Sachin Tendulkar?s current form is way off his peak run and his age, 40, suggests a bounce back may not be possible. As with many subcontinental greats, his retirement, when it arrives, will be overdue
Say goodbye, those days are gone, And we can?t, keep holding on These Hard Times, Matchbox 20
Tendulkar has turned 40. Yet another landmark for him and hence yet another occasion to discuss his retirement. That Tendulkar should hang his boots is, according to some, no longer a question mark. But should he please the skeptic critics, or should he play on? In a recent candid ?retirement? interview (March 25, 2012) this is what the possessor of most records as a batsman, the Master Blaster, had to say: ?The day I don’t have it, I will think about retirement? (emphasis added); he reiterated the fact that he considered it ?selfish to retire when on the top, because when you are on the top you should serve the country instead of retiring?; and on him being a trifle too old to play at the highest level, ?Yes, it?s a different body, but there’s a big difference between the mind of a 17- year-old and that of a 37-year-old (sic). The way I see it the glass is half full.? (emphasis added).
Given that the decision is his, when should Tendulkar retire in his own interest and/or in Indian team?s interests? The entire decision making exercise is compounded by the vagaries of human nature. We, especially sportsmen, are reluctant to accept aging, and to leave center stage. Actually, forget sportsmen, politicians and business leaders are the same. While we should retire at our peak or slightly off it, many rarely do ? for the simple reason that it is impossible to ex-ante know one?s peak! Just ask the PM, Manmohan Singh, who should have retired soon after the 2009 election. Thankfully, in cricket, there are the selectors who do the dirty on the individual, but the clean for the country and the sport, by quietly whispering to players that it is time to go. In Tendulkar?s case, partly or almost wholly upon his own insistence and because of his stature, the decision to retire is entirely his. Should the selectors let that happen or should they ask Tendulkar to retire from Test cricket by dropping him?
That is what we attempt to answer in this first of a series of articles on the game about which the famous mathematician GH Hardy said many years ago, ?If I knew I was going to die today, I think I should still want to hear the cricket scores?. We will use data, and analysis, to come to some determination about the vexing issues of cricket. First, the ?optimum? time for retirement.
Our calculus is to look at the great batsmen in Test history and see when they retired. Test cricket has a long history (137 years to be precise) and more than 2,700 players, now from 10 countries, who have participated in this glorious game. There are 74 batsmen who have played at least 100 innings (in the olden days, it used to take more than 10 years to complete this task) and have a career batting average above 40. These 74 make up our sample of great batsmen for analysis. How have these great batsmen handled their retirement? Age is a factor, though Wilfred Rhodes is the oldest cricketer on record, retiring in 1930 at age 52-plus.
The criteria
There are two important yardsticks by which a player should be ?judged?? at least for the issue of retirement. How good is the recent form of the batsman, and how good is his form relative to the contenders? While several indicators are possible, two are most relevant ?a recent, say 20 innings, moving average of runs scored by the batsman and a moving average of share in team score. (It turns out that there is a huge correlation between the two as is evident from Tendulkar?s career graph presented in Figure 1). We will therefore concentrate on the moving average of runs. In addition to the number, it is the trajectory that is important ? in particular, how far is the batsman from his peak.
So, in order to capture the form of the batsman, a 20 innings moving average of his score is taken. It is further smoothed to remove the jags. Let us call this the smoothed moving average or SMA. One of the important results that come out by tracking the SMA is that it provides empirical evidence of dual peaks in the careers of several great batsmen (which makes the decision about retiring that much more difficult). When did the greats in our sample retire? The metric with which we assess their retirement is to look at the SMA relative to the peak SMA achieved by the batsman. For example, Clive Lloyd, who played 175 innings, had a peak value of 59 and at the end of his career, his SMA was 56. So his retirement average was just a fraction, 6 %, below the peak.
Given this metric, is there any variation in how individuals in different countries have retired? Is form at retirement a random walk? Most emphatically not. The chart in Figure 2 summarises how individual countries have performed, and they do provide some guidance, and interpretation, on the optimum calculus. The numbers reported are the average distance at retirement from the peak of batsmen for each country (some batsmen from India included in the analysis like Sehwag have not officially retired).
