Where will today's top Test players finish their careers?

Around six years ago I wrote an article in which I made projections of where the top players of that period would end their careers. Now nearly all those same players are in the last quarter of their illustrious careers. It is time to look at the topic again, factoring in the additional element of World Test Championship cycles that has in this decade.

Batters' selection

The Fab Four batters select themselves. They may all have minor weaknesses, but they are the top batters of the past decade and a half. For a couple of other batters to be included, the criterion is that the selected batter should have the potential to be considered in the same league as the Fab Four at the end of their careers. Only two such players present themselves - Babar Azam and Marnus Labuschagne. Both are around the midpoint of their careers and have done well. Both have had recent slumps in form. However, they have the potential to finish their careers averaging around 50, which is a sign of greatness.

The six batters I've selected for analysis are featured in the order of their debut Tests.

Bowlers' selection

James Anderson, Nathan Lyon and R Ashwin are obvious picks as the top three wicket-takers among active bowlers today. Stuart Broad would have qualified but he retired at the end of the Ashes series. Anderson might retire anytime now, but since he is playing currently, he is included in the mix. Mitchell Starc, entering the final phase of his career, is an obvious choice. Pat Cummins, possibly at around the midpoint of his career, offers compelling evidence of his greatness. The final position rests with someone who has the best strike rate among the bowlers who have come after George Lohmann: Kagiso Rabada is a magnificent bowler, only hampered by South Africa's paucity of Tests.

These are the six bowlers selected.

Data points

The data elements that are used in this analysis are outlined below. All are career values.

1. Career Tests/Runs/Batting Ave/Wickets/Bowling Ave
1a. Home Tests/Runs/Batting Ave/Wickets/Bowling Ave
1b. Away Tests/Runs/Batting Ave/Wickets/Bowling Ave
- Asia Tests/Runs/Batting Ave/Wickets/Bowling Ave
- Rest Tests/Runs/Batting Ave/Wickets/Bowling Ave
2. Recent Form: the data for the last seven Tests

Home matches present no problems. Most players generally do well at home. When there are home matches in the WTC cycle, these figures can be used.

However, away matches do present a few problems. Most players do not have great figures away, and also, players' records in Asia often are significantly different from their records in the other countries.

It is not possible to break this down to the country level, since the number of matches played in many cases is too low for the values to be used with any level of confidence. For instance, Williamson has a sub-par average of 26 in Sri Lanka. We cannot use this value with any confidence given he has only played four Tests there. On the other hand, Smith has played only two matches in the West Indies and averages 141. He might average 20 in his next two matches there.

The bottom line is that we need to use a higher level of consolidation. So I will use the consolidated figures at the level of Home (H); Asia (A) - India, Pakistan, Sri Lanka, Bangladesh, UAE; and Rest (R) - Australia, England, South Africa, West Indies, Zimbabwe, Ireland. This allows the kinks to be smoothed out.

The computation of recent form has its own challenges. Earlier, I used ten-innings/spells. But that was not satisfactory since a ten-innings stretch could end in the middle of a Test. Then I considered complete Tests. I started with ten Tests.

The problem with this number is that India could play ten Tests within a year while South Africa might take two years. Finally I decided on seven Tests. It would require between two and four series for a player to reach this number. The most time any cricketer would take to hit this figure would probably be a year and a half. That seemed fair to me. And the ten Tests are from across locations.

It is a surprising fact that other than Williamson, and to a lesser extent Kohli, the other batters are currently in poor form. That should reflect in the projections.

Data in the tables is as on March 15, 2024, the starting point for projection computation.

Girish TS / © ESPNcricinfo Ltd

The table is self-explanatory. Note how well Smith has done in the Rest group of countries. Note also the huge disparity in home and away figures for Babar Azam and Labuschagne, and Kohli's surprising performance in Asia.

Girish TS / © ESPNcricinfo Ltd

Again, the table is self-explanatory. Points of interest are Ashwin's below-par but acceptable performance in the Rest countries, and Cummins' sub-par and Rabada's very poor performance in Asia.

Methodology

The main currencies for me in this exercise are runs and wickets. These are the only metrics that can be projected with any degree of accuracy and certainty at an individual Test level. To do this, I use the Career Location figures at the beginning of the Test, the Recent Form (only for the WTC 25 cycle), and a randomisation algorithm to bring in the expected variations.

Recent Form is very relevant for the WTC 25 cycle, with Williamson and Labuschagne being perfect examples of the two extremes. Across a number of Tests, this metric will work out very well, rounding the values (4.3 is rounded down to 4.0 and 4.7 up to 5.0; rounding is more important for wickets than runs).

It is clear that averages cannot be predicted with any degree of accuracy, not even by AI-driven methods, since the average depends on whether a batter was dismissed or not. So I will extrapolate averages based on career figures. The average calculations, thus, will be mere calculations and not projections.

