Wednesday, February 21, 2007

Stats: Should "in matches won" be a criteria to judge individual performances?

Yuvraj Singh averages 84 in SUCCESSFUL run chases above 275. Dravid and Ganguly are the only others in his team that average 50+ in similar situations. Sachin averages 42 which, though good, is just half of Yuvi. Incidentally Sarwan is Bradmanesque (118) in big chasing wins and so is Inzy (104). Jayasuriya averages 90 while Brian Lara returns 77.

I am a keen follower of cricinfo's 'The List'. However the purpose of this post is not to praise that particular stats filter but to ...well, not exact bury but question it. I have serious reservations about this 'averages in wins' bit. Why bring a 'win' into the equation? A win is a combination of numerous factors that no individual player has control over - unless it is THAT day of his life. Even Andrew Flintoff struggled to win the match of his life on his own; when the wheels were falling off he needed Harmison to clinch the win.

While analysing a large number of samples in a team game, is it not a given that wins are performance combinations where a high percentage of the team's players do well? The stats filter under discussion then does little more than stating the obvious. It can be argued upon as an indicator of match winning potential but certainly not as a definitive measure of real worth of a player.

Remember Kaif's glorious 87 in that 2002 Natwest Final when India succesfully chased 327? That innings was magnificently well-paced but it still contributed only just over a fourth of the total. You faintly recall more than one match when some guy scored bigger runs in a smaller chase at a handy scoring rate and yet ended up on the wrong side. Worse, some of these guys have scored big runs more often in their careers than Kaif.

Kaif’s match winning abilities are not in question but it is singularly unfair to cast a shadow on those other folks’ performances just because their teammates were not up to scratch in those matches. Even Kaif could have been the tragic hero had Yuvraj not fired along with him on that summer evening of 2002. Going back to the last ball of that Edgbaston cliffhanger, it is comparable to looking down upon Flintoff’s capabilities after Michael Kasprowicz had edged Harmison for a match winning brace just beyond the reach of a legside diving Geraint Jones.

In my view therefore, 'averages in chases' (irrespective of match results) is a far better indicator of performance. We will look at a very recent example that demonstrates the loopholes of that filter in a 4 match capsule.

West Indies chased in all four matches of their
recent series versus India. Three of them were big chases and West Indies could win just one of them. Shiv Chanderpaul played in three matches (two of them big chases), aggregate the highest runs [230 at an average of 115] and yet ended up on the losing side in all of them.

The highest average in a 'winning big chase' - their only win which Shiv sat out - was thus a race between Samuels (98) and Lara (83). Of the two, Samuels scored consistently over the series but Lara featured there inspite of failures in all other matches.

Now this 4 match case study throws up three stereotypes amongst 'players that are discussed’. The types do not necessarily hold true for these particular players over their careers but here they are:

One: Those who perform well in most matches and happen to do so in the 'win' too - like Samuels (heh)
Two: Those who perform well in most matches but unluckily miss out when the 'win' comes, like Chanderpaul
Three: Those who are inconsistent in the long run but happen to score big occasionally - and are hence often the difference between a loss and a win, like - Lara!!

[Needless to say but talking of career performances Lara & Chanderpaul are cat-1 (category one) in both forms of the game while Samuels is firmly in cat-3.]

The whole point of this long analysis: Filtered results for 'Averages in wins' can glorify cat-3 players more than they deserve; similarly it can belittle cat-1 or cat-2 players for little fault of theirs. This is not to say cat-3 players have no place in the game. Far from it… (else what are the Sehwags and Afridis doing in national WC squads?)

Statistics, like they say about electricity and the domestic LPG stove, is a very good servant but a bad master. It is wise to use them with clarity of purpose, judgement and perspective.

So, do we crown 'averages in big chases' as the prime indicator of a batsman's ability to cope with adversity and keep the filter at that?

[to be continued...]

No comments: