Tilak Varma, ₹1.7 Crore and the Mispricing Machine: How the IPL Auction Undervalues India's Bench

Bengaluru, February 2026. A conference room, ten franchises, ₹90 crore in purses, and a nineteen-year-old left-hander on the screen with a base price of ₹20 lakh. My laptop had a ranking of roughly 600 names built on five seasons of ball-by-ball data — a phase-adjusted replacement-value model — and he sat inside the top five of the entire room. The client did not raise a hand at twenty lakh. Mumbai Indians bought him for ₹1.7 crore, and the following season Tilak Varma made 397 runs at a strike rate of 131. Nine months later I was in a commentary box in Adelaide watching India fold to 168/6 and England reach 170 without losing a wicket, and not one of the four live numbers on my screen made it to air.

This is not an auction story. It is the story of a gap between two valuation systems — one that runs inside a franchise purse, and one that should run off the national team's requirements. The gap matters because bench strength in Indian cricket is no longer a sentimental debate. It is capital allocation.

The auction is not a talent market; it is a risk-transfer market. And risk is always priced below talent, especially talent that has not yet been proven in front of people who read scorecards in a different language.

The arithmetic of a mega auction

Ten teams, ₹90 crore each, roughly 600 names, about 200 sold. The market was not pricing who the best cricketer was. It was pricing who the lowest-risk investment was. An analyst on the table asks one question — how many runs or wickets does this slot need from me, and what fraction of that will this man deliver. Everything else is narrative.

One term earns its place here, defined plainly: replacement value is the average output you would get from the cheapest available alternative in the same position, in the same phase, against the same kind of bowling. A batter who beats that line has positive value. That is the whole idea, and it is the least-spoken idea in the room.

Phase adjustment was necessary because a single T20 average hides two entirely different jobs. A 130 strike rate in the powerplay is a loss; in the middle overs it is acceptable; at the death it gives a team almost nothing. I split five seasons into three phases, built a separate baseline for each, and placed him against that baseline. He made the top five because his middle-overs spin matchup was among the best five or six in the country — not just the runs, but the ball-by-ball shot map against turn and length.

Why uncapped Indian batters go cheap

An overseas player is priced on three things: international record, broadcast footage, and established market trust. An uncapped Indian player is priced on domestic scorecards, a handful of televised games, and a scout's eye. All three carry different amounts of information, and where information is thin, bids are thin. That is not a virtue of the market. It is a defect.

The ₹1.7 crore bid was not generosity. It was a discount against the model. In my file, his phase-adjusted value at that season's market averages sat closer to ₹4 crore. Mumbai paid half of it and gave him three seasons of rope. That is the difference between a bid and a development plan.

And here is the arithmetic no auction studio shows. Ten teams, eleven players: 110 starting slots. A maximum of four overseas per XI means more than sixty of those slots are never available to Indians at all. The Impact Player added a twelfth slot, but it added it mostly as a specialist slot, not as a place to grow another all-rounder. The domestic contracted pool runs past fifteen hundred. So roughly seventy to eighty starting jobs for more than fifteen hundred applicants. No team sport on earth carries a sharper selection ratio.

When the ratio is that steep, selection necessarily means discarding good players. And when you discard good players, your model tells you whom you dropped, not whom you picked. India's bench strength is not a talent shortage. It is a selection problem dressed as a talent problem, because selection problems are hard to solve and talent problems are easy to speechify about.

Adelaide, November 2026

I coded the collapse before I had a clean sentence for it. November 10, 2026, Adelaide Oval, T20 World Cup semi-final. India 168/6; England 170/0 in sixteen overs. The scoreboard said India lost. My file said India had already lost, and that the loss was decided not in the fourteenth over but in the tenth.

I was feeding live numbers — rolling required rate, wicket hazard, boundary percentage after the powerplay. What the surface showed was fine and what sat underneath was hollow. Their boundary percentage collapsed over the last five overs alongside the wickets, which means the runs stopped and the rate stopped with them. A chase breaks in one of two ways: you block, or you block by losing wickets. India was the second kind.

A collapse is a regime change, not a mood — the point where required rate and wicket hazard cross and everything after it is a different game from everything before. That night the crossing sat around the twelfth or thirteenth over. What followed was cricket in form, not a contest.

The people beside me in the box had the same numbers. They went unused, because broadcast sells feeling rather than information, and the feeling of a collapse is usually a lagging indicator that arrives fifteen minutes after the fall. I came back from Adelaide believing that if a number does not change a decision, I should write that it did not change one, and name who left it unchanged. Using data to say "the data shows" is rhetoric. Data is used when something moves.

Ahmedabad, November 19, 2026

India were bowled out for 240 in the World Cup final — Rohit Sharma 47, Virat Kohli 54, KL Rahul 66 — and Australia reached 241/4 in 43 overs, Travis Head 137. The match that gets sold two years later as inevitability says something else in the data.

I pulled the stretch from the twenty-fourth to the fortieth over, where India scored 80 and lost four wickets. Boundary percentage in that window was 9.4%. The time left after it was a fifth of the innings. This was not a technical failure. It was an innings whose scoring ceiling on that pitch was 245, not 260, and reaching the ceiling required all-round risk. The team took the risk and lost wickets.

That is the difference between mood reporting and model reporting. Mood says the team froze. The model says the team played shots from a position with no route back, because the required-rate calculation was running off the wrong ceiling. One of those is true, and it is the model. The error was never in what was attempted; it was twenty runs of mispriced ceiling.

My model was wrong too

The most honest part of the Tilak Varma episode is that my error outweighs my hit. I put him in the top five and said so with confidence. His pre-auction T20 sample was roughly forty-five innings, and the televised slice of it was smaller still. My confidence interval was wider than the claim I was making in that room. What I should have said was: he is fifth if you accept that his spin matchup translates against domestic bowling quality — and that if is large.

Second, and more uncomfortable: it does not follow that 397 runs happened because the price was ₹1.7 crore. That is a classic confounder. Had Punjab Kings bought him at the same price that February, would he have received the same role, the same run of three seasons, the same patience in the middle order? Mumbai's development system, its school of playing spin, its tolerance for a young middle-order batter — those are variables separate from price, and without controlling for them, concluding the market was wrong is a naiver claim than the market's.

Third, which I hold more firmly after Adelaide: the auction is a risk-transfer market. A franchise buys a batter's talent and leaves the risk of failure with itself. A low bid does not mean it rated the talent low. It means it declined to carry the risk. Once you see that, the debate about who got paid what loses its meaning, and the debate about who got opportunity begins.

The bottleneck I failed to name

In that room my job was to say, when the hammer moved: sixty names have gone, he is sitting at twenty lakh, we should take him to two crore because every bid after that is a discount on his real value. I did not say it, because in language it stayed a suggestion and never became information inside the decision.

I keep the spreadsheet close because the story usually hides in the footnotes, and in that February the story was a boy ranked in the room's top five, unnamed for two hours, picked up by the one team that also knew how to play him.

So here is the falsifiable call, registered in advance. At the next mega auction, the average winning bid for uncapped Indian batters under twenty will clear ₹1 crore, and the rise will come not from better models but from Impact Player-driven demand for middle-order specialists. If the median stays below ₹1 crore, my conclusion that the market systematically underprices collapses with it, and I will throw out the whole frame and accept that franchises know better than I do. There are worse outcomes.

The question is not how many players India has. It is what logic prices them, and whether that logic shifts next time — or whether we write the same piece again after another 397-run season, marvelling that the market missed one.