A look at the top allrounders in ODIs August 17, 2013

Ryan ten Doeschate on the top!

A detailed look at the top allrounders in ODI cricket based on their batting and bowling skills
43

Ryan ten Doeschate has an impressive record as an allrounder in ODIs. © Getty Images

I started looking at the work I had done over the years and found out that I had never really done an article on ODI allrounders. Test allrounders, plenty, but almost nothing on ODI allrounders. So I decided to complete this gap in my bio-data.

There is no doubt that ODI allrounders analysis is a specialised area requiring a deeper level of analysis than the Test exercise. Wickets are paramount in Tests and a 6 for 75 is invariably considered to be better than 5 for 50. However in ODIs with the bowlers bowling for a limited number of overs, an analysis of 2 for 20 is almost always more effective than 3 for 40 or 4 for 60. Also a 50-ball 30 is more likely to win a match than 75 in 90 balls. These are true in most situations. I agree that defending a low total or chasing a middling score would promote the importance of wickets and runs. Hence it is essential to give due weight to all four facets of batting and bowling.

As normally happens, the first step is to draw up a minimum cut-off point. With lofty ideas I started with 100 wickets and 2500 runs to find that not many qualified. I progressively lowered the bar and finally decided on 50 wickets and 1500 runs. These numbers represent an average playing career of 40-50 matches and would allow the players from the 1980s and associate countries to be considered. There were 65 qualifying players, which is a fair number.

How does one measure the all-round ability of an ODI allrounder? I decided on the following measures.

1. A composite career-level ratings of the four measures: Batting RpI, Scoring rate, Bowling Strike rate and RpO.
3. The trusted AllRounder (A-R) ratio of Batting RpI / Bowling average.
2. A run-equalised delivery per match measure.
4. Run-equalised values for career. This is strictly a longevity measure and is meant to recognise the allrounders who played a huge number of matches.

I will also show a final table classifying allrounders into bowler-centric, balanced and batsman-centric groups.

Let us now move on to the tables.

Allrounder Rating Index
NameCtryODIsRpISc/RSt/RRpORpI-IdxSc/R-IdxSt/R-IdxRpO-IdxA/R-Index
RN ten DoeschateNed 3348.15 87.728.725.0424.0820.1026.1114.8885.17
A Flintoff Eng14127.81 88.833.274.4013.9020.3622.5417.0673.86
SR Watson Aus16034.25 88.536.344.7817.1220.2820.6415.6873.72
IVA Richards Win18740.24 90.247.834.4920.1220.6715.6816.6973.16
V Sehwag Ind25133.76104.345.755.2616.8823.9116.3914.2571.43
DS Lehmann Aus11730.47 81.334.484.8415.2318.6421.7515.5071.13
KJ O'Brien Ire 7228.89 83.334.134.9114.4419.1021.9715.2870.80
Shahid Afridi Pak35921.99114.643.864.6110.9926.2717.1016.2670.63
JH Kallis Saf32137.45 73.039.394.8318.7316.7219.0415.5470.03
GS Chappell Aus 7432.37 75.743.164.0516.1817.3517.3818.5369.44
CH Gayle Win25335.11 84.244.774.7417.5619.3116.7515.8269.44
L Klusener Saf17126.10 89.938.194.7013.0520.6119.6415.9469.24
SR Tendulkar Ind46340.76 86.252.335.1020.3819.7614.3314.7069.17
A Symonds Aus19831.60 92.444.625.0115.8021.1816.8114.9768.76
ME Waugh Aus24436.01 76.943.374.7818.0017.6217.2915.6968.61
Shakib Al Hasan Bng12929.74 78.240.854.3114.8717.9218.3617.3968.54
N Kapil Dev Ind22519.10 95.144.273.72 9.5521.7916.9420.1768.45
ST Jayasuriya Slk44431.01 91.246.034.7915.5120.9016.2915.6668.37
RJ Hadlee Nzl11517.86 75.539.123.31 8.9317.3019.1722.6968.09
SM Pollock Saf30317.16 86.739.973.68 8.5819.8718.7620.3967.60

The first is the allrounder ratings table. I decided that the batting and bowling functions would get 50% weight each. I could not separate the importance of the two bowling measures. Hence those two get 25% each. However I realised that the Batting scoring rate should be valued marginally higher than the Batting RpI. Most followers would agree with this. The quantum of differential: a simple 10%, leading to an actual differential of just below 20%. So this was 27.5% and 22.5% for the two measures. Arbitrary: Of course, yes. I have no problems with this allocation. If any reader wants a different weight, the Excel sheet contains all measures and the readers can work out their own alternate tables.

