Stats Analysis

# Jayasuriya's stunner in Singapore, and other outliers from ODI history

Instances when a player or a team stood out for performing way above, or below, the mean

What is an outlier? A fairly acceptable definition is: "a data point that differs significantly from other observations". However, I am going to use the definition with a little more freedom in this article. I will look for performances that stood out in an ODI, and also performances that stood out across the many years during which the game has been played.
Let it not be forgotten that an outlier can be a positive one (a standout performance that shone among mediocre ones) or a negative one (a forgettable performance in a collection of good ones). I will feature both - often within the same table.
In the table above, a batter's strike rate and runs are contrasted with those of their team-mates. In 1996, in the unlikely location of Singapore, Sanath Jayasuriya scored at a rate of 271 while his colleagues plodded along at 57 - an amazing ratio of nearly 4.8. It was almost like two wildly contrasting games being played at the same time. A likely sequence of deliveries could have been: four dot balls, a single and a four or six. Thirteen years before that, Lance Cairns achieved something similar - a ratio of 4.4 against Australia. Who else but Shahid Afridi next? A ratio of nearly 4.0 in a 2015 match against New Zealand. At Headingley in 1982, Kapil Dev achieved a factor of just over 3.5. Unfortunately, in all four of these cases, the batter's team lost. Since the outlying situation was within the same team, the opposing team name is not provided.
In 2019, chasing a low total, Chris Gayle went berserk, outshining his team-mates by a factor of 3.45. In that instance, he won the match for West Indies. Arjuna Ranatunga's ratio of nearly 3.4 in a 1990 game ended on the losing side. Three other innings are featured with ratios exceeding 3.3; of those, only Sandeep Patil's and Michael Leask's finished on the winning side.
Now for the other end of the spectrum. With a minimum of 25 balls faced by the batters, I look at very low ratios of the batter's strike rate to the team's. Elton Chigumbura and Runako Morton scored no runsin 27 and 31 balls respectively, and finished with a ratio of 0.0. A few other batters scored one run in a fair number of balls and finished with strike-rate ratios of below 0.10. An innings that is not on the table deserves a few lines here. Opening the innings for Pakistan against West Indies in Sialkot in 1986, Rizwan-uz-Zaman scored 4 in 62 balls in an eminently forgettable innings that led to a low score and a loss. Compared to him, his team-mates looked like Shahid Afridis. And this was a strong Pakistan team with Ramiz Raja, Javed Miandad and Imran Khan.
Against Bangladesh in St Kitts in 2009, Andre Fletcher scored 50 of his 52 runs from 11 boundaries. Amazingly, he also played about 30 dot balls: a combination of extreme caution and extreme aggression. Gayle and Brendon McCullum, at Nos. 2 and 3, had identical figures as far as runs and boundaries are concerned, though their overall balls-faced figures were slightly different. Gayle played around ten dot balls and McCullum only seven. Matthew Hayden's innings in Auckland in 2000 was perhaps the most amazing one - 11 boundaries, two singles and 36 dot balls. There are five other instances of batters exceeding 94% in boundaries, including another by McCullum. Can one think of such an analysis without the presence of Afridi? Interesting to note that he played only around eight dot balls in his Colombo innings in 2002 that features on the table.
Now to those batters who appeared to believe that it was a crime to hit more than one boundary an innings. All the players in the bottom half of the table hit a single boundary in their innings, and appear on the table in descending order of their scores. Against New Zealand in 1993, Mark Taylor played 129 balls and hit one boundary. His is the only innings where less than 5% of the runs came off boundaries. Mohsin Khan managed to play 175 boundary-less balls in a World Cup game against West Indies. Not surprising to note that many of these stonewallers' teams lost the games in question.