Any projection that is done at the level of a Test will not work, since certain Tests (for instance, last year's New Zealand vs England Tests) are not part of the WTC cycle. And I cannot go by team, since not all Tests played by a team qualify as part of the WTC cycle.

That leaves me with one viable option: going by player. It also gives me the ability to remove certain Tests from the reckoning for certain players; for example, Ashwin is not an automatic pick for India in away Tests. I have assumed that Ashwin will play in three of five away Tests against Australia. Ravindra Jadeja's seemingly better batting figures will get him the edge over Ashwin's superior bowling numbers, especially away. It is also possible that Anderson will play only one Test away against Pakistan.

Finally, despite their star status, it is entirely possible that players might be dropped. However, I cannot project based on this premise. Hence the assumption is that none of these players will be dropped.

Also, the assumption is that they will not take significant breaks, as has happened with Kohli recently.

And finally, some of these players could retire earlier than projected, and not necessarily at the end of a WTC cycle. Especially if their team is out of the reckoning to make the final in that cycle, they could quit at any time. But things of those sort cannot ever be predicted and so are not factored in here.

This is the distribution of Tests in the WTC 25 cycle that will be used as the basis for the current and future cycles.

Girish TS / © ESPNcricinfo Ltd

WTC 25 cycle: schedule in place

All 12 players in this analysis play in this WTC cycle. First, I determine the batter's or bowler's Test schedule for the rest of the WTC. As on March 15, 2024, Smith's sequence for the rest of the cycle would be H H H H H A A (five home Tests against India and two away-Asia Tests against Sri Lanka). Ashwin's sequence will be H H H H H R R R (two home Tests against Bangladesh, three home Tests against New Zealand, and three away-rest Tests against Australia). Rabada's sequence will be R R A A H H H H (two away-rest vs West Indies, two away-Asia vs Bangladesh, two home vs Sri Lanka, and two home vs Pakistan). The advantage of this method is that the sequence really does not matter since I process each Test individually. I do this in two stages.

First the Runs/Wickets per Test. I determine the appropriate career figures (Home/Asia/Rest) at the level of each Test. The career figures get a weight of 80%. Only for the WTC 25 cycle do I use Recent Form, which gets 20% weight. A preliminary RpT or WpT (Wickets per Test) figure is arrived at. Then a random factor between 0.91 and 1.06 (a spread of 0.15, with a heavier negative - the negative side is accorded 60% weight and the positive side 40%) is applied to the derived value to arrive at a final figure. This value determines the expected Runs/Wickets the player will secure in any given Test. This is then repeated for the next Test, and so on.

At the end of the cycle, in the case of a batter, I add the runs scored to the career values. The number of innings and the not-outs are estimated based on the batter's career values and the appropriate average is determined. At the end of each WTC cycle, I have a fresh career set of values.

Let us say that Smith's next Test is at home against India. His Career-Home RpT is 88.7. His Recent Form across seven Tests is 52.1 RpT. The extrapolation for this specific Test is 81.4 (88.7*0.8 + 52.1*0.2). A random value of, say, 1.027, determined through a faultless algorithm, is applied to this. So Smith is expected to make 84 runs in this Test.

Finally, at the end of the cycle, the career values are updated.

Now for the average. Let us say that Smith is estimated to score 555 runs in the seven Tests in this analysis. His cycle innings/not-out numbers are estimated based on his career values. He will have played 12 innings and will have been not-out twice. That allows us to update his career average and recent form values. A simplified process is followed for the bowlers, incorporating a simpler randomisation algorithm.

For Ashwin's first Test against Australia, his away-rest values (103 wickets in 31 Tests at 3.3) and a Recent Form value of 4.3 WpT will be used. For the first of Lyon's home Tests, against India, his home values (259 wickets in 67 Tests at 3.9) and a Recent Form value of 4.9 WpT will be used.

Anderson will likely have retired by the end of the WTC 25 cycle. He may retire himself, say, at the end of the West Indies tour of England. However, that is in the future and I am not Nostradamus. He will fall just a few Tests short of Tendulkar's 200, but sentiments relating to such landmarks might not play a part in decisions in this regard in England.

Virat Kohli might end his career just short of 11,000 runs © Associated Press

The WTC 27 Cycle: some information available

The Recent Form values as on March 15, 2024, do not make much sense during the years 2025 onwards. So I will use a wider randomisation range for this cycle instead, since there is a greater degree of uncertainty as far as future Tests are considered. Taking into account the phase of the selected players' careers are in, I will use a range of 0.88 - 1.08 (a spread of 0.20, with a heavier negative side) for the players in the last quarter of their careers, and a range of 0.90 - 1.10 (same spread of 0.20, but equal on either side) for the players in the middle of their careers.

