994 | 49% | 28% | 4% | 2% | 11% | 7% |
| 1995 | 62% | 30% | 1% | 2% | 1% | 5% |
| 1996 | 56% | 24% | 1% | 10% | 3% | 5% |
| 1997 | 55% | 25% | 5% | 11% | 2% | 4% |
| 1998 | 53% | 29% | 2% | 11% | 3% | 3% |
| 1999 | 42% | 35% | 3% | 13% | 3% | 4% |
| 2000 | 32% | 36% | 5% | 12% | 3% | 13% |
| 2001 | 39% | 41% | 6% | 8% | 3% | 4% |
| 2002 | 35% | 38% | 7% | 14% | 2% | 5% |
| 2003 | 42% | 32% | 7% | 8% | 3% | 8% |
| 2004 | 39% | 30% | 8% | 9% | 6% | 7% |
| 2005 | 47% | 22% | 12% | 6% | 8% | 5% |
| 2006 | 37% | 30% | 11% | 9% | 9% | 6% |
| 2007 | 46% | 17% | 18% | 4% | 7% | 8% |
| 2008 | 52% | 19% | 9% | 4% | 8% | 9% |
| 2009 | 47% | 19% | 13% | 3% | 10% | 7% |
| 2010 | 51% | 19% | 11% | 4% | 11% | 5% |
| 2011 | 48% | 23% | 12% | 5% | 9% | 4% |
| 2012 | 47% | 20% | 13% | 5% | 9% | 7% |
| 2013 | 48% | 22% | 14% | 3% | 8% | 6% |
| 2014 | 45% | 24% | 16% | 2% | 6% | 7% |
| 2015 | 45% | 26% | 18% | 4% | 6% | 1% |
| 2016 | 46% | 27% | 15% | 6% | 3% | 4% |
| 2017 | 41% | 31% | 14% | 4% | 6% | 4% |
| 2018 | 36% | 32% | 18% | 4% | 5% | 6% |
| 2019 | 33% | 33% | 20% | 2% | 7% | 6% |
| 2020 | 36% | 34% | 17% | 2% | 7% | 5% |
| 2021 | 37% | 37% | 16% | 2% | 4% | 4% |
| 2022 | 36% | 33% | 16% | 3% | 6% | 4% |
| 2023 | 36% | 34% | 18% | 4% | 4% | 5% |
| 2024 | 39% | 36% | 17% | 2% | 3% | 3% |
| 2025 | 40% | 33% | 18% | 4% | 3% | 3% |
| 2026 | 46% | 30% | 15% | 5% | 3% | 1% |
- 2005: Draft cut to 7 rounds
Follow the major junior line down from the top left. The line that climbs to meet it is Europe. The US junior line starts at zero and ends near one pick in six. The chart uses shares because the draft kept changing size: 84 picks in 1969, 293 in 2000 and 224 in 2026.
The 1970s draft was a major junior draft. Through 1978 the only other big source was American colleges. Their share peaked in 1978, at 31.2% of picks. College players were older, about 20 on draft day in 1974-78. Then teams started drafting younger players, and the college share fell: 20.0% in 1980, 5.8% in 1983. It has been below 5% in 18 of the 20 drafts since 2007. Europe came in two waves: a slow climb through the 1980s, then 21.0% in 1990, 31.4% in 1992 and a peak of 41.2% in 2001.
In 2000, 2001, 2002 and 2021, teams took more players from European clubs than from major junior. In 2019 the two were level, at 71 picks each.
Major junior's share did not fall in a straight line. Across the 1985-89 drafts it averaged 40.6%. In 1987 alone it was 33.7%, about one pick in three. That year teams took more than a quarter of their picks out of American high schools. By 1995 the high schoolers had nearly vanished, and major junior was back to 61.5%, about three picks in five. It averaged 56.3% across 1995-98. Then it fell again. In 2000 it was 32.1%, under a third, and teams took more players from European clubs than from major junior. The 2000-02 drafts averaged 35.2%. Since 2018 its lowest year was 2019, at 32.7%, about a third. Its highest was 2026, at 45.5%, just under half.
The three major junior leagues don't send equal numbers. In 2017-26 the OHL supplied 15.8% of all picks and the WHL 13.8%. The QMJHL supplied 8.3%, about half as many.
Detroit found Lidström, Fedorov and Konstantinov in one draft
Europe's biggest jump came between 1989 and 1992. It began on June 17, 1989, in Bloomington, Minn., when the Quebec Nordiques took Mats Sundin of Nacka with the first pick. United Press International called him "the first European selected first in the NHL amateur draft." Sundin was 18, and a military commitment at home was expected to keep him in Sweden for two years. "He's worth waiting for," said Pierre Gauthier, Quebec's chief scout. "If he'd have been playing in North America you would have heard a lot more about him."
Detroit took its first European in round three. The Red Wings took Nicklas Lidström of Västerås at No. 53, Sergei Fedorov of CSKA Moscow at No. 74 and Vladimir Konstantinov, also of CSKA, at No. 221. Lidström played 1,564 NHL games, Fedorov 1,248 and Konstantinov 446. Four of Detroit's 14 picks that year played 1,000 games or more.
