Nations · 6 editions · 17 countries in the long run · 95/96–25/26
The same five questions, asked of 6 nations.
Each edition — Czechia, Denmark, England, Spain, Germany and Norway — measures its pool the same way: how many players reach the strongest leagues per million, how many minutes the home league gives its own under-21s, when the first move abroad comes, how strong the home league is, and what the national-team squad is built from. Side by side, the table shows where the editions differ on the same definitions; the decomposition describes how each gap lines up with three measured channels; the long run dates when each country's presence in the Big-5 most probably changed level. The page is exploratory and descriptive, was not pre-registered and identifies no cause.
Data: each edition's snapshot under data/snapshot/ (FBref tables fetched 14 September 2026) · code: src/nations_compare.py, site/build_nations.py · status: exploratory, not pre-registered
Findings
1
Czech players with 450+ minutes in the Big-5 leagues: 26 at the 2007/08 peak, 10 in 2025/26.
The change-point model’s most probable season for the fall is 2014/15 (posterior 58 %), a level change of ×0.59 (90 % HDI 0.37–1.02); that interval includes ×1.00, so the fall is not distinguishable from no change at the 90 % level. Its most probable season for the other step, a rise, is 2000/01 (posterior 56 %), ×1.67 (90 % HDI 0.74–2.40); that interval includes ×1.00. A break model dates a change in level; it identifies no cause.
The median age of the nation’s Big-5 players was 28 at the 2007/08 peak and 32 in 2014/15, when no player made a first Big-5 season of 450+ minutes.
2
The per-head gap to Norway and Denmark, split over three measured channels (descriptive).
In a ridge fit over 8 countries, against Norway (a gap of 6.98 per million) youth minutes go with +4.22 (90 % bootstrap interval +1.22 to +5.91), league strength with +5.12 (+1.90 to +6.77) and export age with 0.00, residual −2.37; against Denmark (a gap of 10.19 per million) youth minutes go with +7.35 (90 % bootstrap interval +2.17 to +10.27), league strength with +1.67 (+0.65 to +2.26) and export age with 0.00, residual +1.17.
The league-strength coefficient in this fit is negative (−56.5 per unit of the league multiplier): a weaker home league goes with more players in the top-ranked leagues per million, so a peer whose league is weaker than Czechia’s (m_L 0.665) gets a positive league-strength contribution. The export-age channel is zero by construction: Czechia, Norway and Denmark share the same median age at the first top-9 season among recent entrants (22). The split describes an association across few, correlated national measures; it is not a causal accounting.
3
Czechia’s own under-21 nationals’ share of home-league minutes over 6 seasons: a trend of −0.39 pp a season (90 % CI −1.26 to +0.48).
The share by season, 2020/21 to 2025/26: 10.3, 8.6, 9.0, 7.9, 11.1, 6.4 %. An OLS line through the 6 seasons falls by 0.39 percentage points a season (90 % CI −1.26 to +0.48 pp), an interval that includes zero. The last season’s share is a floor (FBref leaves part of the league’s rows without a nationality): at its ceiling of 7.6 % the slope is −0.21 pp a season (90 % CI −0.95 to +0.54). The last step, 2024/25 to 2025/26, for the comparison countries and for other leagues whose share fell by 2 points or more: Norway 10.5 → 11.5 %; Denmark 13.2 → 15.3 %; Poland 12.5 → 9.3 %; Spain 7.8 → 4.8 %.
4
Where Czechia ranks among 9 countries on the pathway measures, 2025/26 (1 = highest share, youngest age, most moves).