It is very difficult to get a high A-R Index. The concerned player would have to be within the top-80% level in each of the four measures. That is what Ryan ten Doeschate has achieved. His numbers are, to say the least, Bradmanesque! In 33 matches, he has scored 1541 runs at an average of 67.00 and captured 55 wickets at an average of 24.13. Individually, these are figures which would put him at the top of the batting tables and the top-15 of the bowling tables. To the sceptics, let me say that he has played eight matches against Test-playing countries and 25 against very tough teams such as Ireland, Kenya and Afghanistan et al. He plays for an average team and does not get as much support as the other big-team players. If he has not played more matches, then that is the problem playing for an associate country. He does not seem to have played in the recent Netherlands-matches. Probably was trying to earn a living. But let us give credit to the best allrounder the world has ever seen: Ryan ten Doeschate of Netherlands.

ten Doeschate is a clear ten points ahead of Andrew Flintoff who has 73.32 points. He has been excellent: 3394 runs at 27.81 and 169 wickets at 24.83 indicate that he has every right to be called the best allrounder ever, only the freakish short-career numbers of ten Doeschate are better. I get the feeling that Flintoff's ODI skills are rather underrated and overshadowed by his Test exploits. Next in the list is Shane Watson. May not be everybody's favourite with his Test woes now but one of the greatest ODI allrounders ever. 4795 runs at 34.25 and 156 wickets at 28.98 illustrate that he is only marginally behind Flintoff. And both these wonderful allrounders have strike rates around 88+.

These are followed by two explosive batsmen but no mean bowlers. Viv Richards must be the most underrated ODI bowler ever. Coming as the fifth bowler, he plundered 118 wickets, albeit at a reasonably high average of 35.83, to go with his 1980s compilation of 6721 runs at 47.83 (and a strike rate of 90.2). A surprise at No.5 is Virender Sehwag. A very much underrated bowler like Richards, he still managed to capture 96 wickets (albeit at 40.14) to go with his 8273 runs at a strike rate exceeding 100. Shahid Afridi is in the eighth place confirming his status as one of the top allrounders ever. And nice to see Kevin O'Brien in the top ten.

Ratio of Batting RpI and Bowling average
NameCtryODIsRpIBow-AvgeA-R Ratio
RN ten DoeschateNed 3348.1524.131.9957
SR Watson Aus16034.2528.981.1818
JH Kallis Saf32137.4531.701.1815
A Flintoff Eng14127.8124.381.1405
IVA Richards Win18740.2435.831.1231
GS Chappell Aus 7432.3729.121.1114
DS Lehmann Aus11730.4727.811.0957
ME Waugh Aus24436.0134.561.0418
KJ O'Brien Ire 7228.8927.931.0343
Shakib Al Hasan Bng12929.7429.371.0125
CH Gayle Win25335.1135.380.9925
SC Ganguly Ind31137.8738.510.9834
MJ Clarke Aus22735.6237.540.9490
Imran Khan Pak17524.5626.620.9226
SR Tendulkar Ind46340.7644.510.9157
WJ Cronje Saf18831.8034.790.9141
AR Border Aus27325.8828.370.9122
L Klusener Saf17126.1029.950.8714
NJ Astle Nzl22332.6738.460.8494

This is a simpler measure. I divide the Batting RpI by Bowling average to get an Index value. This is my favourite measure and tells a lot about a player's all-round abilities. The ratio is superior to the difference. A player with values of 30 and 25, will have a difference of 5 and a ratio of 1.20. A player with values of 40 and 35 will have the same difference of 5 but a lower ratio of 1.14. It is obvious that the first player is a better allrounder since the lowering of 10 in bowling average is more valuable than the dropping of 10 in batting.