Christer Rockström was the club's top European scout, and he credits the idea to Neil Smith, Detroit's head amateur scout at the time. "At the time, Detroit was one of the worst teams in the league," Rockström said in remarks Detroit Hockey Now published in 2025. "It was actually Neil's idea that the organization should start looking in Europe to find talent." Picking his man still meant winning an argument in the draft room. "It was a matter of convincing the guy who had another player in a league in Western Canada, in college hockey, or the OHL in Ontario, that we should choose my guy." That is how it went in Detroit's draft room. Rockström wasn't speaking for anyone else's.
Teams took 38 players from European clubs in 1989. Ten of them made it. Judged by every player taken at those same spots in the drafts around 1989, from any league, about four should have. We picked 1989 because it stands out. One good year proves little on its own. By 1992 teams were taking 45 players from Russian leagues alone, three times the 15 of 1989. That jump came as the Soviet Union broke apart, though we didn't test whether one caused the other. Moscow Dynamo had four first-rounders, among them Alexei Yashin at No. 2, and Sergei Gonchar went 14th out of Chelyabinsk.
Russian leagues surged after 1989, and Sweden became the steady source
Share of all NHL draft picks from each European country, by the country of the league the player was drafted out of, 1969-2026
Sweden11%
Finland6%
Russia and USSR9%
Czechia and Slovakia3%
Switzerland0%
Germany0%
Other Europe1%
Show data
| Year | Sweden | Finland | Russia and USSR | Czechia and Slovakia | Switzerland | Germany | Other Europe |
|---|---|---|---|---|---|---|---|
| 1969 | 0% | 1% | 0% | 0% | 0% | 0% | 0% |
| 1970 | 0% | 0% | 0% | 0% | 0% | 0% | 0% |
| 1971 | 0% | 0% | 0% | 0% | 0% | 0% | 0% |
| 1972 | 0% | 0% | 0% | 0% | 0% | 0% | 0% |
| 1973 | 0% | 0% | 0% | 0% | 0% | 0% | 0% |
| 1974 | 2% | 0% | 0% | 0% | 0% | 0% | 0% |
| 1975 | 1% | 1% | 0% | 0% | 0% | 0% | 0% |
| 1976 | 4% | 2% | 0% | 0% | 1% | 0% | 0% |
| 1977 | 1% | 2% | 0% | 0% | 0% | 0% | 0% |
| 1978 | 3% | 1% | 1% | 1% | 0% | 1% | 0% |
| 1979 | 4% | 0% | 0% | 1% | 0% | 0% | 0% |
| 1980 | 3% | 2% | 0% | 0% | 0% | 0% | 0% |
| 1981 | 7% | 6% | 0% | 2% | 0% | 1% | 0% |
| 1982 | 5% | 2% | 1% | 6% | 0% | 0% | 0% |
| 1983 | 4% | 4% | 2% | 2% | 0% | 0% | 0% |
| 1984 | 6% | 4% | 0% | 4% | 0% | 1% | 0% |
| 1985 | 6% | 2% | 1% | 3% | 0% | 0% | 0% |
| 1986 | 4% | 4% | 0% | 2% | 0% | 0% | 1% |
| 1987 | 6% | 2% | 1% | 4% | 0% | 0% | 1% |
| 1988 | 6% | 3% | 4% | 2% | 0% | 1% | 0% |
| 1989 | 4% | 1% | 6% | 3% | 0% | 0% | 1% |
| 1990 | 3% | 4% | 6% | 8% | 0% | 0% | 1% |
| 1991 | 4% | 2% | 9% | 3% | 0% | 0% | 1% |
| 1992 | 4% | 3% | 17% | 7% | 0% | 1% | 0% |
| 1993 | 6% | 3% | 11% | 4% | 1% | 1% | 0% |
| 1994 | 6% | 3% | 12% | 6% | 0% | 0% | 0% |
| 1995 | 3% | 5% | 11% | 9% | 0% | 0% | 1% |
| 1996 | 7% | 3% | 7% | 6% | 0% | 1% | 0% |
| 1997 | 6% | 4% | 7% | 7% | 1% | 0% | 0% |
| 1998 | 7% | 5% | 8% | 7% | 1% | 0% | 0% |
| 1999 | 9% | 6% | 10% | 7% | 1% | 0% | 1% |
| 2000 | 6% | 6% | 13% | 8% | 2% | 0% | 0% |
| 2001 | 5% | 10% | 13% | 10% | 2% | 2% | 0% |
| 2002 | 8% | 9% | 11% | 7% | 1% | 0% | 1% |
| 2003 | 7% | 4% | 10% | 7% | 2% | 1% | 1% |
| 2004 | 7% | 5% | 8% | 8% | 1% | 0% | 1% |
| 2005 | 7% | 4% | 4% | 7% | 0% | 0% | 0% |
| 2006 | 9% | 6% | 7% | 5% | 1% | 1% | 1% |
| 2007 | 8% | 2% | 3% | 2% | 1% | 2% | 0% |
| 2008 | 9% | 3% | 4% | 1% | 1% | 1% | 1% |
| 2009 | 11% | 4% | 3% | 1% | 0% | 1% | 0% |
| 2010 | 10% | 3% | 2% | 1% | 1% | 1% | 1% |
| 2011 | 12% | 5% | 3% | 2% | 1% | 0% | 1% |
| 2012 | 11% | 4% | 3% | 1% | 1% | 0% | 1% |
| 2013 | 12% | 5% | 4% | 0% | 1% | 0% | 0% |
| 2014 | 14% | 3% | 3% | 2% | 1% | 0% | 0% |
| 2015 | 9% | 6% | 6% | 5% | 1% | 0% | 1% |
| 2016 | 12% | 6% | 6% | 1% | 1% | 0% | 1% |
| 2017 | 15% | 7% | 5% | 3% | 1% | 0% | 1% |
| 2018 | 14% | 7% | 7% | 3% | 1% | 0% | 1% |
| 2019 | 12% | 9% | 10% | 2% | 0% | 1% | 1% |
| 2020 | 14% | 8% | 8% | 1% | 1% | 1% | 0% |
| 2021 | 10% | 8% | 12% | 3% | 1% | 1% | 2% |
| 2022 | 13% | 7% | 9% | 3% | 0% | 1% | 0% |
| 2023 | 13% | 7% | 9% | 2% | 1% | 1% | 0% |
| 2024 | 12% | 8% | 11% | 3% | 2% | 0% | 1% |
| 2025 | 15% | 5% | 10% | 2% | 1% | 0% | 0% |
| 2026 | 11% | 6% | 9% | 3% | 0% | 0% | 1% |
Look at the Russia panel's spike in the early 1990s. It comes back down. Then find Sweden, whose line climbs slowly and keeps climbing.
