March 16, 2014

What can speed guns tell us?

They may provide objective information about the pace of the bowling, but the complete picture is provided by also considering the way batsmen respond to pace

The difficulty of playing Shoaib Akhtar when he was in the mood went beyond measurements of distance and speed © Getty Images

"To be objective," wrote the historians of science Lorraine Daston and Peter Galison, "is to aspire to knowledge that bears no trace of the knower - knowledge unmarked by prejudice or skill, fantasy or judgement, wishing or striving. Objectivity is blind sight, seeing without inference, interpretation or intelligence."

We love objective information precisely because it appears to be indisputably true - true beyond the reach of argument. The speed of the ball is an "objective" measure provided by the ball-tracking package used to enhance the quality of the broadcast. The trajectory of the ball, where it was released from, where it pitched in relation to the stumps, where it was headed, where it was met by the bat, where it might have crossed the plane of the stumps, whether or not it bounced normally - all this information is available too. The data provided by all this technology is systematically verifiable. If you used the same apparatus in the same way, you will get the same readings for a particular delivery.

Speed guns are lethal ammunition for the casual cricket fan. How often have you heard the lament that a bowler is down to the mid-70s? The conclusion from this finding is often of the deadliest kind - the bowler isn't trying hard enough. But quite apart from such simplistic crudeness, speed guns can also be bewildering.

In all the cricket that I have watched, the speed in miles per hour has tended to reveal very little about what is going on between bat and ball. I have seen bowlers deliver bouncers clocked at 81mph that rush batsmen, and short deliveries clocked at 90mph that are hammered by the same batsmen, seemingly with moments to spare. On some wickets, the short, lifting delivery just outside off is pulled effortlessly through midwicket. On others, the same delivery, at the same speed, is fended at. The speed of the ball through the air tells us very little about the contest between bat and ball. What's more, the contest between bat and ball often obscures the true speed of the bowling, especially on TV.

I remember the first time I watched truly fast bowling at a ground. Curtly Ambrose was bowling to Andrew Hudson and Kepler Wessels at the Cricket Club of India in 1993. Wessels left Ambrose's first two balls outside off stump. It seemed to me that he could have done little else. The balls zipped by so fast into Jimmy Adams' gloves that I failed to see how anyone could possibly react to them in time. Adams was standing halfway to the boundary. A couple of balls later, Hudson slapped a shortish ball to the backward-point boundary, after having calmly blocked the previous ball from Ambrose down by his feet. The batsman's response changed my perception of Ambrose's speed. Suddenly, even I began to see the ball a bit better from the North Stand!

The lived experience of watching cricket is often at odds with "objective" measurements like distance and speed. Now, after having watched and played a lot more cricket - on TV, on maidans, in stadiums, on Youtube - and having listened to endless hours of commentary, I think that the speed gun provides an invaluable measure, but only because it provides a reason to look for things one might not otherwise notice - things that reveal much about the contest between bat and speeding ball.

The first over bowled by Shoaib Akhtar in Centurion on March 1, 2003 included six deliveries, each clocked at a minimum of 90mph. It cost him 18 runs, which included the shot of Sachin Tendulkar's lifetime in my opinion (no, not the six, the boundary to square leg the ball after that). It ought to be humanly impossible to play the strokes that Tendulkar did in that over, if you consider the limitations of human reaction. The ball that Tendulkar square-cut for six was clocked at 150.8kph, or 93.7mph. The next two balls that went to the boundary were also quicker than 150kph.

Tendulkar v Shoaib in Centurion, World Cup 2003

From the moment the ball leaves the bowler's hand to the moment it is met by the batsman, the ball travels about 19⅓ yards. This is the distance between the two batting creases. The batsman therefore has at most 0.44 seconds to read the line and length of the ball, decide a response, and perform it. In reality there is less time than this, because the batsman meets the ball at least a foot (or often, two feet) outside the batting crease when playing forward. The speed is measured at the batting end, so the ball is travelling faster than the measured speed for most of its flight. Finally, the batsman has to "read" the delivery, a process that begins with identifying where the ball has been released from.

Experiments studying eye movements of county-level batsmen against bowling at 25 metres per second (by a bowling machine), or 90kph, the speed of the average spinner's faster delivery, showed that batsmen do not trace the exact path of the ball all the way through. Instead, they identify the point of release, and then the eye moves to the expected point of pitching. After the ball pitches the eye tracks its path until the time comes to play the ball. When the batsman is playing forward, he does not actually watch the ball on to the bat (simply because he can't - the bat and the arms are in the way). I have not come across studies like this involving extreme pace.

One of the most interesting ways in which batsmen cope with extra pace is to play the ball partially before it is delivered. Take a look at the screenshots of Tendulkar facing up to Shoaib Akhtar and Abdul Razzaq in that innings in Centurion. Against Shoaib, Tendulkar has almost fully completed a short forward stride before the bowler has released the ball. He does not use this method against the slower Razzaq, maintaining his stance until the ball has been released. Incidentally, Tendulkar also did not premeditate against Wasim Akram and Waqar Younis in that innings.

