The Empty-Stand Residual: A Three-Season Audit of Home Advantage in Asian Cricket

Hook — A Circled Row

There is an old steel almirah in my Delhi flat, and in its second drawer sit clippings of a 2026 newsletter. A November issue of 'Expected Delhi' — two thousand subscribers then. That issue carried a table with twenty-two rows. The nineteenth row has a pencil circle around it that I have never erased. The row said that in one Indian Super League season, Bengaluru FC had scored 27 goals from 22.4 xG — a 4.6 overperformance. Why did I circle it? The number did not surprise me. What surprised me was a narrow column sitting beside it, holding exactly one non-pitch variable: crowd attendance.

Four years later, on 23 March 2026, the Maharashtra Cricket Association Stadium in Pune sat almost empty. I watched from my Delhi desk, matching the scorecard against ball-by-ball data, because a question had been sitting in my notebook for a long time. If part of a home side's advantage comes from the voices in the stands, how much of that advantage survives when the voices are removed? What my script showed that night refused a simple answer. The home side won the match, but the 'home run residual' in my model — extra runs per wicket for the home team after adjusting for pitch, innings and toss — sank below the normal Asian range. A win arrived. The advantage did not.

That gap became my next three years of work. When a stadium empties, home advantage does not die; it changes shape. Some people called it noise. Some said an empty ground makes everything equal. My archive says something else, and proving it cost me 1,284 matches, 612 franchise matches, and forty-seven different pitches across one continent.

Context — Variables Before Opinions

I write my sample and method before any claim, because forgetting them once taught me how expensive correction is. The lesson came in 2026.

That year a new media house hired me to build a Russia World Cup model. Twenty-two days of work produced one number: France at 18.4% title probability, the tournament's highest. It rested on two figures still written on the first page of my old notebook: 0.8 xGA per game and a PPDA of 9.8. France won. Everyone said the model worked. I wrote then that the 18.4% model did not predict France; it predicted my next five years.

What does that mean? It means a successful output creates a private damage — the loss of the limits of your own confidence. Since then every piece I write carries a methodology note of at least five hundred words. When editors ask for hot takes, I send three pages of variable definitions instead.

Let me clear the ground for this piece. My archive holds men's international matches from 2026 to 2026, of which 1,284 were played at forty-seven Asian venues. Add franchise cricket — IPL, BPL, PSL, LPL, ILT20 — 612 matches. Matches played with crowds below five percent of capacity carry a separate badge; Asia produced seventy-eight of them between June 2026 and December 2026.

I tag four environmental variables on every match, because without them home advantage becomes a false number.

First, pitch inheritance — how many wickets spin and pace have taken at that venue over the last three seasons, and whether that matches what curators announced before the game.

Second, travel fatigue — flight hours from the previous venue, plus time-zone shift. In Asia this matters, because Kandy to Chattogram or Karachi to Colombo are distances that break a body's clock.

Third, dew and evening humidity — how wet the ball gets under lights. This has nothing to do with crowds and everything to do with pitch behaviour.

Fourth, schedule density — balls bowled in how many consecutive days. I added this after the Tokyo Olympics, where Pedri played six matches in eighteen days, and my workload model had already flagged the risk.

Core — The Evidence Chain

1. The empty-stand residual. In May 2026, with world sport paused, I sat down with 56 Bundesliga matches played behind closed doors. Two numbers came out: home advantage fell from 0.42 to 0.17 goals per game, and home teams' PPDA worsened by 1.3 units. The second number matters more, because PPDA measures pressing aggression. Fewer goals means fewer points; less pressing means fear.

When the stadiums emptied, the home advantage stayed and stared back.

Across Asia's seventy-eight crowdless matches I found three separate layers, and this separation is the genuinely new part of the research, because it shows the advantage is not one slab.

Layer one — captaincy advantage. Toss, field placement, DRS decisions: a home captain does these with long-term memory of the pitch. In Asia between 2026 and 2026 my home run residual was +6.8 runs per wicket. In the crowdless sample it fell to +4.1. A drop of 2.7.

Layer two — bowling-environment advantage. A home side usually knows the two spinners who understand its own surface; an away side cannot learn those angles in ten days. In crowdless matches home spinners' economy improved by 0.31 runs per over — meaning this advantage did not shrink when the crowd vanished. It grew slightly. The reason is simple: less pressing means defensive bowling is punished less in an empty ground.

Layer three — the run-chase advantage. This is the most fragile. Where the home side batted first, the home residual held nearly intact at +3.9. Where the home side chased, it fell to +1.4. In an empty ground there is no sound off the boundary, and that creates a small but measurable slack in a batter's decision speed.

2. Advantage by format. Here is a pattern I first saw in a Delhi newsletter, long before the data had a name. I used to call it 'format sensitivity' because I could not name it. Now I can.

Home advantage is most stable in Test cricket, because over five days a pitch changes character, and the side that has studied that change for five seasons owns deeper research. My figure for Asian Tests: +8.2 runs per wicket. In the crowdless sample it held at +7.1. A loss of only 1.1.

ODI: +5.4, falling to +3.2 without crowds.

T20 is the least stable, most toss-dependent surface of all — home advantage +2.1, falling to +0.6 behind closed doors, statistically near zero. In a twenty-over game, a crowd's roar pulls decisions forward before a bowler finishes his run-up, and that pressure is invisible in an empty ground.

3. Travel fatigue — what stays when crowds leave. In my 2026 essay I argued that no single variable explains home advantage. That still holds.