Russian leagues went from 6.0% of picks in 1989 to 17.0% in 1992, the biggest three-draft rise for any country in the series. They faded to 1.9-3.8% in 2009-14, then climbed back to about 10% in 2021-26. Czech and Slovak leagues followed a similar arc and never recovered: 7.9% in 1999-2003, 2.4% in 2022-26. Swedish leagues grew steadily, from 7.0% of picks in 1999-2003 to 12.9% in 2022-26.
Swedish leagues have supplied the most European picks in every draft since 2005 except 2021, when Russian leagues edged them. In 2005 Czech and Slovak leagues matched them. Frölunda alone sent 28 players to the draft in 2017-26 and Djurgården 27. In 2026, Frölunda's Ivar Stenberg went second overall. Where those players were born is a separate question; our story on NHL birthplaces follows who actually plays.
The first round held out longest
Detroit found its Europeans in the third round and later. In the first round, where teams expect to find their stars, major junior held on longer. In 1987, Buffalo took Pierre Turgeon of Granby first, New Jersey took Brendan Shanahan of London second and Boston took Glen Wesley of Portland third. Major junior supplied 20 of the 21 first-rounders that year.
Major junior took almost every first-round pick in the early 1970s and well under half now
Share of first-round picks by road, 1971-2024. Each point counts the first-rounders of five drafts together: that year’s and the two on either side
Show data
| Year | Major junior (CHL) | Europe | US junior | NCAA |
|---|---|---|---|---|
| 1971 | 96% | 0% | 0% | 0% |
| 1972 | 97% | 0% | 0% | 0% |
| 1973 | 96% | 0% | 0% | 0% |
| 1974 | 94% | 1% | 0% | 0% |
| 1975 | 94% | 1% | 0% | 0% |
| 1976 | 93% | 1% | 0% | 0% |
| 1977 | 88% | 1% | 0% | 1% |
| 1978 | 90% | 1% | 0% | 1% |
| 1979 | 88% | 2% | 0% | 1% |
| 1980 | 87% | 2% | 0% | 2% |
| 1981 | 85% | 2% | 0% | 3% |
| 1982 | 85% | 4% | 0% | 2% |
| 1983 | 81% | 6% | 0% | 3% |
| 1984 | 76% | 4% | 0% | 6% |
| 1985 | 77% | 4% | 0% | 6% |
| 1986 | 77% | 5% | 0% | 6% |
| 1987 | 76% | 5% | 1% | 7% |
| 1988 | 77% | 4% | 1% | 8% |
| 1989 | 78% | 9% | 2% | 7% |
| 1990 | 67% | 17% | 3% | 6% |
| 1991 | 62% | 23% | 4% | 6% |
| 1992 | 61% | 25% | 3% | 6% |
| 1993 | 60% | 29% | 2% | 4% |
| 1994 | 61% | 27% | 2% | 4% |
| 1995 | 67% | 22% | 1% | 4% |
| 1996 | 71% | 20% | 0% | 3% |
| 1997 | 68% | 23% | 0% | 5% |
| 1998 | 61% | 25% | 1% | 9% |
| 1999 | 56% | 30% | 1% | 10% |
| 2000 | 50% | 33% | 1% | 13% |
| 2001 | 47% | 32% | 2% | 17% |
| 2002 | 45% | 32% | 2% | 17% |
| 2003 | 49% | 28% | 3% | 13% |
| 2004 | 49% | 25% | 5% | 13% |
| 2005 | 51% | 20% | 9% | 10% |
| 2006 | 53% | 21% | 9% | 6% |
| 2007 | 56% | 18% | 11% | 5% |
| 2008 | 56% | 17% | 13% | 5% |
| 2009 | 60% | 15% | 12% | 3% |
| 2010 | 61% | 17% | 12% | 3% |
| 2011 | 63% | 18% | 12% | 2% |
| 2012 | 65% | 17% | 13% | 1% |
| 2013 | 65% | 19% | 12% | 3% |
| 2014 | 61% | 19% | 12% | 4% |
| 2015 | 58% | 23% | 10% | 5% |
| 2016 | 52% | 25% | 11% | 7% |
| 2017 | 47% | 26% | 14% | 7% |
| 2018 | 48% | 28% | 13% | 5% |
| 2019 | 46% | 29% | 15% | 5% |
| 2020 | 43% | 32% | 17% | 4% |
| 2021 | 42% | 33% | 17% | 4% |
| 2022 | 42% | 33% | 15% | 6% |
| 2023 | 43% | 29% | 15% | 9% |
| 2024 | 44% | 31% | 13% | 10% |
The major junior line starts near the top and slides to well under half. Europe and US junior teams climb from the bottom.
Major junior still held about 77% of first-rounders in the late 1980s, while US high schools were taking about a fifth of all picks but few first-rounders. The big drops came with Europe in the early 1990s and again around 2000. After a recovery to 65% around 2012, major junior slid to a low of 42% around 2021. US junior teams never took more than 4% of first-rounders in any five-draft stretch before 2005. They have taken about 15% since 2017.
The 2026 first round split 14 picks from major junior, 11 from Europe, 5 from colleges and 2 from US junior teams.
The American road moved from the high-school rink to junior teams
The other new supplier was American, and it changed address halfway through. On Feb. 23, 1981, Sports Illustrated put a 17-year-old from Peabody, Mass., on its cover: "The Can't-Miss Kid -- Here's Bobby Carpenter." That June, Washington took him third overall out of St. John's Prep, the first American high schooler taken in the first round. A year earlier Buffalo had made Jay North of Bloomington Jefferson High the first high schooler drafted at all, at No. 62. North never played an NHL game. In 1983 Minnesota took Brian Lawton of Mount St. Charles in Rhode Island first overall.
By 1987, teams drafted 68 high schoolers, more than a quarter of the draft. Then the road nearly closed. Teams took 30 high schoolers in 1994 and just two in 1995. American-born picks of every kind fell from 48 to 17 that year. The draft also shrank that year, from 286 picks to 234. That can account for only part of the drop. We found no documented reason for the rest.
USA Hockey started its national team development program in 1996, and the first class arrived in Ann Arbor in 1997. It was not popular at first. "We were fighting against it those first two or three years," Jeff Jackson, who came up with the concept, told ESPN. Greg Cronin, one of his first assistants, described how the local programs that lost players saw the new team: "We were looked at like pirates and the bounty was the player."
In 2006 St. Louis took Erik Johnson of the program's under-18 team first overall, the first No. 1 pick the draft record lists under the program. In 2019 the same team had eight first-round picks, Jack Hughes at No. 1 among them, the most from one club in any draft since 1969. The next most is six, by the same team in 2022. In that 2019 first round, major junior took 13 of the 31 picks.
USA Hockey called Hughes "the NTDP's fifth number one pick in program history." The draft record files only two No. 1 picks under the program, Johnson and Hughes. The others moved on before their draft year. Kane went first in 2007 as a London Knight, and Matthews first in 2016 from Zurich.
High-school picks collapsed in 1995, and US junior teams have out-picked them every year since 2011
Share of all picks drafted out of US high schools, US junior teams and the NCAA, 1969-2026
Show data
| Year | US high school | US junior | NCAA |
|---|---|---|---|
| 1969 | 0% | 0% | 8% |
| 1970 | 0% | 0% | 14% |
| 1971 | 0% | 0% | 19% |
| 1972 | 0% | 0% | 14% |
| 1973 | 1% | 1% | 15% |
| 1974 | 0% | 4% | 17% |
| 1975 | 0% | 3% | 27% |
| 1976 | 0% | 2% | 19% |
| 1977 | 0% | 2% | 27% |
| 1978 | 0% | 1% | 31% |
| 1979 | 0% | 0% | 12% |
| 1980 | 3% | 0% | 20% |
| 1981 | 8% | 1% | 10% |
| 1982 | 15% | 3% | 8% |
| 1983 | 19% | 2% | 6% |
| 1984 | 18% | 2% | 9% |
| 1985 | 18% | 2% | 8% |
| 1986 | 16% | 1% | 9% |
| 1987 | 27% | 2% | 16% |
| 1988 | 23% | 2% | 19% |
| 1989 | 19% | 4% | 19% |
| 1990 | 22% | 2% | 15% |
| 1991 | 14% | 6% | 16% |
| 1992 | 10% | 3% | 3% |
| 1993 | 12% | 4% | 6% |
| 1994 | 11% | 4% | 2% |
| 1995 | 1% | 1% | 2% |
| 1996 | 3% | 1% | 10% |
| 1997 | 2% | 5% | 11% |
| 1998 | 3% | 2% | 11% |
| 1999 | 3% | 3% | 13% |
| 2000 | 3% | 5% | 12% |
| 2001 | 3% | 6% | 8% |
| 2002 | 2% | 7% | 14% |
| 2003 | 3% | 7% | 8% |
| 2004 | 6% | 8% | 9% |
| 2005 | 8% | 12% | 6% |
| 2006 | 9% | 11% | 9% |
| 2007 | 7% | 18% | 4% |
| 2008 | 8% | 9% | 4% |
| 2009 | 10% | 13% | 3% |
| 2010 | 11% | 11% | 4% |
| 2011 | 9% | 12% | 5% |
| 2012 | 9% | 13% | 5% |
| 2013 | 8% | 14% | 3% |
| 2014 | 6% | 16% | 2% |
| 2015 | 6% | 18% | 4% |
| 2016 | 3% | 15% | 6% |
| 2017 | 6% | 14% | 4% |
| 2018 | 5% | 18% | 4% |
| 2019 | 7% | 20% | 2% |
| 2020 | 7% | 17% | 2% |
| 2021 | 4% | 16% | 2% |
| 2022 | 6% | 16% | 3% |
| 2023 | 4% | 18% | 4% |
| 2024 | 3% | 17% | 2% |
| 2025 | 3% | 18% | 4% |
| 2026 | 3% | 15% | 5% |
The high-school line rises through the 1980s and drops off a cliff in 1995. The US junior line takes its place, slowly.