The ?best? retirees are from England, Australia, South Africa and New Zealand. The top batsmen in these countries have retired at around 75% of their peak. In contrast, players from the sub-continent retire further off their peak with average of 65%.
While 65% of peak may seem respectable, the average hides considerably more than it reveals. The best Indian retirees list is led by Mohinder Amarnath ? he retired close to peak form, while the tail is brought up by Sehwag, Tendulkar, Vengsarkar and Gambhir. All four of these batmen have their last average less than half their peak value. In particular, Tendulkar’s rank is a lowly 67 in our list of 74, just behind Sehwag who may have got the boot, and behind Ponting. It does not get much lower than this ? the last batsmen on the list, Gibbs, retired when his form average was just 42% of his peak.
Other ?star? players from the sub-continent also seemed to have retired too late ? like Mohammed Yousuf, Javed Miandad and Jayasuriya. The three retired with a life-time average close to Tendulkar?s, and with an SMA also less than half peak value. Each of these batsmen was a great batsman, did a lot of service for their country, and each apparently stayed way past their sell-by date. That they were able to overstay says volumes about the near feudal nature of Test cricket in the sub-continent.
Sachin ? Time to go?
Tendulkar’s career performance is shown in Figure 1 ? the blue line is his recent 20 innings form (smoothed moving average or SMA) and the red line is an analogous computation of the fraction of team score attributable to Tendulkar. The two indicators follow the same pattern, and the team score pattern is more damning in its evaluation. Just discussing the pattern of form for runs, it is seen that for most of his career, Tendulkar?s scores averaged above 40; his peak average was close to 76, and that happened close to his 300th innings. (From the 280th inning onward, in his next 20 innings, Tendulkar had two double centuries and two centuries!). But his decline from that peak has been dramatic and sharp. In the subsequent 35 innings (since July 2011) he has averaged just 31 with a top score of 94; in the last 20 innings, he has averaged just 21 with a top score of only 81.
Note that the form pattern has two prominent peaks, something exhibited by careers of several great players. Isn?t some bounce from this present trough possible? Of course it is. And what magnitude of the bounce can be expected? To answer this, out of the 74 great batsmen in our analysis, we handpicked batsmen who had career trajectories similar to Tendulkar’s? double peaks, steep declines, retired with a career average of above 50, etc. The list (Figure 2) has its usual suspects ? Rahul Dravid, Steve Waugh, Brian Lara, Gavaskar, Ponting and some not-so-obvious names ? Mohammad Yousuf, Javed Miandad and JB Hobbs. Some of the trajectories are strikingly similar. Half of these eight batsmen did not bounce back and retired trying. In the other half, Dravid had the most impressive comeback, bouncing back almost 90% from his lows, the other three ?Steve Waugh, Gavaskar and Ponting ? had a modest improvement of about 25-30% from their lows. For Tendulkar, such a bounce would translate to an average of about 44 runs from here on?still well below his career average of about 54. Hence, even with this bounce, Tendulkar?s career average at retirement will be lower than what it is today. So is it in his interests to carry on? Does not seem so, especially given his age, the odds of such a comeback are stacked heavily against Tendulkar. To be ranked in the worst five in terms of timing the retirement is one record Tendulkar should not wish to possess.
How about carrying on in the interests of the country? Would it have been selfish of him had he retired earlier or decides to do so now? Looking at the country averages of retirement decisions presented earlier, such ?selfishness? certainly seems to be encouraged by better and more professional teams! For those who still think ?this time it’s different?, the latest Test series against Australia has highlighted some serious upcoming talent which needs to be groomed by giving responsibility and exposure. So no matter how one slices the data, or whose interests one looks out for, the answer is unambiguous?it is high time Tendulkar retired (read dropped by selectors). Gambhir and Sehwag, whose last averages are close to Tendulkar, may have already been told to go. When will the bell sound for Tendulkar?
Surjit S Bhalla is Chairman of Oxus Investments, an emerging market advisory firm, and a senior advisor to Blufin, a leading financial information company. He can be followed on Twitter, @surjitbhalla.
Ankur Choudhary, an IIT Kanpur Computer Science graduate, is a Strategist/Portfolio Manager at Oxus Investments.
