The Mirpur Coefficient: Where Bangladesh's Home Advantage Actually Lives in Asian Conditions

August 2026, Rawalpindi. Bangladesh won a Test series on Pakistani soil for the first time, the opening match by ten wickets. In the twenty-four hours that followed, almost everything written circled three words: history, courage, upset. I was at a desk in Liverpool watching ball-by-ball data, with one question: how much of the wicket-taking in that series came from the slow bowlers, and how much from swing with the new ball and reverse with the old one? The answer did not surprise me. It made the market's story uncomfortable. The accepted framing was that Bangladesh won on spin; the larger share of the wicket pattern was designed by seamers — the ball slightly old, the pitch slightly dry, the batter under first-innings scoreboard pressure. Match reports open with the scoreline. I open with the model's disagreement. I built the Burnley model to hear the mean, not to cheer for it. Bangladesh's Test calendar carries an arithmetic problem no points table displays. Six to ten Tests a year, roughly half at home — meaning three to five home-condition matches per season. At that sample size, the thing called home advantage is statistically almost invisible, yet every decision is built on it: selection, the toss, field settings, even the market's line. Bookmakers exploit exactly that gap when pricing a home series, because nobody reprices a six-match series; everyone reprices last series' narrative. In 2026 I worked on empty-stadium data — the Bundesliga restart and the first six rounds of Project Restart. Home win rate fell from 43.3 per cent to 33.8 per cent; goals rose. Two lessons. First, home advantage is a named variable: crowd, familiar pitch, travel, sleep, even the pressure on a marginal umpiring call. Second, if you do not separate the variables, you credit the wrong one. When the stadiums emptied, home advantage left with the crowd. Mirpur, Chattogram, Sylhet — three grounds, three characters. Mirpur is usually slow and low, helping spin from day one; Chattogram offers new-ball movement and bigger innings; Sylhet's wind and humidity open a reverse-swing window. Treat those three as one condition and the plan is wrong before the first ball. The BPL sharpens the error, because most of it is played at Mirpur — building a national home model off league data is a classic category error. Now the arithmetic. I split home advantage into four layers: toss and pitch preparation, first-innings lead, wicket type by ball age, and session-level run rate. The first layer is the most misunderstood. The toss is a coin, but the pitch is a decision. The home side prepares it before the toss, and pitch character matters more than toss outcome. At Mirpur, batting last is hard regardless of who calls correctly — a constant in the model, not a random draw. Selection errors rarely come from the pace-spin ratio; they come from how extreme that ratio is taken. The second layer, first-innings lead. In my model, of the seventeen Mirpur Tests for which I hold complete ball-by-ball data over the last five years, the side leading on first innings won twelve — about 71 per cent. The 90 per cent confidence interval runs from 43 to 90 per cent. That wide interval is the actual news: we are less certain than we sound. But a lead is not only runs — it is freedom to set fields, licence to move catchers, and pressure applied session by session on the opposing batting order. The third layer, ball age. In Asian conditions the ball's life cycle splits roughly three ways: overs 1 to 15 the new ball, 16 to 40 the spin window, 41 to 80 the old ball and reverse. The market prices Bangladesh as a spin-only side. That leaves new-ball seam movement and old-ball reverse work underpriced. The Croatia position was not faith; it was a mispriced midfield — and Bangladesh's seam attack in Asian conditions is a mispriced asset. The fourth layer, session-level run rate. I treat each session as a separate unit, because the first session's mean and the second's tell different stories. At Mirpur, spinners' economy is usually lowest in the second session, and that is where the match's speed is decided. I have dropped the word momentum; I write session scoring differential instead. Join the four layers and what emerges is this: Bangladesh's home success is driven first by the first-innings score, second by the seam-spin balance across ball ages, and only far below by the crowd. Lose, and we blame conditions; win, and we credit a batsman. The model does none of that. The BPL is a useful laboratory here. Fortune Barishal won consecutive titles in 2026 and 2026, and the accepted explanation was leadership and team chemistry. The league's ball-by-ball data shows something else: death-over economy and powerplay wickets were the two variables that separated teams in the play-offs and the final. Championship stories sell; death-over economy quietly names the hole in a batting order nobody wants to buy. Here is my disagreement. It is easy to say with this data that Bangladesh win at home because of spin. But the share of wickets taken by spinners is an outcome, not a cause. Spinners take wickets because the pitch is dry, because the opposing order is under pressure, because the field is attacking. The pitch is made before the toss, and the toss is a coin. Turning a comfortable correlation into a cause is the oldest disease in Asian cricket commentary. Second caution, aimed at myself. A model is a confession of what you refuse to guess. I do not model crowd excitement, because its effect size in my data sits near zero while measurement error is enormous. I do not chase edges; I build the cage where edges must appear. I pre-register hypotheses before looking at the data — otherwise, after every series, a beautiful story can be assembled, and that is journalism, not analysis. Third caution, the one I voice against myself most often: seeing Asian cricket through London eyes. ECB pitch data, English seam-movement models, UK market lines — none of it transfers cleanly here. Bangladesh's domestic first-class data is thin, scouting is sparse, ball-by-ball recording is incomplete. Part of my judgement is therefore still visual scouting, not numbers, and I write that down rather than hide it. Finally, a workload check that market models almost never make. In Asian conditions, the seamer bowling overs 41 to 80 carries a physical load far heavier than the spinner's. Give a quick bowler long spells match after match and his reverse window narrows next series. That cost never appears on a scoreboard, but it appears across two seasons of data. For the next home series I will watch four things. One, the pitch announcement and grass cover before the toss. Two, the first-innings target score — is 350 still the line at Mirpur, or has it moved. Three, whether a third seamer plays, and who bowls the 41st over. Four, the slope of session-level run rate, especially in the second session. And I will leave one question open: if the bulk of home advantage comes from pitch preparation and a first-innings lead, how good is Bangladesh really once the crowd returns — and do we have the nerve to measure it?