In my dataset, away sides with more than 4,000 flight hours lost 3.2 extra runs per wicket — and that loss did not change with crowd presence or absence. Travel fatigue and crowd pressure are separate things that we lazily bundle into one bag called home advantage.

The bigger find is the two-day gap effect. If a side gets only two days between matches, its second-match strike rate rises 4.7 and its dot-ball share rises 2.9 percent. The body recovers; the time available to read patterns does not.

4. Dew, over-rates, and a misunderstanding. In South Asian evening ODIs and T20s, the chasing side wins 62 percent of the time. Do not read that as a home bonus. In my model it is almost entirely a pitch-and-dew phenomenon, not a home-away label. Curiously, in crowdless evening matches the wet-ball gift stayed almost unchanged — from 62 to 59 percent. Empty or full, dew falls.

This is where a metric caveat is needed, and I have attached it to every piece since 2026: the same economy number says two different things on two different pitches. 7.4 economy at Rawalpindi's flat deck and 7.4 at Chennai's slow turner are not the same figure.

5. The 900-minute threshold and the Pedri lesson. Working the Euro 2026 commission in 2026, I tracked Pedri across six matches. The output: 65 progressive passes, 92 percent pass completion, zero goals. A match reporter would begin at that zero. My model rated Pedri's 8.3 progressive carries per 90 as elite. Pedri won the Young Player award at UEFA Euro 2026 — that is official tournament record. Spain reached the semifinal.

Two rules came out of that tournament and still govern my cricket writing.

Rule one — the 900-minute threshold. I do not judge a young batter or bowler before 900 minutes of field time. At 400 minutes you can see a pattern; whether it survives requires meeting the same situation twice.

Rule two — every eye-test claim carries a progressive-pass or carry map. Pedri proved that a player can sit at the centre of a match without scoring, if you choose the right metric.

My old notebook's last page carries a line I still read sometimes: s prior. It means every new match arrives with a prior in my head. The work is to write it down so you can catch yourself later.

6. Metric to market — the language of numbers on the auction table. What I write now is not analysis; it is translation, from tournament metric to auction table.

A junior on my team built a table last year before an IPL auction — four columns: powerplay strike rate, death-over economy, progressive carries per 90, total overs bowled across three seasons. Then he attached a 'context multiplier' to each column: which pitch, which tournament, what pressure. The same 'death economy of 8.9' pointed to two different valuations, because one bowler did it at Rawalpindi and the other at Chennai and Dubai.

Franchises do this work now, but most still skip the final step — sample context. Two hundred overs on home pitches and two hundred on neutral pitches are not the same experience. That difference becomes ten lakh rupees at auction.

There is a human layer the table never shows. A twenty-year-old who flew four thousand kilometres, bowled six overs on a neutral pitch and conceded seven runs — those seven runs differ from seven runs conceded at home. The auction table files both in the same box. That gap builds careers and destroys them.

Contrarian Angle — Correlation Is Not Causation

Now I argue against myself, because I know exactly where this piece is weakest.

I showed that the home residual fell in empty stadiums. That is a correlation. Correlation is not causation. Three other things changed at the same time, and separating them cost me real effort: project-based workload control, which increased rest; travel restrictions, which reduced fatigue; and bio-secure bubbles, which cut press conferences and practice time.

So is the 2.7-run drop about losing crowds? A large share probably is; a smaller share certainly is not — and I still lack a clean natural experiment to isolate that smaller share.

Second weakness: sample size. Seventy-eight Asian crowdless matches spread across forty-seven venues means no single venue has more than three samples. Venue-level claims under those conditions are cleverness, not science. So I make regional claims, and even there I refuse to push the error range past two decimals.

Third weakness: I avoid direct comparison with data outside Asia. My model's history taught me that mixing variables across continents makes numbers look clean while explaining nothing. Even so, I admit a bias: thinking about India's broadcast economy, I too often push non-Asian leagues aside. The right question is whether the same pitch behaviour that holds home advantage in an empty Lahore stadium holds on synthetic turf in Major League Cricket. Until I know, my conclusions should stay inside a regional boundary.

One more thing about my own profession. Analysts are walking into dressing rooms now. I do not call that bad, but I add a caveat: the rhythm of a match and the rhythm of a spreadsheet are not the same. The spreadsheet always lags, because a match ends in forty-five overs or five days, while the court of statistics convenes much later. That delay has a cost, and nobody refunds it to me.

Takeaway — Not a Conclusion, a Next-Round Signal

At sixty, I have learned that the quietest spreadsheet often has the loudest story.

For the coming season I hold three watch signals, and I keep them as questions rather than declarations, because what I own is a three-season audit, not a crystal ball.

One: on Asia's slow pitches, will the gap between a home spin pair's over-rate and dot-ball share widen? In my count, that gap owes something to the crowdless season even after crowds return, and home spinners will pay that debt.

Two: with T20 home advantage near zero, where did the home pitch go? The answer may sit at the toss — a side that bats first and posts 190 owns an advantage not on the pitch but inside the head.

Three: will the 900-minute threshold reach the auction table? If it does, some franchises will stop paying ten lakh extra for a twenty-two-year-old strike rate. A rising star is a culture, not merely a number — and cultures build on time, labour and patience.

What I can state with certainty is this: home advantage with empty stands and home advantage with full stands differ not in quantity but in nature. One is called pressure. The other is called memory. Anyone who confuses the two will pick the wrong team, overpay at auction, and sell the wrong story on broadcast.