Comparing batsmen: an objective approach (aka Who is a better batsman than who and by how much?)

[Arnab]

I have used the following steps to compare batsmen from different countries and eras.

1. Use an acceptable rating system. This was easy. Pricewaterhouse Coopers has an excellent ratings system for each player going back to the very first test ever played. It makes things easy to compare players of different era, since their ratings are calculated using the same criteria. These criteria include number of runs scored, opposition bowling strength, whether the match was high/low scoring, whether the team won the match or not, etc. I think the PWC ratings is the best ratings system we have.

2. A player's ratings change over the course of his playing life. So, to compare him with other players, we need to calculate his average rating. For this I added all the ratings for each test he has played so far and then divided the total by the number of tests.

For example: Lara has played 96 tests so far. After each test, his rating as a batsman was calculated and published by the PWC team. So there are 96 instances of ratings. These instances can easily be found in the form of a table from the PWC website. When I add them all up in Excel, I end up with a total of 73344. I divide that by 96 to get his average rating = 73344/96 = 764.

This average rating reflects the player’s rating over his entire career. Then I used this average rating to compare him with other players.

[Edited on 30-9-2003 by Arnab]

[Arnab]

Using the methodlogy listed above, I have come up with these numbers for different batsmen(in no order):

Player: Career Avg. Rating

1. Donald Bradman: 855.40

2. Viv Richards: 765.27

3. Brian Lara: 746.00

4. Sachin Tendulkar: 728.82

5. Matthew Hayden: 614.04

6. Ricky Ponting: 644.95

7. Steve Waugh: 666.14

8. Inzamam-ul Haq: 666.00

9. Adam Gilchrist: 738.33

10. Rahul Dravid: 713.58

11. Javed Miandad: 745.15

12. Sunil Gavaskar: 736.85

13. Alan Border: 738.30

14. Gary Sobers: 779.67

I don't how you do it, but this is how I would measure, as much objectively as possible, the greatness of different batsmen.

I will add more to the list. Or maybe Tintin can join.

[Edited on 30-9-2003 by Arnab]

[Arnab]

If you are curious to know where our own Habibul Bashar stands among all these greats, her ya go:


Habibul Bashar: 505.52

[Mahmood]

This rating system do not look good. There is no way Javed Miandad is as good as Lara and better than Tendulkar!

[Edited on 30-9-2003 by Rajputro]

[Arnab]

[Arnab]

Originally posted by Rajputro
This rating system do not look good. There is no way Javed Miandad is as good as Lara and better than Tendulkar!

Raj bhai, have you watched every game played by Javed Miandad? Have you considered the quality of bowlers he had to face? Have you considered how many matches he has won for Pakistan? Have you considered how he has scored runs in hig-scoring/low-scoring matches? All these and several more factors are taken into account in the ratings.

So yes, according to the above methodology, Javed Miandad IS a slightly better batsman than Tendulkar or almost as good as Lara.

One of my conclusions from these figures is that current players are hyped in the media too much. Most evidently Tendulkar.

[James90]

I don't like the positionings of Waugh and Miandad...Waugh is better than Border according to a baggygreen poll and I believe that. Gilchrist isn't that great...he's just a fiery batsman that's great to watch. Hayden is very consistant and has great hand/eye and power but his technique isn't so great now he has confidence. Still he consistantly hits 100s

Also I don't like the distance between Bradman and Sobers....it should be much much more.

[Edited on 1-10-2003 by Habibul_fan]

[Edited on 1-10-2003 by Habibul_fan]

[Arnab]

That's the point. You don't have to LIKE it or DISLIKE it. Because when you start doing that you're being subjective and biased.

The stats are based on hard facts on ground. The facts, i.e. the data extracted from the scorecards, themselves are objective and undeniable. Now, the methodology used by PWC to interpret these data is not 100% foolproof, but it has been seen to work well over quite a number of years. I think it's as close as we can get to objectively measure different batsmen. If you have a better methodology, feel free to write about it.

[Edited on 1-10-2003 by Arnab]

[Arnab]

I don't like the distance between Bradman and Sobers....it should be much much more.

What do you mean? There's a humongous difference of almost 75 points between Sobers and Bradman.

As for your other remarks, remember that we are comparing their performances as batsmen only. Not as leaders. Not their potentials. The ratings are measured using a definite number of factors and they reflect the bearings of just those factors.

A poll, and an internet poll at that, which is basically a collection of biased human opinions, may be indicative of public's subjectivity. But not a player's performance over a long time.

