August 23, 2013

Winning despite a misfiring top order

England's top three wickets have averaged 23.75 per partnership in Ashes 2013, which is their lowest in a series win in over 100 years

Conventional wisdom would suggest that top-order partnerships are crucial to any team that wants to win a Test series: get runs at the top, make it easier for the lower order to build on those runs, and then use those runs to put pressure on the opposition. England, though, have done anything but follow that logic on their way to a 3-0 series lead in the Ashes. In the first four Tests of the series, their top three wickets have batted together 24 times, and the only instance of a century partnership was in the second innings of the first Test, at Trent Bridge, when Alastair Cook and Kevin Pietersen lifted the team from 11 for 2 to 121 for 3, putting together 110. There have been five other century partnerships, but three of them were for the fifth wicket, one for the fourth, and one for the seventh. (Click here for the full list of partnerships for England.) For the top three wickets, the highest has been 110, followed by 73 and then 51 - of the 15 stands of 50 or more in the series for England, only three have been for the top three wickets. Clearly England have won this series despite the top-order batting.

Australia's top-order batting was quite average too in the first four Tests, but it still a touch better than England's, averaging 28.16 for the top three wickets. A big reason for that was the opening partnership, which averaged 48.62 for Australia in the first four Tests; the second-wicket stand averaged 17.25, and the third 19.12. The one century partnership came from Chris Rogers and David Warner in Durham, while there were three other 50-plus stands. (Click here for Australia's list of partnerships.)

However, given that Australia are trailing 3-0 in the series, a partnership average in the late 20s for the top three wickets isn't all that unexpected. However, England's situation is quite unusual, since you wouldn't normally expect a team with such poor top-order numbers to be so far ahead in a series. In fact, England's top-three-partnerships average of 23.75 is their third-lowest ever in a series win (with a 15-innings cut-off, though it's entirely possible that the series numbers could look better after the Oval Test). At the moment, though, it's the poorest for England in more than 100 years, and the poorest ever for them in a series they have won by a margin of two or more Tests.

The two previous instances when England won a series despite poorer displays by their top three wickets were in 1882-83 and in 1907. In the first of those, England beat Australia 2-1 despite their top three wickets averaging 22.73 runs per completed partnership; in 1907, England beat South Africa 1-0 despite a top-three-partnership average of 23.46. In both those series, though, the margin of victory was only one Test.

The table below lists England's lowest partnership average for the top three wickets in series since 1970, and apart from the drawn 1991 series against West Indies, England have lost each time their top order has averaged less than the partnership average in the current series. In 1991, England's top three stands averaged a mere 20.55, with only two century stands, while West Indies' averaged 33.33. In 1989, England's poor performance at the top was punished with a 4-0 trouncing against Australia, while they also lost 3-0 to them in 1979-80, and 2-0 to India in 1986. They also did almost as poorly as they have in the current series in 2000 against West Indies, but ended up winning that one 3-1. Five of their seven poorest such series have been at home, which is also a surprise.

The third table in the column below, though, shows the stability in England's current set-up and their faith in the current set of batsmen: unlike in 1989, when 21 different pairs batted in 33 partnerships, this time around only seven pairs have been used, one of them Tim Bresnan and Alastair Cook, when Bresnan came in as nightwatchman. Otherwise, it has been any two of Cook, Joe Root, Jonathan Trott and Kevin Pietersen batting in the middle for the top three wickets.

England's lowest stands for the top three wickets in a series since 1970 (Qual: 15 p'ships)
Series Tests No. of pairs P'ships Average 100/ 50 stands Series result
Aus in Eng, 1989 6 21 33 20.24 1/ 2 Aus 4-0
WI in Eng, 1991 5 6 27 20.55 2/ 0 2-2
Eng in Ind, 1972-73 5 15 27 20.73 0/ 3 Ind 2-1
Eng in Aus, 1979-80 3 9 18 20.94 1/ 1 Aus 3-0
Ind in Eng, 1986 3 12 18 23.16 0/ 4 Ind 2-0
Aus in Eng, 2013 4 7 24 23.75 1/ 2 Eng 3-0
WI in Eng, 2000 5 11 26 24.16 1/ 3 Eng 3-1
Eng in WI, 1985-86 5 11 30 24.73 2/ 2 WI 5-0
Eng in Ind, 1976-77 5 10 25 24.95 1/ 1 Eng 3-1
Eng in SA, 2009-10 4 7 21 26.00 1/ 3 1-1
WI in Eng, 1984 5 18 33 26.16 1/ 4 WI 5-0

In the table above, there are three instances of England winning series since 1970 despite their top three wickets averaging less than 25 runs per partnership (including the current Ashes series). Overall during this period, there have only been four other such instances by all teams. It's not surprising that three of them are by traditionally strong bowling sides, who have been able to overcome poor batting by the top order with strong performances in the field.

The lowest such average belongs to South Africa, whose top three managed an average partnership of 20.11, with only one 50-plus stand in 18 innings against Pakistan in 2006-07. However, despite the lack of partnerships, South Africa still had reasonable numbers from their top order, with Herschelle Gibbs and Jacques Kallis averaging more than 45. Makhaya Ntini led the bowling effort with 19 wickets at 18.68 as South Africa wrapped up the series 2-1.

Pakistan's batting - especially at the top of the order - has always been their weaker suit, and the bowling has often bailed the team out. That happened in Zimbabwe in 1995, when Pakistan wrapped up a 2-1 series verdict despite lack of partnerships at the top: the highest stand in 18 innings in that series was 58, while 11 out of 18 ended before reaching 20. Yet Pakistan had Inzamam-ul-Haq - who batted at No. 6 or 7 and scored 367 runs at 73.40 - and the bowlers to rescue them. There were also a couple of other series that Pakistan drew despite very low top-order numbers, both of them at home against West Indies: in 1986-87 they managed a 1-1 despite a partnership average of 18.50, and in 1990-91 the result was the same despite a partnership average of 19.52.

Australia swamped New Zealand 3-0 despite pretty ordinary top-order numbers in 1999-2000, while Sri Lanka rode Muttiah Muralitharan's wizardry - 26 wickets in three Tests at 19.84 - to beat Pakistan 2-1 despite the top three wickets averaging less than 25.

Teams* to win series since 1970 with ave stand for top 3 wkts less than 25 (Qual: 15 p'ships)
Team Series Tests P'ships Average 100/ 50 stands Series result
South Africa Pak in SA, 2006-07 3 18 20.11 0/ 1 2-1
Pakistan Pak in Zim, 1994-95 3 18 22.29 0/ 2 2-1
Sri Lanka SL in Pak, 1999-2000 3 18 24.44 0/ 3 2-1
Australia Aus in NZ, 1999-2000 3 18 24.61 0/ 3 3-0
* Excluding England

S Rajesh is stats editor of ESPNcricinfo. Follow him on Twitter