High-school players were a 1980s story: 3.3% of picks in 1980, 27.0% in 1987, then 10.5% in 1994 and 0.9% in 1995. The draft record and the NHL's player bios agree on that dip, so it is not a data error. When American picks came back in the 2000s, they came from junior teams. US junior has supplied more picks than US high schools in every draft since 2011.
Major junior sends the most players to the NHL because it gets the best picks
Is major junior the best road, then? Count the players who made it, and it looks that way. From the 2015-19 drafts, major junior players were 39.9% of the picks but 46.7% of the players who made it. European picks were 29.8% of the picks and 24.2% of those who made it; US junior picks 16.9% and 17.6%.
But teams took 71 of the 150 top-30 picks in those five drafts from major junior, and a top-30 pick makes it far more often than a seventh-rounder, whatever his league. So the fair test compares each pick with other players taken at the same spot.
Take the 1987 draft. Teams took 85 players from major junior, and 20 of them made it. Was that good? It depends where they went. Turgeon, Shanahan and Wesley went one, two and three, and picks that high nearly always make it. A pick between 151 and 210, from any league, made it about once in 22 in the drafts of 1985 to 1994. So each of the 85 gets his own chance from his slot: close to a sure thing for Turgeon, about one in 22 for a 180th pick. Add the 85 chances together and you get about 20 players. Major junior delivered 20, right on its number. We ran that same test on every road: add up the chances each pick had from his slot, then count the players who actually made it. The draft hit-rate story shows how those chances fall from the top of the draft to the bottom.
At the same draft slot, only US high schoolers fell well short
Each bar: for every 100 picks from a road, how many more (above zero) or fewer (below zero) players reached 200 NHL games in 7 seasons than their draft slots predicted; drafts 1969-2019
At the same draft slots, major junior picks made it about as often as expected in every era; US high schoolers made it less often in all three (-9.5, -5.1 and -2.6 players per 100 picks), and Europe's early edge faded to 0.3.
- 1969-84 drafts
- 1985-2004 drafts
- 2005-19 drafts
A bar near zero means a road's picks made it about as often as their draft slots predicted. Most bars sit near zero. The two longest bars below it belong to the high schoolers. Other Canadian picks, a small group, dip too: they came up 3.4 players per 100 picks short in 1985-2004.
Over all 11,374 picks from 1969 to 2019, major junior produced 0.2 fewer players per 100 picks than its draft slots predicted, almost exactly on target. US junior produced exactly as many players as its slots predicted, and other Canadian roads 2.0 fewer players per 100 picks than theirs. So a major junior or US junior pick turned out about as well as any other player taken at the same spot. US high schoolers were the exception. They made it 4.4% of the time where their slots predicted 9.6%, about half as often, across 907 picks. That gap is far too big to be luck. It shows in all three eras. In 2005-19, when far fewer high schoolers were picked, it narrows to 2.6 players per 100 picks, and that gap on its own could be luck.
The 1987 high schoolers show the one road that fell short
One road missed in every era. In 1987, teams took 68 American high schoolers. Their draft slots predicted that four or five would make it. One did.
He was John LeClair, taken 33rd by Montreal out of Bellows Free Academy in St. Albans, Vt. He played four seasons at the University of Vermont, two of them interrupted by injuries, according to College Hockey Inc. He won the Stanley Cup with Montreal in 1993, his second pro season, and later became "the only American-born player to score 50 goals in three consecutive seasons." He played 215 games in his first seven seasons and 967 in all.
Seven more of the 1987 high schoolers reached 200 NHL games eventually, too late for our count: Joe Sacco (No. 71), Kevin Dean (No. 86), Ted Donato (No. 98), Shawn McEachern (No. 110), Damian Rhodes (No. 112), Garth Snow (No. 114) and Jim Dowd (No. 149). So does a seven-season count shortchange high schoolers, who, like LeClair, might spend four years in college first? Only a little. Give every pick 10 NHL seasons and the high schoolers still come up about 6 players per 100 picks short of their slots. Count whole careers and they are about 5 per 100 short. In 1987 alone, 8 of the 68 high schoolers had 200-game careers, about one in eight. Of the 85 major junior picks, 29 did, about one in three. Major junior picks went higher in that draft, which helps them in a raw count. Measured against their slots over whole careers, the high schoolers still fall short.
