Thursday, October 28, 2010

The Allure, and the Pitfalls, of All-time XIs

It seems like the season for "All-time XIs" has returned with a vengeance, thanks in no small part to CricInfo's series of country-specific XIs, followed up with a World XI. This is really the armchair cricketer's ultimate delight. What better way to indulge your fantasies than to cook up a couple of all-time XIs and imagine (or even better, simulate, perhaps with book cricket!) a game between them?

Back in 1994-95, when Usenet was still usable, the newsgroup rec.sport.cricket saw a sustained frenzy of all-time XIs. The leader of the pack then was a respected poster called John Hall, who kicked things off by posting a series of "alphabet XIs" -- e.g., the "A Team" comprising of players whose names start with A, and so on. Not quite 26 of them, because some alphabets got combined to get to critical mass (e.g. the QXYZ team), but great fun nevertheless. Must have taken an enormous amount of research. He followed that up with teams from each decade in which Test cricket had been played, which was another tour de force! Of course, as soon as each team was posted, it drew in the rest of us on rec.sport.cricket, with our own "expert comments" on John's sins of omission and commission!

Naturally, no two posters could agree upon an entire XI, but that is almost besides the point. The allure is in reliving the careers of these all-time greats, debating their relative merits, etc. As a bonus, many a thread also spun off from these topics into discussions of the personalities and careers of those involved, especially the relatively lesser-known cricketers who made it to the all-time XIs because of these artificial constraints of alphabets or specific decades.

There are of course pitfalls in making these sorts of comparisons across generations. One argument goes that the relatively recent players, whom you've had the chance to watch, follow, perhaps even idolize, are more likely to make the grade than earlier ones whom you have only read about. On the flip side though, older players tend to be romanticized with the passage of time. History mostly records their successes while glossing over their failings; hagiographies abound.

Modern players are exposed to a wider set of playing conditions -- far more venues, varying equipment, with/without floodlights, with/without umpire reviews, bowlers using slower balls, doosras etc. -- that didn't exist in the old days. The oldies however had to deal with uncovered wickets, and that too in places like England with fickle weather! They didn't have protective helmets and arm-guards. They had to travel long distances by ship, train, buses... Flip again -- the fielding standards, and on average the catching standards, have improved dramatically. So the oldies might have got away with snicks through the slips etc. that they wouldn't, today. Or would've got boundaries for what would only be a single today...

With so many contradicting signals, how can the armchair cricketer even begin to compare cricketers across eras? Clearly, a direct comparison of statistics doesn't seem fair -- we don't know whether these factors cancel each other out. The written word is also unreliable -- whether it's the cricket correspondent covering a Test match, or an autobiography, or a book like Sunny Gavaskar's "Idols"...

I believe the closest we can get to an objective comparison is if we compare statistics within the same 'era'. This of course requires a clear definition of eras in cricket, and then identifying contemporary cricketers whose careers substantially fall within that era. This was relatively easier in the early days, when the World Wars provided some natural delimiters for eras, and several careers were terminated by those wars. In more modern times, perhaps we can identify two eras based on the teams that dominated those years -- the West Indies era (1970s and 1980s), and the Australian era (1990s and 2000s). Many careers don't fit nicely into those eras though -- a Bishen Bedi, or an Alan Border, for example.

Nevertheless the relative performance of a player, when compared with his contemporaries of the same kind (bat/bowl/keep/allround/captain), seems to be the only reasonable yardstick of greatness, and thus, selection in an Alltime XI. In batting for example, a Test average of 50 was considered the yardstick of greatness during the 1970s-90s, and possibly earlier as well. The very best batsmen of those years -- Sunny Gavaskar, Greg Chappell, Viv Richards, Javed Miandad -- reached and sustained that mark. Very few others were able to sustain a 50+ average over a long career. That benchmark seems to have shifted up by 4-5 points in recent years. Today's greats are all in the 54-57 range -- Sachin, Ponting, Kallis, Mike Hussey, et al. Dravid and Lara are lower, but probably dropped off from that range. Clearly we cannot claim that all of today's greats are significantly better than those of the 1970s (and a 4-5 point difference in batting average is significant!). Thus it makes sense to compare only with contemporaries whose careers overlapped for the most part.

All that is fine for a scientific, or statistically defensible exercise in selecting an Alltime XI. But, all said and done, it's eventually a very personal, emotion-riddled flight of fantasy! So who's to quibble with the critic who prefers a Dravid to a Vishy, or someone who selects a Dhoni ahead of Kirmani as his keeper? To each his own, and the more the merrier, and I'm all out of cliches now!