Since this is a performance oriented measure there are some juxtaposition of players at the top from the previous table. Ten Doeschate is at the top with an index value of nearly 2.0, more Bradmanesque than the previous table, since the next value is Watson's 1.18. Jacques Kallis, Flintoff and Richards complete the top-five positions with index values just over 1.12. Watson and Kallis are separated only at the fourth decimal.

Average Run-equalised delivery per match
NameCtryODIsRunsWicketsEqualised valuePer Match
RN ten DoeschateNed 33 1541 55 3301100.0
Shakib Al Hasan Bng129 3688161 8840 68.5
JH Kallis Saf3211149827020138 62.7
GS Chappell Aus 74 2331 72 4635 62.6
A Flintoff Eng141 3394169 8802 62.4
SR Watson Aus160 4795156 9787 61.2
RJ Hadlee Nzl115 1751158 6807 59.2
IK Pathan Ind120 1544173 7080 59.0
IT Botham Eng116 2113145 6753 58.2
L Klusener Saf171 3576192 9720 56.8
IVA Richards Win187 672111810497 56.1
HH Streak Zim189 294323910591 56.0
Wasim Akram Pak356 371750219781 55.6
DJ Bravo Win146 2495173 8031 55.0
Imran Khan Pak175 3709182 9533 54.5
CH Gayle Win253 874315713767 54.4
ST Jayasuriya Slk4441343032323766 53.5
SM Pollock Saf303 351939316095 53.1
CL Cairns Nzl215 495020111382 52.9
M Prabhakar Ind130 1858157 6882 52.9

This is a measure of the actual performance delivered per match. For this, I valued a bowler wicket at 32 runs. Arbitrary? No way. In the 3403 matches played until 3 August 2013, 42832 wickets were captured by bowlers at a total cost of 1372875 runs. This works to 32.05 RpW. I rounded this to 32. So no one can say this is arbitrary. Of course it has changed over the years. However let me give the readers the following summary.

 
	1971-1984 281 3469 99425 28.66 
	1985-2013 3122 39363 1273450 32.35 
	

So this confirms that for 29 years and for 92% of matches, the average wicket value has been slightly above 32. Just to firm up this number, let me say that, for the period 1985-1989, the average was 32.09.

Ten Doeschate has delivered an equivalent of 100 runs per match. I feel this is one of the most significant figures in the ODI game. A player from an Associate country walked in and delivered the equivalent of a hundred in every match he played. I checked his match performances. He has failed to deliver in either batting or bowling in just two matches out of 33. That means he has delivered in 94% of the matches. I am glad to see the equally talented Shakib Al Hasan in the second position, delivering the equivalent of 68 runs per match. Imagine what these two would have achieved with better support. Then comes the giant, Kallis, with 62.7 runs. Greg Chappell is placed next just behind and then come Flintoff and Watson. Imagine these players guaranteed 60+ runs each time they took the field.

Career Run-equalised delivery
NameCtryODIsRunsWicketsEqualised value
ST Jayasuriya Slk4441343032323766
SR Tendulkar Ind4631842615423354
JH Kallis Saf3211149827020138
Wasim Akram Pak356 371750219781
Shahid Afridi Pak359 730335818759
SM Pollock Saf303 351939316095
WPUJC Vaas Slk324 202540014825
SC Ganguly Ind3111136310014563
SR Waugh Aus325 756919513809
CH Gayle Win253 874315713767
Abdul Razzaq Pak265 508026913688
PA de Silva Slk308 928410612676
CL Hooper Win227 576119311937
N Kapil Dev Ind225 378325311879
Yuvraj Singh Ind282 821111211795
CL Cairns Nzl215 495020111382
V Sehwag Ind251 8273 9611345
ME Waugh Aus244 8500 8511220
DL Vettori Nzl275 211028411198
CZ Harris Nzl250 437920310875

Now the final table. This is the same table as previous one but ordered on the total match deliveries, in terms of run-equivalence. Sanath Jayasuriya stands on top with 23766 "runs". When we talk of Jayasuriya the destroyer we forget the fact that he also captured 323 wickets. A very canny bowler who averaged 0.75 wickets per match. Sachin Tendulkar, again the 18000+ runs over-shadowing the fact that he also captured over 150 wickets, is close behind. He has 23354 "runs". Then comes Kallis, with 20138 "runs". Now we get two Pakistani giants, Wasim Akram, with 19781 "runs" and Shahid Afridi, with 18759 "runs". This is a tribute to the longevity of these outstanding allrounders.