The shortfall sits in the later rounds. High schoolers taken after the 100th pick made it 7 times in 646, about one in a hundred, where their slots predicted about one in twenty. Between picks 31 and 100, 13 of 216 made it against about 32 predicted. In the first round, 20 of 45 made it against about 23 predicted, a gap that could be luck.
We found no contemporary account of why teams drafted so many high schoolers in the mid-1980s, or why so few of the late picks made it.
Europe’s late-round edge was real, and it faded
If one road came up short, did any come up long? Europe did, for a while, in exactly the place Detroit's scout said it would. "They wanted to have the Canadian style, but also be able to steal good talent from Europe in the late rounds before other teams reacted," Rockström said of the Red Wings. "Detroit was early to invest in this. Now everyone takes Europe with the utmost respect. They didn't do that in the early years."
Through 2004, he was right. From 1969 to 2004, teams took 1,048 players from European clubs after the 100th pick. Of those, 90 made it, about 9 in a hundred, where their slots predicted about 6. From 2005 to 2019, 21 of 452 made it, against about 20 predicted. After 2004, a late European pick was worth what any other late pick was worth.
The Europeans of that era were also older. Konstantinov was 22.2 on draft day and Anatoli Semenov, a Moscow Dynamo forward Edmonton took 120th the same year, was 27.3. The typical major junior pick of that era was 18.8. Viacheslav Fetisov was drafted twice, by Montreal at 20.2 in 1978 and by New Jersey at 25.1 in 1983, and played 546 NHL games. Could their age explain the edge?
Size doesn’t separate the roads either
So a player's league barely changes his odds once you know his draft slot. Could it change what a team gets in another way? A team might favour a league because it sends bigger bodies. The draft record lists each player's height and weight, and no road stands out. In every era the roads sit within about an inch of one another: the average skater from the tallest road stood 0.7 inches taller than the average skater from the shortest in 1979-93, 0.5 in 1994-2008 and 0.6 in 2009-23. Major junior skaters were the heaviest in 1994-2008 (200 lb against 193 for Europeans), and every road got lighter after that.
Every road’s skaters stood within about an inch of each other
| Road | 1979-93 drafts | 1994-2008 drafts | 2009-23 drafts |
|---|---|---|---|
| Major junior (CHL) | 72.5 / 195 / 18.8 (n = 1,524) | 73.5 / 200 / 18.5 (n = 1,542) | 72.6 / 188 / 18.5 (n = 1,248) |
| Europe | 72.4 / 192 / 20.6 (n = 557) | 73.0 / 193 / 19.3 (n = 1,035) | 72.8 / 185 / 18.6 (n = 755) |
| US junior (USHL, NTDP) | 73.1 / 195 / 18.6 (n = 76) | 73.3 / 197 / 18.9 (n = 212) | 72.4 / 188 / 18.6 (n = 444) |
| NCAA | 72.7 / 193 / 19.3 (n = 364) | 73.1 / 198 / 19.5 (n = 294) | 72.5 / 189 / 19.4 (n = 107) |
| US high school | 72.8 / 192 / 18.4 (n = 516) | 73.2 / 192 / 18.7 (n = 165) | 72.7 / 183 / 18.4 (n = 200) |
Continues below
| Road | 1979-93 drafts | 1994-2008 drafts | 2009-23 drafts |
|---|---|---|---|
| Other Canadian | 72.7 / 192 / 18.7 (n = 208) | 73.2 / 193 / 18.9 (n = 161) | 73.0 / 185 / 18.6 (n = 139) |
Pick any era and read down its column. No road's skaters stand taller than another's by even an inch. The ages differ more. In 1979-93 European picks were almost two years older than major junior picks, and college picks about half a year older.
McKenna took the newest road, from the WHL to Penn State
The roads are still moving, and McKenna took the newest one. For decades, playing in the CHL cost a player his NCAA eligibility, because some players in those leagues have signed professional contracts with NHL teams. On November 7, 2024, a group of college officials voted to change that. They sit on the NCAA's Division I Council, which writes the rules for the top division of college sports. From August 1, 2025, CHL players could keep their Division I eligibility. Their junior team could have paid them only "actual and necessary expenses," covering the costs of playing and nothing more. Players who sign NHL contracts are still excluded. So a player can now play major junior, then move to a US college before his draft year. This story files every pick under the league of his draft season, so a player like that counts as college.
McKenna was the biggest name to take that road. "It was a super tough decision. There are a lot of really great options out there," he said when he announced the move on SportsCenter. "But me, my family and everyone in my circle decided that the best spot for me next year is Penn State." His agent, Pat Brisson, told ESPN that "the NIL obviously comes into play, but it's not the primary decision of why he decided to go to college." NIL stands for name, image and likeness, the endorsement deals US college athletes are now allowed to sign. Keaton Verhoeff, a WHL defenceman, went to North Dakota the same fall.
