Lovely articles yogesh....I was shocked when i got to know of this report from a popular Indian Daily.
I hope the authors at HW do read your piece which is full proof.

Bad analysis and bad reasoning. Thats the first thing that comes to mind. You are talking only about instances where Tendulkar excelled and India lost. Your argument is that india lost because other players werent good enough. Surely such a reasoning isnt the best to indicate Tendulkar was great. Nor is it good enough to indicate statistics are flawed. What if someone else had performed well, and India still lost? And what about the opposite case where India lost because Sachin didnt perform well? According to your reasoning, will not Tendulkar's stock take a dive in this case?

1. Your sample size is too low.
2. Your sample is not the best fit for a normal distribution.
3. You must have heard of the statistical theorem 'The sum of a variance is equal to the variances of its sum'. Translated, it means the team's performance with its variabilities is equal to the sum of every individual player's variability. It cannot be less than or greater than it (as you have claimed).

Cheers!

Not a good one. You can compare Lara with Sachin and no one else. Ponting, waugh, Dravid are very good batsman but not as great as Brian or SAchin.

Your article is true, but to any "cricket" fan with even basic cricketing knowledge, it is very obvious.

If you take team "results" both Tendulkar and Lara would fare way down among top batsmen in their respective countries.At their peaks they were mostly fighting single handedly. The dravids,gangulys,sehwags etc flowered after the late 90s.And what about the obscenely huge ODI burden on Tendulkar?

For eg. a Richards 50 may have been in a match which WI won, and a big Lara hundred in a losing cause.
How in the world that makes one batsman superior to another is beyond me or perhaps anyone (except perhaps the statisticians,who one is beginning to doubt have ever even held a bat).

If you put even say a Maradona in a very poor team that team will lose most of the time even if Maradona manages to score a goal just about everytime.
So then as per our knowledgeable "statisticians" Maradona wouldnt even qualify as "great"...
I could go on ,but most sports fans will get the drift.

Well, if I were an Indian fan I would certainly hope that Sehwag scores a century because that would presage a victory. The indian team improved considerably under Ganguly and John Wright, it was only then that India managed to draw or win Test Series' away from home. I think where Tendulkar gets a bad rep is that he fails in the second innings when India are chasing a target. it is unfair I suppose, one shouldn't be expected to score a 100 in both innings, or even 50. Cheers.

To Bikal, I agree partially with your analysis. However, it is common to have outliers in statistical data. Those outliers may indicate that the rest of the team did not perform well when Tendulkar is performing well. Or collective performance of Indian team is not enough to counter strong performance by opposition despite brilliance performance by couple of individuals.

you, mate, have hit the nail on the head. is it sachin's fault that his knocks in the 90's that dragged india to decent totals were not defended by less than ordinary bowling. are only chasing knocks 'match-winning'. the laras and the pontings have had great bowlers to back their totals, and sachin for a large part of his career hasn't been fortunate enough to have that luxury. And more ofen than not, he walked in with the team reeling at 10 odd runs for 2, with the weight of expectations on him. read my lips.. greatest batsman ever - sachin ramesh tendulkar.

Bikal,

Which sample size are you talking about ? No: os losing centuries or the players ? Why should sample be the best fit for Normal distribution ?
And the statistical theorem you mention assumes that the variables be uncorrelated. And this is clearly not true here. Runs made by players are correlated.

But i still don't know why these factors play a role in my analysis. All i have done is taken 5-6 best players in the last 25 years, compared their statistics in matches won as Holding Willey and many others have done. Solely relying on this statistic, the conclusion is Dravid is way ahead of Lara. And to whosoever thinks that this is not true, the statistic cannot be considered as a performance measure.

Very nice article. These were my thoughts too when I saw that dumb report. I think this report was not as lopsided as the ICC rankings but it comes close. There is just way too much emphasis on winning though victory is not a one man's effort. In my opinion Sachin and Lara deserve more praise as they have scored when all others failed. They brought some dignity to the team's score unlike the recent England collapse. Its unfair on Sachin really. For me he was, is and will be the greatest player. Lara too :-)

This is the bane of statistics - it can be twisted in any manner to make one's point. But all would agree that Bradman was the greatest ever, statistically and classically. And his word was that Sachin was the closest batsman to him - he said, "Sachin reminds me of me". This is enough of a complement for me. In fact, if you ask Sachin himself, he will probably say that that complement means more to him than any of the statistical analysis of his greatness! So there!