Under-21 share of home-league minutes 6.4 % (a floor; up to 7.6 %), 8 of 8 (highest: Denmark 15.3 %); regular under-21 starters per club 1.06, 8 of 8 (highest: Denmark 2.50); median age at the first season with real playing time abroad 24 (n = 32), 5–7 of 9 from youngest (youngest: Norway 22.5); first moves abroad in the covered seasons 32, 8 of 9 (most: Denmark 112); home-league UEFA-coefficient multiplier ×0.43, 2 of 9; median share of club minutes for exports 50 % (90 % bootstrap interval 22–66 %, n = 26), 3 of 9 by the median; 54 % of the 48 players who went from the home league to a top-9 league since 2020/21 left from Sparta Prague or Slavia Prague. These are descriptive ranks on medians and counts; apart from the minutes share they carry no interval, so neighbouring ranks can swap with a few players; none of them is shown to cause the per-head count.
5
Youth minutes in Germany and England after two academy reforms (descriptive).
Germany after its DFB/DFL academy licensing (2001/02; Honigstein, 2015): its own under-21 nationals played 5 % of Bundesliga minutes the season before, 11 % at the highest point within twelve seasons (2009/10) and 6 % in 2025/26. England after its Elite Player Performance Plan (2012/13; Premier League, 2011): its own under-21 nationals played 3 % of Premier League minutes the season before, 6 % at the highest point within twelve seasons (2019/20) and 3 % in 2025/26. Each is one sequence in one country without a comparison group, so it describes timing and does not estimate an effect of either reform.
Each finding is generated from the numbers below at build time; all are descriptive, and the evidence and its limits follow in order.
A · side by side
Six editions on the same pathway measures, last completed season.
How to read it: each edition's own headline numbers, last completed season; the brightest value in a row marks the edition ranked first on that row.
what the rows mean
Per million counts players with a season in the nine strongest leagues. For a nation whose home league is one of them (England, Germany, Spain) that includes the home league — which is why the rank is against the nation's own peer set, not across editions. Under-21 minutes and regular starters are measured in the home league; the first move abroad is the median age at a player's first season in a headline league; the squad row is the share of the last full national-team squad playing in a top-9 league.
B · the gap to the peers, split over three measured channels (descriptive)
Czechia against Norway (a gap of 6.98 per million) the largest segment goes with home-league strength, +5.12 (90 % bootstrap interval +1.90 to +6.77); against Denmark (a gap of 10.19 per million) the largest segment goes with youth minutes at home, +7.35 (90 % bootstrap interval +2.17 to +10.27).
how the decomposition works, and what it cannot say
Each edition's own decomposition: a ridge regression of players-per-million on youth share, home-league strength and the age of the first move across that edition's eight or nine peers, then a Oaxaca–Blinder-style accounting of the gap to each contrast ((Oaxaca, 1973); (Blinder, 1973)). A channel's share is how much of the gap goes with that channel at the fitted coefficients; the channels can sum to more than the gap and the residual takes the rest. A positive league-strength segment can mean a weaker home league for the comparison country: in Czechia's fit the league-strength coefficient is negative. The split describes an association across few, correlated countries; it does not say what would happen if the youth share rose, which would need change over time within a country.
Six seasons at home — the change the cross-section cannot see
How to read it: the share of each home league's minutes played by its own under-21s, every complete season the pipeline covers; the chips choose the countries (they also drive the long-run charts below).
How to read it: each country's change in that share against its change in Big-5 presence over the same seasons — the within-country picture, one point per country, 16 countries and six seasons: a direction, not an estimate.
The cross-section behind it
How to read it: one point per country (16), last completed season; pick the measure on the x-axis. The dashed line is a plain fit for the direction only; the three measures move together across countries, so the line describes an association.
C · the long run, 95/96–25/26
Who fell, who rose, and when — every country's presence in the Big-5, with two dated steps each.