Bowling-centric, Balanced and Batting-centric
NameCtryODIsRpI-IdxSc/R-IdxBatting IndexSt/R-IdxRpO-IdxBowling IndexRatio
WPUJC Vaas Slk324 4.6016.6221.219.0217.9036.90.57
RJ Hadlee Nzl115 8.9317.3026.219.1722.6941.90.63
M Prabhakar Ind130 9.4813.8523.318.5217.5436.10.65
Wasim Akram Pak356 6.6420.2426.920.7119.2540.00.67
Mudassar Nazar Pak12211.5312.0023.517.1517.6834.80.68
...
CL Hooper Win22713.9817.5631.515.1217.2032.30.98
SB Styris Nzl18813.9218.2032.116.7915.8132.60.99
Shoaib Malik Pak21614.2217.9432.215.6216.5032.11.00
GW Flower Zim22115.3515.4930.814.2916.1630.41.01
WJ Cronje Saf18815.9017.5333.415.9616.8832.81.02
...
A Symonds Aus19815.8021.1837.016.8114.9731.81.16
TM Dilshan Slk26715.7919.7435.512.9215.9728.91.23
IVA Richards Win18720.1220.6740.815.6816.6932.41.26
V Sehwag Ind25116.8823.9140.816.3914.2530.61.33
SR Tendulkar Ind46320.3819.7640.114.3314.7029.01.38

This is a quirky report. I have divided the batting rating total points by the bowling rating total points. This ratio reveals the nature of the player's all-round abilities: ranging from bowlers who could bat through genuine allrounders to the batsmen who could bowl. The first five belong to the bowling allrounders classification. One surprise there! Chaminda Vaas, Richard Hadlee, Manoj Prabhakar and Wasim Akram were top bowlers first and their batting skills were only add-ons. However Mudassar Nazar is a surprise. My conclusion is that, for a batsman, his batting numbers are quite average.

The middle five, either side of a ratio 0f 1.00, are here only because of the way the numbers fell. Shoaib Malik is the only allrounder whose batting total matches the bowling total exactly. The others are all good allrounders.

The third group is quite clear. All five were far superior batsmen than bowlers. Jayasuriya does not appear here because his bowling numbers are quite impressive.

Finally, who do I salute? I will doff my imaginary hat at ten Doeschate, Flintoff, Watson, Jayasuriya and Afridi, in that order. I will end with an additional salute to ten Doeschate.

I have created an Excel sheet with complete details for these 65 players. To download/view the document, please CLICK HERE.

One final comment on ten Doeschate. He can only play the cards dealt to him. Since he made his debut and got a permanent place, Netherlands have played 50-odd matches. As I have mentioned before, he missed playing the last few matches for his country. So 33 matches in six years is the limit for a player like him. Anyhow if any reader does not want to consider Tendo, he only has to take off the top entry in the relevant tables.

To round off the article I have given below what I feel, based on a number of relevant objective measures, the five greatest all-round performances ever in ODI matches.

1. Shahid Afridi's 76 and 7 for 12 against West Indies during 2013. That an all-round performance contains the second best bowling spell ever speaks volume of this superb effort. And not to forget that the 76, made out of a middling total of 224 for 9, and a recovery from 47 for 5, was itself a truly match-winning effort.
2. Scott Styris' 63 and 6 for 25 against West Indies during 2002. This was a low scoring and close match. New Zealand scored 212, mainly due to Styris lovely innings late in the order. Then he captured 6 for 25 to restrict West Indies to 202 for 9. I would say this performance is only bettered by Afridi's recent effort.
3. Richards' 119 and 5 for 41 against New Zealand during 1987. Until Paul Collingwood's effort this remained a unique effort. Richards' 119 comprised of more than 50% of his team total. The next highest score was 48. Four of his five wickets were of top order batsmen.
4. Paul Collingwood's 112 and 6 for 31 against Bangladesh during 2005. Granted that this was against Bangladesh but let us not forget that no one has scored a century and captured 6 wickets in a match.
5. Tendulkar's 141 and 4 for 38 against Australia during 1998. This was a big innings and the four late order wickets arrested Australia push towards a win. This was also the quarter-final of the Wills International Cup.