LOL Yogesh, the problem here is that you say statistics prove that Waugh, Ponting and Dravid were better than Lara, AND becuase somebody subjectively believe that this shouldn't be the case thats why statistics is wrong and doesn't prove anything.

Can you see the absurdity of your argument, certainly I can

Besides HW article was not talking about the most talented which Sachin is undoubtedly but was about greatest Indian cricketer.The matches Dravid had major contributions are also involving Tendulkar, and that is a LEVEL PLAYING FIELD.HW article had written down parameter, so counter the methodology not the results..Just that the results are not acceptable by you doesn't mean they are wrong.And also comparisons should be cricketers of same era and same team when comparing on HW parameters

Please go ahead and publish your framework and please don't feel hurt..

Amit Varma wrote on July 4, 2006 on cricinfo that "In terms of getting his side results, Dravid is India's greatest batsme

@VJ Ashleigh Mallet swore blue until the end that Trumper was better than Bradman.

@ DVC, Was it ashley mallett ? 'cos he played way after these two guys. But yes, there have been lot of players who've considered Archie Jackson/ Stan McCabe/ Trumper to be better than Don.

@ Shalabh,
You are right. I should have a few things more clearer.

1) Whomsoever thinks Dravid/Waugh/Ponting aren't way ahead of Lara, this statistics is simply not the right way to compare players.

2) I said ability is not quantifiable. Thats what is my main grudge more than tendulkar being 4th or anything else.. But may be it has been lost in my writing !! its my mistake.

3) I've accepted friends who've told that these are the criteria i look for in a batsman & so X/Y/Z is better than Sachin. I'm perfectly fine with it. But what i am trying to criticise is the theories that so & so statistics characterize match-winning ability, greatness etc. They are not simple things characterizable by a few statistics. They are much complicated things and my rant is against quantification of such qualities. By their very name, qualities can't be quantified.

You start with subjectivity & end with a subjective opinion of Amit Verma :)

@shalabh,rohan,VJ..
You've missed the point.
If Bradman himself had been in the IND team instead of tendulkar in the 90s, or in the WI team in place of Lara..he would still have ended up on the losing side routinely.
That would then place Ponting,Hayden and co. higher than Bradman as per the current "methodology".

Further if you simply insert say Akram and Waqar in the 90s indian team, then tendulkars runs would have resulted in more "team wins"..so pushing his rank up...Effectively,a case of the bowlers influencing the batsman's ranking.

And what "level playing field" are you talking about?tendulkar has been playing since 1989.as "Rohan" mentions the sehwags,dravids etc flowered after the late 90s: when india was already a decent team and coincided with a continous string of injuries for tendulkar.

If the bowlers cant get the 20 wickets ,you cant win: Period.

as "rohan" says:if you had put Maradona in the IND football team theyd still lose all the time to ANY top team!

Yogesh- Nicely done.

Bikal,
Cheers! but, come on! You may have the urge to put it down but your fitting of the concept (the theorem),I think, is "perfectly" wrong! Here, Yogesh is comparing the players. Also, I suspect this is not the first time you hide under the sample size shield :)- but they are playing cricket all their permitting life -this is it! not many more samples in this life. You could be less sure your points and be wiser.

I feel that there should be only 3 criteria in rating a batsman :
1)Runs scored (Aggregate and average) against
2)Quality of opposition (the value of runs scored against quality bowlers should carry more weight.Again the bowlers need to be rated not with a "fixed" value,but during various stages of their careers.i.e higher rating at their peaks).
3)Quality of pitch: Obviously scoring runs on a dead track doesnt really reflect on the "quality" of a batsman.More often than not it reflects on the "ruthlessness" of a batsman and his penchant for avoiding boredom .When the bowlers are practically helpless on dead flat tracks, then the onus is on the batsman to get himself out.
Runs scored on "juicier" pitches should be given more weigtage.

