Players who retired on a Rosberg-style high, and those who didn't

Steve Waugh averaged 49.33 in his first 158 Tests, but an average of 86.30 in his last ten Tests lifted his career average to 51.06 Getty Images

On Friday night, when my 12-year-old grandson, Abhay, a committed F1 follower like me, called at 10pm and informed me that Nico Rosberg had announced his retirement, I was stunned. We are both Lewis Hamilton fans, however, for Abhay, Hamilton came second to Kimi Raikkonen. A few days back, we had endured a harrowing race in which Rosberg, because of Hamilton's slowing down strategy, was forced to hold off two racing tigers, in the form of Sebastian Vettel and Max Verstappen, to cling on to second place. Rosberg won the World Championship title, virtually in the last corner of the Abu Dhabi track. We both agreed that Rosberg was a true champion and deserved his narrow win.

Now we heard that Rosberg had forsaken about 100 million dollars - the salary earnings of a top driver for four years or so - and quit immediately after reaching the pinnacle. His World Championship win became more poignant. Let me wish him every happiness off the track, with his wife and daughter. Nico, you have risen immeasurably in the eyes of many F1 followers.

The next day, I was battling a severe cold, and sleeplesly ruminating on Rosberg, when the idea dawned on me that I had never looked at the end of the careers for Test players. We had vaguely talked of Kapil Dev and Sachin Tendulkar playing a few Tests too many and Sunil Gavaskar playing fewer than he should have. But a purely objective and numbers-based analysis? No. So I shelved my about-to-be-edited article and started on this one, which looks at the way players' careers have ended. I assure readers that there will be no subjective statements and this will be pure analysis, and the conclusions similar.

Once the criteria are set, this is an article of graphs and tables, with very little commentary.

As far as the end of the career is concerned, I will look at the last ten Tests. I know this 5% of Tendulkar's long career, or 20% of Everton Weekes' short one. But, barring war-interruptions or long absences towards end of the career, this represents a year's worth of cricket and works well.

In addition, I set a rather higher cut-off than I usually do. For batsmen, it was 4000 Test runs and for bowlers, 150 Test wickets. I wanted the players to have played a minimum of around 50 or so Tests. Sydney Barnes is the one exception. I had a collection of 124 batsmen and 99 bowlers who met the criteria. Out of these, there are 11 active batsmen and 11 active bowlers who were excluded. This left me with 113 batsmen and 88 bowlers.

The basis of analysis is simple. For the batsmen, I determine the batting average up to the end of the Test that was ten Tests before the end of his career, and compare this value with the batting average in the last ten Tests. Let me call this the LTTPF (Last Ten Tests Performance Factor). If LTTPF is greater than 120%, the batsman ended his career in the zone, improving his batting average. If this value is between 100% and 120%, the batsman ended his career very well. If it is between 80% and 100%, he ended his career slightly below par. If it is between 60% and 80%, he ended his career way below par, denting his average significantly. Finally, if the LTTPF is below 60%, the batsman ended his career miserably and his averaged dropped quite a bit.

I have used the batting average since this is the most followed and accepted measure. The RpAI (Runs per Adjusted innings) would have been a better measure, but it would also have taken the emphasis away from the topic. For bowlers, I did a similar calculation, except that the division is the other way around to ensure that an LTTPF value of above 100 represents a positive condition. Thus, it can be seen that the comparisons are not across players but only within a single player's career.

This graph portrays batsmen who finished their careers strongly. In other words, with an LTTPF of above 100%. Steve Waugh, Mark Taylor, Ian Chappell, Mahela Jayawardene, Gavaskar, Don Bradman, Mohammad Azharuddin and Michael Hussey are the prominent batsmen who finished strong.

This was news for me. Steve Waugh had the best finish to his career among all batsmen. At the end of his 158th Test, his average was 49.33. In his last ten, he scored 863 runs at an average of 86.30 and improved his average to 51.06. That is a Rosberg-like exit. Gary Kirsten was averaging 43.05 at the end of his 91st Test. He scored 1003 runs at an average of 66.87 and finished his career at the respectable 45.27. Desmond Haynes moved from 6768 at 41.27 to 7487 at 42.30, thanks to his last-ten Test burst of 719 at 55.31.

