Special Project Baseball

The Expansion Penalty

It began as one fan's question about his team. It turned into a study of what it costs to be born into baseball late, and how much of that is arithmetic, geography and money.

Question from Patrick Ishmael, a longtime Royals fan Analysis by Claude (automated) Data MLB seasons through 2025 · 30 franchises · 65 folded clubs Status Published October 2026 Jump to the Royals ↓

How this started

This began as a personal project. This project exists because Patrick Ishmael is a longtime Kansas City Royals fan, and the original question was simple: are the Royals unusually bad compared with other teams, expansion and not? His own working conclusion, before any of this analysis, was that they are middling: a poor all-time record, partly because they arrived late; not always in the playoffs; but strong in World Series appearances and wins for a club of their kind.

Testing that meant first answering a bigger question, what does it even mean to have been an expansion team? As a class, MLB's 14 expansion franchises have losing records and the 16 older clubs have winning ones, so the Royals' record means little until that gap is understood. Is it survivor bias (the old clubs are the ones that did not fold), head start (New York and St. Louis were big cities when baseball was a regional game), the start-up hole every new club digs, or money? The answers follow, and the Royals get their own section.

Who wrote this

This is not Patrick's writing. It is an automated analysis, run by Claude at his request on research he began, and should be read as a working study rather than a finished argument. Every figure is computed in the accompanying workbook with live formulas, so each can be checked and re-run.

What we found

Being an expansion club costs about 2% of win percentage even after the start-up hole is removed, and the hole itself is deep. Throughout this page, percentages are percentage points of win percentage: 1% is .010, about 1.6 wins over a 162-game season. Survivor bias is real but belongs to the 19th century. Market size mattered a great deal before 1961 and since then works mainly through payroll.

1 · Survivor bias

The old clubs won partly because they were the ones left

Include every club that ever played in the majors and the picture changes. In the 1870s–1900 era the 30 clubs that survived to today played .537 ball, against .465 for the 65 clubs that folded. That is a real filter: weak clubs disappeared. But it ends by 1901. In the closed 16-club league (1901–60) the survivors were an average .500, because the league's standings must sum to .500.

Pooled win percentage by group

Dashed line = .500. Hover or tap a bar.

Reading it: survivor bias explains the 19th-century glow around the old names, not the 1961–2025 gap. Since 1961 the older clubs have beaten the expansion clubs by almost 3% on their own merits (.511 vs .483), no filter required. No expansion club has ever folded.

2 · The hole

The first decade is a real hole, but not the whole gap

A new club loses heavily at the start, and then even a .500 club never recovers the games already lost. The measure below is each club's win percentage relative to .500 by age. Expansion clubs run about 7.7% below .500 in years 0–4, 3.2% below in years 5–9, and then roughly level.

Expansion-club win percentage versus .500, by age

70 club-seasons per five-year age block. Each bar = average across 14 clubs.

Established clubs average about +0.6% over the same calendar span. A club has to get past age ten before the comparison is fair.

So judge each club after the hole. Below, each club's all-time record (hollow) is moved to its hole-free record (filled): after season ten for expansion clubs, and 1961–2025 for the older clubs.

All-time versus hole-free win percentage

Sorted by hole-free record. Red = expansion, navy = established.

ExpansionEstablished

Without the hole the average expansion club is .491 against .511 for the older group (all-time: .482 vs .508). The hole closes only about a quarter of the 2.6% gap; the rest is genuine. Four expansion clubs are at or above .500 once the hole is removed, and two are above the older-club average. Tampa Bay ranks third of 30.

The question that started this

Are the Royals unusually bad? No: middling, with an unusually quick start

This study began as a check on Patrick's own team, as the introduction explains. Against the 13 other expansion clubs and the 16 older ones, Kansas City is a middle-of-the-pack franchise whose record looks worse than its trophy case. They did not dig much of a hole, though. Their first five seasons left them 17.5 games under .500, among the shallowest holes of the 14, and they were above water by year nine. The long middle is what costs them.

First ten seasons, club by club

Pooled win percentage over each expansion club's first ten seasons. Sorted high to low.

Reading it: only two of the 14 expansion clubs were above .500 across their first ten seasons: the Royals (.518) and the Diamondbacks (.505), and Arizona's came with a title in year four. The other twelve were all below .480, and the average expansion club was .446. This comparison is among expansion clubs only. Established clubs have no comparable first decade to measure.

Season by season, one to one

Cumulative games above or below .500, counted from each club's first season (year 1 = debut). Pick a club to highlight, and a second to compare.

