Draft Price vs Death-Over Economy: The Ledger the BPL Would Rather Not Reconcile

On February 21, 2026, in a Dhaka hotel ballroom, the second round of the BPL draft was underway. Names climbed the screen; representatives of the seven franchises tapped at tablets below. In my hands was a blue notebook, the direct descendant of the one I filled at Khulna District Stadium in 2026 when I first wrote a model by hand. A name appeared: a 34-year-old right-arm pacer, 41 wickets in the current domestic T20 season, an economy of 6.42 in the death overs. He went for 4.5 million taka. Two minutes later a 23-year-old opener came up: strike rate 138.4, but below 20 in three of his last six innings. He went for 2 million.

The difference is not in the money. The difference is who got counted, and who never did.

This piece rests on the scorecards of 318 domestic T20 and Dhaka Premier League matches, counted ball by ball, with a cut-off of January 31, 2026, and on one admission I print in every article: whatever my model cannot see, I write down too.

Context: three markets inside one draft

The BPL player draft is not one market but at least three. The first is the constrained market inside the board-set salary cap, where the gap between top and bottom prices is several times over but nobody has the freedom to raise a bid. The second is the franchise's own scouting estimate, a coach's memory, a video clip, a phone call at two in the morning. The third, and the most invisible, is the availability market: which pacer is fit now, whose knee is sore, who will miss the whole season because of a national-team series.

These three markets set the price together, but each keeps its books separately. That is where the trouble starts. A tournament's economy rests mainly on the second and third markets, scouting judgement and fitness risk, yet nobody opens the first book at decision time, which is the domestic performance evidence itself.

No international provider systematically charts bowling-phase data for Bangladesh's domestic T20 or the Dhaka Premier League. Powerplay economy, death-over economy, separate strike rates against spin and pace: none of it sits in one place. Data that does not exist does not get priced. I built the model by hand, because the league deserved to be counted. No provider would chart it, so the counting became a kind of prayer.

Core: what the ledger says and what the screen says

My draft value ledger carries 123 listed players, and two numbers for each. One is the final price. The other I call the Impact Index, on a scale of zero to ten.

The Index is simple to build, but each step carries an admission. For batting I take deviation from phase-based par strike rate, reconciled over by over and weighted by balls faced, so that a man with four innings does not outweigh a man with ten. For bowling, deviation from phase-based par economy, weighted the same way. I keep the spin-pace split separate, because a spinner's 7 economy in Comilla is not a pacer's 7 economy in Sylhet. The error margin is plus or minus 0.09, and fitness, dressing-room temperament and a possible national call-up are three things my model cannot see at all.

Across the scorecards of 318 matches, the correlation between draft price and Impact Index is just 0.28. In other words, what a player was paid and how much he actually changed a match are related, but so weakly that price cannot be used to infer performance. Among pacers the correlation is 0.41; among batters, 0.19. The bowlers' market at least looks at some evidence. The batters' market looks at an impression.

That 34-year-old pacer, Rafiqul Hasan, has the highest Impact Index of the 29 pacers in my ledger, at 8.7. Across his 78 death overs the economy is 6.42, against a league pace average of 9.31 in the same phase. The sample is large enough that this is not an eight-over fluke. Yet in price he ranks eleventh among pacers.

On the other side sits Sajidul Islam, a 23-year-old opener with an Impact Index of 5.2. His sixes per innings are high, his highlight reel is handsome, so his price lands mid-table. The market's eye is caught by the visible boundary, not the marginal value.

The ledger says something else that cuts against received wisdom. In Bangladeshi domestic conditions, a pacer's impact peaks between ages 29 and 33, not 24 to 27. Younger pacers bowl faster, but where to put the ball in a death over is learned over years, and that learning shows up in the model. Still, pacers over 30 carry a price roughly 2.3 times my model's estimate, because franchises pay a risk premium for experience.

The wage structure makes it plainer. One franchise spent 38 percent of its draft budget on its four most expensive picks, two of whom sit near six on my ledger, below ten. In a domestic tournament the budget is finite; one wrong price does not lose you one player, it loses you three.

Every number is a person who never got to explain themselves. To put a name in the ledger is not just to add data. It is to claim an innings.

Contrarian: correlation is not cause

This is where I have to stop. A correlation of 0.28 does not mean the price is wrong and my index is right. Two separate things appear to move together, most likely because a third thing moves both.

Take the 34-year-old pacer at 4.5 million. That price probably does not come from his performance in general but from squad construction. Five of the eight top-order batters in his franchise's group are left-handers, and his death-over economy against left-handers is 5.9. That is performance, but matchup-specific performance, and it looks small inside a general index.

The opposite possibility must stay open too. My model knows nothing about knees, temperament under pressure, or who gets along with whom in a dressing room. Behind that 2.3-times premium on pacers over 30 there may be information my notebook does not hold, information no scorecard will ever show. In my sample the premium is not explained by performance, but a failure to explain is not proof that the premium is wrong.

One more trap needs avoiding: seeing a league nobody charts and assuming it must be good. Domestic data is thin, samples are small, and my ledger is no more than 318 matches. A player who is not counted is not good and not bad. He is simply uncounted.

Takeaway: the signal for the next window

In the next transfer window I will watch one thing, not price but the retention list and the shape of release clauses. If a franchise retains a low-Impact, high-price player, then the market is paying for information my model cannot hear.

The question is for the league, not the franchise: before the next draft, will anyone publish a death-over economy table? If not, I will open the notebook again, and this time, counting ball by ball, the name of the spinner Nazmul Haque may join it, a man with a powerplay economy of 6.88 who sat at the very back of the price list.