It was the 832nd six of IPL 2026, and the season is producing them at a record pace – one every 12.3 balls, up from one every 16.3 balls in 2022. A 32% jump in four seasons. The boundary rope, in T20 cricket, has effectively moved closer.
And yet, not all of those sixes mattered equally. Some, like Washington’s, won matches. Others were already “priced in” by the time they cleared the rope. By the end of this article you’ll have a number for every one of them – including that match-winning six, which turns out to be more ordinary than it sounded.
Why a six can be worth less than six
Here’s the intuition. A run-projection model looks at the current score, the overs remaining, and the wickets in hand, and estimates where the innings is heading. It’s the same logic a commentator uses when they say, “They’re on track for 220” – just formalised into a number that updates after every ball.
A team is 190 for 2 after 18 overs. The model projects them to finish around 215. The batter hits a six. The score is now 196 for 2. But the model doesn’t now project 221. It projects about 219, because it had already expected roughly two runs off that ball that went for six – the typical yield at that stage of the innings. The six added six to the scoreboard but only about four to the projection. Two of those runs were already priced in.
Think of it like a stock. If the market expects a company to report $5 in earnings and it reports $5, the stock doesn’t move. If it reports $8, the stock moves – but the move reflects the “surprise margin”, not the absolute number. The six is the earnings report. The delta is the stock move. What matters is how much the result exceeded expectations, not the result itself.
A six in the death overs, when the par-score model already anticipates fast scoring, shifts the projection less than a six in the powerplay does, when the model doesn’t yet have that run rate built in. Both are six runs on the scoreboard. They’re not six runs of trajectory change.
This is what we mean by delta: the change in projected innings total a ball produces, given the situation it was hit in. If you’ve followed ESPNcricinfo’s Smart Stats, the idea is similar – measuring what a player contributed in excess of what you’d expect from an average player in the same situation. Delta is our version of that idea, tuned to IPL data. The specific framework matters less than the principle: not all runs are created equal, and context determines value.Two specific sixes from this year’s IPL make the point concrete. M Shahrukh Khan was chasing 200 against Mumbai Indians. Nine overs in, his side, Gujarat Titans, were 58 for 5 and sinking. The par-score projection had them finishing on just 136.
Shahrukh hit Allah Ghazanfar for six over long-on off the last ball of the ninth over, and the projection jumped to 147. That single ball changed his team’s expected total by 11.4 runs – nearly twice the scoreboard value. Because the six landed at a moment when the Titans’ innings was collapsing, it briefly reset the trajectory. (Titans eventually folded for 100 – the six was a moment of defiance, not a rescue. But the model didn’t know that yet, which is why the delta was so large. That’s the point: delta measures how much the ball changed expectations at the moment it was bowled.)
A day later, a ball into the 19th over of Sunrisers Hyderabad’s match against Delhi Capitals, Sunrisers were 226 for 2 and freewheeling – the model already projected a final total of 240. Heinrich Klaasen hit Mukesh Kumar back over his head for six. Score: 232. But the projected final only moved from 240 to 242. The delta was 1.5 runs. The six added six runs to the scoreboard but only 1.5 to the trajectory, because the model already expected heavy scoring from that situation. (Sunrisers finished on 242 – the projection was almost exactly right.)
The crowd cheered the same way both times. They were not the same shot.
The shape of 832 sixes
There were 832 legal-ball sixes in IPL 2026 by the time I pulled the data last night. The mean delta-per-six was 4.85 runs. Half the sixes fell between 4.1 and 5.5. The full range was from 0.32 to 11.4.
Two things are worth noting in this distribution. First, the centre of mass sits below six. Most sixes are worth around five projected runs, not six. This is the stock-market logic from earlier in this article: the model expected about 1.5 runs from a typical ball. A six delivers six runs. The trajectory change is the difference: about 4.5. The scoreboard adds six. The projection moves by less, because 1.5 of those runs were already priced in.
Second, the right tail is real but thin. The sixes worth genuinely more than six runs are the ones hit at moments where the batting innings was in trouble – during collapses, rebuilds, tight chases – and a single boundary jolted the trajectory upward. About 30 sixes this season are above seven runs. Another 30 are below three. Most of the action is in the middle.
Death sixes are already priced in
The folklore says the death overs are when sixes matter most. But in terms of trajectory change, they matter least – not because they’re bad, but because the model already expects fast scoring in overs 16 through 20. A death six confirms what the projection anticipated. A dot ball at the death would actually hurt the trajectory more than a six helps it. Here’s the average delta per six by innings phase in IPL 2026:
A powerplay six is worth, on average, 5.06 runs of projected total. A middle-overs six is worth 4.94. A death six is worth 4.29. Death sixes are the least valuable category by a clear margin.
Last night’s match-winner sits exactly on that line. Washington’s six off Stoinis came in at 4.30 runs – almost identical to the death-six average. The shot won the game; Gujarat Titans’ projected innings total barely moved, because the model already expected fast scoring in the 20th over, and a six just delivers what has already been priced in. Match value and trajectory value are not the same thing, and the most decisive ball of the night was, by trajectory, an utterly typical death six.
It’s worth noting that the death overs also have the highest six rate per ball: 9.6%, versus 8.5% in the powerplay and 7.4% in the middle. Batters hit sixes most often in the phase where each six is worth the least. That’s not irrational – it’s optimal play when wickets don’t matter – but it means the phase that generates the most sixes per ball is the one where each six moves the needle the least.
Why? In the death overs, the model expects about 1.7 runs per ball – the highest of any phase, because everyone is slogging. A six delivers six runs, so the surprise is about 4.3. In the powerplay, the model expects about 1.6 runs per ball. Slightly less expected, slightly more surprise. But that’s only half the explanation.