Disclaimer: There is a problem when we look at the innings without boundaries. There are 59 instances of batters scoring 50 or more runs without hitting any boundaries. I fear that in most of these cases, we do not have all the data needed. Take the 1988 Asia Cup ODI between Bangladesh and Pakistan: Moin-ul-Atiq scored 105 in 117 balls, but his fours and sixes are not listed. In the same match, Ijaz Ahmed is shown to have scored a hundred with nine fours and four sixes. As such, I decided to exclude such innings from consideration. It is possible that there are genuine instances of fifties with zero boundaries, but I cannot confirm that. I have taken the view that if at least one boundary is shown, that scorecard has to be assumed to be correct.
The next set of outliers belongs to a grey area of records: matches where one batter accounted for an inordinately high percentage of the team score. I have also presented the next-highest score and the ratio between the two innings. In order to avoid fast cameos such as Brendon McCullum's 80 (out of 95) dominating this chart, I have put in a condition that at least six wickets should have been captured.
Viv Richards' 189 at Old Trafford in 1984 has remained on top of the best-innings table for the past four decades and I can confidently say that it will never be unseated. Richards scored nearly 70% of his team score in this game. And given the next-highest score in the innings was 26, his support ratio is a huge 7.3. A tribute to, inarguably, the greatest innings of all time. That knock led to a comfortable win for West Indies, but David Warner's magnificent innings, in second place, was in vain. Then comes Kapil Dev's 175 at the 1983 World Cup, often the only innings to be talked about in the same conversation with that of Richards. Kapil's share of India's innings was over 65 and his ratio was fractionally higher than that of Richards.
Tony Ura of Papua New Guinea sounded the bell for the smaller cricketing nations with a magnificent 151 out of a team total 235 in a World Cup Qualifier match against Ireland in 2018. A 25 from the No. 9 was all the support he got. Jaskaran Malhotra's 173 (out of 271) for USA against PNG last year was an even greater innings, coming as it did from No. 5. The next highest score was 22.
Now we move on to the bowlers. We start with the Balls per Wicket (BpW) measure. The comparisons are between the team and an outlier bowler. Andy Bichel had a BpW value of 8.6 in his magnificent 2003 World Cup spell against England, while the other Australia bowlers, headed by Glenn McGrath and Brett Lee, needed 240 balls to take the other wicket that fell. That is a huge ratio of 28. In a totally different situation, Darren Lehmann bowled three balls, claimed two wickets and secured a ratio of 25. In Trinidad in 1989, Kris Srikkanth needed only nine balls per wicket, while his more illustrious bowler colleagues needed 212 balls for the other wicket that fell.
Against Zimbabwe in 2003, Fidel Edwards took six top-order wickets in seven overs while his colleagues needed 25 overs for the other wicket. In 2017, Akila Dananjaya took six India wickets in ten overs while the other bowlers managed all of a single wicket in 206 balls. The three other bowlers on this list had ratios of over 20, and all were casual bowlers. Not featured are the next four bowlers - Mohammad Hasnain, Joshua Holder, Hasan Ali, and Ravi Rampaul - who all had similar figures of five wickets in 60 balls, while their team-mates needed 240 balls for one wicket.
At the other end of the list, in a World Cup game against Namibia, Brett Lee needed 36 balls for one wicket, even as the other bowlers picked up a wicket about every five balls. The ratio was a very low 0.2. Trent Johnston needed 60 balls for his single wicket against Netherlands, while his colleagues took 13.5 for each of theirs. Now come a slew of surprises. Look at some of the other bowlers who needed around four times more deliveries for their one wicket than their team-mates: Imran Khan, Trent Boult (twice), Muthiah Muralidaran, and Wasim Akram.
The next measure of comparison is the Runs per Over (RpO). First, we will look at the bowlers who were stingy as Scrooge in contrast to their team-mates. In September this year against New Zealand, Sean Abbott bowled five overs and conceded one run, while his compatriots, Mitchell Starc and Josh Hazlewood included, conceded 2.89 runs per over. That is a ratio of nearly 15. Steven Finn is the only other bowler with a ratio above 10. He achieved that in 2015 against Ireland, conceding 0.33 runs per over as against the other bowlers conceding 3.67. Courtney Walsh's famous spell of 5 for 1 in 27 balls in 1986 comes in next - a factor of nearly 10.
