July 26, 2014

Test streaks: 52 and 27 matches long

A look at the best batting and bowling streaks in Tests
63

Muttiah Muralitharan got more wickets and Imran Khan averaged better during their peak years than Sydney Barnes - considered the best bowler by his contemporaries - did in his best years
Muttiah Muralitharan got more wickets and Imran Khan averaged better during their peak years than Sydney Barnes - considered the best bowler by his contemporaries - did in his best years © PA Photos/AFP

A few years back I had done an analysis on streaks in Test cricket. I had done this work based on a notional number of consecutive Tests as the streak. Towards the end of the article there was an excellent suggestion that I could use 52 Tests as the basis. Unfortunately since this was a late suggestion, I could not do proper justice to the idea. Also, many current readers would not have seen the earlier article. Hence I have decided to revisit the Test streak scene.

Another important fact is that during the four years, top batsmen like Hashim Amla, AB de Villiers, Kevin Pietersen, Michael Clarke, Alastair Cook, Ross Taylor et al, had their highs and would surely make a significant impact on the tables. On the bowling front, we have Dale Steyn, Saeed Ajmal, Mitchell Johnson, James Anderson, Graeme Swann who have ruled the roost. Hence this article will, I am sure, break new ground.

For batsmen, 52 Tests forms the undisputed cut-off. An Air Guitar for guessing why. The colossus played 52 Tests. Any comparison will be against this number. So I will select only batsmen who have played a minimum of 52 Tests. There are 151 batsmen who qualify. My apologies to those who miss out: to name a few, Everton Weekes, Frank Worrell, Dennis Amiss and Jonathan Trott.

For bowlers, there is an equally dominant figure: Sydney Barnes. He played in 27 Tests only and captured 189 Test wickets, an average of seven wickets per Test. There are 158 bowlers who qualify. So we have a decent population in each category. Let us not forget that each batsman has many streaks of 52 Tests within his career. Sachin Tendulkar had 148 such streaks. Muttiah Muralitharan had 106 qualifying streaks. This means that we have a lot of data available.

Any player's career could be shaded by various textures. Years lost due to war, injuries, non-selections, boycotts, self-imposed exiles, WSC-type series, bans, non-bowling injuries, easy runs/wickets, tough runs/wickets. We have to take all these in our stride and work on the basis that these factors even out across the streaks of between four and 20 years. So this data has a lot going for it. It deserves a lot of respect.

I will be analysing a huge amount of data. However, to make the task of the reader easier, I have created only four relevant tables. For each batsman and bowler, I pick up the best streak embedded within his career. So I reduce tons of data, so to speak, into four tables. Further I will feature in this article the top ten and bottom five players in each table. The complete tables are available for downloading and viewing. The analysis is current up to and including India's historic win over England at Lord's.

Test batting streaks: Runs accumulated in 52 Tests
Batsman Team Tests Runs RpT Best 52-Test streak Runs RpT Worst 52-Test streak Runs Ratio
DG BradmanAus 52 6996134.5 176(1928) - 303(1948)6996134.5 176(1928) - 303(1948)6996100.0
RT PontingAus16813378 79.61595(2002)-1819(2006)5857112.61324(1996)-1590(2002)3068190.9
BC LaraWin13111953 91.21542(2001)-1816(2006)5576107.21299(1995)-1541(2001)3944141.4
SangakkaraSlk12511593 92.71804(2006)-2048(2012)5518106.11504(2000)-1776(2005)3764146.6
GS SobersWin 93 8032 86.4 443(1957) - 636(1968)5468105.2 529(1962) - 738(1974)4410124.0
JH KallisSaf16613289 80.11619(2002)-1856(2007)5311102.11318(1995)-1563(2001)3340159.0
YousufPak 90 7530 83.71513(2000)-1844(2007)5247100.91412(1998)-1726(2004)3744140.1
TendulkarInd20015921 79.61365(1997)-1631(2002)5236100.71127(1989)-1364(1997)3534148.2
L HuttonEng 79 6971 88.2 281(1947) - 386(1954)5114 98.3 276(1946) - 376(1953)4643110.1
JayawardeneSlk14611506 78.81709(2004)-1968(2010)5102 98.11447(1999)-1699(2004)3559143.4
ML HaydenAus103 8626 83.71520(2000)-1723(2004)5092 97.91688(2004)-1904(2009)3966128.4
GA GoochEng118 8900 75.41049(1986)-1260(1994)5025 96.6 760(1975)-1044(1986)3201157.0
SM GavaskarInd12510122 81.0 683(1971) - 856(1979)5007 96.3 863(1979)-1025(1985)3380148.1
...............
HA GomesWin 60 3171 52.9 822(1978)-1045(1986)3021 58.1 839(1979)-1070(1987)2658113.7
CG BordeInd 55 3061 55.7 465(1959) - 634(1968)3021 58.1 459(1958) - 631(1968)2925103.3
JC AdamsWin 54 3010 55.71188(1992)-1523(2000)2995 57.61208(1993)-1527(2001)2857104.8
A RanatungaSlk 93 5105 54.9 955(1983)-1293(1995)2990 57.51103(1988)-1394(1998)2653112.7
GW FlowerZim 67 3457 51.61285(1995)-1625(2002)2872 55.21240(1993)-1581(2001)2643108.7

I have presented the best 52-Test streak and the worst 52-Test streak. This will enable the readers to see the fluctuations in a batsman's career. It is obvious that the best and worst streaks will be wide apart for players with long careers.

It does not take rocket science to know who would lead the table. As day follows night follows day, one DG Bradman, the "Boy from Bowral" is certain to be there at the top with 6996 runs. No need to describe Bradman other than to say that he stands supreme and in a separate zone of his own. This streak, also coinciding with his career, lasted just under 20 years. If those seven years had not been lost, would Bradman have reached 10,000 runs at 100 or ended his career with 6000 at 90? There are no answers.

What is important is the collection of modern greats who come after Bradman. Ricky Ponting's purplest of patches lasted just under five years and he accumulated nearly 6000 runs. Against a career RpT (Runs per Test) value of 80, he scored over 110 runs per Test during this period. He was within 17% of Bradman in his aggregate. This is a magnificent achievement considering that most measures place the second-placed batsman at around 40% below Bradman.

He is followed by another modern great. Brian Lara scored over 5500 runs during his heyday of five and a half years. Since his career RpT is itself a high 91, the accumulation of 107 does not seem that high. Another significant fact is that this was achieved virtually towards the end of Lara's career.

Kumar Sangakkara's recent form has been outstanding and this is shown by his six-year accumulation of just over 5500 runs. Garry Sobers completes the top five with an aggregate of nearly 5500 runs. However this took over ten years, indicating the paucity of Tests during the '50s and '60s.

Jacques Kallis, Mohammad Yousuf, Sachin Tendulkar, Len Hutton and Mahela Jayawardene complete the top ten. Hutton is the only batsman belonging to the earlier generation. Three other batsmen, Matthew Hayden, Graham Gooch and Sunil Gavaskar, are featured here since they accumulated over 5000 Test runs during the 52-Test streak.

It can be seen that where the batsmen have had careers close to 52 Tests, their career RpT and the best streak RpT are close. However, where the batsman has played in many Tests, such as Tendulkar, Ponting et al, there is a wide gap between the values.

The years that the concerned player took to complete this streak is of interest. For the older players we have periods exceeding 20 years. Bradman was close to this figure. Some of the modern players have achieved this in four years. For the record, Jack Hobbs took 21 years to play 52 Tests. Hayden took only four years to play the same number of Tests.

A tweak was necessary to present a meaningful bottom-five batsmen. The last part of the table is full of allrounders and wicketkeepers. Hence I have selected five players who played purely as batsmen. Grant Flower accumulated only 2872 runs in his best streak. The other four batsmen - Arjuna Ranatunga, Jimmy Adams, Chandu Borde and Larry Gomes - also aggregated only around 3000 runs in their best streak of 52 Tests.

Ponting's best streak is nearly twice as productive as his worst streak. It is of interest to note that only Daniel Vettori, with a ratio of 250%, has a wider variance. This indicates the topsy-turvy nature of Ponting's career. Most other players with long careers are around the 150% mark, with Kallis' 159% being quite high. Steve Waugh has a ratio of 170%. It is interesting to note that Tendulkar's best streak starts right at the end of his worst.

Test batting streaks: RpFI-MA across 52 Tests
Batsman Team Tests Runs Career RpFI-MA Best 52-Test streak Inns Unf-Nos Runs Streak RpFI-MA
DG BradmanAus 52 699689.50 176(1928) - 303(1948)804699689.50
RT PontingAus1681337847.721595(2002)-1819(2006)929585766.50
GS SobersWin 93 803251.33 443(1957) - 636(1968)885546863.27
SangakkaraSlk1251159355.221804(2006)-2048(2012)913551861.82
YousufPak 90 753048.961513(2000)-1844(2007)873524761.44
TendulkarInd2001592149.491365(1997)-1631(2002)884523660.81
BC LaraWin1311195351.851542(2001)-1816(2006)930557659.96
RichardsWin121 854047.55 767(1976) - 987(1984)760451459.39
JH KallisSaf1661328948.441619(2002)-1856(2007)915531159.26
JB HobbsEng 61 541054.55 102(1909) - 194(1930)853489759.23
ML HaydenAus103 862648.571520(2000)-1723(2004)916509258.53
de VilliersSaf 93 724047.701871(2008)-2119(2014)824464458.25
R DravidInd1641328847.731515(2000)-1765(2005)865488358.21
JayawardeneSlk1461150647.731709(2004)-1968(2010)894510258.18
MiandadPak124 883247.72 945(1983)-1130(1989)731424058.14
...............
CG BordeInd 55 306132.60 465(1959) - 634(1968)927302134.07
AJ LambEng 79 465633.87 978(1984)-1163(1991)913303333.91
BE CongdonNzl 61 344830.71 583(1965) - 818(1978)962317133.65
GM WoodAus 59 337431.11 816(1978)-1021(1985)993308131.96
GW FlowerZim 67 345728.411285(1995)-1625(2002)982287229.71

This table is ordered on the batting measure, "Runs per Fulfilled Innings - Milind-Ananth" (RpFI-MA). This is Milind's version of handling the non-fulfilled innings and has been discussed a few times already in my earlier articles. He has referred to this measure as 'ยต' in his blogs. To recapitulate in a simple manner, we expand, in an indirect manner, all not out innings which are below the Out-Average to the Out-Average. This ensures that scores like 5*, 15* and 20* do not lower the RpI figures drastically. This method takes care of the problem in Batting Average with middle order batsmen remaining unbeaten in 15-20% of the innings they played in.

