Falling short: Has Bumrah's bouncer lost its bite?

That was the ESPNcricinfo ball-by-ball commentary when Jasprit Bumrah hit Dean Elgar with a bouncer in the Johannesburg Test.

You would have expected no less from Bumrah on a pitch with as much up-and-down bounce as that one. Over the past few years, Bumrah has troubled many batters in a similar manner, courtesy his hyper-extended elbow and delayed release point.

Of course, Bumrah is much more than his freak anatomy. He is smart. He is incredibly skilled. If you are in any doubt, watch his slow yorker to Shaun Marsh at the MCG in 2018. Or his spell in the second innings of the Antigua Test in 2018, where he took 5 for 7, swinging the ball both ways. With an action tailor-made to swing the ball in to right handers, he had learnt to make the ball leave them in the air as well. In just two years of playing Test cricket. Most bowlers take years to swing the ball both ways.

However, something is amiss at the moment.

After all, you don't expect a bowler of Bumrah's calibre to not take a single wicket from 17 overs on a pitch with as much uneven bounce as in Johannesburg, especially in the second innings. He did beat the bat often enough to have picked up more wickets. But he didn't. All he got was one wicket in the whole Test from 38 overs? And he is a bowler who strikes every 50 balls.

But a bit of digging gives us surprising results. Of the seven fast bowlers in the Johannesburg Test - barring Mohammed Siraj who hurt his hamstring in the first innings and didn't bowl at full pelt thereafter - Bumrah caused the least trouble to batters with deliveries that were pitched shorter than good length. According to ESPNcricinfo's length data, Bumrah induced false shots in 20 balls out of the 101 (19.8%) he bowled on short or short-of-good-length areas in the Test. Shardul Thakur comes in next, inducing false shots 22% of the time. Mohammed Shami was at 23.5%. Expectedly, the South Africa bowlers were a lot more difficult to handle with their higher points of release owing to their heights.

The more surprising fact was that this Test wasn't an exception for Bumrah.

Ever since his comeback after the stress fracture in February 2020, Bumrah has been less threatening with his short balls according to our data. Prior to the injury, Bumrah induced false shots off 26.3% of deliveries that he pitched shorter than on good length against batters in the top seven in Tests. Since his return, that percentage has come down to 17%. That's a drop of a whopping 35.4% - from troubling top-order batters once every 3.8 balls before injury to once every 5.9 balls since. The trendline in the chart below shows how his short balls have become increasingly less threatening. The inflection point - to apply the term loosely - came in the first Test on his return from injury, in Wellington. This was the least he has troubled batters with short balls - only three out 39 he bowled shorter than good length troubled the batters. Perhaps the pitch was flat and perhaps Bumrah, understandably so, wasn't at his best.

Now, for Bumrah, that is not the most productive length anyway. He often uses it as a weapon to push batters back, and then lure them into playing shots at fuller deliveries. Only 34 of his 107 Test wickets so far have come from shorter deliveries. A far cry from Neil Wagner, for example, for whom it is a bread-and-butter length: 123 of 232 Wagner's Test wickets have come from short deliveries. In fact, Bumrah clocks in at 31.8% for wickets from short balls, which ranks only tenth among the 24 fast bowlers - since his debut - who have taken 50 or more wickets.

There's strong evidence that Bumrah could have lost some bite in his short deliveries, when we look at his averages against top-order batters before and after the injury. Before the injury, he took 17 wickets off shorter balls at an average of 18.1 and a strike rate of 46.2. Since his comeback, he has taken nine wickets at 42.2 apiece and it takes him 109.5 short balls to get a wicket.

His fuller deliveries have also become less incisive. It is likely he has lost some pace when bowling fuller lengths too. But the numbers when bowling full haven't come down by as much as they have when bowling short. Before his back injury, Bumrah took 30 top-order wickets from balls landing on good length or further up at an average of 20.4, striking every 46.4 balls. After his injury, that average has gone to 31.9 for 23 wickets, and the strike rate has been 66.9. This could be because fuller deliveries are more likely to trouble batters, because of swing and seam, than short balls that lack pace to discomfort batters.

There's more evidence pointing to a lack of zip in Bumrah's short deliveries. The lower the pace on the ball, the more comfortable batters are in playing shots off them square of the wicket. An analysis of five shots, namely the pull and hook on the leg side and the cut, dab and steer on the off provides that evidence.

According to the data with ESPNcricinfo, batters attempted 107 shots to Bumrah's short deliveries before his injury. Batters were in control only 52.3% of the time. Since Bumrah's comeback, 106 similar shots have been attempted by batters, and they have been in control 71.7% of the time.

Moreover, batters are more comfortable than earlier leaving his short balls too. Before injury, only 28.8% of his 787 short balls were left alone. This number has increased to 36.7% since. And it's not because Bumrah has been wayward. Out of his 787 short and short-of-good-length balls before injury, 89.2% were in line with the stumps or outside off. That percentage has not changed since, with 89% of his short balls being on target.

All this does suggest that Bumrah may have lost some pace and that could be the reason that his numbers aren't as good as they used to be. In his first 12 Tests, he had taken 62 wickets at an average of 19.24. In 14 since his comeback, he has 45 at 28.75. These are decent numbers by any measure. But they aren't a patch on the numbers Bumrah racked up before his stress fracture.