New Zealand thumped India 2-0 in the recent Test series; according to my Team Performance Rating system, the scoreline was 139-61. Whether New Zealand would have struggled if India had given them a target of 200 in Christchurch is debatable. As is conjecture whether the Wellington Test was ever in a position where India might have saved it.
Setting aside the failures of the Indian top-order batsmen and the finely planned and executed thousand-cuts strategy of the New Zealand pace bowlers, a significant difference between the sides was how the New Zealand late-order batsmen batted. Batting was tough for everyone in the series, but New Zealand's lower order combined good defence and judicious attack to make the difference.
Did that difference cost India the series? I doubt it. But without it, the results would have been much closer - maybe more so in the second Test than the first, which was a rout. These two Tests provide the latest data in my analysis of late-order batsmen across the ages.
Let me first define a late-order batsman. In principle, it is anyone who bats at No. 8 or lower. So my analyses start after the fall of the sixth wicket.
I will look at three aspects of late-order batting: the score at the fall of the sixth wicket, the runs added to the team score after the fall of the sixth wicket (partnerships), and the runs scored by qualifying lower-order batsmen (LOBs). (Information on balls played is available only for the past 20 years or so and is hence not considered.)
Let me first highlight certain points related to the late-order batting situations.
1. A comprehensive analysis of all the scores at which the wickets fell across the completed innings of the 2387 Test matches in question has revealed the summary values listed below. The values indicate the percentage of team score reached at the fall of each wicket. The difference between this value and 100.0%, shown in brackets, is the expected contribution of the rest of the wickets. Thus it can be seen that, at the fall of the sixth wicket, the expected average contribution from the last four partnerships is a little over 26%. Although 40% of wicket resources are available, the expected returns are much lower.
FoW-1: 11.7% (88.3%)
FoW-2: 24.1% (75.9%)
FoW-3: 37.2% (62.8%)
FoW-4: 50.1% (49.9%)
FoW-5: 62.1% (37.9%)
FoW-6: 73.6% (26.4%)
FoW-7: 82.1% (17.9%)
FoW-8: 89.6% (10.4%)
FoW-9: 95.4% (4.6%)
2. Not all six-down positions qualify. For instance, 167 for 6 when the target is 180 will not make it if the chasing team wins. But it will if the chasing team is bowled out and loses the match; 167 for 6 when the target is 270 will always get in, and so will 167 for 6 to 180 all out and 167 for 6 to 321 all out.
If an innings total of 400 is reached before the sixth wicket falls, that innings will not qualify. I am not interested in 450 for 6 becoming 480 all out, nor in 450 for 6 becoming 550 all out. The dynamics change completely when the team has already reached a huge score before the late order gets in. These innings are excluded to prevent diluting the final values.
3. The runs added are obviously important, but the match situation when the sixth wicket fell is equally relevant. Let us compare two cases. In each, 100 runs are added by the last four wickets. In the first, the score moves from 150 for 6 to 250; in the second case, from 350 for 6 to 450. Ignoring all other considerations, the first set of late-order partnerships is far more valuable than the second one.
4. This difference is identified by determining the share of the late-order partnerships in the team score. In the first case, it is 40% and in the second, 22.2%. Any value at or above 40% is very good, and anything below 12.5% is poor. Anything in between ranges from "way below par" to "way above par". The percentage of the team score contributed by the late-order partnerships, rather than the percentage of score at the fall of the sixth wicket, is considered since the latter is open-ended and difficult to evaluate: 20 for 6 becoming 100 will have a value of 400%.
5. The context of the match is not taken into account. This is a macro analysis and I am interested in determining overall numbers - by period and by team. I have gone anecdotal only while describing the best and worst late-order partnerships.
6. At this stage, I am more interested in the late-order partnerships than "runs scored by late-order batsmen". Out of a total of 100 runs added by the last four partnerships, 30 might have been scored by batsmen 8-11 (that means there was a top-order batsman scoring 70, if we assume zero extras). In another case, 60 runs could have been scored by batsmen 8-11 (maybe there was a No. 7 batsman scoring 40 or less). I have derived the data but will not dwell too much on this difference. The team needs are paramount.
7. It is possible that there was a nightwatchman and Nos. 6 and 7 are at the wicket after the fall of the sixth wicket - as India were in Christchurch. I have not taken cognisance of such fine distinctions. The team was at 90 for 6 but could add only 34 more - that is what matters. I have only taken care not to include top batsmen's runs (for example: Hanuma Vihari's and Ravindra Jadeja's contributions) as runs by late-order batsmen.
I will explain all these with my illustrations of the eight innings in the two New Zealand-India Tests.