I have given the example of Bradman to illustrate this calculation. The average of the 70 innings in which he was dismissed is 83.83 (5868/70). His 10 not-outs are 37*, 299*, 103*, 144*, 102*, 56*, 127*, 57*, 30* and 173*. Out of these ten innings, six are higher than 83.83 and are considered "fulfilled innings". It is clear why. These innings have gone past the Out-Average. The other four innings - 37*, 56*, 57rh and 30* - are indirectly extended to the Out-Average, by tweaking the number 4 to a lower value 2.147(180/83.83). The value of RpFI-MA for Bradman is 6996/(76+2.147) which works out to 89.50. Hats off to Milind's awareness for maximum computing correctness.

It should be noted that the RpFI-MA value is derived for each streak and the best one presented here.

The first entry is the expected one. 99.94 would have been more easily recognisable. But let us start looking at Bradman's RpFI-MA figure carefully since that would appear quite frequently hereafter. The RpFI-MA is 89.50. Ponting continues to be in second place with an excellent RpFI-MA of 66.5. This is about 25% off Bradman's average and firmly places Ponting's streak on a fairly high pedestal.

Then we have changes from the earlier table. The increased number of not-outs, which are inherent in any middle order batsman's career, propels Sobers, with a RpFI-MA figure of 63.3 into third place. Sangakkara remains in the top five with 61.8 and the top five is completed by Yousuf, with 61.4.

Next in the list are the two modern legends: Tendulkar and Lara, who have almost identical figures either side of 60. Viv Richards is a surprise entry into the top ten. Kallis and Hobbs complete the top ten.

Grant Flower is comfortably propping up the table with a fairly low value of 29.7. Again, let me remind you that I have ignored allrounders and wicketkeepers. We also have four other batsmen who promised a lot but did not deliver much. Especially Graeme Hick.

Test bowling streaks: Wickets captured in 27 Tests
Bowler Team Tests Wkts WpT Best 27-Test streak Wkts WpT Worst 27-Test streak Wkts WpT Ratio
MuralitharanSlk1338006.01670(2003)-1820(2006)2057.61195(1992)-1359(1997)1013.7203.0
SF BarnesEng 271897.0 65(1901) - 133(1914)1897.0 65(1901) - 133(1914)1897.0100.0
YounisPak 873734.31151(1990)-1268(1994)1776.61442(1999)-1617(2002) 823.0215.9
SK WarneAus1457084.91615(2002)-1763(2005)1736.41405(1998)-1556(2001) 933.4186.0
RJ HadleeNzl 864315.0 959(1983)-1072(1987)1646.1 710(1973) - 873(1980)1184.4139.0
MD MarshallWin 813764.6 991(1984)-1100(1988)1616.0 837(1978) - 988(1984)1094.0147.7
DK LilleeAus 703555.1 790(1977) - 908(1981)1605.9 876(1980) - 955(1983)1294.8124.0
CV GrimmettAus 372165.8 195(1930) - 251(1936)1595.9 166(1926) - 234(1934)1385.1115.2
DW SteynSaf 733715.11830(2007)-1951(2010)1575.81913(2009)-2056(2012)1294.8121.7
R BenaudAus 632483.9 431(1956) - 507(1961)1575.8 347(1952) - 433(1956) 732.7215.1
Imran KhanPak 883624.1 909(1981)-1058(1986)1545.71067(1987)-1182(1992) 782.9197.4
A KumbleInd1326194.71574(2001)-1724(2004)1535.71247(1994)-1387(1997) 873.2175.9
AK DavidsonAus 441864.2 449(1958) - 537(1963)1515.6 372(1953) - 483(1959) 923.4164.1
...............
BracewellNzl 411022.5 891(1980)-1093(1988) 712.61000(1984)-1138(1990) 592.2120.3
RazzaqPak 461002.21584(2002)-1809(2006) 672.51487(2000)-1716(2004) 481.8139.6
TE BaileyEng 611322.2 386(1954) - 440(1957) 662.4 334(1951) - 412(1955) 441.6150.0
RJ ShastriInd 801511.9 962(1983)-1054(1986) 632.31063(1986)-1177(1991) 371.4170.3
CL HooperWin1021141.11364(1997)-1553(2001) 451.71096(1988)-1174(1991) 110.4409.1

For bowlers also, I have presented the best 27-Test streak as also the worst 27-Test streak. It is to be expected that the bowlers will have much wider fluctuations between their best and worst streaks. This can be seen later.

What do we have here? We would have expected Barnes, with his career average of seven wickets per Test, sitting comfortably at the top. No, not this time. Murali, that Lankan wizard, has comfortably overtaken Barnes and is perched on top with 205 wickets in 27 Tests during the 2003-2006 period. An extraordinary WpT figure of 7.6.

Do I hear the words Bangladesh and Zimbabwe and home conditions in Sri Lanka? I am a firm believer in the axiom that a Test wicket is a Test wicket. There have been weak teams right through the 137 years of Test cricket. Surely Bangladesh and Zimbabwe are no worse than many teams during the 1930s, 1950s, 1960s? Even many established teams were poor. How many 5-0 drubbings have teams endured? Not to forget the uncovered pitches during the first few decades of Test cricket. So let us keep our views upright and not try to belittle one player or group of players. One day I should do an analysis to prove the point I have briefly talked about.

The next five bowlers are contemporary greats. How else can one describe Waqar Younis, Shane Warne, Richard Hadlee, Malcolm Marshall and Dennis Lillee? All these bowlers have accumulated in excess of 160 wickets in their prime stretch of 27 Tests. A special mention should be made of Marshall, who has averaged around six wickets per Test in his prime despite the presence of so many world-class bowlers in his team.

The top ten is completed by Clarrie Grimmett, the surprise package - Richie Benaud - and Steyn. Benaud is indeed a surprise. To have accumulated over 150 wickets is praiseworthy and should let us revise our opinion of him.

Bill O'Reilly's 27 Tests took over 14 years since he played one Test after the war and this Test completed the tally of 27 Tests. Erapalli Prasanna took over 11 years to play 27 Tests. Mitchell Johnson, Tony Greig and Andrew Flintoff took less than two years to play the 27 Tests.

As has happened often, Ravi Shastri and Carl Hooper prop up the table with very poor aggregates below 65 wickets. Not one of these five is a pure bowler.

Wilfred Rhodes is in a freak situation. He went through a 27-Test streak with 95 wickets and another 27-streak with 15 wickets. Sobers had very striking contrasting streaks with 104 and 24 wickets respectively. Kallis has similar figures. The first real surprise is Ian Botham. He picked up 146 and 57 wickets respectively in his best and worst streaks. Murali picked up 205 and 101 wickets respectively. Thus it can be seen that about ten bowlers have ratios more than 200%. Surprisingly, despite the non-bowling Tests, Imran Khan finishes with 154 and 78 wickets respectively.

Test bowling streaks: Bowling average across 27 Tests
Bowler Team Tests Wkts Avge Best 27-Test streak Wkts BowAvge
Imran KhanPak 8836222.81 909(1981)-1058(1986)15414.85
SF BarnesEng 2718916.43 65(1901) - 133(1914)18916.43
MD MarshallWin 8137620.95 991(1984)-1100(1988)16117.09
Waqar YounisPak 8737323.561151(1990)-1268(1994)17717.32
JC LakerEng 4619321.25 383(1954) - 464(1959)12617.49
J BriggsEng 3311817.75 23(1886) - 62(1899)11817.64
MuralitharanSlk13380022.731670(2003)-1820(2006)20518.15
CEL AmbroseWin 9840520.991144(1990)-1257(1994)14018.16
RJ HadleeNzl 8643122.30 959(1983)-1072(1987)16418.46
AV BedserEng 5123624.90 317(1949) - 389(1954)14918.56
AK DavidsonAus 4418620.53 449(1958) - 537(1963)15118.95
Wasim AkramPak10441423.621134(1990)-1268(1994)14919.24
IT BothamEng10238328.40 806(1977) - 881(1980)14619.30
SM PollockSaf10842123.121380(1997)-1475(1999)13019.38
AA DonaldSaf 7233022.251320(1995)-1437(1998)14819.72
RGD WillisEng 9032525.20 780(1976) - 836(1978)11719.89
...............
RafiqueBng 3310040.761512(2000)-1832(2007) 9136.37
AF GilesEng 5414340.601606(2002)-1728(2004) 8136.38
CL HooperWin10211449.431364(1997)-1553(2001) 4536.60
DVP WrightEng 3410839.11 263(1938) - 326(1950) 9039.39
RJ ShastriInd 8015140.96 962(1983)-1054(1986) 6339.44

I was almost certain that the average of Barnes would be the best, even though he was upstaged in the wickets tally by Murali. But there is another bowling giant who firmly shuts the door on Barnes. Imran Khan captured 154 wickets in the five-year period around early eighties at an unbelievable average of 14.85. Yes, that is true. This is the sort of average the demon bowlers around the turn of the 19th century used to get, bowling on uncovered wickets against teams new to Test cricket. I have carefully pondered over this. In my opinion the numbers of Murali and Imran Khan are two of the greatest bowling achievements ever. Imran's career figures changed from 128 at 29.45 to 282 at 21.48 during this memorable period.

The next three bowlers are Marshall, Younis and Jim Laker. All are wonderful bowlers. After Johnny Briggs appears Murali, who has an average of 18.15. Not bad for a spinner. Both Younis and Akram find a place in the top 12. What a period for Pakistan. Between 1990 and 1994, they had these two stalwarts picking up 320 wickets at an average of around 17.

The usual culprits prop up the table, including the ever-present table-proppers, Shastri and Hooper.

There is a case for increasing the bowling cut-off to a higher number of Tests. However, I am satisfied with this: after all 27 Tests have produced 150 wickets and more.

What do we conclude? Bradman was supreme, there has been no doubt about it and this analysis confirms this view. In his long career, scarred by six years of war and deliberate bodyline attack, he accumulated nearly 20% more runs at 25% higher RpFI-MA value than the next player who was playing at his prime. This one sentence confirms that his zone was beyond all other players' zones. This is the one analysis where there is a smaller gap between the master and the next best, Ponting. He is only around 20% off. Let us give him due credit. Lara and Sangakkara also stand out.

