The Mirpur Spreadsheet: When Home Advantage Becomes a Variable

I began with the live thread and ended with a broadcast truth. In the second T20I of the Bangladesh–Australia series at Mirpur's Sher-e-Bangla Stadium, Bangladesh's openers took 47 off the powerplay. The live thread had already started chanting "unbeatable at home." But my scorebook was blinking with an uncomfortable comparison: the same batting line-up posted a powerplay run rate of 6.8 on Australian soil, and 7.8 at Mirpur. One run per over is enough for some to declare a home advantage—but the spreadsheet remembers what the stadium forgets.

For five years I have run three separate spreadsheets: one for Mirpur, one for Chattogram, one for the Sydney Cricket Ground. Each has the same columns—powerplay run rate, dot-ball percentage, spin economy, death-over run rate, and wickets per match. The reason is simple. I do not trust the eye test until the data signs the same sheet. And across five years of sheets, one line keeps repeating: home advantage is not a single number. It is a variable whose value swings with four separate inputs—venue, pitch, travel, and crowd.

The interest did not arrive by accident. In 2026, when the A-League resumed in empty stadiums, I pulled the data from 24 matches. Home teams' expected goals (xG) fell from 1.45 to 1.12, while away teams' pressing intensity (PPDA) improved from 12.1 to 9.8. When the crowd leaves, a large slice of home advantage evaporates. Empty seats taught me that home advantage is a variable, not a myth—and I now carry that lesson into cricket.

So let me write the question down, because I do not write without a template: does Bangladesh's T20 batting and bowling genuinely differ at home from away—and if so, what share of that gap is pitch, what share travel, what share crowd, and what share simply the opposition's quality? I pre-register the variables, align the baselines, then apply the context coefficient. The conclusion comes last.

The first table is Mirpur's pitch profile. Across the last three seasons, the average first-innings T20 score at Mirpur is 148; the average second innings, 129. Chasing here means carrying a structural deficit of roughly 19 runs. Dot-ball percentage is 41% in the first innings and 48% in the second. That is not the pitch; that is the pitch's age. Mirpur's surface typically slows after the toss, grips more for spinners, and keeps low. The side batting first gains an easy truth: its batting happens in the best conditions.

Here lies a trap I note every toss. People read winning the toss as a tactical victory. Yet in 22 of the last 30 T20Is at Mirpur, the chasing side lost. The toss is an unfair coin here. Some say "the captain should have bowled first after winning the toss"—but the data says batting first is statistically the better bet at Mirpur, not merely instinct.

The second table is Bangladesh's home-away split. Over two years, Bangladesh's T20 powerplay run rate is 7.8 at home and 6.8 away. Death overs (16–20): 8.9 at home, 7.6 away. Spin economy: 6.1 at home, 7.4 away. Yet one number hides in the gap that nobody reads: wickets per match at home is 6.8, away 6.3. Bangladesh scores more at home, but loses more wickets too. This is not a win-loss ledger; it is a trade-off.

Here the model cracks against the standard narrative. The narrative says, "At home, Bangladesh builds spin pressure." The model says, "At home, Bangladesh builds spin pressure, but simultaneously raises its own batting risk, because a slower pitch makes singles hard, forcing a tilt toward boundaries." The second sentence is far more useful, because it feeds into the next match's decisions.

I deliberately keep a gap between the live thread and the final data. During a match, what I post is a timestamped hypothesis. After the match, I reconcile ball-by-ball and broadcast data and revise that hypothesis. This discipline has saved me many times. In one Mirpur match I thought live that Australia was falling behind in the powerplay because its run rate was under seven. Later I saw its dot-ball percentage was only 33%, with eight wickets in hand. It was not behind—it was building a base. A number is a witness; a trend is a confession.

The third table is the crowd effect. I pulled this from my 2026 empty-stadium coefficient, so it is a transfer—from football to cricket, as a context coefficient, not a colonial imposition. The result: with a full crowd, the home side's run rate gains roughly 0.6 to 0.8 runs per over, and spin economy saves 0.4 to 0.6. In an empty stadium, about 70% of both edges vanish. This is not just an emotion story. The crowd is a practical factor—fielders' drives, spinners' length, the umpire's subconscious margin. That is why I never lock the crowd into a box; I make it a slider that moves between 0.0 and 1.0. A packed Mirpur is a slider near 1; an empty one, near 0.3.

Now the part where I disagree with the common read. Home advantage is often treated as a quality of a team—"they are fearsome at home." My model says no. My model says that whoever plays at Mirpur will see its powerplay run rate rise by 0.5 to 0.7, purely from pitch and environment. The edge is written on the venue, not the team. That distinction changes pre-match forecasts.

Travel is another under-discussed variable. When Australia lands in Dhaka from Sydney, the time zone shifts eight hours, and its death-over economy in the first T20I drifts from 9.2 to 10.4. In my count, travel fatigue adds roughly 0.8 to 1.1 runs to death economy. Not huge—but in the first two matches of a series it is often decisive. By the third match, the coefficient falls close to zero.

The link is clear here. Part of Bangladesh's home success at Mirpur is team skill, part pitch, part crowd, part the opposition's jet lag. Without separating these four, what we do is not data analysis—it is convenient storytelling. And a number is a witness; a trend is a confession. The trend says the Mirpur edge is not in Bangladesh's control—it is in Mirpur's.