So,basically its the runs scored ,against the quality of bowling ,and type of pitches.Thats ALL.

Again ,the result will be a little vague.but then all stat based results are.

But using team results to rate an individual player is ridiculous.

Rohan, you probably missed the team situation. But the reason one cannot rely a lot on statistics is that these four factors can never be satisfiably measured statistically. Simple eg :
Gambhir made a double century against an attack with Brett Lee in Delhi. Tendulkar & Laxman made against an attack with Brett Lee in MCG & SCG in 1999. Those were the first two matches of Lee & he was steaming in & blasting out batsmen. Whereas in Delhi, he was decent without threatening. No insults to Gambhir. But to me those two centuries will rank much higher. This can never be quantified in statistics. Thats why one can never give a definitive statistical definition of a great batsmen.

@ Rohan, rating against the quality of opposition can't be done fairly. To say a player is not as good as another because they didn't face a particular bowler is erroneous and unfair to players in teams with strong bowling attacks. And if you are judging by the opposition, you would also have to include bat technology, size of the ground they scored their runs on (small in WI, big in Aus) and a host of other determining factors. Averaging performances over a sustained period is perhaps not the most accurate but it's worked so far. No one (sanely) argues that Bradman isn't the greatest, nor do they argue the talents of Richards, Hammond, Lara etc. An average over 50 over a long period determines a batsmens place in the upper echelons, everything else is just debate. And do we REALLY want to know who is the "best" ever? To know once and for all takes all the fun out of it. Why do we never question who the best bowler is? But using team results to rate a player is indeed ridiculous!

Yogesh
I fully agree with you!! Perhaps i am not being clear.
Donald in the mid 90s made lara look like a club batsman.But in the 2003 WC lara(and practicaly everyone else)smashed a pedestrian Donald.
The bowler may be same,but the "bowling"will be different at diff stages of a bowlers career.

A good recent eg. would be Ponting's 101 and 99 against SA in Mel. Ponting batted very well ,probably as well as he ever has.
Aus ended up losing the match.
A couple of years back with Warne and Mcgrath they may well have won a similar match.
But either way, whether Aus win or lose ,that should not take away from Ponting's batting performance/ability.
That is what i am trying to get across.
In some combined 280 test matches and 500 innings both Tendulkar and Lara have played one so called "match winning" innings each(the recent tendulkar 103, and the lara 153).

But how does that reflect on their true batting abilities? One brilliant individual in a poor team can only do so much.

To me,personnally,it is daft to compare Tendulkar with the Sehwags,Dravids and Laxmans.
Sure,they have had their moments.
However,no batsman has so completely captured a generation as completely as Tendulkar.
Of course,that alone is not a criteria for "greatness"...
But combine it with his batting purity,lack of weaknesses,the sheer "genius" of his strokes,his balance,the incredible pressure he has sustained for almost two decades,his performances in both ODIs and tests(yeah,yeah i know tests are the "ultimate"
,but no batsman worth his salt in the modern era would sit out ODIs at his peak.It would be taken as a "weakness".),his consistency,the way he has carried and inspired his team and teammates( players have all bloomed under his selfless support and guidance)......etc etc etc etc
And incredibly you get a once in a blue moon player who simply transends statistics!!
Using numbers to quantify the Tendulkar effect and his greatness is almost insulting to the "Little Master"!

Even Tendulkar's career reflects this : Before 2001, India won 17 of the 73 matches he played. He averaged 59+ in matches won and overall 57. Only 4 out of his 22 centuries were in matches won. Post 2001 - India has won 24 out of 69 matches he played. He has averaged 54+ in matches won and 47 overall. 7 out of his 13 centuries have been in matches won in this period.

So much so for number of centuries in victories !!

PS : Matches against Bangladesh/Zimbabwe have been excluded.