Murali, on the aggregate front, and Imran, on the average measure, have upstaged Barnes, considered by many to be one of the greatest, with career figures expected to be unreachable. That too, in modern conditions. Maybe all of us must give the modern bowlers their credit. We tend to put the batsmen on a pedestal and do not give the bowlers their right due.

An interesting add-on observation from Milind. If we define an alternate measure of consistency of a batsman as the ones with the lowest ratios, which would be close to 100%, between best and worst streaks (min 78 Tests considered - to have a career at least 150% of the streak length), then the top batsmen are given below.

NJ Astle (106.9%), MS Dhoni (109.4%), L Hutton (110.1%), GC Smith, N Hussain, Richardson, RJ Hadlee, GS Chappell, IR Bell, A Ranatunga, Tillakaratne, A Flintoff, Barrington, MW Gatting, Abbas, AJ Lamb, RB Kanhai (114.8%). A nice nugget: thank you, Milind.

The names that appear here which were well-placed in my earlier Consistency analysis are Hutton, Ken Barrington and Rohan Kanhai.

I have uploaded the Table containing the four full tables. To download/view this file, please CLICK HERE.

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

Comments have now been closed for this article

  • on August 1, 2014, 11:19 GMT

    Nice insights Gerry. To add on, to manage a team like Pakistan which had 25 captains in the playing XI is indeed a task. Not just Imran, but Pakistan and SL in general, lose out to Indian viewer power. As I had observed in my previous comment as well, biased commentators add on a few inches to people's greatness. Ravi Shastri was talking to Dravid in a TV show about captaining in Tests, ODIs and T20 match. Dravid has never captained in T20 International, but the way question was framed, equated IPL to a T20 International. My point is not about that, but how the paid media deludes the general public into believing what they want them to. The recent one is "sessions won in a test". No matter how many sessions a team wins, who wins the match? These things take away the beauty of tests. So if India wins 14 / 15 sessions and lose 20/8 on day 5 post tea, commentators would keep talking about how India "dominated" and was "unlucky" to sustain viewership. Why dont they promote good cricket?
    [[
    Ranga,
    Interesting point about Sessions in Test cricket. What is being shown in Star Tv is a half-baked binary working. There is no doubt that England won all sessions and no doubt that India did not have 2 hours dominance. However back in 2001, I did the Sessions analysis differently.
    I would look at what happened in the session objectively using runs scored, wickets captured, runs saved, the quality of balls bowled and most importantly the partnership which is unbeaten at the break.
    There is a big difference between two sessions with lunch score of 100 for 2 at lunch. If the batting team had been 90 for 0 and lost 2 wkts towards end of session the momentum is with the bowling team. If they had been 10 for 2 and a blossoming partnership of 90 at lunch, the momentum was with them. I would use all these and work the session at 37-63 or 65-35 or whatever. I would then add these progressively as the match progresses and discuss at the end of each day. The advantage is that this would enable the analysts to distinguish between three close sessions and three dominating sessions.
    Theoretically let me present a scenario. India moved to 200 for 6 at lunch. They would give give this session to England. Then they moved to 270 for 9 at tea. Again to England. Leading by 14-0. After tea, India did a South Africa and finished at 330 for 9. Session to India. 14-1 and a drawn Test.
    Ananth
    ]]

  • Cool_Jeeves on July 31, 2014, 9:04 GMT

    2/2... This seems to show that Imran was practically a one-man army. Qadir's averages are not great. Akram was good but nowhere near what he became later. Sarfaraz had faded. Qasim / Tauseef were brief participants. During this time, Pakistan drew with West Indies at home and away. Windies were not at their best, with Holding and Garner not playing in these series, but still stronger than any other team by a huge margin. Pakistan relied on Imran heavily in these tests. When India were blanked 3-0 in 1982-83, Imran said that to maintain pressure he had to bowl in long spells, so he rarely bowled at express speeds after the first spell. As a consequence of long spells, he developed a stress fracture, else his numbers may have been even better. One can rarely have seen a better example of leadership should personal example and action. If Sobers was all genius, Imran was sheer strength and will-power. I would give Imran the edge in their respective peak periods.
    [[
    Thanks, gerry, for some nice insights on Imran. I feel that he is not often given his due. Player skills is one thing. But the Captaincy is something else. I would place Imran amongst the best three captains of all time. Not just on the field but the way he worked tirelessly for neutral umpires. When the talk on greatest cricketers come, we tend to talk of Bradman, Sobers, Tendulkar et al. We never seem to bring in Imran.
    Ananth
    ]]

  • Cool_Jeeves on July 31, 2014, 8:59 GMT

    Hi Ananth, Nice work. Plenty of comments on Imran, and especially his bowling and average. But there are some issues worth highlighting, as I followed his career obsessively right from 1976-77 Sydney Test.

    Imran's home and away averages in this period hold some clues. His home average is 12.9 for 102 wickets, and away average 18.7 for 52 wickets. On closer scrutiny, you will find that the scores in many home matches during this period, specifically involving West Indies, were very low. In fact, 7 tests and only 1 century (Richards 120) in 1981+1986, and scores like 53 all out and 77 all out.

    But this link from statsguru throws some more light: http://stats.espncricinfo.com/ci/engine/stats/index.html?class=1;filter=advanced;orderby=bowling_average;qualmin1=10;qualval1=wickets;size=200;spanmax1=31+Dec+1986;spanmin1=1+Jan+1981;spanval1=span;team=7;template=results;type=bowling;view=host

    ...

  • on July 29, 2014, 13:53 GMT

    Right observation Ananth - "On the evidence of the comments so far, the readers have also matured. They are looking beyond one player". Having said that, the influence of TV is so very huge, by and large. Many people (who comment in this space - not just these articles, but in gen) do get hugely influenced by what the commentators say. It influences to an extent, even who they idolize. In general, I tend to agree that it would not be unfair to Sir Gary to mention Niazi in the same breath as Sobers. Captaining Pakistan is the toughest, particularly for a bowling captain. Particularly when <i>not</i>playing against India. In fact, to me, We had 3 bands of allrounders - Sobers/Imran, Kapil, Botham types and then you had Flintoff, Cairns, Hadlee , Akram types.
    [[
    Probably Imran goes ahead of Sobers on the third area: captaincy.
    Ananth
    ]]
    I also don't really see why people cant appreciate Murli or Lara (or even Hadlee) - Being lone rangers in a team game is not that easy. I don't believe that the opportunities to score are more when you play in a weak team.
    [[
    I would also appreciate the bowler who stands out in a quartet, like Marshall, than pull down the one who shone in a less strong team. The first is a positive attitude and the later is negative.
    Ananth
    ]]

  • kamranwasti on July 29, 2014, 7:18 GMT

    Excellent piece. A striking feature is the way teams tend to demolish opposing teams when they have two bowlers on a hot streak - Wasim, Waqar aside, you have Garner and Marshall doing the same for the West Indies around the same era and they destroyed every team for less than 100 at least once. I had always wondered where would Imran's show over these five years stood. I think he had just one ordinary return (quality-wise) when he had 3 for 112 against Sri Lanka in Faisalabad on a dead wicket on his return to test cricket. Othewise, he had exceptional returns in every single outing - not a single ordinary return even when he took 2 for 84 at Lord's which Wisden assesses as a display of magnificent stamina on a flat wicket. I think I will have to format this table on an excel sheet and find out such combinations.
    [[
    Kamran, I have various display patterns for my database and when I see Imran's 27-Test streak on one screen, I am awestruck. As you say, one has to search deep for an average performance. And the opponents: Aus, Eng, Ind, Slk and Win. And everywhere. I stand by what I said in my piece. This streak of Imran and Murali's 27-Test streak are two of the greatest ever cricketing achievements.
    Ananth
    ]]

  • on July 28, 2014, 3:47 GMT

    IMO using Tests for batsmen is still an artificial fabrication. Bradman may have got to play 104 innings in his 52 Tests vs the 80 he actually played. The Bradman standard is simply "80 innings at 99.94". Using 80 innings, runs accumulated in them and the standard means of calculating average is the simplest and cleanest means of comparison to Bradman.
    [[
    I strongly believe that a Test is the most logical unit for measuring a player's career and related aspects. That is the base for a result. Having said that I find nothing wrong in having innings as basis. However I do not think there will be a material change to the results.
    And I also strongly believe that the batting average is biased very strongly towards the middle order batsmen and the RpFI addresses these concerns. If we take the top-20 averages, the % of not outs runs thus: 12.5, 9.8, 10.0, 10.7, 11,5, 6.2, 7.9, 11.4, 13.1, 6.9, 9.5, 10.9, 12.6, 11.3, 10.0, 2.6, 11.1, 11.2 and 7.7. This is a way-out range spanning 2.6% to 14.3%. And down below, Chanderpaul, Andy Flower, Misbah and Waugh have NO% in excess of 17.
    Tradition has to be respected. But nothing set in stone. When the game itself has undergone so much of a change, what is the wrong in addressing a serious issue and coming out with a very compelling solution. Maybe the RpFI deserves a separate article.
    Also we will then have 80 innings as the basis for batting and 27 Tests as the basis for bowling.
    Ananth
    ]]

  • SLSup on July 27, 2014, 6:09 GMT

    Good to see Barnes & Murali paired for once. While it is true SOME "dilute" Murali's stats by citing his record in the Sub Contient the FACT remains Murali took 188 wickets in 36 Tests OUTSIDE the Sub Continent at 5.22 wickets PER TEST (only the BEST bowlers achieve that). Murali has 6.30 wickets PER TEST in Sub Continent conditions. Barnes took all his 7 PER TEST in AUS/ENG/SA in conditions similar. Murali gets his due by thoughtful folks like yourself. Someone said SOME moden greats didn't face Waqar, Akram, Donald, Walsh, Ambrose et al but the fact remains Don Bradman didn't face them either nor Murali, Warne, Holding, Garner, Marshall, or Steyn. Sangakkara has played against same bowlers that Sachin, Rahul, Mahela, Kallis, Ponting, Yousuf, and Lara did but averages 70 in the last 82 Tests. On Not Outs. A debutant scoring 44*, 10*, 72*, and gets out for 45 is said to average 169! Lie. It's 42.25. An Inning must count as completed when a batsman walks-in past the boundary line.
    [[
    We are considering all innings as completed innings, subject to the tweak for the not out innings below the out-avge. You will see that the RpFI factor has a lot of going for it.
    Ananth
    ]]

