Watch enough T20 cricket, and you’ll notice batting orders barely resemble a fixed list anymore. A player who opens one week might bat five the next; a finisher promoted up the order during a chase might be sent back down again two games later. In Test cricket the top seven is treated almost as sacred; in T20, it’s closer to a live variable that shifts with match-ups, conditions, and gut feel.
Part of that flexibility is simply necessary. A 120-ball innings gives captains far less room to correct mistakes than a Test innings does, so getting the sequencing right — pushing an aggressive player up during a chase, or holding a finisher back for the closing overs — genuinely changes outcomes in a way it rarely does across five days. Anyone tracking ICC World Cup odds through a tournament will have seen how quickly in-play pricing moves the moment a captain sends out an unexpected batter at three or four, because the market treats those calls as meaningfully predictive.
The Importance of Data
What’s newer is how much data now sits behind those decisions. A 2026 modelling study built on more than a thousand IPL ball-by-ball records used a three-phase player profile — splitting each batter’s output across powerplay, middle overs, and death overs — combined with simulation across tens of thousands of possible innings to work out the batting order that actually maximises a team’s win probability, rather than simply their expected total. Applied retrospectively to a real IPL fixture between Mumbai Indians and Kolkata Knight Riders, the model found the optimal batting order would have lifted Mumbai’s win probability from 52.4% to 56.5% — and that the order Mumbai actually used that day was the second-worst of all six realistic permutations available to them. The specific culprit was a death-overs specialist who had the best strike rate of any available option in that phase but was sent in at number six and ended up facing only two balls.
That’s a strikingly large gap for what looks, on paper, like a fairly ordinary lineup decision — over four percentage points of win probability lost simply through sequencing, without changing a single player in the XI. It also captures something coaches have long suspected intuitively but rarely had hard numbers for: a genuinely excellent finisher provides almost no value if the match situation never actually gives them the ball count to use their skill.
Who (or What) Should Decide Batting Orders?
None of this means batting orders will ever be fully solved by a model — conditions, momentum, and the specific bowler at the crease all still matter enormously, and no simulation captures the psychological edge of a set batter staying in.
But it does explain why franchises are increasingly willing to shuffle an order mid-innings rather than sticking rigidly to a pre-match plan: the data now backs up what felt, until recently, like pure instinct. For more on how modern outlets are covering the numbers behind the modern game, News Examiner’s wider sports coverage is worth a browse.