Highlight Compare with

Other expansion clubs

The Royals bottomed out at 28 games under .500 in year two, climbed above zero in year nine (1977) and kept building to about 80 games over .500 in 1989 (year 21). Most other expansion clubs are deep below zero for most of their first 25 years. Arizona is the only comparable curve, and it started above .500. Kansas City's curve is a shallow hole, a fast climb, and a slow giveback after 1989.

What the numbers say about Kansas City

Record: .476 all-time, 11th of 14 expansion clubs and 28th of 30 once the first ten seasons are removed (.468). The Royals' problem is the stretch after the early years, not the start.
Two eras: 1976–85 they made the playoffs seven times (.528 over the 1970s, .529 over the 1980s). Then came a .468 decade in the 1990s and a .415 decade in the 2000s, with no playoff trip from 1986 to 2013.
Trophies for their class: two titles (1985, 2015) and four pennants (1980, 1985, 2014, 2015), third among the 14 on both. They are in the playoffs less often than most (8th per season played), but they convert them.
Market: a metro of 2.25M puts them second smallest among expansion clubs. Since 1995 they run about 4.3% below what that market predicts, 13th of 14, so size alone does not excuse the record. Payroll is the likely channel, and the 1990s and 2000s are where it shows.
Verdict: middling for a franchise that had the best first decade of any expansion club, with a World Series résumé better than the standings suggest.

3 · Population, geography and money

Market size mattered before 1961. It now works through payroll.

Each team-season is compared with its league average, against how big its home market was relative to the other clubs that year. The pull of size was strong in the closed league and has faded steadily.

Wins gained per doubling of market population (162-game season)

Each bar is a regression over the era shown.

In the 16-club era a market twice as large was worth about 4 wins a year. Since 1961 it is about 1, and since 1990 about 1.4. Adding payroll to the model drives the market coefficient to essentially zero: bigger cities spend more (correlation .42), and spending wins (payroll versus win percentage, .36). A payroll twice the league norm is worth about 7.5 wins a season, and market size adds nothing beyond it.

Metro population versus all-time win percentage

Metro population per club on a log scale. Hover or tap a dot.

ExpansionEstablished

Across clubs, doubling the metro area goes with about +2.4% of win percentage (correlation 0.35, so it explains about an eighth of the variation). The Yankees are the outlier: after adjusting for market and expansion status they still sit about 4.3% above everyone else.

Relocation does not settle it either. Across the ten moves, win percentage versus the league averaged 3.8% better in the ten seasons after than the ten before. Moves to a larger city improved by 6.4% (two moves) and the rest by 3.2%, so a new market is not obviously the cause. These are tiny samples and descriptive only.

4 · The decade index

Best teams of each decade after accounting for all of it

The index is actual win percentage minus the win percentage expected given market size, expansion status (including the first-ten-years flag) and, from the 1990s on, payroll. It is expressed in wins per 162 games. Choose a decade; "payroll-adj" periods also correct for spending.

Payroll is unavailable for 1980–84, so the 1980s are market-adjusted only. Payroll for 2017–2025 is from press compilations (Spotrac, AP, theScore) and should be treated as approximate.

5 · All 30 clubs

The ledger

Click any column to sort. "Over/under market" is the average win percentage in 1995–2025 versus what the club's market size alone predicts.

Method, limits and sources

Records: Lahman/Baseball Databank through 2025 and a coordinator-maintained spreadsheet (through 2026) of Baseball-Reference figures. The 2026 season is in the spreadsheet but excluded from the regressions.
Folded clubs: 65 major-league clubs from 1871–1915 (National Association, American Association, Union Association, Players' League, Federal League and early NL clubs). The median club lasted one season. Negro leagues are not included.
Market population: decennial census city populations, interpolated, with clubs sharing a city splitting it equally; 2020 metro-area figures for the cross-section. Several figures (San Francisco series, early Los Angeles, Boston 1960, Montreal series, a few 2020 metros) come from general reference and are flagged in the workbook for verification.
Payroll: Lahman salaries 1985–2016; 2017–2025 from press compilations. Opening-day versus end-of-year definitions differ between sources.
Limits: these are associations, not causal estimates. Payroll and market size are strongly correlated, so the split between them is soft. Fourteen expansion clubs is a small sample and the results for any single club are noisy. Postseason success is shown but not weighted into the decade index.

The numbers on this page were extracted from the workbook by script and checked against an independent recalculation.

Download the workbook (.xlsx)