  • harshthakor on July 27, 2014, 4:21 GMT

    Very creditable research Ananth.I hail Imran Khan's great performance in his peak years and it is a revelation studying his statistics which overshadow any bowler in terms of average,including Hadlee or Marshall. However to me a factor that should have been taken into amount was the strike rate and not just wickets per test.Marshall had an outstanding overall strike rate of 46.7.Another factor is how much a bowler had to carry the brunt of the attack.Richard Hadlee was the best example who from 1978-88 captured 330 wickets in 60 tests for a weak bowling unit..In contrast Malcolm Marshal bowled with 3 of the greatest fast bowlers of his day and still was more prolific than anyone of his time in terms of total haul of wickets.To me it is Waqar Younus who is statistically the best of the post-war paceman with his amazing strike rate of 42.3. Dennis Lille is very unlucky as World Series cricket intervened whee he captured 79 scalps in 15 games.He may well been close to the top.
    [[
    Lovely quartet of comments, Harsh. I will feature only this.
    For bowlers I have always felt that wickets are very important. The bottomline is that 5 for 100 is almost always considered better than 3 for 25 (other than when the team is defending 100!!!). Let me put it this way. With the enlightened readers I have, I will conclude that they are capable of tweaking the analysis, at least in their minds, the strike rate. Let me see whether I can do a completely new analysis with strike rates in place of bowling average.
    Ananth
    ]]

  • jw76 on July 26, 2014, 20:10 GMT

    Interesting, thanks, but it would have been easier to follow using conventional averages. A 50-Test streak would not only have been a very convenient round number, but would also have still further improved Bradman's record with his first and last Tests lopped off
    [[
    I agree with the validity of both your points. But the minute we convert 52 to 50, we run the risk of readers complaining and asking for the few Tests in which the batsmen did not bat to be taken off and so on. The 52 contained quite a few failures as would other batsman's 52 test streaks. In fact for Bradman himself there were two Tests he did not bat in and that forms another 50-Test streak.
    I think the RpFI, if you take the trouble to understand the same, addresses almost all problems with Not outs. If I say that Chanderpaul, with his 24 not outs, has a higher batting average than Lara, with 2 not outs, in their respective streaks, you would understand the futility of the batting average.
    Ananth
    ]]

  • on July 26, 2014, 16:21 GMT

    Wilfred Rhodes began and ended his Test career as a left-arm spinner, but in the middle years he was Jack Hobbs's partner as opening batsman. I would guess that his poor bowling returns would coincide with the period when he was batting at 2.
    [[
    Yes, that explains the 600% ratio.
    Ananth
    ]]

  • on August 1, 2014, 11:19 GMT

    Nice insights Gerry. To add on, to manage a team like Pakistan which had 25 captains in the playing XI is indeed a task. Not just Imran, but Pakistan and SL in general, lose out to Indian viewer power. As I had observed in my previous comment as well, biased commentators add on a few inches to people's greatness. Ravi Shastri was talking to Dravid in a TV show about captaining in Tests, ODIs and T20 match. Dravid has never captained in T20 International, but the way question was framed, equated IPL to a T20 International. My point is not about that, but how the paid media deludes the general public into believing what they want them to. The recent one is "sessions won in a test". No matter how many sessions a team wins, who wins the match? These things take away the beauty of tests. So if India wins 14 / 15 sessions and lose 20/8 on day 5 post tea, commentators would keep talking about how India "dominated" and was "unlucky" to sustain viewership. Why dont they promote good cricket?
    [[
    Ranga,
    Interesting point about Sessions in Test cricket. What is being shown in Star Tv is a half-baked binary working. There is no doubt that England won all sessions and no doubt that India did not have 2 hours dominance. However back in 2001, I did the Sessions analysis differently.
    I would look at what happened in the session objectively using runs scored, wickets captured, runs saved, the quality of balls bowled and most importantly the partnership which is unbeaten at the break.
    There is a big difference between two sessions with lunch score of 100 for 2 at lunch. If the batting team had been 90 for 0 and lost 2 wkts towards end of session the momentum is with the bowling team. If they had been 10 for 2 and a blossoming partnership of 90 at lunch, the momentum was with them. I would use all these and work the session at 37-63 or 65-35 or whatever. I would then add these progressively as the match progresses and discuss at the end of each day. The advantage is that this would enable the analysts to distinguish between three close sessions and three dominating sessions.
    Theoretically let me present a scenario. India moved to 200 for 6 at lunch. They would give give this session to England. Then they moved to 270 for 9 at tea. Again to England. Leading by 14-0. After tea, India did a South Africa and finished at 330 for 9. Session to India. 14-1 and a drawn Test.
    Ananth
    ]]

  • Cool_Jeeves on July 31, 2014, 9:04 GMT

    2/2... This seems to show that Imran was practically a one-man army. Qadir's averages are not great. Akram was good but nowhere near what he became later. Sarfaraz had faded. Qasim / Tauseef were brief participants. During this time, Pakistan drew with West Indies at home and away. Windies were not at their best, with Holding and Garner not playing in these series, but still stronger than any other team by a huge margin. Pakistan relied on Imran heavily in these tests. When India were blanked 3-0 in 1982-83, Imran said that to maintain pressure he had to bowl in long spells, so he rarely bowled at express speeds after the first spell. As a consequence of long spells, he developed a stress fracture, else his numbers may have been even better. One can rarely have seen a better example of leadership should personal example and action. If Sobers was all genius, Imran was sheer strength and will-power. I would give Imran the edge in their respective peak periods.
    [[
    Thanks, gerry, for some nice insights on Imran. I feel that he is not often given his due. Player skills is one thing. But the Captaincy is something else. I would place Imran amongst the best three captains of all time. Not just on the field but the way he worked tirelessly for neutral umpires. When the talk on greatest cricketers come, we tend to talk of Bradman, Sobers, Tendulkar et al. We never seem to bring in Imran.
    Ananth
    ]]

  • Cool_Jeeves on July 31, 2014, 8:59 GMT

    Hi Ananth, Nice work. Plenty of comments on Imran, and especially his bowling and average. But there are some issues worth highlighting, as I followed his career obsessively right from 1976-77 Sydney Test.

    Imran's home and away averages in this period hold some clues. His home average is 12.9 for 102 wickets, and away average 18.7 for 52 wickets. On closer scrutiny, you will find that the scores in many home matches during this period, specifically involving West Indies, were very low. In fact, 7 tests and only 1 century (Richards 120) in 1981+1986, and scores like 53 all out and 77 all out.

    But this link from statsguru throws some more light: http://stats.espncricinfo.com/ci/engine/stats/index.html?class=1;filter=advanced;orderby=bowling_average;qualmin1=10;qualval1=wickets;size=200;spanmax1=31+Dec+1986;spanmin1=1+Jan+1981;spanval1=span;team=7;template=results;type=bowling;view=host

    ...

  • on July 29, 2014, 13:53 GMT

    Right observation Ananth - "On the evidence of the comments so far, the readers have also matured. They are looking beyond one player". Having said that, the influence of TV is so very huge, by and large. Many people (who comment in this space - not just these articles, but in gen) do get hugely influenced by what the commentators say. It influences to an extent, even who they idolize. In general, I tend to agree that it would not be unfair to Sir Gary to mention Niazi in the same breath as Sobers. Captaining Pakistan is the toughest, particularly for a bowling captain. Particularly when <i>not</i>playing against India. In fact, to me, We had 3 bands of allrounders - Sobers/Imran, Kapil, Botham types and then you had Flintoff, Cairns, Hadlee , Akram types.
    [[
    Probably Imran goes ahead of Sobers on the third area: captaincy.
    Ananth
    ]]
    I also don't really see why people cant appreciate Murli or Lara (or even Hadlee) - Being lone rangers in a team game is not that easy. I don't believe that the opportunities to score are more when you play in a weak team.
    [[
    I would also appreciate the bowler who stands out in a quartet, like Marshall, than pull down the one who shone in a less strong team. The first is a positive attitude and the later is negative.
    Ananth
    ]]

  • kamranwasti on July 29, 2014, 7:18 GMT

    Excellent piece. A striking feature is the way teams tend to demolish opposing teams when they have two bowlers on a hot streak - Wasim, Waqar aside, you have Garner and Marshall doing the same for the West Indies around the same era and they destroyed every team for less than 100 at least once. I had always wondered where would Imran's show over these five years stood. I think he had just one ordinary return (quality-wise) when he had 3 for 112 against Sri Lanka in Faisalabad on a dead wicket on his return to test cricket. Othewise, he had exceptional returns in every single outing - not a single ordinary return even when he took 2 for 84 at Lord's which Wisden assesses as a display of magnificent stamina on a flat wicket. I think I will have to format this table on an excel sheet and find out such combinations.
    [[
    Kamran, I have various display patterns for my database and when I see Imran's 27-Test streak on one screen, I am awestruck. As you say, one has to search deep for an average performance. And the opponents: Aus, Eng, Ind, Slk and Win. And everywhere. I stand by what I said in my piece. This streak of Imran and Murali's 27-Test streak are two of the greatest ever cricketing achievements.
    Ananth
    ]]

  • on July 28, 2014, 3:47 GMT

    IMO using Tests for batsmen is still an artificial fabrication. Bradman may have got to play 104 innings in his 52 Tests vs the 80 he actually played. The Bradman standard is simply "80 innings at 99.94". Using 80 innings, runs accumulated in them and the standard means of calculating average is the simplest and cleanest means of comparison to Bradman.
    [[
    I strongly believe that a Test is the most logical unit for measuring a player's career and related aspects. That is the base for a result. Having said that I find nothing wrong in having innings as basis. However I do not think there will be a material change to the results.
    And I also strongly believe that the batting average is biased very strongly towards the middle order batsmen and the RpFI addresses these concerns. If we take the top-20 averages, the % of not outs runs thus: 12.5, 9.8, 10.0, 10.7, 11,5, 6.2, 7.9, 11.4, 13.1, 6.9, 9.5, 10.9, 12.6, 11.3, 10.0, 2.6, 11.1, 11.2 and 7.7. This is a way-out range spanning 2.6% to 14.3%. And down below, Chanderpaul, Andy Flower, Misbah and Waugh have NO% in excess of 17.
    Tradition has to be respected. But nothing set in stone. When the game itself has undergone so much of a change, what is the wrong in addressing a serious issue and coming out with a very compelling solution. Maybe the RpFI deserves a separate article.
    Also we will then have 80 innings as the basis for batting and 27 Tests as the basis for bowling.
    Ananth
    ]]