A powerplay six actually produces a delta of 5.0 – bigger than the simple 6 minus 1.6 arithmetic would give you. The extra comes from compounding: after a six in the powerplay, the model revises its estimate for all 80-plus remaining balls slightly upward, because the batting team still has most of its innings ahead of it. A death six has no future to compound into – the innings is almost over. So a powerplay six earns a bonus from the future; a death six doesn’t.
Thirty-seven sixes is a lot of sixes. Sooryavanshi is 15 years old, opens the batting for Rajasthan Royals, and clears the rope in the powerplay against international new-ball bowlers more or less on demand. The volume is real and the volume is rare.
But notice the third column: his mean delta-per-six is 5.05. That is good – better than the league mean of 4.85 – but it isn’t extreme. He is hitting a lot of sixes at a slightly-above-average leverage. His sixes total about 187 runs of projected match value across 37 events. That’s the season-leading total for sixes specifically, but the per-six leverage is closer to typical than the volume suggests.
The volume itself is the story for a 15-year-old; the leverage profile is the more interesting reading.
The leverage view
Re-sort the same population by mean delta-per-six (with a ten-six minimum so the floor is sturdy), and the leaderboard rearranges:
Mitchell Marsh is the leverage leader: 11 sixes at a mean delta of 5.88. He is followed by Cameron Green (ten sixes, 5.63), Shubman Gill, Jos Buttler, Prabhsimran Singh, and Ishan Kishan. Sooryavanshi sits eighth on this list, at 5.05.
A fair objection: Marsh’s high per-six delta might not mean he’s a “better” six-hitter. It might mean his team is frequently in trouble. And the data supports that reading: when Marsh hits a six, the average projected innings total at that point in his team’s innings is 173 – well below the league average of 198. When Sooryavanshi hits one, the average projection at that moment is 209. Marsh’s sixes move the needle more because there’s more needle to move. His team is behind; a six from behind creates a bigger trajectory shift than a six from ahead.
This is the honest version of the volume-vs-leverage distinction. High per-six delta is not “clutch.” It is “opportunity created by adversity”. Both Marsh and Sooryavanshi are hitting sixes; one is doing it while rebuilding, the other while piling on. The delta captures the difference in context, not the difference in skill. Both are valuable. They are valuable in different ways.
Volume and leverage are not the same axis.
Three signature shapes
Three batters this season illustrate the three places sixes can come from.
Sooryavanshi is the powerplay pure-play. Twenty-nine of his 37 sixes have come in the first six overs; eight were in the middle. He has hit zero sixes in the death, because Rajasthan don’t have him there – he opens, hits, and gets out before the death arrives. His phase profile looks almost nothing like Klaasen’s or Tim David‘s. It looks like Travis Head’s, which is what you’d expect from another opener in the same era.
Klaasen is the middle-overs specialist. Thirteen of his 19 sixes came in the middle overs, six at the death, none in the powerplay. The middle-overs sixes are at 4.45 on average. The death sixes are at 3.85 – substantially below the league mean. Klaasen’s overall mean delta-per-six is dragged down by the death sixes, even though those death sixes are exactly the moments his fan reputation is built on. The metric is unsentimental: a six in over 19 with the score already at 235 mostly just confirms what the projection already had baked in.
David is the death-overs pure-play. Twelve of his 16 sixes have come in the death. He hits them at a mean delta of around 4.5. He is doing the thing he is paid to do, and the metric values him for it, but the per-six numbers are below what Marsh or Green are producing in the middle, where the innings is still up for grabs.
None of the three is wrong. Sooryavanshi is the volume leader and his per-six leverage is fine. Klaasen is a middle-overs specialist, whose death sixes are scoring lower because they often arrive too late to matter. David is doing his job. The point is that the three of them, with very different roles and very different phase profiles, are all hitting sixes that move the projection by roughly the same amount per six.
What this isn’t
The argument is not that some sixes are good and others bad. Every six is a six. They all count for six on the scoreboard, they all win matches, they all earn their place in the highlight reel. The argument is more careful than that: sixes vary in how much they move the projected outcome of an innings, and the variation tracks systematically with phase, score, wickets in hand, and game state.
A few practical implications. First, comparing six tallies across batters with different roles is misleading. It over-credits sixes hit in big totals and under-credits sixes hit under pressure. Sooryavanshi vs David vs Marsh is not a like-for-like comparison – the per-six delta makes that explicit.
Second, “six-hitting” is not one skill. Clearing the rope at 230 for 2 is not the same as clearing it at 90 for 5. The first confirms a trajectory; the second creates a new one.
Third, the volume leaderboard is a measure of opportunity as much as skill. Sooryavanshi is a remarkable cricketer – a 15-year-old hitting 30-plus sixes against international bowling is not normal. But he is the best volume hitter in the powerplay in IPL 2026. That is the precise claim. “Best six-hitter” requires more context.
And a broader observation: the death six looks spectacular but mostly confirms what was already happening in the innings. The powerplay six looks routine but actually changes where the innings is headed. Batters are right to hit sixes at the death – it’s the optimal strategy when wickets don’t matter. But the evaluation of those sixes should reflect that a death six is confirming a trajectory, while a powerplay six is creating one.
Six-hitting isn’t scarce in IPL 2026. Sixes that move the needle are.
Vishal Misra, one of the founders of CricInfo in the 1990s, is vice-dean for Computing and AI at Columbia Engineering. The delta framework used in this analysis was developed jointly with Anand Rajaraman at San Francisco Unicorns, where Misra is a minority owner