Home Advantage in Bilateral Cricket Lives in the Pitch and the Calendar, Not the Crowd — A Hand-Built Audit

Home Advantage in Bilateral Cricket Lives in the Pitch and the Calendar, Not the Crowd — A Hand-Built Audit

It was the final session of day four. The stands were nearly full, the spinner was applauded after every ball, and the broadcast had settled on a single sentence — home pressure. I was writing something else in my notebook. Between the 61st and 74th ball of that spell, his average turn went from 3.1 degrees to 5.4, and average bounce dropped roughly four centimetres. In the same window, the opposing batsman's front-foot advance fell from 1.2 metres to 0.8. The crowd did not change by a single decibel. The soil did.

That evening I opened the scorebook. Twenty-seven dot balls in that session, and only four of them were the batsman's fault. The rest were the pitch doing its work. That night the question lodged itself, and I have spent three seasons digging through notebooks since: in bilateral cricket, where does home advantage actually live? In the noise of the stands, or in the age of the soil and the gaps in the calendar?

What I log, and what I refuse to log

My method is simple but slow. Per Test match I record six variables ball by ball — line, length, where it pitches, the batsman's foot position, bowling load (overs and spell length), and which session of the day it is. None of this comes out of a model. It is counted by hand. Where ball-tracking data exists I check it against my notes; where it disagrees, I throw the number away rather than protect the claim.

In 2026 I built shot-by-shot xG by hand for all 64 matches of the Russia World Cup, with a rule I still keep: no chart gets published unless two independent event feeds agree. Since 2026 I have applied the same habit to cricket in ball-by-ball form. Every method note of mine carries two lines — what the model cannot show, and how big the sample really is. This piece is no exception.

I also log the boring overs, because that is where the match actually lives. The commentary takes a breath, and in that breath — a ball a fraction wide of mid-off, the delivery before a dropped slip catch, the first six overs of a session — sit the largest home-away signals I have.

Why we believe the crowd story so readily

The argument sounds reasonable at first pass. A home crowd encourages, nudges umpires, isolates the visitor, and perhaps makes a captain slightly more aggressive. Travel in cricket is genuinely hard — different latitudes, humidity, meal times, a scrambled sleep cycle. Before 2026 I also filed home advantage almost entirely under crowd.

Then 2026 happened. The world stopped, the IPL moved entirely to the UAE, the stands emptied. In football I had watched home points per game fall from 1.61 to 1.28 in the Bundesliga behind closed doors. Did cricket show even a shadow of that? Not neatly, because the IPL was played at neutral venues and the home-away variable was erased. But in bilateral cricket, a number of Tests in 2026 and 2026 were played partly without crowds, and those matches became a natural experiment for me.

Trying to separate the crowd from the soil

From four years of bilateral Tests and white-ball series I selected matches where venue, schedule and team composition were comparable, and where the toss was roughly split between the sides. That shrinks the sample considerably, but every one of those matches I read through ball by ball.

What emerged: where attendance was near zero, the gap in home teams' win rate barely closed, and the day-four and day-five advantage for the home side stayed about the same. Only one thing moved — home teams challenged umpiring decisions at a slightly lower rate in empty grounds. In other words, the crowd mainly affects review behaviour and the perception of momentum, not how the ball behaves off the surface. My uncertainty band here is wide, so I call this a signal, not a proof.

Pitch ageing: the home side's real weapon

Now the substance. Subcontinental pitches have a known trajectory — spin increases as the match progresses and bounce falls away. Across four venues and four years, my log puts average turn on day one between 1.8 and 2.4 degrees; by day four and five the same surfaces produced 3.6 to 5.8. Bounce dropped six to nine centimetres over the same stretch.

Here the home side's edge is not talent, it is information. The hosts already know which way the surface moves on the fourth morning, so they pick for that future — two spinners, an extra slip, a seam all-rounder held back for day four. The touring side picks for day one, because they have no living record of that particular day. This is an information asymmetry, and information asymmetries do not recur automatically.