  • SLSup on July 27, 2014, 6:09 GMT

    Good to see Barnes & Murali paired for once. While it is true SOME "dilute" Murali's stats by citing his record in the Sub Contient the FACT remains Murali took 188 wickets in 36 Tests OUTSIDE the Sub Continent at 5.22 wickets PER TEST (only the BEST bowlers achieve that). Murali has 6.30 wickets PER TEST in Sub Continent conditions. Barnes took all his 7 PER TEST in AUS/ENG/SA in conditions similar. Murali gets his due by thoughtful folks like yourself. Someone said SOME moden greats didn't face Waqar, Akram, Donald, Walsh, Ambrose et al but the fact remains Don Bradman didn't face them either nor Murali, Warne, Holding, Garner, Marshall, or Steyn. Sangakkara has played against same bowlers that Sachin, Rahul, Mahela, Kallis, Ponting, Yousuf, and Lara did but averages 70 in the last 82 Tests. On Not Outs. A debutant scoring 44*, 10*, 72*, and gets out for 45 is said to average 169! Lie. It's 42.25. An Inning must count as completed when a batsman walks-in past the boundary line.
    [[
    We are considering all innings as completed innings, subject to the tweak for the not out innings below the out-avge. You will see that the RpFI factor has a lot of going for it.
    Ananth
    ]]

  • harshthakor on July 27, 2014, 4:21 GMT

    Very creditable research Ananth.I hail Imran Khan's great performance in his peak years and it is a revelation studying his statistics which overshadow any bowler in terms of average,including Hadlee or Marshall. However to me a factor that should have been taken into amount was the strike rate and not just wickets per test.Marshall had an outstanding overall strike rate of 46.7.Another factor is how much a bowler had to carry the brunt of the attack.Richard Hadlee was the best example who from 1978-88 captured 330 wickets in 60 tests for a weak bowling unit..In contrast Malcolm Marshal bowled with 3 of the greatest fast bowlers of his day and still was more prolific than anyone of his time in terms of total haul of wickets.To me it is Waqar Younus who is statistically the best of the post-war paceman with his amazing strike rate of 42.3. Dennis Lille is very unlucky as World Series cricket intervened whee he captured 79 scalps in 15 games.He may well been close to the top.
    [[
    Lovely quartet of comments, Harsh. I will feature only this.
    For bowlers I have always felt that wickets are very important. The bottomline is that 5 for 100 is almost always considered better than 3 for 25 (other than when the team is defending 100!!!). Let me put it this way. With the enlightened readers I have, I will conclude that they are capable of tweaking the analysis, at least in their minds, the strike rate. Let me see whether I can do a completely new analysis with strike rates in place of bowling average.
    Ananth
    ]]

  • jw76 on July 26, 2014, 20:10 GMT

    Interesting, thanks, but it would have been easier to follow using conventional averages. A 50-Test streak would not only have been a very convenient round number, but would also have still further improved Bradman's record with his first and last Tests lopped off
    [[
    I agree with the validity of both your points. But the minute we convert 52 to 50, we run the risk of readers complaining and asking for the few Tests in which the batsmen did not bat to be taken off and so on. The 52 contained quite a few failures as would other batsman's 52 test streaks. In fact for Bradman himself there were two Tests he did not bat in and that forms another 50-Test streak.
    I think the RpFI, if you take the trouble to understand the same, addresses almost all problems with Not outs. If I say that Chanderpaul, with his 24 not outs, has a higher batting average than Lara, with 2 not outs, in their respective streaks, you would understand the futility of the batting average.
    Ananth
    ]]

  • on July 26, 2014, 16:21 GMT

    Wilfred Rhodes began and ended his Test career as a left-arm spinner, but in the middle years he was Jack Hobbs's partner as opening batsman. I would guess that his poor bowling returns would coincide with the period when he was batting at 2.
    [[
    Yes, that explains the 600% ratio.
    Ananth
    ]]

  • its.rachit on July 26, 2014, 13:17 GMT

    i have always believed that Imran never quite got his due ... IMHO, he is up there with Sobers and Bradman as the greatest cricketer ... average of 22 with the ball and 37 with the ball comfortably compares to Sobers with 57 and 34 ... even better than him according to me ... Imran averaged 19 with the ball an 51 with the bat between 1982-1992 ... comfortably matching the 99.94 of bradman as a combined contribution ... somehow due to his being an all-rounder, the bowler got lost some where .. he averages better than wasim,waqar and lillie ... neve quite understood why these 3 are universally put above him as bowlers .. Ananth your comments on this ???
    [[
    I would say, as a bowler Imran was right up there with the best. Add the batting average of 37 (7 above Srikkanth, let me remind readers), he is certainly in a class of his own. A class which had five players. So I must say i agree with you.
    Ananth
    ]]

  • Anand_S on July 26, 2014, 12:46 GMT

    Ananth: Another out of the box article. Thanks a lot for this one. Just out of curiosity. If you consider the "best" 52 test matches for batsmen, how do they compare with the Don? I mean do you think if Ponting or anyone else in ther "best" 52 matches could have surpassed Don in average/aggregate? Similarly for the bowlers. I am not trying to depreciate the efforts of Bradman or Barnes by this exercise. Infact this will show why they were so great. Imagine if we can finally come up with Bradman is better than all the best innings of Sachin or Ponting. That will show what a daylight exists between him and others!!!
    [[
    Will do and post a result comment.
    When I did the work, Anand, I found that this will be a pointless exercise. The 5 batsmen above 8000 runs in their 52 best Tests is given below.
    Lara: 8655, Ponting: 8608, SRT: 8607, Sanga: 8539, Kallis: 8110
    This is understandable since Tendulkar has scored 51 100s and he is guaranteed these number of Tests at or well above 100. His 52nd best Test stands at 119. Sanga's stands at 99 and Lara's 92. When you realize that Lara had so many big innings, it makes sense that he scored above 170 RpT in his 52 best Tests. So this is really not a meaningful exercise.
    Probably what is of interest is, how many of these 170 players have scored above 6996 in their best 52 Tests. I need to dome more work on that.
    Ananth
    ]]

  • on July 26, 2014, 10:06 GMT

    A "peer ratio" of the batsman's ( or bowlers) best 52 match streak vs. the rest of the field would equalize several factors. For eg. it would be clearer that between 1957-1968 Sobers was clearly the standout, Richards from 1976-1984 . Tendulkar in the mid to late 1990s etc. How far ahead of the general pack was Ponting in his best 52 match streak in the mid 2000s.
    [[
    Again, a few years back I had done a Peer analysis and it was very well received. I could do a complex task of taking this specific 52-Test (27-Test) streak and doing a peer analysis for this duration. This is quite tough since the period is unique to each player. But will put this into my to-do tray.
    Ananth
    ]]

  • Cool_Jeeves on August 2, 2014, 8:49 GMT

    https://www.youtube.com/watch?v=5n9Pslhkp9c

    Please see this link at 1:13

  • Cool_Jeeves on August 2, 2014, 7:16 GMT

    Hi Ranga, Yes, the sessions idea grates. We have Ravi Shastri et al despreately searching for something interesting to do, and this comes in handy. One more irrirant is the way in which highlights for 1-2 minutes are shown - the camera will suddenly zoom, suddenly pan, show from the top, show from square leg, create a lot of movement on the screen. One more example of contempt for viewers. Regarding Dravid, I cannot believe that he can lead RR and be on the payroll's of India Cement (as it is, I was disillusioned when he said six months back in a conference that IPL is the main hunting ground for talent, and that there was nothing wrong with India's fast bowling - on such efforts is the future of our cricket banking. Imran - watch this footage, at 1.13, a legendary dismissal. We saw it live from behind the WK, it was awesome. In a separate sequence of the same interview (unable to trace that link) he says he was at his peak between 1978-82, but played only in 5-10 tests.

  • on August 1, 2014, 18:29 GMT

    Lots of talk on Imran. Well i am from Pakistan and i find this heartwarming. People might find this surprising but here in Pakistan almost everyone talks only about Wasim and never quite recognize Imran's bowling achievements. Many Pakistani fans believe that Wasim was not just one or two steps,but miles ahead of any other bowler when it came to reverse swing.Well i for one believe Imran might just have been a step ahead of even Wasim at his peak when it came to reversing the old ball,and Imran could do it at a decent pace as well.Not all great bowlers have set aesthetic standards of fast bowling and set statistical records,but Imran did.
    [[
    Yes, Imran tends to lose out even amongst Pakistani followers possible because of his foray into politics. Wasim and Waqar have remained in the public eye on the cricketing front.
    Ananth
    ]]

  • Cool_Jeeves on August 1, 2014, 4:44 GMT

    Ananth, Apologies in advance. I have recently noticed that discerning cricket viewers dont really bring Tendulkar in while discussing greatest cricketers etc. (I am keeping my personal bias out of this). Bradman, Sobers, Richards, Imran were all leaders. They may have or may not have been the best tacticians, but they were leaders of stature. Tendulkar is not placed in that category. I personally would not rate him anywhere near the others mentioned here either, for similar reasons. His aggregates are very good, he is #1 in terms of longevity, his technique against all but the fastest bowlers was excellent (he took his eyes of the ball too often against the fastest quicks and kept getting hit on the head as a consequence), but to call him as one of the greatest 5 cricketers of all time is too much. Senior commentators in public may call him so, but it is partly for the sake of appearance.

  • STNS on July 30, 2014, 9:07 GMT

    I fail to understand how people say that Imran did not performed with the bat when he performed with the ball like Kapil & Botham etc..

    In these 27 tests Imran not only averaged 14.85 with the ball, but averaged 39.70 with the bat as well, which is absolutely outstanding, and better than the career batting averages of all the bowling all-rounders in the hjistory of the game.

    To me Bradman, Sobers & Imran are surely there in the top 5 cricketers of all time, if not top 3.

  • kamranwasti on July 30, 2014, 6:19 GMT

    Dear Ananth,

    I cannot find Venkatesh Prasad anywhere. Before his career receded in the last few games, he was one of the best Indian quick bowlers.
    [[
    Kamran, I also put in a limit of 100 wickets and Prasad just missed that mark. I was trying to avoid casual bowlers like Mark Waugh. Could have lossened the cut-off criterion a bit.
    Ananth
    ]]
    Heath Streak of Zimbabwe - exceptional for one of the consistently weakest teams ever.

