T20 is a whole other ball game and numbers

Sir Donald Bradman is widely regarded as cricket’s best ever batsman. His oft-cited batting average of 99.94 is possibly the most famous statistic in the game today. Sure, his detractors may point to his having played in a limited set of venues and conditions, against fewer opponents, and so on; but even after accounting for all such factors, he still towers over his contemporaries, who too faced the same conditions. So, is the point done and dusted? Can we all go home now?

Not quite. Yes, without a doubt, he is the best Test batsman, but he had to largely contend with only one format — the original one (and, of course, with first-class fixtures of varying lengths). While we can debate till the cows come home about how he (or W.G. Grace or Viv Richards) would have fared in today’s era, the fundamental problem with following the game as a fan is that the lens through which the game is viewed and analysed hasn’t kept up with the demands of the newest format, T20.

Take for instance, the batting average. The batting average, quite simply put, is the runs scored by the batsman per dismissal. The bowler has a similar figure (the bowling average) as well, which tracks the runs conceded per dismissal effected. Defining the average in this fashion is intuitive as it is easy to see that the elite performances tug the same quantity (runs per dismissal) in opposite directions for the batsman and bowler. For long, a batting average of 50 has been regarded by fans as a measure of greatness; on the other hand, there isn’t a widely quoted benchmark for the bowlers, but it is safe to say that it is a number less than 30. The batting and bowling averages are widely used in ODI cricket as well — Michael Bevan’s and, presently, Virat Kohli’s stratospheric numbers dominate fan chatter like no other statistic. But the batting average by itself isn’t a useful metric in the T20 format — and that is essentially the issue with T20 cricket.

T20 cricket isn’t cricket as we have always known it. Articulated excellently by freelance T20 journalist and analyst Freddie Wilde in a long, freewheeling chat:

First-class cricket is defined by the struggle for survival and the struggle for wickets. One-day cricket is defined by the struggle for runs within the struggle for survival, and the struggle for containment within the struggle for wickets. T20 cricket is defined by the struggle for runs and the struggle for containment.

The point made about first-class cricket holds true for Test cricket as well. The average balls per dismissal in Test cricket is ~67 (or 11 overs); on the other hand, for a T20 match of a mere duration of 120 balls, the corresponding metric is ~17.5. When a batting team uses less than 70% of its resources on average, it only goes to show that the wickets are overvalued in the T20 format and hence the batting or bowling average loses its primacy. Staying power is only important to a point, after which the batting horsepower dominates in T20. Unfortunately, cricket analysis in the mainstream hasn’t caught on to this maxim. Hence, new statistics and metrics are needed to distill, understand, and interpret T20 and its skillful exponents. Here are some of a few that we hope will make an appearance in TV studios and water cooler discussions in the near future.

Runs Per Over: We start off with an easy one. Strike rate, confusingly, refers to both the runs per 100 balls (for the batsman) and the balls per wicket (for the bowler). What is more, the corresponding runs per ball metric for the bowler has been traditionally called economy rate even though it is a simple multiple of the batting strike rate. Instead of fiddling around with essentially what is one number, it is better to switch to a Runs per over (RPO) metric which intuitively conveys how much the chasing team needs to score per over to overhaul the target — comparing a batsman’s RPO to the required RPO will be so much simpler than dividing the batsman’s strike rate by 16.67.

Adjusted run rate: A batting side’s run-rate doesn’t stay at the same level throughout the innings; rather, it has a strong relationship with the over number or stage of the innings. In an average match, the teams take 1-2 overs to “settle” and start smashing in the powerplay, followed by consolidation between the 8^th and 15^th overs, and going hammer and tongs in the last five. Hence, it is only fair to compare batsmen and bowlers along with their peers after having adjusted for this inherent variation; the bowler or batsman’s RPO has to be weighted based on which stage of the match the player has batted or bowled in. Keeping this in mind, a bowler who concedes 8 runs per over at the death is far more valuable than a bowler who concedes 7.5 in the middle overs even though conventional statistics say otherwise.

Busy-ness index: What kind of a batsman is the player who has come out into the middle? The Gayle kind (deals primarily in dot balls and boundaries) or the Kane Williamson kind (more running between the wickets)? How does this vary according to the length or stage of the innings? How long does it take for a batsman to “get going”? This could apply to the bowlers too in terms of runs conceded. With this, one can easily know what to expect (i.e. what the player has done) in the average game.

Specific matchup analysis: Is there a specific pattern to a batsman’s or bowler’s method? Did Sehwag relish facing spin? How does ABD bat against left-arm spin on average? Is Raina good against the short ball? How many balls have been defended, attacked or attempted for boundaries? Does Bumrah bowl differently to a tailender? Seasoned fans and commentators might be aware of these patterns, but codifying and recording insights such as these will only add nuance to analyses of the methods of top-class players.

Win predictor: Andre Russell set the stage alight in the first half of the IPL with some brutal hitting in the end overs. While these were superlative T20 performances, the average viewer had no idea how unlikely the outcome was considering the match situation. The rarity of an event is often seen as a marker of greatness (eg. Hanif Mohammed’s triple hundred after following on, Botham’s Headingly show in the 1981 Ashes, Anil Kumble’s ten-for at the Kotla and so on); such events often find rich representation in the ballads of cricketing folklore but it would be great for viewers in the early part of their fandom to have a tangible measure of how improbable the situation or event in front of them is, thus giving them a contextual view of the proceedings.

With such granular metrics in place, the viewer’s experience would be made so much richer. Imagine the new levels of fandom, water cooler discussions and geekery this would set up.