Having come up with a formula for baseball that I was happy with, cricket was my first attempt to see if the concept could be applied to other sports. Like baseball, it has a detailed statistical record that goes back to the 19th Century, and it tracks individual numbers for all players, allowing me to put off the problem of team defense for another day.
Cricket fans will no doubt wonder what business an American has passing judgement on the greats of this game, and there will some results that will be blamed on my ignorance (Graham Gooch over Viv Richards?), but I hope a glance at my methods will alleviate any doubts.
Two decisions will stand out in particular. One was to use first-class statistics rather than Test numbers, which are the usual gold standard. This was mostly to get a big enough sample size. Don Bradman, for example, played 234 first-class matches, which is already pretty low. Judging him on just 52 Tests would border on the absurd.
Second, I used runs per match to determine the offensive rating, not runs per wicket, which is the practice for batting average. I find that some high-profile middle-order batsmen (e.g. Steve Waugh) have a very high number of not-out scores, which drives up the batting average. Using runs per match put openers and middle-order batsmen on even footing.
Offense was thankfully straightforward, since everything a batsman does goes right on the scoreboard. In this one instance, I use only a single rating for offense, simply dividing a batsman’s runs by the average for the season.
The first defensive number is wickets taken: catches, stumpings and, for bowlers, wickets bowled and batsmen trapped LBW. Defense 2 is for bowling; runs given up per wicket lower than 1½ times the season’s average.
The complication came when figuring out limited-overs matches. Because the actual number of overs in “List-A” matches has varied over the years, I use runs per 50 overs rather than runs per match. The bigger problem is that each bowler is limited to no more than one-fifth of the overs, which means that star bowlers are not able to dominate as they do in the longer format. Modern bowlers, who generally play twice as many limited-overs matches as first-class, end up with fairly low 160-game averages as a result.
To limit the effect, Apples & Oranges weighs first-class matches 4 times more valuable than limited-overs matches (or Twenty20, which becomes a third category) when figuring out the 160-game average. But it’s still a problem. There’s also a slight change to figuring out the defensive numbers. Defense 2 is based on runs given up per over rather than per wicket, and all wickets taken by a bowler count in figuring out Defense 1.
Lastly, I include the unofficial World Series of Cricket as its own separate “season.”
Bradman’s offensive average of 45.28 is up there with NFL quarterbacks — no other batsman is over 40. I’d say that without World War II, he’d have gotten his overall rating past 30, but by all accounts he was breaking down physically when the fighting broke out, and it was the enforced rest from the hostilities that allowed him to get back in shape in the late ’40s. He may be the only star of that era, in any sport, who’s career was actually prolonged by the war.
Gooch, as I said, will probably raise a few eyebrows. He played for a long time and scored a lot of runs … I make no claims about his artistry.
When he retired, Brian Lara was fractionally ahead of Sachin Tendulkar, essentially tied. Some 300 games later, Tendulkar ended up 1.35 points clear.
Because of apartheid, Barry Richards and Graeme Pollock played very few Tests, another reason for going with first-class numbers. Their limited international record, including the World Series of Cricket, would indicate they belong with any of their peers.
|All-Rounders||Matches||Total||Per 160||Sqr Root||Sum||Off||Def|
|Young Jack Hearne||647||84.06||20.79||9.17||29.96||26.52||15.05|
Just defining what an “all-rounder” is can get tricky. Some are brilliant at batting and bowling; some are great at one skill and merely good in the other; and some aren’t world-beaters in either, but the combination makes them great in aggregate. For my purposes, I picked players that bowled between 1 and 10 balls for every run they scored in first-class cricket, and bowled between 10 and 40 overs per match.
As a group, they are the most valuable athletes in any sport, and W.G. Grace comes out as the top athlete by my method, just edging out Pelé. This is mostly because he dominated a still-developing sport in its infancy, but his numbers are absolutely jaw-dropping when put in context. No one exceeded his peers quite like Grace.
Wally Hammond is never really listed as an all-rounder. He wasn’t one at Test level, but he was in county cricket. If I had listed him as a batsman, he’d be No. 1 there, and he would have edged out Bradman on the strength of his bowling. That didn’t seem right. He did bowl 51,000 balls in first-class cricket, or about 22,000 more than Jacques Kallis.
|Bowlers||Matches||Total||Per 160||Sqr Root||Sum||Off||Def|
A few of the bowlers could be all-rounders by the standard I outlined above. I felt, though, that their production didn’t merit that sort of classification. Offense among bowlers varies a lot more than defense among batsmen, so their ability to hit plays a bigger role in their final rating.
Muralitharan gets hit the hardest by the one-day format. His first-class defensive average is 38.29; in List A matches, it’s 19.69 … half. And that is entirely due to not being able to bowl more than 20 percent of his team’s overs. But he only played 233 first-class matches, compared to 453 one-day matches, and another 164 in Twenty20. He was still a star, but the conditions in which he played prevented him from really dominating like he could have.
Glenn McGrath comes in really low. Partly, this is due to his abysmally low offensive average; just as a bowler, he outranks several of the people above him. He also played very little outside Tests and ODIs; he was essentially a full-time Australia cricketer. This meant two things: He had the best defense in the world behind him, and he never had to face Australian batsmen. All in all, I think he rates as one of the most fortunate athletes ever.
|Wicketkeepers||Matches||Total||Per 160||Sqr Root||Sum||Off||Def|
Pity the wicketkeepers. Even more than baseball catchers, their own job destroys them. Once Kumar Sangakkara is done for sure, we’ll have a borderline 30/29-point player to add to the list. But, man, it’s not a deep pool.
To do list: