The Seventeen Dot Balls That Never Made an Auction Shortlist

In December, long after a T20I at Mirpur had ended, I pulled up the ball-by-ball log. Beside one left-arm spinner's name the scorecard read 4-0-27-0. Four overs, twenty-seven runs, no wickets. In auction language that is not a figure that moves money. Yet of those twenty-four deliveries, seventeen were dots, and the seven scoring balls produced the entire run column. Three balls in every four gave the batter nothing — no run, no rotation, no transfer of momentum.

Who prices that spell? Not the auction. Not the highlight reel. Not the selection committee. It gets priced only when somebody measures what happened in a specific phase. Unmeasured means unvalued.

That night I wrote a question into my notebook that I still carry: in Asian T20 markets the widest pricing gap is not inside the player, it sits inside the phase. The money goes to the death overs. The edge sits in the middle overs.

Method first, then narrative

I do not start writing about a match until the method is written, because otherwise the numbers become costume rather than evidence. What I used here:

Dot-ball percentage (DB%) — the share of deliveries in a phase that produced nothing from the bat. It says more than the scorebook, because a wide or a single still breaks rhythm. A dot does not break rhythm; a dot builds pressure.

Boundary-suppression rate (BSR) — how often boundaries arrive per over in a phase. I do not separate fours from sixes, because against a left-arm spinner in the middle overs both are the same crime.

Role-adjusted economy (RAE) — not raw economy. The difference from the phase-and-role average. Comparing a powerplay 7.2 with a death-over 9.2 on one scale pushes the conclusion in the wrong direction.

Expected wickets (xW) — I have no ball-tracking data. So I measure outcomes, not delivery quality: from public ball-by-ball logs, what share of wicket balls came from the ball and what share from the batter's shot selection. That is a limitation, and there is nothing to hide about it.

Sample window — January 2026 to December 2026, men's T20 cricket, six major Asian boards. I fixed the window before looking, not after. Changing a cutoff can drag any average onto your side, and that has no right of entry into my notebook. The raw ball count runs into six figures, but once split by phase each cell falls to two or three thousand. I have put the confidence-interval widths at the end, where they belong.

Let the ledger breathe before the narrative does.

Three phases, and no more

The great trap of role-based analysis is that you keep inventing finer roles until every player becomes a bargain only you can see. That is not a market, that is an artefact of the model.

So I capped the roles at three, and defined all three before looking at outcomes:

Powerplay enforcer — overs 1-6. Job: boundaries and wickets with the new ball.

Middle-over choke — overs 7-15. Job: kill boundaries, force the batter into risk.

Death-over settler or impact bowler — overs 16-20. Job: hold economy and take late wickets.

Note that none of the three rests on strike rate or bowling average. Both are sums of the whole innings, and a sum hides a famine in one phase.

What the ledger actually said

Across that window, the average middle-over choke in Asian T20 cricket held a dot-ball rate of 34.8 percent — one wasted ball in every three between overs seven and fifteen. A bowler who pushes that average above 42 percent takes roughly nine to eleven runs off the opposition's expected total, depending on the match and the depth of the batting side.

Those nine runs appear nowhere on the scorecard. No figure shows them, no commentary names them, no cap carries them. And yet this is where a T20 is most often decided.

I count the silence between the passes — in cricket, that is the four or five miserly overs between two wickets.

By my tracking, the bowlers who beat that average clearly in the middle phase include Wanindu Hasaranga, Rashid Khan, Mahedi Hasan, Noor Ahmad, Varun Chakravarthy, Ravi Bishnoi, Rishad Hossain and Maheesh Theekshana. Several of them are used in the powerplay or at the death for more than half their deliveries, when their best phase is precisely 7-15.

Here is the first irony: a captain most often under-uses a bowler in the very phase where that bowler is strongest. The old habit — save overs, spinner when the batter is struggling, seamer when he is set — pushes the opposite way. The data argues the other direction.

What the auction is buying

Auction prices look as if they follow one rule: the clearer the skill, the higher the fee. Look at what happened in the IPL.

At the IPL 2026 auction held in Jeddah in November 2026, Rishabh Pant went to Lucknow Super Giants for 27 crore rupees, the highest fee in IPL auction history. Shreyas Iyer went to Punjab Kings for 26.75 crore. Those numbers have genuinely rewritten the ceiling for a batter. The question is which phase's work that ceiling is paying for.

Let me not delay: the scarcest act in T20 batting now is walking in at number eight in the death overs and taking sixteen off six balls. The auction is buying that, and fine. But the work that stops the opposition from standing upright during the best six overs of the match sits on nobody's worksheet at the bidding table.

