Friday, January 17, 2014

block WHACK block block WHACK

Just a quick follow-up to the Kallis post - I'll try and keep it brief, as the fraction of my already minuscule blog readership that has any interest in this stuff must be vanishingly small, indeed the odds are it's probably just me. But, hey, any complaints? Get your own blog.

The criticism of Kallis as a bit of a stodgy batsman and not the most exciting to watch is a bit harsh, given his huge value to the South African team in the role he was asked to perform, but contains a grain of truth. His overall career strike rate of a fraction under 46 (note to non-stats-buffs: this is the number of runs scored per 100 balls faced) is low compared with his contemporaries like Ricky Ponting (59), Sachin Tendulkar and Kumar Sangakkara (both around 54), though higher than the proper stonewallers like Atherton and Boycott (both mid-30s).

So it seems a bit paradoxical that Kallis ranks second on the six-hitting list, almost as if he had a bit of a split personality while batting and would occasionally go berserk and whack a couple into the stands before settling down to blocking everything again.

So I therefore propose a new statistical measure which compares the number of sixes hit by a batsman with his overall scoring rate. You can't just divide one by the other, though, goodness me no, as since the sixes count is cumulative that would disproportionately favour people who've had very long careers. A fairer way would be to weight the formula with the number of innings batted in, as follows:
VMSI = (number of sixes hit) / (overall strike rate x number of innings)
VMSI stands for Violent Mood Swing Index, which is what I'm calling it. I mean, what else would you call it?

As before, let's restrict it to people with over 3000 runs to get rid of some of the freaky statistical outliers. Here's the top 25:

NameMatchesInningsRunsStrike rateSixesVMSI
CL Cairns (NZ)62104332057.0987146.53
NS Sidhu (India)5178320244.3338109.90
CD McMillan (NZ)5591311654.9554107.99
A Flintoff (Eng/ICC)79130384562.0482101.67
Misbah-ul-Haq (Pak)4678308741.223299.53
MS Dhoni (India)81127434259.747598.85
Imran Khan (Pak)88126380747.525591.86
JR Reid (NZ)58108342833.843390.29
AC Gilchrist (Aus)96137557081.9510089.07
BJ Haddin (Aus)5491300758.614788.12
CH Gayle (WI)99174693359.909086.35
JH Kallis (ICC/SA)1662801328945.979775.36
ML Hayden (Aus)103184862560.108274.15
CG Greenidge (WI)108185755849.026773.88
KP Pietersen (Eng)104181818161.728172.51
CL Hooper (WI)102173576250.276372.44
CH Lloyd (WI)110175751557.777069.24
BB McCullum (NZ)82141468460.765968.87
IT Botham (Eng)102161520060.716768.55
IVA Richards (WI)121182854069.288466.62
WJ Cronje (SA)68111371444.633366.61
BC Lara (ICC/WI)1312321195360.518862.69
Mohammad Yousuf (Pak)90156753052.395162.40
V Sehwag (ICC/India)104180858682.239161.48
HH Gibbs (SA)90154616750.264760.72

A few things to note:
  • Chris Cairns is out in front by quite a startling margin, by virtue of his remarkable sixes per innings rate and his brisk but not startling strike rate.
  • Kallis comes in 12th, but note that he is first among batsmen with over 7000 runs, fourth among batsmen with over 4000 runs, and that of the people above him only Sidhu, Misbah and Gayle are full-time batsmen, all the others being all-rounders of one kind or another who typically batted at number 7 or lower and were given a bit more licence to swing the bat.
  • A few names you wouldn't expect pop up, like the late Hansie Cronje and Navjot Sidhu of India. Sidhu in particular was notorious for (among other things) periods of scrupulous defence punctuated by some furious smiting, generally of spinners. In fact if you follow that link you'll see the first line of the Cricinfo biography reads "Navjot Singh Sidhu's cricket had a schizophrenic touch to it". QED, I'd say, since that's pretty much exactly what this index is a measure of.
  • The men who do the breakneck scoring every day of the week, in particular the likes of Adam Gilchrist and Virender Sehwag, appear at 9th and 24th respectively in the list, undone by their high scoring rates.
  • The reason the list seems to favour relatively modern players (John Reid is the only man on the list to have played Tests before 1965) is mainly because stats like exact numbers of balls faced and boundaries hit are not reliably available for older matches. 

No comments: