The Trescothick factor, and Kumble's profligacy

With partnerships of 76, 40 and 49 in three of their four innings, Andrew Strauss and Alastair Cook have done a fair job of opening the batting so far in the series against India, but Michael Vaughan and the rest of the England think tank will surely feel more comfortable if a fully fit Marcus Trescothick was to return to his favourite position at the top of the order. It's been a year since Trescothick played his last Test, against Pakistan at the Oval, and in the 13 Tests England have played after that, Cook and Strauss have only had modest success. They've opened together in 25 innings, and have managed four 50-plus stands - with a highest of 95 against India - and an average of 34.40.

Allowance must be made for the fact that 40% of those innings were against the mighty Australians (against whom the pair averaged 27 in ten innings) but the pair's failure to put together a single century stand in 25 tries is in stark contrast to what Trescothick achieved with various opening partners over a six-year period. Combining with Mark Butcher, Strauss and Vaughan, Trescothick put together 19 century stands for the first wicket in 142 tries - that's one every seven-and-a-half innings.

Trescothick's most successful pairing was with Butcher, but they batted together only six times; among batsmen who paired up with him for a reasonable duration, his stint with Strauss was the most prolific: in 52 stands they went past 50 on 20 occasions. It hasn't helped England's cause that Strauss has been in a batting slump lately - in his last 21 innings he averages a dismal 28.

As the table below shows, during Trescothick's time England's openers had solid stats against all teams. They only averaged less than 40 per partnership against Australia, but even those numbers were respectable - an average of nearly 37 with ten 50-plus stands in 30 innings.

The other area in which the current England team is suffering when compared to those in the recent past is lower-order batting. Chris Tremlett, Ryan Sidebottom, Monty Panesar and James Anderson have all proved themselves to be more than competent with the ball through the summer so far, but their efforts with bat in hand haven't been as successful. Since the 2005 Ashes, the contribution of England's last five is among the lowest in terms of runs per wicket; only Bangladesh have done worse. In the period between 2003 and September 2005 - a phase when England won 23 out of 36 Tests - their last five averaged nearly 21, and were fourth among all teams, after New Zealand, South Africa and Australia.

The magnificent Zaheer Khan took all the honours at Trent Bridge, but his match-winning performance - and the fine support act by RP Singh - took the focus off what was a surprisingly lacklustre display by the metronomic Anil Kumble.

In the end he did wrap up the tail to finish with six wickets for the match, but his second-innings figures of 3 for 104 in 25 overs was a rare instance when he went at more than four runs per over. In innings in which he has bowled more than 20 overs, only ten times has his economy-rate exceeded that mark. Six of those have been against Pakistan, and four of them in one series: on India's tour in 2005-06, the pitches for the first two Tests were so flat that teams piled up mountains of runs and the bowlers suffered. At Faisalabad, Kumble ended up conceding more than four in both innings. The track for the third Test was more bowler-friendly, but the fast bowlers got much more out of it than the spinners, and Kumble leaked 151 in 37.1 overs.

Apart from a couple of times when he has gone for over four an over in the 1990s, all the instances have occurred after 2004, suggesting that while the variations in his bowling have given him more wicket-taking options, especially abroad, he is also more prone to straying in line and offering the batsmen scoring opportunities, as Vaughan found out during the second Test. Given Kumble's improved performances overseas during this period, though, it's a trade-off the team will gladly accept.