Yes, I partly agree. While match-winning ability is an important factor, I also believe that a batsman should not be so heavily penalised for runs scored in losses due to the mediocrity around him.

nice to see your analysis.but i think you never think the team. if ponting and hayden got the team with the bowlers like mcgrath and warne, lara with ambrose and walsh, sehbag with khan and sharma, ok convince us that who was with sachin? This is really poor analysis. if like this analysis could get place in cricinfo i will promice every one i will never try to open this site
cheers
sam from australia

nice to see your analysis.but i think you never think the team. if ponting and hayden got the team with the bowlers like mcgrath and warne, lara with ambrose and walsh, sehbag with khan and sharma, ok convince us that who was with sachin? This is really poor analysis. if like this analysis could get place in cricinfo i will promice every one i will never try to open this site
cheers
sam from australia

i fully agree with u yogesh..seems like you did something which any responsible cricket-lover should do.. and as for those who criticise Sachin..go and bang your head against the wall!! maybe that will bring some sense into you!!

talking of sheer ability ..
1.ask abdul qadir .. how he felt on the day of sachin's ODI debut [which got rained-off]... for those who dont know that is actually the fastest ODI 50 ever..53 off 16 including 28 of a single abdul qadir over .. it was probably 6,0,4,6,6,6..dn remember exactly!!

2.do any of those who call themselves as "cricket" fans - and still criticise sachin - pay any heed to Sir Don Bradman??

Dravid's match-winning knocks which really helped India win were the ones in Adelaide [Laxman got a 100 and agarkar too 6/41.. otherwise India were 85/4 at tea on day II.. so dravid alone didnt win us that game]!

At Headingley Dravid batted on a green top to get 148 [bangar made 68 off 234 .. remember?? kumble got 8 wkts in the match ..sachin got 92 and sourav 99 .. team-effort??]

In the west Indies in 2005, dravid won the test single-handed [no sachin, no sourav in that team]..dravid got 81 & 68 ..lara got single-digit in both innings!!

does this make sense??

Dear sir u invest lot for IPL make little more by investing on Grahm Napeir and Kiror pollard Plzz

Most of the arguments supporting the author are emotional and driven by the person's prejudices. Yogesh - Your grouse seems to be the fact that HW rated Tendulkar 4th. They have publsihed a methodology and come up with a result. You are free to come up with your own methodology which can then push Tendulkar to 1. Bikal certainly does have a point in all his posts. Secondly, your analysis neither proves or disproves HW or their approach. Whats the point being made here?

We should count the number of matches a batsman saved along with the wins. I think, in both cases, they should be multiplied by (1/team's-bowling-ability).
I hope these two additions will make things better

For all his "greatness" gavaskar and sachin come from the same school of CYA. Records matter. Statistics are prime!

Please do not bring in Brian Lara here for comparison. All these statistics would have gone up if he had a better team like Australia or India. Man of the match says that that particular person is the best performer of that match but dont forget the fact that only that person is not winning you the match. It's after all the team effort. Brian Lara was not having as good team like him to have these statistics go up

Very good article indeed!! It isn't a mistake of a genius that he did not have the bowling attack in his team to take 20 wickets when he was at his prime in the 90s.
But its not for nothing that he's called a genius cos we are in 2009 now and he is still arguably one of the best in the world!!
If that doesn't speak volumes i will never know what does... :)

That is what i have been saying all through my life! Whenever sachin plays no other player plays well.

@sharath
So its taken you all your life to realise how great tendulkar is.
He has scored some 80 plus international hundreds,140 fifties and getting to 30000 runs.
So that means everytime he has scored some fifty plus runs noone else has played...
i.e Some 220 TIMES!!
Some team: Team India,ha?

@Vanchy, My Grousse is not Sachin is rated at 4. It is rather saying that the methodology is "the" way to evaluate greatness and more generally against using statistics as "the" tool to order players. And as i've said it in some comments earlier too that whosoever thinks that Dravid/Waugh/Ponting isn't way ahead of Lara, the parameters are not good enough. And there are lot of people who would think so. Bikal might have a point, but i don't see it. It seems like many people he to is a bit obsessesed with Normal distribution and statistics isn't just about Normal distribution !!! And the best player is always based on emotions he stirs up in you.