My last two articles were quite heavy and hyper-analytical. In fact, one of my readers said that those two together could be the basis for a PhD thesis. So, I have deliberately made this one semi-anecdotal, light on analysis, and quirky. The objective is to have fun, both for me and you, the readers. The theme of this article is a favourite of mine - streaks. I have done quite a lot of work on this fascinating subject earlier.
Over 15 years ago, I did my first cut on this theme. I took as the base, Don Bradman's 52 Tests and Sydney Barnes' 27 Tests. Then, I analysed the other players in terms of what they achieved in this equivalent block of Tests. I repeated this a few years back. The results were amazing. Batters cut the gap to Bradman, and Barnes was even overtaken. Then in 2016, I looked at the streaks from a different angle. I concentrated on the number of successful streaks, of varying lengths, a player had in his career and determined the best a player did have.
This is a fresh attempt and looks at a 20-match block of Tests as the streak. The fixed segment of a career makes for effective comparisons. For each player, I determine the best streak he had, from different points of analytical view, and pin that down. I do this for all the qualifying players and present the top ten players for each of those data views. You will be really amazed at some of the featured performances. It opens hitherto unknown facets of players' careers.
A bowler like Muthiah Muralidaran had 114 20-Test streaks. James Anderson had 167 such streaks. The Little Master, Tendulkar, had 181 streaks, Bradman had 33, and Barnes, eight. I identify each such streak and determine what was the best for each player. It does not matter if the player had many highs, as Bradman had. Only the best is considered. So, the final featured groups are the creme de la creme.
Why 20 Tests? Why not more, or less? It is not as arbitrary as it seems. Twenty Tests represents a significant slice of a player's career, especially in the first 75 years of Test cricket. It would normally represent between four and six series and between three and five years. Sufficient time for all form-related peaks and troughs to be ironed out. And we can be certain that any performance in such a slice of career has to be given its importance. And in today's world, it would cover an entire World Test Championship cycle comfortably.
For the bowlers, my metrics are wickets per Test (same as wickets), bowling average, percentage of team wickets, and number of five-wicket hauls. I also have a composite measure combining the number of wickets and bowling average since the two are totally independent measures. This measure merges these two disparate measures. For batters, I look at runs per Test (same as runs), weighted batting average (WBA), percentage of team runs, and number of hundreds. Of course, the common feature of all these measures is the 20-Test streak.
Now. let us move on to the tables. A few words on the cut-offs for selection. The cut-off for the bowlers is 20 Tests and 100 wickets. The 100 wickets is changed to ten five-wickets hauls for the penultimate table. For the batters, it is 20 Tests and 2000 runs. The 2000 runs is changed to ten hundreds for the last table. For the fielders, it is 100 dismissals. For a change, let us look at the bowlers first.
The bowlers
Those who read my early article with the 27/52 Test streaks would have been amazed at seeing that there was someone who matched Barnes. And that was Murali. He continues to do so in this analysis as well. In his purple patch at the turn of the millennium, he took 152 wickets in 20 Tests, averaging 7.6 wickets per Test. Barnes ended his career with an average haul of 7.45 wickets per Test in his last 20 matches. Where he would have ended if Archduke Franz Ferdinand had not been assassinated in Sarajevo in 1914 is anyone's guess. And maybe my own educated one, on another day.
Then comes the magnificent Waqar Younis, who averaged 6.65 in his peak period. He was virtually unplayable in the early-1990s. Meanwhile, R Ashwin was majestic between 2015 and 16. He also averaged 6.65 wickets per Test. An interesting feature was that nine of these 20 Tests were played outside India. Finally, that master bowler, Richard Hadlee, took 6.5 wickets per Test during his golden period in the mid-1980s. In the next quintet, four are spinners - Shane Warne, Richie Benaud, Clarrie Grimmett and Rangana Herath. Dennis Lillee is the sole fast-bowling exception.
No one should be surprised that the 19th-century left-arm spinner Johnny Briggs averaged 10.40 in 20 Tests. Let me also say that if I did not have the 20-Test cutoff, George Lohmann, one of England's greatest bowlers, would have been the top, with a sub-10.0 average. Let us take some time to marvel at the average posted by Pakistan's Imran Khan in the early 1980s. It is an unbelievable 14.06. And when a bowler captures 19 for 90 in a single Test, we can be sure that he will have a low average in that period. England offspinner Jim Laker, in the mid-1950s, picked up wickets for fun. His average was just short of 15.0. His spinning compatriot, Gary Lock, might have missed out in 1956. But he made up elsewhere and finished at just above 15. The top five is rounded off by Barnes, with an average of 15.45.