Note that in all these examples, Tendulkar is batting within the first 20 overs of the Indian innings. These are not the premeditated movements of the slog.

Tendulkar v Razzaq

A few days before that game in Centurion, Tendulkar faced Andy Caddick and James Anderson (middle left below) in Durban. Especially against Caddick, Tendulkar used different methods. Occasionally he would take a small stride down the pitch, along a line of middle and leg; at other times he would step across his stumps well before the ball was released.

Tendulkar v Caddick and Anderson

A few months later Tendulkar faced Pakistan in Rawalpindi. Here he used a different trigger movement against Shoaib and Mohammad Sami, the two genuinely quick bowlers in the line-up. As opposed to the forward stride he used in South Africa, here Tendulkar took a short step back.

The ball that dismissed Tendulkar in Centurion was banged into the pitch by Shoaib, but Tendulkar did not take a stride at all before the ball was pitched. Shoaib had returned for a new spell and Tendulkar, who was batting with a runner by then, was caught on the crease, fending off a rising delivery. In recent Tests we've seen a number of batsmen have this problem against Mitchell Johnson.

Against Caddick, the premeditated stride helped Tendulkar play balls from off stump or just outside through square leg, off both front and back foot. Commentating during this game, Ian Botham repeatedly observed that the deliveries Tendulkar put away through square leg were "on the pads". They were nothing of the kind. Most of the boundaries Tendulkar scored off Caddick were to well-pitched deliveries on off stump. In almost all cases, Tendulkar would have been out lbw if he had missed the ball. The only thing that would have saved him from being lbw in some of these cases is that the impact was likely to have been outside off stump.

These are all examples from limited-overs cricket where scoring runs quickly is important. In the two games against Shoaib that I have brought up, India were chasing five and a half and six and a half runs per over respectively. Against Caddick, India batted first, and were presumably trying to set England a difficult target. Scoring at such rates against high-quality extreme pace (or even medium-fast pace) requires these innovations from the batsman.

I have little doubt that Tendulkar played Shoaib differently in Multan in 2004. At Headingley in 2002, Tendulkar played Caddick and Flintoff differently. He used a pronounced back-and-across movement to counter Flintoff's line of attack from round the wicket early in his innings. Against Caddick, his first movement was typically forward, but only after the ball had been released.

Tendulkar is also not the only player to use these different methods. I keep looking for high-quality videos to see if Brian Lara or Steve Waugh used different approaches at different times. One unforgettable stroke by Lara against a screaming Waqar Younis inswinger on a length in Sharjah started as a cover drive but changed, in mid-stroke, to an astounding on-drive, all because the ball swung just enough that it would have found the inevitable gap between bat and pad that the full-blooded cover drive involves.

That was a quick inswinger from Waqar in 1993. But on that flat Sharjah pitch (Pakistan made 280-odd, and West Indies chased successfully, with Keith Arthurton and Phil Simmons scoring even quicker than Lara did!), where Lara managed to make not one but two strokes to it, it didn't seem that quick.

It is humanly impossible to respond to a 95mph delivery at a distance of 19 yards after it has been released. Most of the time, batsmen either minimise their response (see Yuvraj Singh's response to a 98mph thunderbolt from Shoaib Akhtar here, with the score 177 for 4), or offer a quasi-predetermined response. It is not entirely predetermined, because the anticipation, if you will, is limited to expecting a certain length and line, not necessarily to determining what stroke the batsman will play in response.

Sometimes batsmen premeditate to a great extent. An ugly example would be the premeditated slog across the line of a good-length ball. The success rate for full premeditation is not high. The success rate of partial premeditation, especially for a player like Tendulkar, is very high (judging by his consistency). Against high pace, at least some premeditation is essential, especially if the batsman wants to score quickly against this type of bowling.

It is one of the peculiarities of cricket that the speed of bowling becomes apparent only by the way batsmen respond to it. The reality of extreme speed has to be cheated in order to be addressed successfully. The speed gun provides an alternative verifiable fact about the speed of bowling - alternative, that is, to the speed that is apparent in the way the batsman faces the bowling. At its best, it is one more insight into the intricate contest between bat and ball.

In his 25 years as an international cricketer, Tendulkar has probably never been asked in an interview about the changes in his footwork against a particular bowler over a period of time. But when he practised in the nets, I bet that this sort of thing dominated his discussions about batting with his colleagues.

Perhaps, if that crucial half-second was teased out more systematically on the professional broadcast, we would get better questions, better interviews, and more discussion of the actual cricket. Tools like the speed gun provide the opportunity for this, but a lot of thought is required to go from the speed-gun reading to a conclusion. Blind sight is a myth. To really see, one has to think things through, especially when those things are "objective" facts.

The pictures used in this post have been extracted from extended highlights videos available on You Tube.

Kartikeya Date writes at A Cricketing View and tweets here