Harmless

For a few days now, a friend of mine has been saying that England must pick Steve Harmison for Edgbaston. He has been on fire in County cricket, he's saying, and he's a matchwinner. He isn't alone in this - Vic Marks, one of the more astute and engaging cricket writers around, has written two separate articles in the Guardian calling for Harmison's inclusion.
Now, I refresh my Twitter homepage, and here again is my friend, tweeting "They better play Harmless!"
I think they mustn't. Here's why.
In 30 Test matches since the first Test of the 2005 Ashes at 2005, Harmison's figures show that he's been, well, harmless.
He's picked up 94 wickets, at the average of 38.53 and a strike rate of 68.1. Both his five wicket hauls in this period came in the same Test, at Old Trafford against Pakistan just over three years ago. Measure that against an overall average of 31.78 and strike rate of 59.3.
Even in the 2005 Ashes, the supposedly talismanic Harmison's series average and strike rate were behind those of Simon Jones, Andrew Flintoff and Matthew Hoggard - the Matthew Hoggard whose role in that triumph is forgotten against the backdrop of Harmison hitting people on the helmet in a Test England lost.
Let's compare the career trajectories of the two H-men, shall we?
In the period I mentioned above - Harmison's harmless phase - Hoggard took 88 wickets in 26 Tests, averaging 31.98 and striking every 59.9 balls, with two five-wicket hauls. Not much worse than an overall average of 30.50 and strike rate of 56.00.
Even so, after both were dropped after bowling badly in a terrible defeat to New Zealand at Hamilton in March 2008, it was Hoggard who suffered the worse fate, disappearing entirely from the Test side. Harmison, meanwhile, has played four Tests in three series in three different continents, doing nothing to suggest he'll give Australian batsmen nightmares.
Hoggard's 'nip', an intangible quality that he apparently cannot do without, is gone forever, it seems, while Harmison's fiery pace and steep bounce can be summoned up as soon as he breathes the air of Northumberland.
Agreed, Hoggard, at 32 is older than Harmison by two years, but this isn't really a 'bring back Hoggy' blog post. It is, instead, a blog post that has veered entirely out of track, but attempts now to return to what it started out trying to say - no, sorry, Harmison shouldn't play.

Lara, Chanderpaul and alternate universes

Does anyone have a smaller percentage of not out innings (apart, of course, from batsmen who've never been not out) than Brian Lara? Six not outs in 232 innings. Interestingly, within the six not outs are contained two of Test cricket's immortal innings - the 153 and the 400.
Even without statistics, no one would deny the claim that Lara has played in some pretty underwhelming sides. Statistics sharpen this picture, and starken (does this word even exist?) the contrast with batsmen from other teams.
Number of not outs is a reasonable measure of the effect playing for a good or bad team can have on a batsman. But it isn't foolproof; for instance, Shivnarine Chanderpaul, who has played alongside Lara for much of his career, has 32 not outs in 121 Tests.
On the other hand, number of declarations is a direct measure of a team's batting strength. During Lara's 131-Test career, West Indies declared fifteen times.
Let's apply the same measures to two other batting greats, who averaged over 50 like him, and whose careers ran parallel to his:
Sachin Tendulkar has played 147 Tests so far, and has 27 not outs in 261 innings. In those 147 Tests, India has declared 47 innings closed.
Steve Waugh played 168 Tests, batted 260 times, was not out 46 times, and Australia declared 65 times during his career.
What might have happened if fate had stuck Lara in the Indian or Australian line-up?
Nobody can say with any degree of certainty that batsman A would bat the same way if he were part of team X instead of team Y. It's still an intriguing thought.
Something tells me Lara would have averaged over 60 in a better batting line-up. Everything about his game - his method, his personality, his range of shots - would have ensured him top dog status in any line-up.
That leads me to another intriguing thought: Would Chanderpaul average as much, and be as hard to dismiss, if he was Indian or Australian?
The answer, I think, is yes - provided the selectors pick someone of his aesthetic otherness. He'd love being away from the limelight, he'd be the perfect guy for domineering types to bat in partnership with, and he has so much hunger for runs that it wouldn't diminish one bit even if he was batting in a team where every other batsman was Bradman.
What do you think?

How do you beat Pakistan?

Simple. Bowl badly when Younis Khan and Mohammad Yousuf are batting together.Younis's second-wicket partnership of 200 with Fawad Alam in the second Test against Sri Lanka at the P. Sara Oval was his third 200-plus partnership in a losing cause. The only other batsman in Test history to figure in three losing 200-plus stands? Mohammad Yousuf.
Younis and Yousuf are the only pair to figure twice on the list of losing 200-plus stands. The highest partnership ever in a lost Test? Younis Khan and Mohammad Yousuf, 363 for the third wicket against England at Headingley in 2006.

Soft batsmen, or average bowlers?

India's batsmen, historically, have scored a greater proportion of 'soft' runs than batsmen from anywhere else.

I base this statement on numbers. Dodgy they may be, and liable to scornful ripping apart by statisticians, but numbers, nevertheless. And there's an eerie pattern to all this as well.
Let's not get ahead of ourselves, though. Let me describe how I arrived at my conclusion.
First, I made a lot of assumptions, based solely on intuition.
My first assumption: drawn matches = batting-friendly conditions. Historically, I assumed, bat-a-thon draws numerically overwhelm rain-affected draws and games saved on spiteful wickets by tremendous batting.
So drawn Tests = batting-friendly wickets, and result Tests = sporting, or bowler-friendly wickets.
Runs scored in result matches, therefore, equate to hard-fought runs. This, obviously, ignores knocks played with a 'what the hell, we've lost; let me bash it around a bit' mindset.
Once I made these assumptions, I calculated (on June 16 2009, so recent series like the Ashes and Sri Lanka-Pakistan don't figure), for the eight major Test teams in history - Australia, England, India, New Zealand, Pakistan, South Africa, Sri Lanka, West Indies - the ratio of centuries scored to number of Tests played.
I then calculated the centuries per Test ratio for each country, ignoring drawn Tests.
I subtracted the second ratio from the first.
And discovered that India has had, through Test history, the biggest difference between centuries per match and centuries per result match. That is, in result Tests, Indian batsmen have suffered the steepest drop in century-production from their average rate than batsmen from any other team.
Here's the table (click on the image to see a bigger, more readable table):
This either means that Indian batsmen have scored a lot of soft runs in helpful conditions, or it means that India's bowlers have let Tests drift into draws in conditions where bowlers from other teams have bowled match-winning spells.
Look at the averages of India's great bowlers - Mankad, Gupte, Bedi, Chandrasekhar, Prasanna, Kapil Dev, Kumble, Srinath - high 20s or low 30s without exception. India hasn't had anyone with a Jason Gillespie-esque mid-20s average, leave alone someone with an average approaching Curtly Ambrose or Sydney Barnes. Is Gillespie a greater bowler than Anil Kumble? No. Has he been part of better bowling attacks? Most definitely.
So I decided to do the same calculation for this century - the 2000s. India has regularly won Test matches abroad in this period, a period that's included a number of memorable matchwinning centuries, and seen the emergence of a pretty decent bunch of bowlers.
Here then is the eerie pattern I referred to before. India has done no better or worse in this period than through its history, and the clinching number is exactly the same as in the previous table:
0.24, the most significant number in the history of Indian cricket.