This graph shows batsmen who finished their careers reasonably well, with LTTPF values between 75% and 100%. Greg Chappell, Rahul Dravid, Ricky Ponting, Adam Gilchrist, Neil Harvey, Mark Waugh, Kumar Sangakkara, Brian Lara, Allan Border and Geoffrey Boycott are among the 39 batsmen in this group. As a group, these batsmen did not dent their averages too much since they maintained a reasonable performance level.

The third graph, which has the highest number of batsmen, features those who performed at levels below 75% - below par to awful. It is possible to ignore the top few batsmen in this graph since they were nearer the 75% mark. Garry Sobers, Martin Crowe, Shivnarine Chanderpaul, Zaheer Abbas, VVS Laxman and Kevin Pietersen fall into this category.

Of special interest would be the players in the lower part of this graph. Tendulkar's average dropped from 55.18 at the end of his 190th Test to 53.79 at the end of his career. His last ten Tests yielded 388 runs at 27.71. But we ought to be surprised to see the other top batsmen who have finished at levels below 60%. Weekes, Mohammad Yousuf, Virender Sehwag, Herbert Sutcliffe, Gundappa Viswanath, Inzamam-ul-Haq, Len Hutton, Dilip Vengsarkar, Colin Cowdrey, Jacques Kallis, Michael Clarke and Graham Gooch keep Tendulkar company. Only three batsmen finished below 40% - Ian Healy, Paul Collingwood and Mudassar Nazar.

Below are the tables containing the last-ten Test performances of the top ten batsmen and the bottom ten batsmen.

Barnes was exceptional, going into an extreme zone in his last ten Tests, capturing 88 wickets at 10.68, converting his career average from 101 wickets at 21.45 to 189 at 16.43. He is the only cricketer in history to have an LTTPF value exceeding 200%. Jim Laker moved from 153 wickets at 22.83 to 193 at 21.25, thanks to the 40 wickets he took at 15.20 in his last ten Tests.

Similarly, Clarrie Grimmett, Alec Bedser, Curtly Ambrose and Joel Garner all had a good harvest of wickets at sub-20 averages, significantly improving their career averages. Chris Martin, Iqbal Qasim, Sobers and Chris Cairns used their last ten Tests to bring down their rather high averages. Kapil's is an interesting case. His average was standing at 29.75 when his last ten Tests started. Even though he took only 20 wickets in these ten Tests, well below par in terms of wickets, he was quite accurate and had an average of 27.55 during this period, which did not do any harm to his career average.

A fair proportion of bowlers fall into the 80-100% LTTPF category. Starting with Malcolm Marshall (very close to 100%), the group includes Shaun Pollock, Bob Willis, Andy Roberts, Courtney Walsh, Ian Davidson, Lance Gibbs, Richard Hadlee, Michael Holding, Shane Warne, Saeed Ajmal (whom I consider retired), Shoaib Akhtar, Mitchell Johnson and Imran Khan. They ended reasonably well, even though slightly below par.

The third group contains some famous names. Let us ignore those who have LTTPF values exceeding 75. Look at the last six bowlers. Mushtaq Ahmed, Ian Bishop, Ian Botham and three leading Indian spinners, Bishan Bedi, Anil Kumble and S Venkataraghavan. All these six bowlers had averages exceeding 48 and dented their career averages in their last ten Tests.

Muttiah Muralitharan was the antithesis of Kapil. He captured 44 wickets in ten Tests, a rich haul indeed, but at the cost of around 35 runs per wicket. One could say that Murali finished his career quite well. Waqar Younis also finished with an average of 35 and a poor strike rate. It is interesting to note that the other two leading Indian spinners, Harbhajan Singh and Bhagwath Chandrasekhar, are those in the list with LTTPF values below 75%. It suggests that Indian spinners lose their edge, flounder for a few Tests and then quit.

Below are the tables containing the last-ten Test performances of the top ten bowlers and the bottom ten bowlers.

It is difficult to draw any conclusions from this analysis. It's clear that we can't predict how players' careers will end. How can one explain why Ian Chappell and Chris Martin had great finishes to their careers while Clarke and Botham finished theirs poorly? One can only do a post-mortem.