Wellington
India: From 132 for 6 to 165; 33 runs added - 20.0% (Just below par). 26 runs by the late-order batsmen
New Zealand: From 216 for 6 to 348; 132 runs added - 37.9% (Just short of very good). 92 runs by the late-order batsmen
India: From 148 for 6 to 191; 43 runs added - 22.5% (Just below par). 14 runs by the late-order batsmen
New Zealand: 9 for 0. Innings not included
Christchurch
India: From 197 for 6 to 242; 45 runs added - 18.6% (Just below par). 26 runs by the late-order batsmen
New Zealand: From 153 for 6 to 235; 82 runs added - 34.9% (Good). 71 runs by the late-order batsmen
India: From 89 for 6 to 124; 35 runs added - 28.2% (Just above par). nine runs by the late-order batsmen
New Zealand: 132 for 3. Innings not included
The bottom line is that the Indian late-order batting was just below par (not anywhere near poor, as is being portrayed) and New Zealand's late order was good. Look at the overall comparisons: New Zealand's 36.7% as against India's 21.6%. Significantly different but not as impactful as other factors.
Now that the foundation has been laid, let us look at the best and worst late-order batting performances, done in an anecdotal manner, using a couple of tables and a short trip across the years.
A look at the last column will confirm that in most of these epochal late-order partnerships, late-order batsmen play only a supporting role: their average contribution is 30%. Only once in the list above has this value come close to 50% and only once has the late-order contribution exceeded 100 runs.
Australia's recovery at Old Trafford in 1888 was mind-boggling: 7 for 6 became 70. Only 63 runs were added, but this formed a huge 90% of the final team score. John Lyons (32) and Charlie Turner (26) were the heroes, but unfortunately, in losing cause.
In the 1998-99 Asian Championship Test in Kolkata, Pakistan collapsed to 26 for 6 and were rescued by Saleem Malik (32), Moin Khan (70) and Wasim Akram (38). They reached 185 and beat India despite conceding a lead of 38 - thanks to an all-time great innings of 188 by Saeed Anwar. The value of the first-innings recovery was felt three days later.
In Port Elizabeth in 1992-93, India, trailing by 63 runs after the first innings, went down to 31 for 6 and looked like they would concede an innings defeat. They were rescued by Kapil Dev's magnificent 129 and a trio of 17s by the late-order batsmen. That they still incurred a big loss is incidental.
At the Old Wanderers in 1927-28, South Africa moved from 26 for 6 to 170 thanks to Denys Morkel (29), Cyril Vincent (53) and Shunter Coen (41).
In 1969-70, Ian Redpath (63) and Graham McKenzie (24) led Australia to a match-winning recovery from 24 for 6 to 153.
No one who watched the Karachi Test of 2005-06 can forget the great innings of Kamran Akmal (113) and the contributions of Abdul Razzak and Shoaib Akhtar to complete a first-day recovery from 39 for 6 to a match-winning 245 against India.
At The Oval in 1967, Asif Iqbal made a classic 146 and a record stand of 190 with Intikhab Alam, although their efforts came in a losing cause.
Clem Hill's 188 at the SCG in 1897-98 has never fallen out of the top 20 in the Wisden 100 and the GW25 (the rating system I created for best Test innings of all time) tables over the past 20 years. From 58 for 6, Hill participated in valuable stands for each wicket to lift Australia to 323, which proved to be a match-winning score.
Harry Graham's 105 (which features among the top 25 batting performances in my ratings) and Albert Trott's 85 let Australia move from 51 for 6 to 284 in a huge innings win in Sydney in 1894-95.
Although Sri Lanka eventually lost, Arjuna Ranatunga's 51 helped them recover to 97 from 18 for 6 against New Zealand in Kandy in 1983-84.
Against England in Christchurch in 1929-30, New Zealand slumped to 21 for 6 but were rescued by Roger Blunt's unbeaten 45 and contributions to Nos 9 and 10.
Finally, in Lahore in 1955-56, Pakistan, after slumping to 111 for 6, managed to add a whopping 450 for the last four wickets, which included a stand of over 300 between Waqar Hassan (189) and Imtiaz Ahmed (209).
Now for the worst ever performances by late-order batsmen.
These are huge collapses. There is no need to get into too many details, especially but here are a few interesting sidelights. At The Oval in 1957, West Indies lost their last four wickets without adding a run. In Harare in 2013, Bangladesh got to 134 for 6 and then added no more runs in the rest of the innings. New Zealand, playing Pakistan in Auckland in 2000-01, managed to add a single run for the last five wickets. At the SCG in 1998-99, Darren Gough's hat-trick ensured Australia added only four runs for the last three wickets.