    Also, I am quite surprised to find out that there is no sub-25 Indian bowler.

    The most interesting aspect for me is the line of the West Indian pace bowlers:

    During the mid-70s they were a good team and they had Roberts at his peak. Till 1982, Holding and Croft had their peak runs and from then on you had Garner and Marshall taking over before Ambrose and Bishop had their runs. As long as they had two fast-bowlers at their peak, the West Indians were a great team. The moment it got reduced to one (Walsh, post 1996), they declined. This is how important it is for them to have great fast-bowlers.
    [[
    The West Indian domination was more through bowlers than batsmen unlike the Australian domination which was through strengths in both departments.
    Ananth
    ]]

  • SLSup on July 29, 2014, 5:01 GMT

    @AshwinizXI talks about "aesthetic delights" and "chuckers" and Anantha rightly identifies it as an "extremely unfair comment". I've always had issues understanding the issue of elbow-flex. It is not that their SPEED is an issue cos there are "legitimate" bowlers who bowl much faster. It cannot be where they PITCH the ball cos ALL bowlers must pitch on the same pitch. It cannot be the turn itself because bowlers accused of "chucking" turn it to differet degrees and there are "legitimate" bowlers who spin more than the so called "chuckers". It cannot be their LENGTH cos all bowlers must work within the same pitch lengh. WHAT is it an unfair advantage to a "chucker" in AshwinizXI's world that his "aesthetically" pleasing variety cannot generate to the same extent, if not more? There are human beings who loves the game enough to want to play it at the highest level and all SOME see in them are their disabilities.

  • on July 28, 2014, 15:35 GMT

    "Where did we get into this top order against lower order argument. "

    Because of this - "And I also strongly believe that the batting average is biased very strongly towards the middle order batsmen and the RpFI addresses these concerns".

  • AshwinizXI on July 28, 2014, 9:07 GMT

    Statistics is so cruel. It puts an aesthetic delight and an eyesore (arguably a chucker) in the same bracket.
    [[
    Extremely unfair comment.
    Ananth
    ]]

  • on July 28, 2014, 8:12 GMT

    Also,I fully agree with your view of "Test is the most logical unit for measuring a player's career and related aspects" but ONLY when self-referencing a players career. L.e inn per test, Not outs etc are naturally equalized. Otherwise, when comparing to other players with different such stats it is an apples to oranges comparison. Which leads me to ask - Are the bowling innings for Barnes also different from the other bowlers ? Again this will lead to skewed results, even when considering bowlers.

  • on July 28, 2014, 8:08 GMT

    In general ,the best batsmen do better higher up the order. This is how I see it - Good batting tracks and bad batting tracks affect all batsmen in a team equally. However, Top order batsmen get disproportionate use of a good batting track to score big runs. Other batsmen are deprived of this luxury. On the other hand , on a poor batting track all may fall down equally. It may take a bit more patience to see off the new ball. But the best batsmen have the technique and ability to do so. So, the best batting positions for the best batsmen purely in terms of run scoring are at the top of the order. Middle order positions are more pivotal and are a bridge between the top and lower order .
    [[
    Where did we get into this top order against lower order argument.
    Ananth
    ]]

  • on July 28, 2014, 8:07 GMT

    A look at the career figures of the batsmen on the RPFI list. Averages in batting pos 1-3 and pos 4-10.

    Batsman No.1-3 4-10 DG Bradman 104 91 RT Ponting 56 42 GS Sobers 59 57 Sangakkara 60 31 Yousuf 31 53 (3 inn.s from 1-3) Tendulkar N/A 54 BC Lara 60 50 Richards 62 44 JH Kallis 50 57 JB Hobbs 56 86 ( 4 inns. From 4-10) ML Hayden 51 N/A de Villiers 37 57 R Dravid 52 54 Jayawardene 50 50 Miandad 51 52

  • on July 28, 2014, 5:09 GMT

    If one does a "High score " analysis - we may find that the big scores ( 250+ runs) are scored mostly by batsmen in Positions 1-3. Or otherwise middle order batsmen coming in relatively early on a good track ( This would generally mean the Top batsmen are relatively poor batsmen).
    [[
    All these, and more, are covered in the HSI article which moved on from "High score" to "All scores".
    Ananth
    ]]

  • on July 28, 2014, 5:01 GMT

    There are numerous reasons , such as given above, why the best batting positions for the best batsmen are actually as high up the order as possible. The lower order batsmen actually miss out tremendously on run scoring opportunities.(And then we decide to "penalise" them further by different treatment to NOs).

    In short - When coming to an alternative to the calculation of conventional batting average numerous factors have to be accounted for . This is an extremely nuanced theme. And broad brushes and ideas such as Milinds - though baby steps in the general direction are still far from viable alternatives.

  • on July 28, 2014, 4:57 GMT

    Re "I strongly believe that a Test is the most logical unit for measuring a player's career and related aspects". My view is that this is only the case when measuring "consistency" or batsmen will similar ratios of inns per Test , DNBs etc. If you " also strongly believe that the batting average is biased very strongly towards the middle order batsmen " then this ignores the fact that it is the batsmen in Positions 1-3 who get to fill in their boots on good batting tracks.
    [[
    And on seaming and bouncy bowler-centric tracks.
    Ananth
    ]]
    On a good batting track by the time the middle order batsmen get in ( say Steve Waugh) the top order batsmen have already put up a good total. No chance for Steve to do so unless he decides to declare at his own convenience, or otherwise for the most part he will try to score quick runs. Conversely on a poor batting track ,with the top order out the middle order requires to do a shoring up job.

  • Biophysicist on July 28, 2014, 4:51 GMT

    Dear Anantha: Your analysis is quite interesting and shows that even when you take the best streaks, none of the other batsmen measure anywhere near Bradman. However, in my opinion, to make a fair comparison of batsmen with Bradman and bowlers with Barnes, the only way would be to take the first 52 tests for the batsmen and first 27 tests for the bowlers. I am sure then none of the others would be anywhere near.
    [[
    The first 52/27 might be correct in some aspects but we are also trying to see, even if we give the other batsmen the best of conditions, the best out of 100+ such streaks, how far they are off.
    Ananth
    ]]

  • on July 28, 2014, 4:47 GMT

    Dear Mr. Narayanan: You are the best writer as long as you tell me that Imran Khan - albeit for a few yrs - was the best bowler in the world :)....Kidding aside, very informative and well researched. Thank you.

  • on July 28, 2014, 4:46 GMT

    Looking at your reply - I am not quite certain that you do get it. You have mistaken the trees for the forest. Runs accumulated in say 100 innings cannot be compared with runs accumulated in 80 innings. Pretty straightforward one may have thought. Also the Bradman benchmark is centered around "99.94". All other stats are essentially addendums to this central stat.
    [[
    I would only say one thing. If I had thought of 80 innings to start with I would have gone ahead and completed the article on that basis. However in my mind I always had the tail-end response to my four year old article: 52 and 27 Tests. I sent a short answer then highlighting Ponting's numbers. However I always wanted to do a complete article on 52/27. Anyhow thanks for your nice comments. I request you to take the "benefiting specific batsmen" monkey off your back. I have gone past that way back.
    Ananth
    ]]
    This leads me to another thought - although this is more concerned with the prior consistency articles. Instead of comparing a batsman's consistency relative to his own long term batting average - we may also compare the consistency of a batsman relative to his own team score per Test. i.e The variation of a batsman's score relative to the team score on a Test by Test basis.
    [[
    Offhand I would say that the HSI (High Score Index) work I did a few months back covers this aspect.
    Ananth
    ]]

  • on July 28, 2014, 4:15 GMT

    To elaborate - The Bradman standard ( to the general cricketing public) is simply "99.94". This is calculated by the 6996 runs accumulated over 80 innings using NOs in the conventional manner. That's it. All comparison should be to this benchmark. There was also another good article by Kariteya Date where he once again discussed Not outs. Wherein he tried to give different treatment to Top order and Lower order batsmen. However, with this method consideration is not given to the fact that lower order batsmen rarely get to fill in their boots on good batting tracks. Just a superficial observation that they have fewer NOs and hence these should be discounted is incorrect. In general the conventional method of calculating average works pretty well. One must restrain the tendency to modify systems simply to accomodate a few batsmen with relatively few Not Outs. The fewer NOs also mean maximum runs scored etc etc. The debate is extremely fine tuned . The current method is relatively ok.
    [[
    Now I get it. This is your grouse. Let me assure you that this is not to accommodate a few batsmen. In reality the ostrich-in-sand attitude is practiced by followers who refuse to recognize the problem. Anyhow I have answered in depth. Also what are "systems". Over the past 30 years how many rule changes have been there. Every concievable aspect of the game has been changed. Why not the analytical part.
    Ananth
    ]]

  • on July 28, 2014, 4:09 GMT

    WOW!!! Anant, have always been an admirer of your work. Read everything you wrote, but, this post, converted me to a fan. The same way I was converted to a Harsha Bhogle fan from admirer. Amazing analysis. Have been a huge admirer of Courtney Walsh. Expected him to feature in the list somewhere ...
    [[
    Walsh is there somewhere in the middle with 134 wickets and an average of 20+.
    Ananth
    ]]

  • SLSup on July 28, 2014, 1:18 GMT

    @Worldcricketlover talks of Ishant Sharma in the same breath as Imran. When I read Ishant had taken 7 wickets in the last ENG Test I had to look at the video of the wickets he'd taken and I wasn't surprised at what I saw. Except for perhaps 2/7 wickets Ishant himself was dumb-founded in the middle of the pitch! He was looking around at others to validate that he did take another wicket! It was almost like he wasy saying "Is this really happening" (cos I didnt DO it!).

  • SLSup on July 28, 2014, 0:46 GMT

    Anantha; RpFI is quite a fascinating take in attempting to derive a "completed" inning. I like it better than allowing a batsman easy runs to prop-up what an average score means. I somehow prefer an inning that a player partake into be considered from get go per the cricketing laws. In my exmple of the debutant (prior post) if the average is 169 after being dismissed only once out of 4 innings actually considered DONE once a game is over (per cricketing laws) then the average must be 42.25 per Inning. Then, there is the qusestion of how RiFI applies in the case of the hypothetical I've sighted. It's hypothetical but... it's also REAL! Haha. Hope you see my point.