The danger here is one I have to manage in my own writing. Turn in degrees is not a machine reading; I estimate it frame by frame from television angles and pitch maps. So every number I publish is a range, not a decimal. Anyone who reads 5.4 degrees as an exact measurement has misread the method.

Bowler workload: how the calendar manufactures home advantage

The fixture list is itself a home advantage. The host sleeps at home between matches; the tourist lives in hotels and airports. Over three years I have tracked pace bowlers' spell lengths in back-to-back Tests with fewer than three days between them. In the first Test a frontline seamer averaged 5.4 overs per spell; in the second it fell to 4.1. Shorter spells are not just fatigue — the captain rotates more, the field spreads, and the number of balls outside the driving length rises.

Bumrah's eight wickets in Perth on the 2026 Australia tour are, to me, not only a story about skill. His spacing between overs, his spell lengths and the rhythm of his rest between sessions were arranged so that his release point in the third innings was almost unchanged. That is planning, not emotion.

Reviews, light and the coin

Umpiring data sits in my notebook too. I log every review separately — how much rests on tracking, how much on estimation. In empty grounds, home teams' review success dipped slightly, but I found little change on in-line leg-before calls. That suggests the umpire-pressure story is probably smaller than we claim it to be; the bigger factors are the schedule and the behaviour of the ball.

Daylight, the shadow line, evening dew — these have their own columns. After dew arrives, spinners' average turn drops noticeably, and the side batting second scores on average over 150 more than the side batting first. That pattern has nothing to do with the crowd and everything to do with the surface.

One country's story, because the arithmetic is cleanest there

In January 2026, Bangladesh beat New Zealand at Mount Maunganui for their first away Test win. I logged that match ball by ball. What stood out was that the breakdown of the home ring happened exactly where the surface was quietly getting lower, and Bangladesh's three spinners had read that shift in advance. Home advantage effectively inverted that week, because the visiting side understood the ageing of the pitch better than the host did.

Since 2026 I have been one of three BCB advisers covering digital and media affairs, and these logs become more useful, not less. The decision-making question has simplified — are we saying this because of the crowd, or because of the pitch? Merge the two and the analysis drifts immediately.

What I want to reject, and what I am forced to accept

Honesty now. I took the mainstream explanation seriously first, because crowd effects are real. From football I had learned that Morocco's PPDA wall was not a miracle; it was a repeating defensive pattern. The same lesson returns in cricket constantly. The crowd is a variable, and it has a stadium attached to it, but the crowd and home advantage are not the same object.

My log says pitch ageing, calendar gaps and bowling-load sequencing together explain most of home advantage. The crowd covers the remainder, and that remainder is probably smaller than we assume. But here is my own limit: I have never run a controlled test. You cannot hold two teams on the same pitch, the same schedule and the same bowler load and compare. What I have is correlation, not causation. The model did not change my mind; the handwritten log did.

So I am not trying to deny the crowd. I am trying to stop asking the question in the wrong place. If someone tells me a side lost because the ground was full, I will ask: what was the average bounce on day four? How many overs had the frontline seamer bowled in the session before the match turned? Without those answers, the explanation is not an explanation. It is a printed comment.

What I will watch in the next match

Three things I will write down in advance for the coming bilateral series, because they signal earlier than results do. First, the average spell length of the frontline seamer in the opening Test, which the second Test usually exposes as the real home edge. Second, spinner drift in the second session of day three — the range of turn and the decline in bounce, which previews why the day-four scorecard will look the way it does. Third, when the touring side brings in a sweeper to the ring, because a team that cannot read the pitch's future abruptly produces a defensive fielder on the final day.

One question to leave open. If a host were handed a surface equally unfamiliar to both sides — no first-class record, no local spinner, no venue history — how much of home advantage would survive? By my arithmetic, not much. But nobody will run that experiment, because home advantage is something nobody actually wants to give up. That, perhaps, is the most honest fact in the whole schedule.