In Melbourne in 1978-79, Australia collapsed from 305 for 4 to 310 following a stand of 177 runs. In Melbourne in 1990-91, England imploded from 147 for 5 to 150. Against Pakistan in Wellington in 1964-65, New Zealand slid from 261 for 5 to 266.
Let us now look at the period-wise analysis.
During the last decade, the West Indian team has a high 32.3% of late-order partnerships crossing 40% of their team scores. However, this is mainly because West Indies were in trouble quite often in the 2010s. As such, when a team is at 100 for 6, it is not too difficult to score 40-plus runs. This is substantiated by their average sixth-wicket score, which is 175 - lower than Zimbabwe.
India lead, reaching an average of 211 at the fall of the sixth wicket. The much-maligned Indian late-order batsmen have been quite good: over 30% of the late-order contributions have crossed 40% of the team scores. In other words, the team has been bailed out by their late-order batsmen more often than we may think. They also have the highest average partnership.
Australia have the highest average score by late-order batsmen (47), which forms well over half of their average partnership value. That means their late-order batsmen are counterattackers, while India's, with 28 out of 84, are supporting batsmen.
England's 77.1% came when they slumped to 102 for 6 (and 102 for 7) against Pakistan at Lord's in 2010 before Jonathan Trott and Stuart Broad added 332 runs for the eighth wicket. They finished on 446 and won the match comfortably by an innings.
In the first decade of the millennium, Australia had an amazing average of 252 at the fall of the sixth wicket. New Zealand had the highest late-order partnership average of 46 - and they needed those contributions, since their average score at the fall of the sixth wicket was a low 186. Only Zimbabwe and Bangladesh had lower averages. Over 30% of New Zealand's late-order partnerships exceeded 40% of the team score. The Zimbabwe late order led on the runs-scored metric.
The best late-order performance was achieved by Pakistan against India in 2006. This has already been covered.
Australian batsmen reached an average position of 220 runs at the fall of the sixth wicket in the 1986-1999 period. South Africa, readmitted a few years into that era, had the highest late-order-partnership average of 95 - quite a high figure indeed. However, their average score at the fall of the sixth wicket was not great - 191. In terms of scores, the late orders of England and West Indies comfortably outscored their higher-order counterparts.
The strong West Indian side made its presence felt in the sixties and seventies. Their top six wickets averaged an impressive 232. Like in the previous period listed, English and West Indian late-order batsmen scored the highest number of runs, though they did not outplay their top orders in this period. In the few Tests they played before their ban, South Africa achieved a very high average for late-order partnerships - but only across 12 innings; Pakistan's late-order-partnership average of 80 is probably more impressive.
In the post-war period, West Indies' batting was very strong, as evidenced by their average score of 228 at the fall of the sixth wicket. New Zealand's top order only scored 143 on average. Both New Zealand and India had to get over 28% of their innings rescued by their late-order batsmen. In terms of late-order partnerships, Australia led. Don Bradman was in the side for the first few matches of this period.
In the years leading up to World War II, Australia and England led in almost all segments. The top-order batting of the other teams was quite average and they all had to be rescued often by the lower orders.
It is unbelievable that across 140-plus years, 200-plus locations, more than 3000 players, nearly 9000 innings, and changes in almost every aspect of the game, most of the key summary values are in a tight band of about 10%. The average partnership for the last four wickets ranges from 69 to 76, with three periods having the same value of 76. The average, 74, has a 2.7% variation on the higher side. How does one explain this? I can only conclude that this is the power of a large population size.
The average of runs scored by the late-order players is in a still tighter band - 35 to 38, only around 8%. We are talking of a range of players from Chris Martin to Daniel Vettori, Bhagwath Chandrasekhar to R Ashwin, and Nuwan Pradeep to Chaminda Vaas. And they contributed between 35 and 38 runs on average. Amazing.
The average score at the fall of sixth wicket shows slightly more variation - from 186 to 208. This figure was highest during the first period of this century, and lowest during the first, long period, from 1877 to 1939.
The overall percentage value of late-order contributions is 27.1% (74/273), which is quite close to the already discussed 26.4%, the average percentage of runs added for the last four wickets. The slight difference is because the 26.4% includes all innings that came to a conclusion. The 27.1% is derived after excluding innings that went past 400.
The inescapable conclusion we can draw from these numbers is that there is very little change in the numbers across periods. Today we get to watch plenty of late-order partnerships in the visual media. There was a time when we would only have heard of them or read about them the next day. The impact was probably much lower. So it is possible that we have been under the impression that there are more effective late-order batting efforts nowadays. That is not true.
Email me with your comments and I will respond. This email id is to be used only for sending in comments.