Now for the interesting part. Jasprit Bumrah, very recently, was unplayable, mostly outside India. He picked up 103 wickets at just over 15.8. It was no wonder that India prospered during this period. Unlike the WpT table, this part of the table is dominated by pace bowlers, with Johnny Wardle, Waqar, Glenn McGrath, and Hadlee completing the top ten.
The next table features the barometer of the player's contribution to the team's cause. It is a share of the team wickets during these 20 Tests. As expected, Murali leads the featured players. He took nearly 47% of the team wickets. In order to get the true impact, I considered all the wickets taken by the team, not just the bowler wickets. So, Murali probably took in excess of half the bowler wickets. Hadlee follows with 43%. Maybe a reflection of the responsibilities these two bowlers carried. But look at Waqar, in a strong bowling team, taking nearly 42% of the team wickets. Herath followed in the footsteps of the master, with just under 40%. As did Barnes, more than a century back. An interesting lower half. Four Indian bowlers get featured.
This time I have looked at the number of five-wicket hauls during the streak. One could say that this is a different facet of the bowler's contribution to his team. Barnes, as he finished his career, averaged nearly one five-wicket performance per Test. This is impossible to even imagine - the domination he exerted over the bowling scene. Waqar's 16 five-wicket hauls in 20 Tests is no less amazing, considering the times. Then we have three bowlers, Muralid, Hadlee and Ashwin, who finished with this feat 15 times. It is nice to see a few new names in the second half. Yasir Shah, Alec Bedser, Malcolm Marshall, and Richie Benaud.
The last bowler table is a special index. Why do we need one? Let me explain. Remember the fixed span of 20 Tests, making comparisons valid? Bowler A takes 80 wickets at 15.0. Bowler B takes 100 wickets at 18.0. Bowler C takes 120 wickets at 22.0. Who has done better - without delving into all those complex calculations I did and shared with the readers in my last article. Bowler A might not have taken many wickets, but he has a fantastic average while C has taken lots of wickets but has conceded more runs per wicket.
To arrive at an answer, I devised a little trick. With the safe knowledge that the base matches are the same, and that more the wickets, the better the bowler, and lower the average, the better the bowler, I calculate the Bowler Performance Index (BPI) as "Number of Wickets / Bowling Average". A gets 5.33 points, B gets 5.55 points, and C gets 5.45. This makes B the first among these three equals. Note that this cannot be used across players' careers since the number of Tests vary widely among the players. That was one reason why I did not use this in my analysis last month. In any case, all these metrics were handled separately there.
Barnes gets a BPI of 9.64 points for his phenomenal streak. Murali follows closely with 8.82 points. In third place is Imran, with 8.18 points. Quite close is Waqar, with 8.10 points. Then there is a surprise inclusion in the form of Briggs, with 7.88 points, just pipping Hadlee, with 7.80. Marshall, Bedser, Laker, and Ashwin complete the top ten.
Barnes, Waqar and Hadlee are the only bowlers who appear in all five tables. Murali and Ashwin miss out on the Bowling Average table.
The batters
Barnes and Murali shared the leaderboard in the bowling section, closely followed by Waqar and Hadlee. One would expect the batting tables to be dominated by Bradman. That is true, in spades. In the runs scored (Runs per Test) table, Bradman is ahead by the proverbial mile. Between 1934 and 1948, he averaged just over 160 runs per Test.
Suffice to say that Bradman straddled the years either side of the war like a colossus. But, understandably, the gap is reduced since we are looking at the best streak for each batter. Mohammad Yousuf averaged an excellent 130-plus during his prime in the early part of the millennium. This was matched exactly by Kumar Sangakkara a few years later. And again, nearly matched by Steven Smith a few years after that, and by Garry Sobers a few decades earlier. These five batters proved that during any batter's best periods, Bradman could be made to look human.
Then come an array of modern greats - Ricky Ponting, Mahela Jayawardene, Graeme Gooch, Virat Kohli, and finally Brian Lara. All of them averaged more than 120 runs per Test. Sunil Gavaskar, Viv Richards, and Tendulkar are in the next group of five batters.