Let me say something contrarian, because my template demands it. Many assume Chattogram or Sylhet's pitches are more batting-friendly than Mirpur, so home advantage there is smaller. My data hints at the opposite. Chattogram's second-innings average is higher than Mirpur's, yet the home side's win ratio is nearly identical—because the edge there comes not from the pitch but from wind and dew. Once evening dew sets in, spin grip fades, and the side batting first knows it. The formula shifts with the venue, but the formula remains. That is my portable comparative framework—the same template applied across formats, leagues, and nations, with the context coefficient swapped. The framework travels; the cause does not.

Now the part most readers misread: correlation versus causation. Bangladesh plays well at Mirpur—true. Bangladesh plays well at Mirpur because Mirpur is Mirpur—only a possibility. I know three traps that nearly catch me every time. First, spreadsheet absolutism: my model says home run rate is 7.8, so I take it as truth. But cross-checking with ball-tracking and match reports reveals a chunk of that run rate came from fielding errors and soft dismissals. Model output is provisional; without video and report cross-checks, the number is half-true. Second, template lock-in: the urge to force every match into the same five sections is in my blood. But some matches break the template—a rain-shortened match has no meaningful powerplay data, because the innings are short. There I switch metrics to Duckworth-Lewis par scores and resource usage. Breaking the template is not weakness; it is honesty. Third, context-coefficient overfitting: the more variables I add, the smoother the story. So I now pre-register variables, run holdout tests, and publish sensitivity analyses. If four variables still explain only 40% of the variance, I write that down.

The fourth trap is my own creation: live-thread anchoring. My method begins with the live thread, and that early imprint often lingers in my analysis. In one Mirpur match I posted live that Bangladesh's middle overs had stalled. Later, reconciling ball-by-ball, I saw the stall was deliberate—they were saving wickets for the death, then took 52 off the last four overs. My live conclusion was wrong, and I corrected it. I now keep the live log and the final analysis in separate files, so the early imprint does not contaminate the final truth.

None of this means home advantage is imaginary. It means it is a measurable, shareable, shifting thing—which is exactly what makes it usable. A coach who treats home advantage as a fixed trait cannot use it. A coach who treats it as a four-variable slider can reshape training and team combination before every series. Take set-pieces. Working with Western Sydney Wanderers in 2026, I saw their set-piece xG rise from 0.18 to 0.31 per match in empty stadiums—because the routines had to change when the crowd's roar could not be relied on. Cricket's equivalent is powerplay field placement and death-over yorker planning. With a crowd, bowlers instinctively attack more; in an empty ground that courage dips, so slower balls and length variation must increase.

I will concede one limitation myself, because a caveat block is part of my template. My home-away split rests on a 28-match sample—small. One brilliant series can flip the whole trend. So I write a confidence band beside every number—home powerplay run rate 7.8 ± 0.9. Without the band, the number looks confident but misleads. Another limitation is opposition quality: over two years, the average ranking of teams Bangladesh hosted was lower than those it visited. So part of the home success is simply weaker opponents. Running a simple regression with opposition ranking and home venue as independent variables, home venue contributes about 0.4 runs per over and opposition quality about 0.6. More than half of home success is a story of weak opponents, not the pitch.

Here is the information gain the reader did not have. Mirpur's home advantage is a double-count: the pitch helps the home side and simultaneously drops the visitor into unfamiliar conditions. Separating the two reveals that much of what we call "home magic" is really the visiting side's adaptation lag. That means the most effective way to shrink home advantage is not only strengthening your own team, but venue-specific preparation in advance. This yields a concrete tactical decision. A team visiting Mirpur must replace sweep-and-slog against spin with straight drives and rotation-based approaches in the powerplay. A home side must raise its cutter-cum-yorker mix at the death, because on a slow pitch batters struggle to time a low-riding delivery.

And here is a larger responsibility. It is easy to use data to tell people Bangladesh is good at home, because Mirpur's crowd is shouting. It is hard to use data to show where the edge is actually leaking. I write for the second, not the first. A number is a witness; a trend is a confession—and the trends here say that what we call magic is really a calculation, some of whose inputs are in our hands and some are not.

The spreadsheet remembers what the stadium forgets. Years from now, Mirpur's stands will still tell the story of a famous win; my column will say what share of it came from the pitch, what share from the opposition's jet lag, and what share from simply a good day's batting. The gap between those two stories is my job. I began with the live thread and ended with a broadcast truth—which says Mirpur helps Bangladesh, but never the same way twice. That variability is the real news, and it is why I never measure home advantage once and stop. The match ends, but the model keeps playing.

What is the signal for the next round? Watch two things in the coming series. One, how well the side batting first holds its spin economy in the second innings—if the opposition's spin economy drops below seven at the death, the pitch is not slowing as before, and the home slider will not stay at 1. Two, the visiting side's dot-ball percentage in the first match—if it stays above 45%, the travel coefficient is still working, and the home side's win probability is data-supported. One question remains unanswered in my spreadsheet: if the crowd is the biggest slice of home advantage, do we keep series played in empty or half-empty grounds on a separate baseline, or treat them as the new normal? Whoever answers that first will write the next decade's home-advantage model.