In the second half of the table, Len Pascoe conceded six and a half runs per over against India in 1981, while his team-mates gave away a mere 1.69. That is a ratio of 0.26. Roger Binny's extravagant spell of eight runs per over against Pakistan in Sharjah was nearly four times that of his fellow bowlers. Danish Kaneria conceded a huge 12 runs per over while his fellow bowlers conceded just over three against Zimbabwe in the 2007 World Cup. The table contains all bowlers whose ratio is 0.3 or below.
The table above is about teams, and it straddles the grey area between being an outlier and being a record. The first is a list of high number of no-balls. These three teams conceded over 10% of the runs they gave away overall through no-balls. All these instances were in the 1990s. With the introduction of free hits as a punishment, no-ball numbers have come down drastically. Two Pakistan teams and one West Indian side, with their collection of aggressive fast bowlers, conceded a high number of no-balls. However, it is worth looking at how many runs these teams conceded - 150 and under in each case. In other words, they were aggressive and fast, and achieved their objective. The Pakistan teams won their matches defending low targets.
Now for wides. Defending 149 in a tri-series final against England in 2000, South Africa went all out, bowling 21 wides (18.9%) but dismissing England for 111. In another tri-series two years earlier, India's bowlers had a bad day, but only as far as wides were considered. They conceded 21 wides, but dismissed Bangladesh for 115. In an ODI in Brisbane in 1996, West Indies conceded 18 wides but dismissed Sri Lanka for 102. A common thread through this list of matches is that all the teams pitted against these seemingly profligate bowlers were dismissed for sub-200 scores, and most matches resulted in wins for the teams that conceded large numbers of wides.
Onwards to a more general topic: the percentage of extras conceded. The Carlton & United Series match in 2000 between Australia and India was surely an outlier on this count. Australia conceded 32 extras out of an India total of 100; most of these extras were leg-byes and wides. The Australian bowlers conceded 68 runs in 36.3 overs and 32 in extras. A match straight out of The Twilight Zone. There are three other matches on this table in which 40 or more extras were conceded - all featuring comfortable Pakistan wins. So it would seem that one way to win ODIs is to go all out, not caring for line and length.
We have already looked at boundary-percentage values for individual batters. Now we look at that metric from the teams' point of view. In a 2015 World Cup game against Scotland, Australia scored over 81% of their total in boundaries, chasing a total of 131 in a flurry of fours and sixes. Four years later, West Indies rattled through a similar chase, against England, mainly in boundaries. The third match followed a similar pattern - New Zealand scored 118 in just over eight overs, Martin Guptill blasting about three runs a ball in his innings of 93. The next game is something else. Batting first against West Indies on New Year's Day in 2014, New Zealand scored 283 in 21 overs and hit no fewer than 22 fours and 22 sixes - 220 runs in boundaries.
Now on to the other end. Sri Lanka hit a single four in their innings of 189 against Pakistan in Sharjah in 1996. That boundary was hit by Roshan Mahanama. In a 1983 World Cup match against India, West Indies scored 228, which was adorned by a single six - and that came during a tenth-wicket partnership - by Joel Garner. Miandad scored the single four of Pakistan's innings of 140 for 9 against West Indies in 1989. Franklyn Rose did the honours for West Indies a few years later. Malcolm Marshall was the six-hitter in West Indies' innings of 182 in India in 1983. Rounding off the table is Pakistan's World Cup semi-final match in 1983, where two fours were hit.
The disclaimer here is the same as the one for Table 2 in this article. The 30 team innings of 100 or more runs that feature zero fours or sixes have been excluded.
Now for match-level numbers. We will look at diametrically opposing performances by teams. First, we look at the RpW values across the two teams for both innings. The second-innings dominant situations are presented first. Only wins with no wickets lost are featured, and I have put down a runs-per-wicket value of the actual score. Also, the RpW ratio is notionally fixed at 10.0 in all these cases. The ten matches in which the chasing teams chased 200-plus totals for the loss of one wicket are not featured.