I use the Weighted Batting Average (WBA) as my main metric to measure the batting performance since it is a very fair method to determine the Batting Average. For those who are not aware of it, the concerned methodology assigns fraction values, between 0.0 and 1.0, to all not-out innings, and determines the WBA. It equalises the values for all batters, irrespective of their not-out innings percentage, which could vary from 1% to 25%.
I must sound like a broken record. Bradman is on top with a streak WBA of nearly 110 during the same run-fest period of 1934 to 1948. This is 17 runs more than his career WBA. I feel that this is one area where the differences begin to narrow down. We can surmise that the gap would be much wider for the other batters. Let us see. Ponting, whose career WBA was just short of 50, has a much bigger 20-Test WBA of 84.3, an increase of nearly 35 runs.
For Sobers, the 20-Test WBA is nearly 80 and his career WBA is 54. Yousuf comes in with figures of 78 and 50. Jayawardene completes the top five. Helped by his stupendous 380, his streak WBA is 77, as compared to a career figure of 48. Smith, Kohli, Richards, Andy Flower (very happy to see him here) and Sangakkara complete the top ten. All of them had WBA values of nearly 75 and more.
The next category is the share of the runs scored by the batter out of the team runs in these 20 Tests. Who else but Bradman, who scored 28.4% of his team runs in the same peak period, tops this list. Andy Flower (not surprising, since he was playing for a weak team), George Headley, and Len Hutton, in the post-war period, all scored more than 25% of their team runs. The top ten is completed by Root, Hobbs, Lara, Sobers, Smith (despite the strong Australian batting line-up), and Gooch.
The final batting table is one on the hundreds scored by the batters in the 20-Test blocks. As expected, Bradman leads the table, having scored 14 hundreds. Yousuf and Ponting scored 12 hundreds each. We can see clear evidence of the narrowing of the gap in some of the tables. They are separated by the average value of the hundred. As are the nine batters who scored 11 hundreds. I am happy that finally Tendulkar and Gavaskar have found places in the top ten of a table.
Bradman leads in all four tables and is way ahead of the others. Smith is the only other batter who appears in all four tables. Quite a few other batters appear in three of the tables.
There is no need for a composite index for the batters as we did for the bowlers. The reason is that in both the key measures, Runs per Test and WBA, we have Runs in the numerator and Tests/Inns in the denominator. As such, it is generally true that the more runs a batter scores, the higher his RpT and average will be. Barring, of course, the variations that occur when normalising innings. As such, I have not created any index.
The fielders
As a bonus, I have extended this concept to the fielders also. When I saw that there were wicketkeepers who had averaged over four catches per Test, I was interested in knowing whether anyone crossed 100 dismissals in 20 Tests. Hence this table.
Well, I was spot on. Alex Carey effected 101 dismissals at 5.05 per Test during that wonderful phase he excelled in, between 2023 and 2025. Brad Haddin just missed the 100-mark. In a purple patch between 2011 and 2014, he effected 96 dismissals, averaging 4.8 per Test. Quinton de Kock follows next with 4.75 dismissals per Test. Mark Boucher was just behind, with 4.7 dismissals per Test. Then comes Adam Gilchrist, with 4.65. It is nice to see the great Rod Marsh in the list as well. One thing that strikes me is the fact that all the ten keepers are from fast-bowling-dominated countries. There are no entries from the spin-friendly subcontinent. The highest value for the subcontinent keepers is from MS Dhoni, who averaged 4.15 dismissals per Test around the 2010 period.
Understandably, the top 54 fielders are wicketkeepers. However, I was curious to know who was the leading non-wicketkeeper. That turns out to be Stephen Fleming, who took 51 catches in the late 1990s, averaging just over 2.5 catches per Test. Steven Smith follows him closely, with just under 2.5 catches per Test in his catching prime. This period partly overlaps with Smith's batting peak, making us admire the immense contributions that he made in the period between 2015 and 2019.
The full lists
As I normally do, I have made available the complete list of all qualifying players. There are three text files - one each for bowlers, batters, and fielders. You could download and peruse these here.
Talking Cricket Group
Any reader who wishes to join my general-purpose cricket-ideas-exchange group of this name can email me a request for inclusion, providing their name, place of residence, and what they do.
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 will be permanently blocked from sending in any feedback in future.
