Tip to India's opponents: get Rohit Sharma out for under 15 runs

An analysis of the opener's ODI career using the survival curve - which shows you the exact distribution of a player's scores

Himanish Ganjoo
Following a shot at captaincy in 2013 (in the IPL), a man who had struggled for over half a decade despite being labelled talented by all and sundry turned a new page. Rohit Sharma had averaged 30 in ODIs since 2007, not quite living up to his billing as a worthy successor to Sachin Tendulkar despite some handy knocks. His shift to opening unleashed a ruthless beast, and he has since then churned out a sense-defying three double-centuries. He averages 63 since after the last World Cup, the highest for an opener with 20 innings or more.
Yet, following Rohit's career can be a frustrating hobby, with his behemoth innings peppered with damp squibs in between. When he doesn't go big, his manner of dismissal often seems to betray a lack of technique, as he falls to the laterally moving ball; or a lack of game awareness. Without getting into the data, it also seemed like his big innings hid away his numerous failures - scores that were not only low but also wasted precious deliveries in the Powerplay, owing to the way he constructs his ODI innings.
In data like cricket scores, where a few very high figures can inflate averages, the exact "distribution" of scores becomes essential. This is the survival curve, which tells us about the batsman's chances of passing a given score.
The shape of the survival curve tells us about the spread of scores: flatness in one area of the curve means the batsman is less likely to get out in that range of scores. The average condenses all information about a career into one number; the curve splits the details open.
Looking at Rohit's survival curves, his career as an opener is revealed to be a story of two halves, split by the 2017 Champions Trophy.
Both curves show a steep fall early in the innings. In fact, between the 2013 and 2017 editions of the Champions Trophy (both included), half of his innings end at or before the measly score of 29, although he averages 55.24. This indicates an inflation of the average by a low number of very high scores. Rohit is a feast-or-famine batsman - after the median, his survival curve flattens out, which signals extreme difficulty in getting him out once he crosses that barrier.
After the 2017 Champions Trophy, we see Rohit Sharma 3.0, if you will: more dangerous when he goes big, but also more consistent: his median sees an upward shift of six runs, now at 37, and his average is 65.78. The devil in the detail is that he is now more likely to get out very early: his survival curve dips before the 15-run mark. After that, it flatlines, going much flatter. He still gives teams a window to get him out, but the width of that window has shrunk: attack him before he gets to 15 runs and you have a great chance of seeing his back. After this point, he "settles", is less likely to get out, and goes big.
We can use medians to look at the "skewed-ness" of the distribution of runs made by a batsman. The more the difference between median and mean, the more his numbers are inflated by very high scores. Let's look at the ten highest-scoring openers since the 2017 Champions Trophy, sorted by median.
Rohit, with his new, higher median, is much more consistent, but he displays the highest jump, and his runs-per-innings value is the highest among the lot. He has an RPI of 98.24 in innings that go past his median.
Aaron Finch is similar is terms of numbers, but with a slightly higher median. Quinton de Kock has been the perfect opener in this period: he scores a lot, and he scores it often.
In the period under consideration, the average score for an opener is a high 35 runs. One way of looking at the quality of an opener is to check two things: how often does he go past this average score, and how high does he go once he crosses it?
Here again, Rohit lies in a league of his own: he crosses the magic level of 35 half the time, which puts him bang in the middle of the table, but once he does, he scores 98 runs on average, much more than anyone else.
All other openers score between 60 to 80 runs on average, but Martin Guptill scales the 35-run barrier only about 40% of the time. Colin Munro crosses it only once in four innings, and scores just 61 (the lowest) once he does. Jason Roy and Finch are the most frequent in crossing the barrier.


While talking about limited-overs cricket, especially in today's slam-bang era, strike rates cannot be left out of the conversation. The median divides the collection of a batsman's scores into two halves. How fast does a batsman score in innings that are lower than the median compared to the ones higher? This comparison tells us how a batsman's striking picks up after he crosses that barrier of settling in.
England's new template of high scoring is on display: both their openers strike in the high 110s once they cross their medians, starting out around a strike rate of 90.
Here again, Rohit stands out for his enormous transition. His strike rate before his median is a paltry 55.2, but once we get to his scores in the upper half of the distribution, it shoots up to 103. This is, by far, the highest jump in scoring rate compared to his peers at the top of the order.
In the first ten overs of the innings he strikes at 78.1, and he survives past the tenth over 53% of the time. On the other hand, his partner, Shikhar Dhawan, survives past that point in 47% of his innings, but strikes at 92 in the Powerplay.
Dhawan is the initial aggressor, allowing Rohit time to settle and float past his median. After that, Rohit transforms into a monstrous run scorer, difficult to get out, averaging a Bradmanesque 100 runs per innings and striking at better than a run a ball.
Data up to and including the World Cup game on June 5, 2019