My next article will again be the first of a kind. I will apply the ball-by-ball data to a single event, in this case the Ashes series, and come out with the highlights of head-to-head confrontations.

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

  • dummy4fb on August 26, 2013, 6:22 GMT

    Ananth - Interesting article - and an eye opener as well. We remember a couple of years back you made an article on all-rounders (in tests) and we realized that out of 80+ tests, an all-rounder gets to display both his facets in the same match very rarely (less than 5%) and think of it - on paper, 80 tests = 160 ODIs (opportunity-wise). So guess some messages identifying true all-rounders rather than pretenders are sort of short-sighted. Sachin did at least once what Kapil couldn't do even once (100+4w). From an ODI perspective, balance is tilted towards Batsmen who could bowl (Jayasuriya became a true batsman after 60-70 games). Wasim Akram or Ian Botham could never crack the ODI all-rounder leagues.

    An aside, though this doesn't make up the number qualification, deSilva's 107 + 3 wkts surely should be an all-round great, coming in a crunch match (summit clash). Unlike the Debils who had nothing to lose in '83, Lions had a lot to prove in the '96 WC.
    [[
    Once context comes in, De Silva's effort would move up quite rapidly.
    Ananth
    : ]]

  • Charith99 on August 23, 2013, 14:58 GMT

    as always very nice article ananth. regarding tendoeschates selection as a top all rounder i completely agree with ananth, His stats are really impressive. To all my fellow readers who are questioning his stats what do you guys suggest he should do , he certainly can't say i will score less and get less wickets because its Kenya or Afganisthan.He can't pick and choose whom to play against.If his opposition was always easy then a lot of other dutch players would have being on this list too. Personally my vote goes to sanath. over 13000 runs and over 300 wickets plus during his younger years he was a brilliant fielder plus an excellent captaincy record my only wish is that he should have retired earlier and stayed out of politics.
    [[
    Jayasuriya's bowling stats are extremely impressive. He could bowl 10 overs in every match he played. He bowlred just over 5.5 overs par match.
    Ananth
    : ]]

  • alarky on August 22, 2013, 16:16 GMT

    Ananth, Congrats again on an outstanding piece of work! And I'm convinced that all your work is "NAME BLIND"! You do the research and the chips fall where they may! You've proven that Ryan ten doeschates' career so far is "a Lilliputian in the land of giants" story! Hence, we need to be more appreciative of him, knowing that his status is solely determined by genuine statistical facts, and not bias media hypes and exaggerations. Ryan doesn't have those guys with the megaphones in the commentary boxes, or those with the golden type writers in the media houses. Hence, only work like yours could show case him to the world! So, to hell with those who are disappointed that they didn't see the name that they want at the top, and asking for weighting system to meet their desires! Some great players owe part of their glorification to performances against the weakest teams: Eg SRT made 5 of his so called intern'l (100s) vs Namibia and Kenya! Would they like a weighting system that deny him any?

  • dummy4fb on August 20, 2013, 13:09 GMT

    I should also add that even when a team is all out they still have scored some runs on average, 182.8592 (1st innings where match is completed) so value of 1 wicket is equal to (Total runs scored in 1st innings - 182.8592*number of innings)/total wickets not taken = (753480- 182.8592*3288)/7408 = 20.55. I think I like this number better than my previous 13 as that was too weighted towards the relatively rare occasions where a team lost very few wickets (the average is 7.7)

  • dummy4fb on August 20, 2013, 12:41 GMT

    Ananth, the 13 from my earlier post is that for each additional wicket lost a team scores approximately 13 less runs. The gap is largest between losing 9 wickets and being all out as you would expect as not all balls available have been faced. I don't have the figures to hand now but the average 1st innings scores (where a game has a definitive result can be found using data from statsguru). The price of a wicket is often underestimated in limited overs cricket despite the fact a team rarely wins if they are all out. An all rounder also only helps his team if he is bowling or batting. Hence Flintoff coming in at 6 or 7 means he is less influential with the bat than Kallis or Watson and the likes of Lehman, Waugh, Sehwag are very uninfluential with the ball as they bowl very few balls per match. This is where your ratio of batting RPI and bowling average is fairly good but simplified for batting contribution to a match but is blind to bowling contribution to a match.
    ON ROAD. HENCE UNABLE TO RESPOND
    ANANTH

  • dummy4fb on August 20, 2013, 9:01 GMT

    @Michael Counsell your formula for bowling value seems well thought out. if i am correct 13 is the amount by which the eventual total score is reduced when a wicket falls. a wicket taken in the final over is of no value while at the start it is of high value. even if u pick a top batsman in the 40th over the eventual score will only vary by 10-15 so 13 seems good. by valuing balls per match wrt rpo along with wickets u have valued all essentials of bowling. for batting i think it shud be calculated per innings rather than per match, also the batting value seems flowed. when richards and kallis have the same value u know smoething's wrong. richards has higher avg, str rate, rpi and lower not out %. the best bowler has a value of 35 and best batsman 25.5, batting allrounder rchards has bowling value of 12 while bowling allrounders have values of 6 so u have not given equal weightage to bat and bowl.

  • gandabhai on August 20, 2013, 8:29 GMT

    This is all good work you are doing sir but its a bit namby pamby , Please do an analysis to show which countries benefit the most from umpiring decisions . You have a moral duty to conduct a fair study of this .

  • dummy4fb on August 20, 2013, 1:40 GMT

    Anantha Great analysis. One suggestion, though. For calculating delivery per match, you have awarded a weightage of 32 / wicket. But you have disregarded the cost at which the wicket came and the run per over conceded, which is more important in an ODI than in a test. My suggestion is as follows for converting wickets into runs: Bowling contribution converted to runs per wicket =32+ ( 32- (Bowling average ( Runs per wicket ) X Runs conceded per over/6)) This will separate A conceding 24 Runs/Wicket @ 4 Runs/over from B conceding 30 Runs/Wicket @ 4 Runs/over from C, conceding 30 Runs/Wicket @ 5 Runs/over.

  • Engle on August 20, 2013, 1:08 GMT

    Another performance which will stand the test of time and secure it's place in cricket history was Aravinda DaSilva's unbeaten century in the WC final vs Aus, plus his 3 wkts when no other bowler could muster more than 1, and 2 catches to boot. When the pressure is on, the stakes are high and the world is watching, then performances on the highest stage of competition remain indelibly embedded in one's consciousness.

  • MilPand on August 19, 2013, 20:10 GMT

    Not entirely related to this article but there are certain methods that I prefer to avoid.

    1. Measuring a player's contribution based on one-dimensional 'run equivalence' has been done for IPL MVP analysis too - see this rediff article - http://www.rediff.com/cricket/report/ipl-stats-gayle-continues-to-be-the-most-valuable-player-purnendu-maji-srinivas-bhogle/20130522.htm

    There are benefits but I prefer a neutral measure which transforms BOTH batting and bowling figures.

    2. Carefully selected weights are fine but sometimes there are special numbers that boost a player's overall rating. See the rules for IPL fantasy league - http://cricketat.info/the-official-fantasy-iplt20-com-fantasy-league-2013/#.UhJ155L2a8A - There is a milestone bonus for 25 runs. So scoring 24 or 49 vis-a-vis 25 or 50 is considerably worse. Similarly bowling 5 dot balls in an over fetches 1 point for each dot ball whereas the 6th dot ball is worth 15 points.
    [[
    Again, Milind, when I strictly limited myself to Career numbers and career numbers only, all these extremely relevant and priceless measures are out of reach. That would be a different type of analysis which would require even more data such as ball-by-ball data.
    Ananth
    : ]]