How to read it: players with 450+ minutes in a Big-5 league that season, per million; a diamond is a dated step (hover for size and certainty); a dashed rule is a documented reform — a date, not a cause. Every fit was checked against a convergence rule written down on 30 September 2026, before the refit (four chains, 2000 tuning steps and 2000 draws, a stricter step size only where needed; max R-hat ≤ 1.01, bulk and tail effective sample size ≥ 400, no divergent transitions); a country whose fit fails has no dated steps here. The diagnostics are in the table below.
convergence of every country's fit
| Country | max R-hat | min bulk ESS | min tail ESS | divergences | stage | rule |
|---|---|---|---|---|---|---|
| Austria | 1.002 | 1 403 | 1 427 | 0 | 2 | passes |
| Belgium | 1.002 | 1 864 | 2 691 | 0 | 2 | passes |
| Croatia | 1.002 | 2 862 | 2 981 | 0 | 2 | passes |
| Czechia | 1.004 | 2 321 | 1 830 | 0 | 2 | passes |
| Denmark | 1.003 | 2 265 | 2 458 | 1 | 2 | fails: no dated steps |
| England | 1.003 | 2 022 | 2 346 | 0 | 1 | passes |
| France | 1.001 | 2 754 | 2 719 | 0 | 2 | passes |
| Germany | 1.003 | 1 730 | 2 128 | 0 | 2 | passes |
| Hungary | 1.002 | 3 438 | 3 361 | 0 | 2 | passes |
| Italy | 1.002 | 2 723 | 3 665 | 0 | 1 | passes |
| Netherlands | 1.003 | 2 121 | 2 669 | 0 | 2 | passes |
| Norway | 1.002 | 1 327 | 1 169 | 0 | 1 | passes |
| Poland | 1.002 | 2 612 | 2 438 | 0 | 2 | passes |
| Portugal | 1.004 | 2 196 | 2 392 | 0 | 1 | passes |
| Slovakia | 1.002 | 3 352 | 3 343 | 0 | 2 | passes |
| Spain | 1.006 | 1 198 | 1 162 | 0 | 1 | passes |
| Switzerland | 1.003 | 1 484 | 1 204 | 1 | 2 | fails: no dated steps |
Stage 1: target acceptance 0.95; stage 2: 0.99, used only for fits that failed stage 1. Protocol: design/nations-refit-protocol.md in the repository.
how we know, and the two reform dates
FBref's season tables for the Premier League, Serie A, La Liga, Bundesliga and Ligue 1, from the first season all five are covered. Each country gets the report's two-step change-point model ((Adams and MacKay, 2007), in a batch, two-break setting). For a Big-5 nation the count includes its own league, so its series is mostly about how international that league became. Reform markers: Germany 01/02, DFB/DFL academy licensing (Honigstein, 2015); England 12/13, Elite Player Performance Plan (Premier League, 2011). A marker cannot tell whether a series moved because of a reform; it shows the timing, beside countries without such a reform.
every country's steps as a table
"Shape vs CZE" is the correlation of the per-million series with Czechia's — the countries whose long run is shaped most like Czechia's, whatever their level. Each step comes with its posterior and interval in the table; a step whose interval includes no change is not distinguishable from none.
When a country's players arrive, and how young
How to read it: the same countries as above; age at a player's first Big-5 season of 450+ minutes (three-season median, because a small nation sends two or three a year), or the under-23 share of the nation's Big-5 minutes, or the count of first seasons; the chart shows whether a rise came with younger arrivals and how the age of Czechia's arrivals moved around its fall.
Youth minutes in the Big-5 leagues themselves
How to read it: the share of each Big-5 league's minutes played by its own under-21s (age at 1 July), season by season — the one youth series the data carries back this far.
D · what this can and cannot say
- Can: put 6 nations on one ruler; describe how a gap lines up with measured channels; date when a country's Big-5 presence changed level; show whether a reform preceded a change.
- Cannot: attribute a change to a reform (no counterfactual, no randomisation); separate channels that move together; see youth minutes in the small leagues before FBref covers them.
- Next: the same decomposition within a country over time, and a synthetic-control read of the two dated reforms against the countries that did nothing — both need more seasons than the small leagues have yet.