    THEN there are umpires who screw or scew averages by giving out when it's not and not giving out when it is! Even DSR is now tweated in favor of THE UMPIRES where batsmen no longer get the benefit of the doubt (as in the case of Mahela's dismissal @ SSC yesterday). IF Bowden was SO sure why review SO much and give OUT!

  • Tansen on July 27, 2014, 18:02 GMT

    I am not sure if you mentioned it, but Imran also did not bowl in two of the Tests within his best 27 Tests. If you have counted those within the 27, then his numbers quoted above for wickets and average would be a bit off.
    [[
    Consecutive Tests are consecutive in letter and spirit. Any further tweaking is unnecessary. Then Bradman's 2 Tests have to be taken off and so on.
    Ananth
    ]]

  • Nuxxy on July 27, 2014, 17:40 GMT

    @Ananth: If you look at, for example, South Africa, it is often described as the most difficult place to bat, based on having the lowest batting average of the big 10 countries. If you weighted players based on overall average (or possibly a better batting metric), which players would be very affected, and which wouldn't be?
    [[
    It is essential that such analyses covering 137 years of Test cricket are not unnecessarily tweaked by strictly short term variations. There might have been times when touring SA was difficult, there are times when it was easy. No single team has had consistently great bowlers across a very long time. This is an analysis of player streaks, that is all. Whether they travelled to Bangladesh or South Africa they Tests are part of one such streak.
    Ananth
    ]]

  • on July 27, 2014, 16:45 GMT

    The Test measure is ok for measuring consistency - However for runs accumulated etc number of innings played is more accurate. Bradman played 80 innings in his 52 Tests. How many innings did the other batsmen play in theirs? For runs accumulated, averages etc innings must be used.
    [[
    The RpFI value is based on innings.
    Ananth
    ]]

  • P.N.Sudarshan on July 27, 2014, 11:21 GMT

    Hi Ananth The phenomenon of Imran Khan is amazing but he is never mentioned in the same breath as Sobers or even a Botham. Is it because that his bowling and batting exploits were 2 distinct "careers" and never really overlapped. As always enjoyed your article and the objectivity of your analysis. Perhaps like the magical 52 and 27 analysis, is there one you could do for all-rounders?
    [[
    Can do. Need a magic number like 52/27. The problem is that there have been only around 10 world class all-rounders. So the real analysis can only be done amongst these. Four years back I did an analysis of Test all-rounders. I may have to re-visit that theme but incorporating many new insights recently acquired.
    Ananth
    ]]

  • Bilal_Choudry on July 27, 2014, 10:40 GMT

    must say I am surprised Waqar Younis didnt top the list I remember him picking up wickets at an insane 10 runs per wicket .. infact he was picking up so many 5fors it wasnt special anymore ... e.g. an ODI in Sharjah the MoM was given to Ijaz for an 80 odd run innings ... waqar got 6 for 20 odd ... but then again waqar got such figures in all games

  • Harry31j on July 27, 2014, 7:45 GMT

    One thing that has always annoyed me is that when people hold Lara in greater esteem than Tendulkar or Ponting, they say it's because Lara had to carry his team on his own. But same is true for Tendulkar in 90s. So what I'd like to know, based on the best streak of 52 you've given Anand is how did teammates compare to the great during that streak. Say how many runs did Dravid or Laxman (or whoever next best batsman for the team was) make during Tendulkar's purple patch & vice versa. Same for Hayden & others during Ponting's & vice versa.
    [[
    This question will be effectively answered when I do the Peer analysis. By Peer analysis I mean both within the team and across all teams.
    Ananth
    ]]

  • Worldcricketlover on July 27, 2014, 7:09 GMT

    Well with all due respect to Imran , his most of wickets are on English pitches. Ishant's subcontinent records is not best but his records other then subcontinent is much better . I am not comparing Legends Imran with Ishant. Bradman records seems so great against one opposition but all records after that were not good as Mighty WI pace battery has held everyone. Fr example Sangas test runs in SL is great as SSC any team can score 900 in a test . In My book Steyn and Murli are best because they have ability to win in any condition. Imran has best action for fast bowler as technically Gavaskar has best technique
    [[
    Your first statements are questionable. Imran captured wickets on his home grounds quite frequently. To say that Ishant's record outside the sub-continent are better is such a travesty of facts that it is not worth further look. I think you have been unduly influenced by his last Test heroics too much. Please present supporting figures.
    Ananth
    ]]

  • aman15 on July 27, 2014, 6:48 GMT

    Interesting article. if you see both batting and bowling lists for the respective streaks of 52 and 27 tests respectively most of the usual suspects are there. However there is one name that has truly surprised in the bowling list w.r.t wickets captured in a particular streak of 27 tests and that is richie benaud. Benaud obviously was a fine bowler and a wonderful allround cricketer but still. Wonderful reading, as always.

  • SLSup on July 27, 2014, 6:27 GMT

    Kartikeya Date refers to Waqar having a best 'Barnes Score' based on Barnes as the be-all and end-all of bowlers. That's per the strike rate. I also see @harshthakor refers to strike rate too. POINTLESS. Lillee is a great bowler by ANY standards yet even during days that I idolized Lillee I knew he took an aweful lot of tail-enders! THAT will help your strike rate. Just run through them. : )
    [[
    One reason I am always slightly uncomfortable with strike rates is this factor. I think the Bowling average is, in my opinion, the most balanced and self-contained of all measures. It does not have the weakness of not outs and incorporates the two bowling components, viz., WpB and RpB in equal measure.
    Ananth
    ]]
    My reading on Sydney Barnes informs me he also exerted a lot of reverse swing at the turn of the 20th Century. I wonder how.

  • harshthakor on July 27, 2014, 5:04 GMT

    Ananth maybe you could make an equivalent list for great batsmen who did not play 52 tests like Everton Weekes,George Headley and Greame Pollock.A little strange that Walter Hammond does not figure in the top bracket who almost averaged 60.
    [[
    Hammond's RpFI is a very creditable 56.29 which puts him just out of the top dozen.
    Ananth
    ]]
    Today's batsmen are at an advantage with easier batting conditions and weaker bowling attacks.Many more tests are played which had advantages and disadvantages.We must give credit to modern greats like Kallis who are so prolific when so much and so many diverse forms of cricket are played.Infact in many ways it is unfair to compare the batsmen of the 1970's with those of post 1990. In the final word I take nothing against Imran,arguably the best cricketer after Bradman and Sobers with Viv Richards,Jack Hobbs ,Shane Warne or Sachin Tendulka and statistically at his peak arguably the best of all paceman.However it is also about the impact ,artistry and entertainment and not mere statistics.At his peak Ian Botham was close to the very top.

  • harshthakor on July 27, 2014, 4:51 GMT

    To me Ananth as far as batting is considered Bradman is on another street while the best runner-up is morally Brian Lara.Imagine the brunt of the batting of a weak team that Lara had to carry repeatedly .Lara had the best average percentage score of the team's total since George Headley and could turn games or champion a crisis more than even Tendulkar at his peak.Ponting had the advantage of playing for a great side.Sadly the careers of Viv Richards and Greg Chappell were curtailed by World Series Packer Cricket.Adding the figures of those games both these stars would have come close to the top.No batsmen changed the complexion of a game more than Viv Richards and morally from 1976 -81 he was the best after Bradman. Great credit to Graham Gooch for edging even Gavaskar proving he was a truly great player.Dilip Vengsarkar from 1986-88 averaged more than 86 and also outscored Gavaskar in his peak years from 1978-80.Sangakaara and Jayewardene favoured by playing weaker teams.
    [[
    Gooch started with a pair and improved considerably. The 456 runs certainly helped.
    Ananth
    ]]

  • harshthakor on July 27, 2014, 4:35 GMT

    Above all Imran was outstanding against the best team of his day,the West Indies with 80 scalps which is remarkable.Arguably he was the best cricketer of the 1980's with his great leadership qualities..However Ananth it is also possible that had he played in the years that he missed as a pace bowler from 1984 his strike rate and overall average may not have bettered.His feats were herculean but Imran was not a great bowler and great batsmen in the same period of his career like Sobers and Botham. Other factors came into the game from the 1990's like restriction on bouncers,more placid tracks,neutral umpiring etc and the game became more loaded in favour of the batsmen.In Pakistan several lbw decisons were wrongly given which went in favour of pakistani bowlers.Another factor is that now many more tests are played in addition to one day and t-20 format.. Fascinating that Glen Mcgrath ,arguably the best of all pacemen does not make the list which Ambrose,Wasim or Waqar do.
    [[
    McGrath captured 143 wickets and the best streak average was 20.5. Both values are just off the featured bowlers list. That is all.
    Ananth
    ]]

  • imran786c on July 26, 2014, 23:31 GMT

    I am always a big fan of Imran Khan...the only one who came close to Sobers, but in my humble opinion, even better given his leadership. As many have commented, one of the few who took the responsibilities of the captaincy and reached heights few could reach. My minor query is... how many of the 27 tests, at his peak, did IK actually bowled and what was his bowling average then and also WpT.
    [[
    During his peak of 27 Tests, Imran did not bowl in two of the Tests. His numbers are there in the tables.
    Ananth
    ]]
    Another of his great legacy was the set up of the Pakistani fast bowling factory....since he nurtured the 2 Ws, Pakistan have never had a shortage of quality fast bowlers.

