Format Contamination: Indian Cricket's Selection Debate Is Fought on the Wrong Pitch

I stayed in my seat at Uppal after the last ball. The notebook was open, the phone was off, and the broadcast truck was hauling its cables back towards the container. Hyderabad had beaten Andhra by one run. Ricky Bhui's 64 would be filed as "match-winning" by the next morning. My notebook had something else — fourteen dot balls across the last three overs, one missed run-out, and two field placements nobody noticed.

Two men behind me were arguing about a nineteen-year-old batter. One of them, a district coach, said: "His T20 strike rate is over 140, push him up." I didn't turn around. But the sentence kept circling, because the same logic runs every week on a television panel, in a newsroom planning meeting, and in whatever version of the selection meeting leaks.

So the claim goes first: Indian cricket's most expensive error is not a shortage of data. It is that numbers from one format get inserted into decisions about another, and then the decision is defended as data-driven. The Test selection argument is not fought on Test numbers. It is fought on whichever numbers happen to be available, and the available ones are almost always T20.

Look at what a young batter's public record actually consists of. Fourteen to seventeen IPL games a year, a handful of bilateral T20Is, the occasional domestic one-day tournament, and two or three Ranji knockouts if he is lucky. The broadcast shows only the first part. The Ranji knockout is the real scouting desk, but it is not televised, it does not generate clips, and its strike rate does not circulate on Twitter. The consequence is that the player facing the biggest decision of his career — a Test decision — owns a data set that is mostly not from Test cricket.

In 2026, when I was a junior producer at a Hyderabad digital sports startup, Gujarat won their maiden Ranji title by beating Mumbai by five wickets in Indore. My editor sent my first script back with one line: start with the argument, not the scorecard. I rebuilt it as "Mumbai's aura lost to Gujarat's data" and spent ninety seconds proving it with two numbers — Priyank Panchal's season aggregate of 1,310 runs, and Gujarat's 125-run chase under Parthiv Patel. That video taught me that the verdict comes first. It also taught me something I did not register at the time: the Ranji knockout data exists; it simply does not travel into the ecosystem where decisions are made.

The imbalance did not build itself. IPL broadcast rights sell in the billions, advertising rates climb in the last five overs, and a newsroom measures itself by whatever clip travelled overnight. When all the money flows towards the short format, the analysis flows with it. The data desks won the argument about how to measure things years ago; the commentary box never noticed, because it still reads the numbers that appear on the screen. Across years of watching matches from the press box and the second tier, the fastest change I have seen is this — the numbers multiplied, the understanding did not.

Now the rule that governs everything else and gets broken first in Indian cricket discourse. Averages, strike rates and economy rates are not comparable across formats; every metric has to be quoted with its format attached. A Test average and a T20 average are different units, like kilometres and nautical miles. A Test batter gets eight to twelve innings a year. A T20 opener gets thirty to forty. That means a Test judgement built on a small sample is always fragile, and the small sample cannot be topped up with the big one. When someone says "his strike rate is X, so bat him at three in Tests," he is substituting one measurement for another and does not know he is doing it.

The freshest example is the argument that returns every season — Suryakumar Yadav's T20I strike rate, which has sat above 160, against Virat Kohli's ODI average, which sits around 58. Both numbers are correct and both are useless when stacked against each other. Kohli's ODI average is not evidence of his Test technique, and Suryakumar's T20 strike rate does not tell you whether he will leave a ball outside off stump. This error is more common in India than elsewhere because a player's public career here begins with the IPL, not with Ranji.

The second filter is the home-away differential, cricket's largest structural inconvenience. Pitches in the subcontinent favour spin, and a home average inflates accordingly. The honest test of a batter is not his home average but his away average — Australia, England, South Africa, New Zealand. Those innings tell you whether the technique holds. In recent years I have watched every strong domestic season end the same way: the conversation reaches the home-away split and stops, because most people in the conversation never pulled the split. Home numbers hide weaknesses, and on Asian surfaces they hide the most.

The third problem is sourcing. An auction price is not a cricketing valuation, but in print the two share a column and the reader has no tool to separate them. The sum a franchise pays at auction is a market estimate, not a sporting one. A player can go for fourteen crore and be technically exposed the following season; another can go for twenty lakh and score three hundreds in a Ranji knockout. Both facts can be true at once because they belong to two different markets. Above that sits another layer — aggregator blogs recycling stale figures without a primary citation. When a financial number arrives without a traceable source, the professional default is "pending verification."

Now the number nobody quotes. What I carried out of Uppal was not a strike rate; it was a pressure map. How many dots in the 40th over, how many run-out chances missed, how many times a fielder stood in the wrong place and conceded a single. Those figures do not appear on the front page of any stats site because no broadcast graphic is built from them. One-run finishes are decided by nerves, not arithmetic — and nerves are measurable, just not inside the broadcast agenda. Andhra did not lose that game because the required rate climbed. They lost because fourteen balls went unscored in the last three overs and nobody came up and took the risk.

Which raises the real question: what should the correct data set for Test selection be? Mine has three components — Ranji knockout performance, A-team tours abroad, and averages in overseas conditions. All three are environments where the bowling is close to international standard and the conditions are hostile. All three are also environments with no broadcast. The most necessary data is the least visible, and the least visible data does not get a seat in the selection argument. That is not a failure of journalism; it is the direct output of broadcast economics.

Now the part where I argue against myself, because this claim is easy to make sound strong, and sounding strong is not the same as being right.

The strongest case against me is that skills are not entirely format-locked. Most of India's best Test batters of the past decade had strong white-ball records — Virat Kohli, Rohit Sharma, Rishabh Pant. If white-ball performance did not predict Test success, why does that cohort exist? It is possible that the people building Test claims out of IPL numbers are accidentally correct. And there is a second cost to clean format separation: you miss the batter whose technical base is sound but whose short-format numbers are ordinary because he was never used in the right position.

The counter-stat that hurts my argument most is this: across a large cohort of top-order batters, the ability to score quickly in T20 and the ability to play the ball late in Tests are not statistically unrelated. The mechanism is obvious — a batter who can hit a length ball over cover can usually also delay contact. So my claim is probably too clean. The real problem is not format-mixing; it is that we do format-mixing without declaring it.

So I will write my terms down in advance. If a five-year data set appears showing that Ranji knockout averages predict Test success no better than IPL strike rates do, my entire frame collapses. I do not think it will, but if it does, I will be the first to write that I was wrong.

One more accounting item. When I reach for AFL list management or NBA load management, the analogy snaps in a specific place. Indian cricket runs on central contracts, players cannot exit to overseas leagues, and injury authority sits with the board. In the NBA, load management means a franchise deciding to rest its most valuable asset. Here that decision is split across the coach, the selection committee and the board. The parallel holds only to the extent that availability is a strategic asset in both systems — everywhere else it breaks, and that has to be conceded in the second paragraph, not the footnotes.

So what comes next? I think the centre of the selection argument moves within three years, and the movement comes from weights, not from players. The moment a franchise or a selection committee publishes a format-weighted index — one that gives Ranji knockouts more weight than the IPL — the first fight will be about the coefficients, not about the players. And that fight will be worth having. The argument should be about what we weight and how much, not about whose strike rate is higher.

The day Indian cricket accepts that one format's number is not another format's proof, half the heat leaves the selection debate — and the other half gets a lot more accurate. My call: by 2029, at least one IPL franchise will rebuild its scouting model around Ranji knockouts and A-team tours. If it does not happen voluntarily, the market will force it.