For the post-2025 WTC cycles, I do not know the sequence of Test series. So I do all my calculations at the macro level: x Home Tests, y Asia Tests, and z Rest Tests. For want of any firm data, I will assume that the WTC cycle 2025-27 will have the same structure and fixtures as WTC 2023-25. This is as good an assumption as any: India will not play Pakistan and the big-three teams will play more Tests than the other teams. And there will likely be a flurry of retirements at the end of this cycle - of Smith, Kohli, Williamson, Joe Root, Lyon, Ashwin, and Starc.

The WTC 29 Cycle: up in the air

Specifics about this cycle are very much open to speculation. It is not known if the WTC will continue in the same format or become, say, a single-year or single-area format. But I have to make assumptions and I will make them on the basis of the current status quo, including the number of matches and home-away splits. I will use a wider range of 0.85 - 1.10 (spread of 0.25), while randomising, for this and the next cycle. Cummins will possibly retire at the end of this cycle.

The WTC 31 Cycle: way up in the air

The comments about the previous cycle will apply here too. This is a totally unknown WTC; it is valid to ask if it will even exist. The randomisation will be on the same 25-point spread used in the previous cycle. There will likely be a full stop to the careers of the remaining three players - Babar, Labuschagne and Rabada - at the end of this cycle.

WTC finals

Finally the elephant in the room - the four WTC finals. Which teams are going to contest these? Since I am not running analyses of the Tests in the WTC cycles, I cannot predict the possible finalists. We have some idea for the 2023-25 cycle. New Zealand, Australia and India seem likely finalists. Who do I select among these three? And what do we do about the next three cycles? We have no real grounds to base our predictions on. Some of these players could score a few more runs or take a few more wickets than are accounted for in this projection.

Non-WTC Tests

This is another small elephant in the room. At this stage, we are not even aware of Tests scheduled between Afghanistan, Ireland, Zimbabwe and any of the WTC teams, especially the six countries involved in this analysis. It is possible that some Test series might be planned after this article sees the light of day. I cannot do anything about it.

Recurring injuries, T20 commitments and South Africa's relative paucity of Tests mean that Kagiso Rabada could finish his career with less than 500 wickets © AFP/Getty Images

Final projections

Smith's recent loss of form, partly compounded by his self-propelled move into the opening spot, means that his total number of runs stays below 12,000 and his average just under 57. My expectation that Smith will end his career with a 60-plus average have gone with the wind. Williamson gets to just under 11,000 runs with an average under 55. With a series of tough away Tests, he will do well to maintain this average.

Kohli's average falls below 49, while Root just misses reaching 50. However, the relatively large number of Tests that England play allows Root to go well past 14,000 runs and move to second place on the all-time runs scored table. Kohli just about crosses 11,000 runs.

As has been said already, the four WTC finals could see some of these batters playing in one or more of them and scoring a few more runs. But Root is too far behind Sachin Tendulkar to have any hope of catching up to him.

Labuschagne crosses 9000 runs at an average close to 50, and Babar nearly gets to 8000 runs at an average below 48. The recent form of both these batters is not good.

Let me make it clear that the averages are not predictions but calculations based on their current career figures. It is quite possible that Root and Kohli finish above 50. It is unlikely that Smith and Williamson would move up or down significantly.

Anderson finishes just short of 200 Tests but is in second place on the all-time Test wicket-takers' table comfortably, with just over 730 wickets. Will he play any Tests beyond the current cycle is the $64,000 question. He might very well play a home series or two that will move his tally up, but not so much as to take him to the 800 mark. That will require around 15-20 Tests at least.

Lyon and Ashwin finish almost at the same level - either side of 650 wickets. Lyon does not overtake Shane Warne but Ashwin manages to go past Anil Kumble. And Ashwin maintains his five-plus wickets per Test standard.

Starc and Cummins finish around the 470 mark. Rabada just misses out on reaching 500 wickets, but there is a big question mark there. I am not sure whether he will play all the planned Tests. The recent New Zealand tour fiasco could well be repeated in years to come.

In almost all cases, the bowling averages remain at the current levels.

Conclusions

Root and Anderson comfortably end as the players with the second best aggregates in their respective areas. Tendulkar and Muralidaran stay on top, by decent margins. This lot of players was truly world-class; their successors do not seem to come close. Even if he plays on on until the WTC 31 cycle, Labuschagne does not even reach 10,000 runs and Rabada misses going past 500 wickets. In Rabada's case, it is quite possible that he may not reach the projected figures, since he is likely to miss a few Tests.

Email me your comments and I will respond. This email id is to be used only for sending in comments. Please note that readers whose emails are derogatory to the author or any player will be permanently blocked from sending in any feedback in future.

Anantha Narayanan has written for ESPNcricinfo and CastrolCricket and worked with a number of companies on their cricket performance ratings-related systems

Stats feature

Where will today's top Test players finish their careers?

RELATED ARTICLES