[Edited on 1-10-2003 by Arnab]

[James90]

You love PWC, eh

[Arnab]

I love good statistics in general, not just in sports, because it almost always cuts through people's bullshit opinions.

[Edited on 1-10-2003 by Arnab]

[asifr]

Arnab wrote about Miandad: "Have you considered the quality of bowlers he had to face? "

Miandad against West Indies 16 tests 834 runs Average of only 29. Miandad basically failed against the best bowling attack of that time. He is in the league of Dravid,Kallis,Hayden etc and not Lara and Tendulkar. Also I have heard he wasnt LBW in Pakistan for the first 9 years of his career. Maybe Tintin can confirm this.

[Arnab]

Meh. Lara and Tendul's average against Pakistani and South African bowling attacks are also around 30-31. What does that say?

[asifr]

Has South Africa been the best bowling attack of the 90s? I thought I made it clear that its the performance against the best bowling team that counts the most. Hence my mentioning of WI in relation to Miandad.

[asifr]

Also Lara avegraes 36 agaisnt SAF and Tendulkar 38. Thats isnt 30-31.
Miandad averaged 29.

[Arnab]

sigh OK how about their averages against PAK? :P

Or do you NOT consider pakistan the best bowling attack in the 90s?

Also, do you think there has been any bowling attack in the 90s that was as good as the Windies attack in the 80s? Probably not.

[Edited on 1-10-2003 by Arnab]

[asifr]

"do you NOT consider pakistan the best bowling attack in the 90s? "

Absolutely not. The one with the most variety yes..but not the best. It has to be Australia. Hands down.

"do you think there has been any bowling attack in the 90s that was as good as the Windies attack in the 80s? Probably not. "

Maybe not, but that doesnt give a player "better than Lara and Tendulkar" an excuse to crap against them. Gavaskar and Greg Chappell made runs against WI. Why couldnt Miandad?

[Arnab]

Ok think of the avg career ratings like this.

It's an OVERALL rating.

The opposition bowling strength IS considered. It is measured by the ratings of opposition bowlers.

The runs scored IS considered.

But, also, whether he made those runs in conditions where most of his peers failed to score decently IS considered and the ratings WERE adjusted accordingly.

And other factors too. A LOT more. All those play a role in the ratings.

Maybe when Javed Miandad played against WI, Pakistan as a whole also had a low scoring match. May be all his compatriot's scored even worse. Or may be the opposite happened. May be the pitches were all bowler friendly and produced a low scoring match. Maybe sometimes Javed failed while others did good. ALL these stats ARE taken into account in the rankings.

[asifr]

"But, also, whether he made those runs in conditions where most of his peers failed to score decently IS considered and the ratings WERE adjusted accordingly. "

What ratings? I just posted some cold hard facts. Also just because his peers failed doesnt give the excuse for Miandad to flop. That is what makes Lara such a great batsman..scoring runs when all others around him are flopping left,right and center. Thats what separates the good batsmen from the great ones.

"May be all his compatriot's scored even worse"

Majid Khan and Salim Malik are a few of his compatriots who performed much better against WI than him.

[Arnab]

Maybe not, but that doesnt give a player "better than Lara and Tendulkar" an excuse to crap against them. Gavaskar and Greg Chappell made runs against WI. Why couldnt Miandad?

I couldn't pass this up. Looks like I do have some time to respond to these.

1. According to the car avg ratings Javed is not better than Lara. But he is slightly better than Tendulkar.

2. If Lara and Sachin had to face the WI attack, chances are that they would have crapped against them too. Who knows? Who cares?

All these "what if"s and "but"s, "would have"s, etc. are irrelevant when we are trying to measure a player's overall career contribution and performance. All that matters is the fact on ground and how we intepret them.

Think about it.

[Edited on 1-10-2003 by Arnab]

[Arnab]

What ratings?

The PriceWaterHouse Coopers ratings. Did you even read my first post? I am beginning to get slightly irritated now.

[Edited on 1-10-2003 by Arnab]

[Arnab]

I couldn't help but dig out Javed Miandad's stats against WI.

Yes his average of 29 sucks. But he wasn't a total doozie...he has scored 4 fifties and 2 hundreds against them. Let's look at the circumstances of those fifties and hundreds and some other significant scores as well.

1. His first fifty against Windies was in Faisalabad, Second test, 8-12 December, 1980. In that innings, Pakistan scored a total of 176 runs in the first innings. Javed Miandad scored 50, one-third of his team score. An extremely valuable knock.

2. His second fifty against the Windies came in the next test in Karachi, 22-27 december. Again in the first innings, Pakistan was blown out by the devastating Windies bowling for a meagre 126 in the first innings, which included six ducks. Javed Miandad scored 60 runs in that innings, one-half of Pakistan's total.

3. Same series. 4th test in Multan, 30 Dec 1980 - 3 Jan 1981. Pakistan again allout for 166 in the first innings. Who is the top scorer? You guessed right: Javed Miandad. Scored a team highest 57, one-third of his team's contribution.

4. Lahore test 1987. Pakistan all out for 131 in the first innings. Javed scores team highest 46.

5. Karachi test 1987. Pakistan all out for 239 inthe first innings. Javed scores team highest 76.

6. Georgetown test, April 1988. Pakistan scores 435 in the first innings to take a match winning lead. Javed scores 114, the only century in the whole match.

7. Port of Spain test, 1988. Pakistan set a monumental target of 372 runs in the fourth innings to win. Pakistan scores 341/9 and the match is drawn. Top scorer inthe fourth innings? Javed Miandad with 102 runs. As long as he was on the crease Pak still had a chance to win.

8. Bridgetown test 1993. Javed in his twilight. Pakistan following on in the 2nd innings scores 262. Javed top scores with 42.

Asif wrote: That is what makes Lara such a great batsman..scoring runs when all others around him are flopping left,right and center. Thats what separates the good batsmen from the great ones.

As you can see, my friend, the same is true for Miandad against Windies, too.

-------------------------

The opposite is also true. In other matches, he didn't score many runs while Majid Khan, Wasim Raja, Saleem Malik, Inzamam, etc. scored.

I hate to sound repetitious but: All these ARE taken care of in the ratings.


[Edited on 1-10-2003 by Arnab]

[Edited on 1-10-2003 by Arnab]

[asifr]

I must commend myself for the patience I am displaying. Ok here goes.

Arnab wrote:"2. If Lara and Sachin had to face the WI attack, chances are that they would have crapped against them too. Who knows? Who cares? "

Firstly,what makes you think that the chances were that they would have crapped against that attack as badly as Miandad did? Gavaskar and Chappell certainly didnt. They were great bats,who are in the same category as Lara and Sachin. Miandad as I have said before belongs in the Dravid,Kallis,Hayden category. BTW the WI bowling average over Miandad's playing days was 27.7

The Australian bowling attack average since Tendulkar made his debut is also 27.7. So one can say that these two attacks are comparable. You may find that to be blasphemy of the highest order but I dont. Sure the WI pace quartet was fearsome etc etc but so is the Aussie bowling juggernaut. And against this attack Tendulkar and Lara have been averaging over 50 each. Something which Chappell and Gavaskar did against the WI attack and Miandad barely avged above half of what they were avging against WI.


" As you can see, my friend, the same is true for Miandad against Windies, too."

So you include his 46's and 42's as well to show that? The fact is that Miandad averaged 29 against the WI's mate. When BCL scores, he scores big to help his team and not just 30s and 40s.
Miandad failed against the WI big time so how can anyone call him a great? To be great, you gotta do well against the best, something which Javed did not.

As far as the ratings go, I am not talking about them. I dont have much faith in PWC or any other ratings. It is very difficult/nearly impossible to come up with a near perfect rating system in cricket which is fair to every batsman. A BCL 222 against SL was rated lower than a Richardson 94 against SL by PWC. Go figure.

[Arnab]

A BCL 222 against SL was rated lower than a Richardson 94 against SL by PWC.

Do you know why? Have you asked them why? Have you heard anything called "scaling by factors"? If you're unsure what it means, then email the PWC team. If you can't understand why they were rated like that then I can't continue this conversation. Because you have to understand how the ratings work. Otherwise, these debates will always be pointless. After you throughly read and understand how the ratings work, I am more than sure you will look at your own posts in this thread and will be able observe where your thinking is flawed.

I CAN write humongous, detailed, point-by-point replies to all the issues(mostly redundant) you raised in your last reply, but they will all be in vain since you don't even know how the ratings work. I can't waste my time like that.

[Edited on 2-10-2003 by Arnab]

[asifr]

Its a pity that you need a ratings system to formulate your opinions. And nice sidestepping to the facts I presented in my 3:17 post. I will take that as a compliment. Goodday