The WHL's commissioner, Dan Near, expected few to follow. "The idea of someone going to the NCAA before their draft year will be occasional," he told ESPN. He pointed to the 2025 first round, with 21 of its 32 picks from Canadian junior and five from US colleges. The draft record has the same count: 21 from major junior and 5 from the NCAA.
The first draft under the new rule had two college picks from major junior
| Pick | Player | College | Major junior before the draft |
|---|---|---|---|
| 1 | Gavin McKenna | Penn State | WHL, 133 games |
| 9 | Keaton Verhoeff | North Dakota | WHL, 75 games |
The table has two rows. The 2026 draft had 11 NCAA picks, 2 of them former CHL players, including the first overall pick. Among the 35 NCAA picks of 2021-25 there were none.
College hockey's climb to the top of the draft started before the rule did. From 2021 to 2026, seven of the 30 top-five picks came out of US colleges: Owen Power, Matty Beniers and Kent Johnson in 2021, Adam Fantilli in 2023, Macklin Celebrini and Artyom Levshunov in 2024, and McKenna. From 1969 to 2020 it was 14 of 260. At the old rate, six drafts would have produced one or two. Seven is far too many to be luck. But six of the seven came before the rule changed, so the rule didn't cause it. Five of the seven had played in the USHL or for the development program before college, by their NHL.com career records. Those players count as college picks. The reverse happens too: a player drafted out of the USHL who then plays college hockey counts as US junior, so college hockey shapes more NHL players than its draft share shows.
For the draft as a whole, it is too early to see the rule. NCAA picks were 4.9% of the 2026 draft against 3.0% on average in 2017-25, and major junior's share rose to 45.5%. Two players who crossed over in one draft are too few to judge the rule by. The 2027 draft and those after it will show more.
From one European pick in 1969 to 67 in 2026
Salmelainen went to Toronto that summer to coach at Carl Brewer's hockey school, and learned on the trip that St. Louis had picked him. "The man himself didn't know he had been drafted," Risto Pakarinen wrote for NHL.com's Finnish site in 2024 (our translation). "He didn't even know what the draft meant." Brewer talked him into staying, and he became the first Finn to sign an NHL contract. He never played an NHL game.
McKenna has already worn three labels: WHL star, Penn State freshman, first overall pick. Our data files him under college. Five decades of drafts say it hardly matters which label sticks. All 51 players taken first overall from 1969 to 2019 made it, whatever league they came from.
The 2026 board took 102 players from major junior, 67 from European clubs, 34 from American junior teams and 11 from colleges. Two of those college players had come from major junior. One of them went first.
Share this article
Draft and records tools
Live Edgehalla tools on the same subject.
Show the work
- Tests written down
- Not in advance (exploratory)
- Results computed
All 13,031 picks from the 1969-2026 drafts (records.nhl.com), each mapped from its draft-day league to one of seven routes. Outcomes for the 11,374 picks of 1969-2019: 200+ NHL regular-season games in the first 7 seasons, compared with the rate for the same overall-pick range and draft era (baseline eras 1969-76, 1977-84, 1985-94, 1995-2004, 2005-11, 2012-19, so the 1987 picks are measured against all picks of 1985-94 in the same pick range); checked in three separate eras and with 10-season, whole-career and age-25 windows. Small groups (one draft, one route, one pick range) use the same baseline and are listed in slot-checks.csv; the 1987 and 1989 drafts were chosen because they stand out, so they illustrate the pooled results and are not tests on their own. NCAA top-five picks: US junior before college = any USHL, NTDP or NAHL regular-season games before the draft season on the player's NHL.com career record. Gaps are in percentage points of hit rate, which the article reads as players per 100 picks. League names were cleaned by hand: 288 spellings map to seven routes; 76 picks list no league and sit in "Everything else", about a third of them (24) from the 2000 draft. Our database copy of the 2017-26 drafts matches the NHL records on every one of 2,214 picks. Window limits: the 2020-26 drafts have not had seven NHL seasons, so the games comparison stops at 2019.
Continues below
Players drafted twice count once, at their final selection. Only NHL regular-season games count; playoff games and games in the WHA (the World Hockey Association, a rival major league of 1972-79) do not. What this can't test: why the roads shifted. The Soviet collapse, the development program's founding in 1996, scouting budgets and draft-age rules line up in time with some of the changes, but lining up in time doesn't show cause. Rockström's account is one team's reason, not the league's. The 1995 drop in American picks is unexplained. The games measure can't separate a league that develops players better from one that is simply scouted better. Seven seasons is short for a high schooler who goes to college first, which is why 10-season and whole-career windows were also checked. The college rule change has one draft behind it.
