Outcomes of Scientific Australian study on batting greats

Bradman the best ever

"Sehwag greater than Sachin". WTF? Sure, I'd have Seaweed in my side any time, but is he better than SRT???

So it took 4 years for these guys to do the study?? They've used stats only up to 2006, and we all know how farked the Wall has been since then. As the article says, in that time Dravid's average has fallen from 58 to 53. And he is ranked in front of TLM.

Surely their formula could spit out new ranks after each test match ...

Or is the report a Karticle written by a Wall fan and cut off at that point to improve his rating?

Maybe Vani K Borooah is Karti.

Would you listen to anythiung two economists said? I doubt it!

Ah, but I loves me stats, bl.

lulz. kambli india's 5th greatest bat ever. 13th best ever in the world

Aye, TO, that does tend to put the study into perspective. i.e. it's trash.

Perhaps Josh forgot to read that these ratings are late by 4 years.

Had he read that I'm sure he won't have posted this.

As long as BCL is greater than Tenduckar, then all is good in the world.

Rest assured ... all IS good in the world

SG wrote:Perhaps Josh forgot to read that these ratings are late by 4 years.

Had he read that I'm sure he won't have posted this.

But that's what makes the recent publishing very odd, SG. As MB says, surely their ranking formula/program could've been rerun with more modern data prior to the release of the study. It makes the study just a little ridiculous and some of the findings just a little ludicrous.

Does anyone have a copy of the full article. The journal seems to be subscription only and not for ordinary punters.


In any case, it's hard to see what it adds over zcore's opus in 2000.

Where's Yuvi in the list ?

philcric wrote:Where's Yuvi in the list ?

Well Bradman was number 1. So i'm guessing he was 0.

JGK wrote:Does anyone have a copy of the full article. The journal seems to be subscription only and not for ordinary punters.


In any case, it's hard to see what it adds over zcore's opus in 2000.

you just need to give a name and email id to get it

its not particularly interesting

Charlie Davis's (zscore's) Best of the Best was an excellent read if you are a stats geek. And I agree with JGK, this study probably adds nothing.

It'd be good to see a later addition of zeddy's book come out that included some of the prolific stats of the 2000-2010 decade.

Just found the article - pretty sloppy this part:

"Nor did we, in arriving at our assessment of the world’s top fifty batsmen,
make any attempt to allow for the quality of the opposition against which they
played. There has, arguably rarely been a fiercer bowling attack than Jardine’s
body-line team and, in this context, McCabe’s innings of 187 at Sydney in
December 1932 – hailed by Bradman as the greatest innings he had ever seen –
must mock his lowly position in the rankings (joint 14th)."

I'm pretty sure Bradman didn't even attend that match as he was in the contract dispute with the ACB at the time.

So it's basically not worth the paper it's written on then?

skully wrote:So it's basically not worth the paper it's written on then on the bog roll ?

Bradman .... phurt ...

Academics (if they are) are allowed to have fun too.

But like asking AB for an assessment of the state of the economy.

The study rewards consistency, but is being consistent a good measure of a batsman though?

Would you prefer a batsman who scores 50, 50 & 50, or one who gets 0, 0 & 150? I don't have the answer - maybe the all or nothing batsman would win more games while the consistent batsman merely contributes and depends on his team mates for support? Or do you have to carry the second batsman in 2 matches out of 3?

To illustrate this - to players with quite similar records, which is the most effective?

The consistent batsman:
Stephen Fleming - 189 innings, 7172 runs at 40.06, 9 x 100, 46 x 50

The 'all or nothing' batsman:
Aravinda de Silva - 159 innings, 6361 runs at 42.97, 20 x 100, 22 x 50

Then there are the myriad other factors - the nature of pitches, quality of opposition, relative strength of your own team, scoring runs against lesser teams vs performing in pressure matches, length of career, etc.

To illustrate a couple of these points further lets compare Mike Hussey to Tamim Iqbal.

Length of career - I think to be considered a great you have to take into account how long they performed at the top level. It is easy for Jardine or Davis to sustain a high average over a few matches, much harder for the Tendulkar's and Waugh's who played for 20 years. Imagine if Mike Hussey had suffered a career ending injury in January 2008 - he would have played 20 Tests for a career batting average of 84.80. And i'm sure the statisticians would have put him second to Bradman. Someone with cricket knowledge though would know that a batsman not good enough to be selected for his country until the age of 30 (i.e. not considered within the best 6 batsmen in Australia at any one time), is not even worthy of the accolade 'great'.

Strength of team - Another thing in Hussey's favour is that he was playing for a team with a very strong batting line-up. He would never be fast-tracked into the team at 19 like a Tamim Iqbal. If Iqbal was Australian he would have only broken into the national XI in 2010. For arguments sake we assume that this transition does not affect his level of ability, the 19-20 year old Iqbal who averages 27.63 in 12 Tests was still good enough to play for Bangladesh. However if he was Australian he would have only broken into the team a couple of years later, because his rivals are better than that - and his Test Career would consist of 7 matches, averaging 59.78, rather than his Bangladesh record that includes both the novice player and the one who has come of age.

This demonstrates how batsmen for good teams will have a better average, as they are likely to only break into the team when they are close to their peak. The 25 year old Hussey was still learning his trade in the Domestic game, while the likes of Mark Waugh, Damien Martyn and Darren Lehmann played for Australia. If he had been born Bangladeshi with the same level of ability he would have played 60 more Tests, scored 4,000 more runs and averaged 5 runs less.

And that's not even to mention that the batsman in the good team often begins an innings at 200/2 against weary bowlers, compared to the struggling team who often are at 20/2 with the strike bowlers fresh and targeting them as the key man.

Gary 111 wrote:

This demonstrates how batsmen for good teams will have a better average, as they are likely to only break into the team when they are close to their peak. The 25 year old Hussey was still learning his trade in the Domestic game, while the likes of Mark Waugh, Damien Martyn and Darren Lehmann played for Australia. If he had been born Bangladeshi with the same level of ability he would have played 60 more Tests, scored 4,000 more runs and averaged 5 runs less.

Absolutely. Look at someone like Steve Waugh who got picked as a 19-year-old and kept his place in the side because Australia were crap at the time. Had he not broken through until 1993 he'd have an average (over the course of 120 tests) of about 57.

Had he then retired a couple of years earlier (while still at his peak) he'd have averaged over 60 (over the course of about 100 tests ie. a substantial sample size).

Stats can be helpful, but they need to be looked at in context and there's no magic mathematical forumla that will spit out a magic "context number".

That's why I find all these ratings a load of bunk. Ranking batsmen can only be done manually by a human. You can use the stats for reference, but you can't make them do the work for you.