When Bangladesh scored 278 against South Africa in Kimberley in 2017, they must have felt it was a competitive total. Only to see, 200 minutes later, a scoreboard featuring two batters with hundreds against their name and the score at 282 for 0. Quite similar to the situation India found themselves in three years later against Australia. In the 1992 World Cup, Pakistan laboured to 220 for 2 and West Indies scored 221 without loss, with Brian Lara retiring hurt at 88. In Hamilton in 2009, New Zealand scored 270 and India replied with 201 without loss with Virender Sehwag going on the rampage. India won that rain-interrupted game comfortably.
In Bulawayo in 2018, Pakistan's imposing score of 399 for 1 and the fact that they dismissed Zimbabwe for 155 gave them a huge ratio of over 26. Sri Lanka's huge win over India in Kingston in 2013 fetched them a ratio of nearly 20. New Zealand lost two wickets in scoring over 400 against Ireland, which lowered their ratio to around 18. And so on. Until we come to the last match: a prosaic Australian score of 301 for 6 produced an RpW of only 50, but the demolished of Namibia for 45, giving them a ratio of 12.5.
Onwards to the matches in which the teams were chalk and cheese in terms of Runs per Over (RpO). Bangladesh were clueless in Queenstown in 2007. They batted nearly 38 overs to score 93 at 2.46 and then watched New Zealand score the required runs in six overs at, believe it or not, 15.8 runs an over. That is a ratio of 6.4. Earlier the same year in Cape Town, Pakistan took 45 overs to score 107 and then allowed South Africa to reach 113 in 14 overs. That is a ratio exceeding 3.4. England's two-runs-an-over total of 88 was overhauled by Sri Lanka in a trice.
In the second half of this table, I have listed matches in which the team scoring faster batted first. The first match features Jayasuriya again. His 189 helped Sri Lanka reach 299, and India were then dismissed for 54. One could say that this enormous upset was re-enacted 21 years later when England demolished India in the T20 World Cup semi-final last month. The RpO ratio was nearly 3.0 in that Sharjah game. In 1986, West Indies inflicted similar carnage on Sri Lanka, also in Sharjah: 248 meets 55. In 2012, South Africa scored over 300 and then dismissed Sri Lanka for 43. Both these matches had RpO ratios near three.
For the final table, we will look at the match as a whole. And I will walk the line dividing outlier from record again. These are mostly records. However, one could say that these are also outliers if we consider the history of ODI games.
The first table is one of high RpO values. These are absolute values and not ratios. The first game has already been featured in the high-boundary-percentage table. New Zealand and West Indies combined to score no fewer than 407 runs in 42 overs - for a match RpO of 9.7. That is some scoring rate for New Zealand considering West Indies did not exactly push the envelope in the game. The second match is one of the greatest ODIs ever played. A score of 434 (the first time 400 was breached in an ODI) was chased down with one ball to spare. The total of 872 runs in 599 balls made for an RpO of 8.7. In 2011, India scored 187 in 23 overs and England ran the target down with nearly an over to spare.
The second section of the table is in the order of match RpW values. The 1992 World Cup game between Pakistan and West Indies leads the way: 441 runs for two wickets makes for an RpW value of 220, which makes this match the leader by the proverbial mile. In 2013, Australia's huge total of 359 for 5 was chased with contemptuous ease by India, with seven overs to spare for the loss of a single wicket. The match RpW was 120. Eight other matches had an RpW value greater than 85.
And now, arguably the greatest outlier of all. Who would have thought that an outstanding white-ball player (with sterling red-ball credentials, to boot), who could have been the chief coach with any T20 franchise in the world, maybe even the global head, chose to become the Test coach of England? Yes, I am referring to McCullum - the outlier extraordinaire.
Talking Cricket Group
Any reader who wishes to join the general-purpose cricket ideas exchange group of this name that I started last year can email me a request for inclusion, providing their name, place of residence, and what they do.
Email me your comments and I will respond. This email id is to be used only for sending in comments. Please note that readers whose emails are derogatory to the author or any player will be permanently blocked from sending in any feedback in future.

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