Old Pitch, New Audit: Phase-Adjusted Run Expectancy in Asian Domestic T20

Old Pitch, New Audit: Phase-Adjusted Run Expectancy in Asian Domestic T20

On 14 March, in the left corner of the scorers' box at Sher-e-Bangla National Cricket Stadium in Mirpur, I was logging the last ball of the 16th over when one number in my handwritten ledger went red. The home team's run expectancy at the end of that over stood at 7.9; five overs earlier, at the end of the powerplay, it had been 9.6. In those five overs no batter was dismissed, nobody new came in, and the bowling attack never changed. One thing changed: the age of the ball and the moisture of the pitch. The scoreboard does not show that. The ledger does.

That 1.7-point slide in run expectancy found no place in the match report. The home side won, and nobody in the dressing room discussed it. To me it was the loudest signal of the evening — and this article is the audit of that signal.

Context: My Data Window and Claim Tiers

I opened my xG ledger in 2026; the 2026 World Cup wrote its own audit. I carried that discipline into cricket, but I never force football units onto it. Cricket's native units, for me, are three: phase-adjusted strike rate (PASR), dot-ball pressure (DBP), and a wickets-per-pressure-delivery index (WPI). I have logged all three by hand at Mirpur, the Zahur Ahmed Chowdhury Stadium, the Sylhet International Cricket Stadium and Colombo's R. Premadasa Stadium since 2026.

The data window here has four parts: 86 Bangladesh Premier League matches across the last three editions logged ball-by-ball, 22 domestic first-class matches, 13 matches from the 2026 Asia Cup, and 31 Asian bilateral T20Is from 2026-24. That is 152 matches. It is not a large sample, and I am not hiding it.

I split claims into three tiers. Exploratory — a single match or venue observation, not a basis for prediction. Gated — sample above 30, but wide confidence intervals. Audited — reproducible, venue-controlled, and repeatable elsewhere. The central claim here is gated; the sub-claims are exploratory, and I have labelled them where relevant.

Core Analysis: When the Phase Changes, Run Expectancy Moves More Than Strike Rate

In my ledger, across Mirpur's last three BPL seasons, the home side's middle-overs (7-15) PASR is 118 against 131 for the away side. A 13-point gap might read as weak home batting. But run expectancy per over in the same window differs by only 0.5 — 6.9 against 7.4. The strike-rate gap is large; the run-expectancy gap is small, because home sides in Mirpur's middle overs eat more dot balls while losing fewer wickets. Dot-ball pressure and wicket preservation offset each other. An analyst pricing batters on raw strike rate alone gets this wrong.

The mechanism is pitch age. Sher-e-Bangla's square reuses the same strip repeatedly through a tournament, and the older the ball, the more grip it takes. In my spin log, the dot-ball percentage climbs from 38% in the tournament's first week to 46% in its last. That is exploratory-tier, but the direction held across three editions. Asian domestic T20 forces teams to play two distinct batting environments inside a single tournament. That is a modelling problem, not a moral one.

The dew question in the second innings is subtler. Of 63 evening chases I logged at Mirpur, 38 succeeded — 60%. In afternoon matches the rate is 47%. That number appears to support the dew story. At the gated tier, though, chasing sides manage a middle-overs PASR of 124 against 119 for comparable first-innings sides — a five-point gap. Dew does change grip, but the full chasing edge does not come from dew; a large share comes from post-toss decisions and from how aggressively the powerplay is played.

Venues differ internally too. My log gives middle-overs PASR of 133 in Chattogram, 127 in Sylhet, 119 in Mirpur. Put those three side by side and the same batter's value can swing by up to 14 points when the venue changes. After Wanindu Hasaranga's 3/21 in the 2026 Asia Cup final in Dubai, many concluded it was a spinner's pitch. Dubai's strike-rate-to-economy profile is in fact far more batting-friendly than Mirpur's; Hasaranga's figures reflect the timing of his overs after the powerplay, not the surface. On 11 September 2026 Sri Lanka made 170/6 and beat Pakistan by 23 runs — that bowling mapping explains more to me than any pitch theory.

From personal experience: when I log these three units, I keep two extra columns beside every ball — which strip it landed on, and how many overs earlier that strip was used. In the 14 March match at Mirpur, the story lived in those two columns, not on the scorecard.

The Home-Advantage Coefficient: Pitch, Not Crowd

In football I watched 92 behind-closed-doors Bundesliga matches in May 2026 — the home win rate fell from 43.2% to 21.7%, and home advantage dropped from 1.43 to 1.18 points per game. Empty seats did not just change the noise; they rewrote the home-advantage coefficient.

Run the same test in cricket and the result differs. Across 40 Asian T20 matches in 2026-21 played at neutral venues or with restricted crowds, the home win rate was 52%, against 54% in my full 152-match set. A two-point difference is statistically meaningless. Cricket's home advantage in Asia is a pitch-curation and practice-access effect, not a crowd effect — that is my biggest new finding. A curator who knows the home spinner keeps delivering the edge even with the stands empty. This is exactly why football's crowd-effect framework cannot be pasted onto cricket.

Translated to the market, the implication is direct: BPL auctions and Asian franchise contracts still price batters on raw strike rate. As a Transfer Market Administrator, my first question is always the same — whose decision does this number change? If a middle-order batter's raw PASR is 142 but his Mirpur-controlled PASR is 119, his true value sits roughly a tier below his market price. A franchise using venue-adjusted PASR buys more runs on the same budget.

Contrarian Angle: Three Things That Could Break This Story

First, my own weakness. The whole analysis is observational, not randomised. Teams playing in the final week of a tournament are, on average, better sides — survivor bias hides there. Pitch age also correlates with heavier spin usage, because matchups clarify as the event progresses. Separating ageing pitches from greater spin deployment is genuinely hard. Correlation and causation do not part cleanly here.

Second, the dew narrative is routinely overused. In my log, second-innings advantage swings wildly match to match: 60% one evening, 38% the next. Announcing a dew theory off a single match means making a claim whose confidence interval touches zero.

Third, treating Asia as one data environment is the biggest error of all. Mirpur's humidity, Colombo's breeze and Dubai's dry deck behave differently. Transplanting one venue's model onto another guarantees mispricing. Without venue-adjusted strike rate, pricing a batter in any Asian domestic tournament means deciding on half the picture.

I also have not yet added umpiring variables — LBW review rates, no-ball calls, team-level DRS usage. Those interact with the pressure-delivery index and can shift run expectancy; they enter my next model version (v3.1).

Takeaway: What to Watch Next Round

In the first two weeks of the next edition, note one number — spinners' dot-ball percentage. If it climbs past 45% from 38%, expect middle-overs run expectancy to collapse, and sides that do not bank powerplay wickets will sink. Second, do not treat batting first after winning the toss at an evening Mirpur match as automatic; my log rates it the most contested tactic in the format. And before sitting at the auction table, franchises might ask one question: what is our venue's PASR — and are we paying for raw numbers or venue-controlled ones?