| Test | Result | Verdict |
|---|---|---|
| Major junior supplies a smaller share of the draft | 75.6% of picks (1969-73) to 39.4% (2022-26); first round 96% to 44% | Holds |
More tests below
| Test | Result | Verdict |
|---|---|---|
| Major junior picks out-produce their draft slots | -0.4 / 0.0 / -0.3 points across three eras (n = 5,511); 1987 alone: 20 of 85 made it against 20.2 expected | No difference |
| European and NCAA picks out-produce their slots | Europe 2.1 (SE 0.6), NCAA 1.8 (SE 0.8) points in 7 seasons; 1.5 (SE 0.7) and 2.6 (SE 1.0) in 10; 1.7 (SE 0.8) and 2.5 (SE 1.0) over careers; 0.8 and -0.7 by age 25 (a window that shortchanges older draftees) | Small edge |
| Europe’s late-round edge (picks after 100) persisted | 1969-2004: 90 of 1,048 (8.6%) against 5.7% expected; 2005-19: 21 of 452 (4.6%) against 4.4% | Faded after 2004 |
| US high-school picks under-produce their slots | -5.2 points (n = 907, 5.8 SE); negative in all three eras; -5.5 over 10 seasons, -4.9 over careers, -3.4 by age 25 | Holds |
Continues below
| Test | Result | Verdict |
|---|---|---|
| The high-school shortfall is in every part of the draft | Picks 1-30: 20 of 45 against 22.5 expected; 31-100: 13 of 216 against 32.3; 101+: 7 of 646 against 32.3 | Later rounds; first round could be luck |
| Routes differ in size at the draft | Average skater heights within 0.7 inches across routes in every era | No real difference |
| College picks took more of the top five before the NCAA rule change | 7 of 30 top-five picks in 2021-26 (6 of them in 2021-25) against 14 of 260 in 1969-2020 (about 1.6 expected at the old rate); 5 of the 7 had US junior games first | Real; predates the rule |
| The NCAA rule change has moved the draft | 2 of 11 NCAA picks in 2026 were former CHL players; NCAA share 4.9% | Too early |
Frequently asked questions
What percentage of NHL draft picks come from the CHL?
About 39% in the 2022-26 drafts (OHL, WHL and QMJHL together), down from 76% in 1969-73. In 2022-26 major junior still supplied more picks than any other road, though European clubs out-picked it in 2000, 2001, 2002 and 2021.
Do CHL or NCAA draft picks make the NHL more often?
Barely, once you account for where they were picked. Compared with how often players taken at the same draft slots reached 200 games within seven seasons, major junior picks from 1969-2019 produced 0.2 fewer players per 100 picks than their slots predicted and NCAA picks 1.8 more players per 100 picks than theirs. European and NCAA picks produced about 2 more players per 100 picks than their slots predicted, a small edge that also shows over 10 seasons and whole careers; US high schoolers produced about 5 fewer.
How many NHL draft picks are European?
About 33% of picks in 2022-26 were drafted out of European leagues, led by Sweden (12.9% of all picks), Russia (9.5%) and Finland (6.4%). The peak was 41.2% in 2001.
When did CHL players become eligible for NCAA hockey?
The NCAA Division I Council voted on November 7, 2024 to let CHL players keep Division I eligibility, effective August 1, 2025. Players who sign NHL contracts remain ineligible, and Division III rules did not change.
More in Bodies
- Verdict: Data story
The 1,000-Game Club Had Its Biggest Year. Getting In Is No Easier.
- Verdict: Data story
Nobody Drafted Martin St. Louis
- Verdict: Data story
Picks Six to Ten Got Safer. After Pick 30, Nothing Did.
- Verdict: Data story
How Much Lighter Did the NHL Really Get?
- Verdict: Data story
The NHL Is Getting Older Again
- Verdict: Data story
Connor McDavid Got Faster. The League, Not So Much.
Sources and method
- NHL Records: draft history API, every pick since 1963 with amateur club and league (2026 draft shown)
- NHL stats (NHL.com), every season since 1917-18
- NHL stats API: skater and goalie season summaries (games played by player)
- NHL.com player career records (api-web player landing, Gavin McKenna shown)
- NHL.com, Toronto Maple Leafs: Maple Leafs draft forward Gavin McKenna with the first overall selection (June 26, 2026)
- ESPN, Greg Wyshynski: Why Gavin McKenna’s Penn State commitment is a game changer (July 17, 2025)
- College Hockey Inc.: NCAA Division I Council votes to make CHL players eligible (Nov. 7, 2024)
- ESPN: NCAA makes Canadian Hockey League players eligible for Division I (Nov. 7, 2024)
- NHL.com/fi, Risto Pakarinen: Ensimmäinen rikkoo aina rajoja (Tommi Salmelainen, Feb. 14, 2024; in Finnish)
- UPI Archives: Quebec makes Swede first pick in NHL draft (June 17, 1989)
- Detroit Hockey Now, Bob Duff: Former Red Wings scout Rockstrom remembers discovery of Lidstrom (Aug. 5, 2025)
- ESPN, SchoolSports.com: NHL Draft Fast Facts (first US high schoolers drafted)
- ESPN Page 2: Prep stars take (SI) cover (Bobby Carpenter, Feb. 23, 1981)
- College Hockey Inc.: NCAA Pro-File, Tony Amonte and John LeClair (Nov. 28, 2010)
- ESPN, Ryan S. Clark: Inside stories behind the start of the US National Team Development Program (Dec. 21, 2023)
- USA Hockey via OurSports Central: Record-breaking 17 players selected directly from NTDP in 2019 NHL Draft
Research and analysis, not betting advice. Past results do not guarantee future returns; bet only what you can afford to lose.