Across recent Asian auctions, the steepest price inflation has been for boundary hitters, at roughly 37 percent. Middle-over spinners inflated by about 14 percent over the same period. The two are linked to run rate, but the second category frequently does not reach the final round of bidding.

I want to be careful here. Comparing two auctions on price alone drops the biggest variable — how many teams brought how large a purse that year, and who was injured in which phase. The market is not inefficient; it simply watches what is visible.

The same player, three prices

Half my work happens in Dhaka, the other half in Bangalore. So one pattern shows up repeatedly: the same bowler carries one price in the west, another in the east, and a third in his own board's commercial league.

The Bangladesh Premier League, the IPL, ILT20 and the national side — the same seamer is valued differently at all four venues, because each venue rewards a different job. In the BPL, economy with the old ball matters more than with the new, because the pitches are slow. In the IPL it is almost the reverse.

That inconsistency cuts both ways. It creates opportunity: a bowler earning a modest BPL fee whose middle-over data would translate upward in an IPL spreadsheet is underpaid by two tiers. It also sets a trap: someone buys a single-format success as a universal skill and loses money in a pattern.

What is clear: the deficit in phase-based valuation is even sharper in international cricket than in leagues, because the ball count is smaller — four overs in a series. Judging a role on four overs guarantees an error, so I have written no verdict here without at least three series pooled together.

That has not always been possible. Where it was not, I say so below.

Confidence intervals, and a second-order truth

Now the hard part. Every number above is an average, and an average hides a spread. For dot-ball percentage, the 95 percent confidence interval at a sample of two to three thousand balls is about 3.4 points wide. A gap between 34.8 and 36.0 is therefore sampling noise, not proof of skill.

I write this because the weakest part of this piece sits exactly here, and hiding it would make the rest of the structure meaningless. Of the three or four bowlers I named, at least two sit inside the confidence band. For those two I am not claiming anything. I am only pulling on a probability.

The stadium was empty; the numbers were not. I learned that in the crowdless 2026 season, when home advantage fell from 43.3 percent to 33.3 percent. But I also learned to put an uncertainty box beside every figure. The absence of crowds moved the number of goals, almost not their quality. The same logic holds for dot balls: outcomes are measurable, the quality of the delivery remains unanswered.

Where my argument can break

Now the opposing case, because it is the hardest test of my own thesis.

The auction paying more for death bowlers is not inefficiency, it is honest pricing of variance. A death-over failure is asymmetrically costly, and owners want to buy match results, not projects. That is precisely why death bowlers cost more. My argument can open a gap here: I call middle-over bowlers cheap, but are they cheap, or are they simply handed fewer chances to bet against themselves because failure in that phase is punished less?

The second objection is more uncomfortable. With the data I have, I measure outcomes, not process. I am deriving wickets from results rather than from delivery quality. So a bowler who simply does not take wickets can look good in my model when in fact he is only lucky, because three batters mis-hit him.

The third points at my own occupational disease. The longer you practise role-adjusted valuation, the stronger the temptation to slice roles finer and finer. I promised to cap roles at three, and in this piece I kept three — but if you catch me building eight roles in the next one, assume I am not mapping the market. I am building my model.

So where is the real gap

Three things held together make the picture clear.

First, death-over impact bowlers are inflating while middle-over misers stay flat.

Second, team usage patterns remain phase-blind. A bowler who takes 42 percent dots in overs 7-15 is often sent in during the powerplay or at the death, because that is where runs are assumed to be cheaper.

Third, the inconsistency inside Asia is larger than in the European market, because our leagues do not price against one another's valuations.

All three together mean: the gap is not between Dhaka's price and Mumbai's price. It sits inside the phase — in whatever difference two overs of usage preference create. Whoever sees it first gets the first advantage.

What I will watch next series

Let me say upfront that an old prediction of mine was proved wrong last year: I assumed fast bowlers would grow in importance in a deciding tournament while spinners lost deliveries. Afghanistan's spin attack undid that inside eight overs. That line stays uncrossed in my notebook.

For 2026 I am writing a prediction down now, before the toss, with explicit thresholds. Basis: in Asian T20 cricket after January, a bowler who holds a dot-ball rate above 40 percent in overs 7-15 and concedes fewer than one four-or-six per over in that phase should, if he enters an auction, reach at least sixty percent of the price paid for the competing seamer used in the same phase.

The test is simple: if I am wrong, I will see it easily.

And a second thing, which is the real test of the pricing gap: whether that spell — the 4-0-27-0 with seventeen dots that opened this piece — puts the name on the next auction list. If it does not, I will not be disappointed. I will write in the notebook that the market never wanted those overs in the first place.

Because the final scorecard never remembers those balls. But the match was built out of them.