  • Deuce03 on July 26, 2014, 23:01 GMT

    Very surprised to see Mike Gatting mentioned as one of the more consistent batsmen, given that he has a reputation for being one of the streakiest batsmen of all time. I suppose the perceived streaks in his career were shorter than the 52-Test cut-off. The same could be said about Ian Bell, perhaps.
    [[
    He is not portrayed as amongst the "most consistent batsmen. It was just an idea to compare the best and worst streaks and see how much of a variance there was. In his 79-Test career Gatting's best streak produced 4409 runs and his worst streak, 3445 runs. Thus there was only a 14% differential. That is all. Incidentally there is a partial overlap between these two streaks. In fact in my earlier Consistency analysis Gatting does not fare that well.
    Ananth
    ]]

  • Nuxxy on July 26, 2014, 22:38 GMT

    @Anath: Another winner article! Something else for your todo list... Ever thought about doing a comparison of players vs the averages (or other metric of choice) for the countries they have played in? For example, did anyone outdo King Kallis in South Africa (maybe Ponting or Lara)? Who is the real King of England?
    [[
    Comment not very clear. Pl explai what you are looking for.
    Ananth
    ]]

  • on July 26, 2014, 19:29 GMT

    Oh Imran, this guy has so many qualities that every now and then we are given a reminder " this was left to be discussed " :)

  • VISH.R on July 26, 2014, 17:48 GMT

    It would be interesting to see this same data presented in terms of all types of cricket in aggregate not just tests. Limiting to tests takes away from evaluating the total impact and standing as a cricketer and therefore not a complete comparison.
    [[
    Even if I do this analysis for ODIs, I will never mix up these two totally different forms of the game. I leave it the TV channels to come out with gems like 30000th run in all forms of cricket or 100th century in all forms of cricket. I think there are certain conventions to be followed always.
    Ananth
    ]]

  • its.rachit on July 26, 2014, 16:56 GMT

    I think Sindbad No has brought a very valid point without quite realizing the significance of the period ... that period was immediately after the retirement/post peak of Donald, Walsh,Ambrose,Wasim, Waqar, Gough,Pollock (past his prime by 2003) .., there was only Mcgrath during that period who is a true great fast bowler ... no wonder so many batsmen who were good became suddenly great ... and this is precisely the reason why experts rate Sachin and Lara above the rest of the modern greats ... the best examples are Dravid/Ponting/Hayden ...
    [[
    Yes, it might be true. There was a long gap until Steyn arrived.
    Ananth
    ]]

  • abhijitsakhare on July 26, 2014, 16:41 GMT

    i had read your previous article that was published few years back. someone had suggested to do similar exercise for first 52 test matches of career. hopefully you will do it some other day..........
    [[
    We could really select the 52 in any which way. First 52, closing 52, best 52 and so on. I feel that the bext consecutive treak of 52/27 comes closest to what I am trying to do: which is to put Bradman's career in perspective.
    Ananth
    ]]

  • ruester on July 26, 2014, 16:22 GMT

    Surely we will have comments, hailing Tendulker as a god and claiming him as the best ever from Indian fans. We always do! I'm England fan but know Bradman was by far the best batsman by far. Of all the players I have watched, for me Lara was the greatest of his generation.
    [[
    On the evidence of the comments so far, the readers have also matured. They are looking beyond one player.
    Ananth
    ]]

  • rajagopalann on July 26, 2014, 12:24 GMT

    You have to see the inconsistency in ponting's career.I feel without doubt after Bradman, its Sobers without doubt the best batsman

  • smalishah84 on July 26, 2014, 11:54 GMT

    Imran Khan. What a legend. It is a pity that his bowling is so often forgotten due to his leadership qualities in the latter half of his career.
    [[
    What would Imran have achieved if he had not lost those Tests in which he played only as a batsman.
    Ananth
    ]]

  • on July 26, 2014, 10:36 GMT

    The "peer ratio" of best 52 match streak would also give us a look into why Sobers is considered better than Kallis ( at least as a batsman). Sobers dominance over the field during his best 52 match streak from 1957-1968 seems to be greater than Kallis'. Unlike Sobers, Kallis seems to have scored his runs when everyone else was scoring heavily too.

  • Nadeem1976 on July 26, 2014, 10:03 GMT

    Great respect for Imran Khan now in my Eyes. I knew that Marshal and Younas were two of the most lethal bowlers at their Peak but Imran Khan was better than even every one is Eye opening stat. Respect for the legend. No surprise that Ponting was at #2 because he really played like Bradman during the period between 2002-2006. Great to see my beloved Superstars Waqar Younas and Muhammad Yousaf among the top performers in Test cricket. Can we get a List like this in ODI cricket, i want to see where Sachin SRT200 , Viv, MSD and Wasim Akram stands in ODI cricket at their peaks.
    [[
    Yes, it can be done. Although the two reference points are not that clearly visible in ODIs.
    Ananth
    ]]

  • on July 26, 2014, 9:53 GMT

    I also wonder what results you woiuld get if you extrude the best 52 match streak. What do the remainder of their careers look like ? i.e How much effect does this 52 match streak have on the overall career stats?
    [[
    The full table can be downloaded and imported into an Excel sheet. Then the world is at your feet. But I am not a great fan of this since the alternate career of players would then be split into two non-connecting parts.
    Ananth
    ]]

  • on July 26, 2014, 9:48 GMT

    In runs accumulated table the pattern is even more noticeable - Ponting,Lara,Sangakarra ,Kallis, Yousuf, Jayawardene, Hayden - i.e 7 out of the Top 11 had their best streaks in the early to mid 2000s.

    Also as can be seen most modern day batsmen had their 52 match streak in a 5-6 year week window as compared to Sobers 11 years or even Richards 8 yrs, Clearly the modern day batsmen were beneficiaries of having a large number of matches played when batting conditions were good and also when they were in form.

  • on July 26, 2014, 9:39 GMT

    The one pattern which distinctly hits you is that in the batting table Ponting, Lara, Kallis,Sangakarra,Yusuf, Hayden and Dravid i.e 7 out of the Top 13 had their best periods in the early to mid 2000s. Given that several of these batsmen had long careers extending either side of this period - this seems to be not just correlation.
    [[
    Not to forget that Tendulkar's peak straddled 2000.
    Ananth
    ]]

  • on July 26, 2014, 8:11 GMT

    I published measures along these lines last year. I used a 70 dismissal measure for Bradman and a 27 Test measure for Barnes. If you account for strike rates as well, Waqar Younis is the bowler with the best Barnes Score. Kallis, Sobers and Ponting were the three batsmen to score more than 5000 runs for 70 dismissals. This was in 2013. Bradman Standard: http://www.espncricinfo.com/blogs/content/story/668351.html Barnes Standard: http://www.espncricinfo.com/blogs/content/story/671745.html
    [[
    Thank you for the links, Kartikeya. I had not seen your articles and my work about 4 years back was the basis. With the idea of the "fulfilled innings", tweaked by the Milind adjustment, I have been able to take the 52 Tests itself as the basis. Bowling presents no such problems.
    Ananth
    ]]

  • on July 26, 2014, 8:11 GMT

    I published measures along these lines last year. I used a 70 dismissal measure for Bradman and a 27 Test measure for Barnes. If you account for strike rates as well, Waqar Younis is the bowler with the best Barnes Score. Kallis, Sobers and Ponting were the three batsmen to score more than 5000 runs for 70 dismissals. This was in 2013. Bradman Standard: http://www.espncricinfo.com/blogs/content/story/668351.html Barnes Standard: http://www.espncricinfo.com/blogs/content/story/671745.html
    [[
    Thank you for the links, Kartikeya. I had not seen your articles and my work about 4 years back was the basis. With the idea of the "fulfilled innings", tweaked by the Milind adjustment, I have been able to take the 52 Tests itself as the basis. Bowling presents no such problems.
    Ananth
    ]]

  • on July 26, 2014, 9:39 GMT

    The one pattern which distinctly hits you is that in the batting table Ponting, Lara, Kallis,Sangakarra,Yusuf, Hayden and Dravid i.e 7 out of the Top 13 had their best periods in the early to mid 2000s. Given that several of these batsmen had long careers extending either side of this period - this seems to be not just correlation.
    [[
    Not to forget that Tendulkar's peak straddled 2000.
    Ananth
    ]]

  • on July 26, 2014, 9:48 GMT

    In runs accumulated table the pattern is even more noticeable - Ponting,Lara,Sangakarra ,Kallis, Yousuf, Jayawardene, Hayden - i.e 7 out of the Top 11 had their best streaks in the early to mid 2000s.

    Also as can be seen most modern day batsmen had their 52 match streak in a 5-6 year week window as compared to Sobers 11 years or even Richards 8 yrs, Clearly the modern day batsmen were beneficiaries of having a large number of matches played when batting conditions were good and also when they were in form.

  • on July 26, 2014, 9:53 GMT

    I also wonder what results you woiuld get if you extrude the best 52 match streak. What do the remainder of their careers look like ? i.e How much effect does this 52 match streak have on the overall career stats?
    [[
    The full table can be downloaded and imported into an Excel sheet. Then the world is at your feet. But I am not a great fan of this since the alternate career of players would then be split into two non-connecting parts.
    Ananth
    ]]

  • Nadeem1976 on July 26, 2014, 10:03 GMT

    Great respect for Imran Khan now in my Eyes. I knew that Marshal and Younas were two of the most lethal bowlers at their Peak but Imran Khan was better than even every one is Eye opening stat. Respect for the legend. No surprise that Ponting was at #2 because he really played like Bradman during the period between 2002-2006. Great to see my beloved Superstars Waqar Younas and Muhammad Yousaf among the top performers in Test cricket. Can we get a List like this in ODI cricket, i want to see where Sachin SRT200 , Viv, MSD and Wasim Akram stands in ODI cricket at their peaks.
    [[
    Yes, it can be done. Although the two reference points are not that clearly visible in ODIs.
    Ananth
    ]]

  • on July 26, 2014, 10:36 GMT

    The "peer ratio" of best 52 match streak would also give us a look into why Sobers is considered better than Kallis ( at least as a batsman). Sobers dominance over the field during his best 52 match streak from 1957-1968 seems to be greater than Kallis'. Unlike Sobers, Kallis seems to have scored his runs when everyone else was scoring heavily too.

  • smalishah84 on July 26, 2014, 11:54 GMT

    Imran Khan. What a legend. It is a pity that his bowling is so often forgotten due to his leadership qualities in the latter half of his career.
    [[
    What would Imran have achieved if he had not lost those Tests in which he played only as a batsman.
    Ananth
    ]]

  • rajagopalann on July 26, 2014, 12:24 GMT

    You have to see the inconsistency in ponting's career.I feel without doubt after Bradman, its Sobers without doubt the best batsman

  • ruester on July 26, 2014, 16:22 GMT

    Surely we will have comments, hailing Tendulker as a god and claiming him as the best ever from Indian fans. We always do! I'm England fan but know Bradman was by far the best batsman by far. Of all the players I have watched, for me Lara was the greatest of his generation.
    [[
    On the evidence of the comments so far, the readers have also matured. They are looking beyond one player.
    Ananth
    ]]

  • abhijitsakhare on July 26, 2014, 16:41 GMT

    i had read your previous article that was published few years back. someone had suggested to do similar exercise for first 52 test matches of career. hopefully you will do it some other day..........
    [[
    We could really select the 52 in any which way. First 52, closing 52, best 52 and so on. I feel that the bext consecutive treak of 52/27 comes closest to what I am trying to do: which is to put Bradman's career in perspective.
    Ananth
    ]]