• Welcome to the Cricket Web forums, one of the biggest forums in the world dedicated to cricket.

    You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, respond to polls, upload content and access many other special features. Registration is fast, simple and absolutely free so please, join the Cricket Web community today!

    If you have any problems with the registration process or your account login, please contact us.

Garry Sobers v Imran Khan,Test Cricket:Poll

Who was the better Test cricketer: Imran or Sobers?


  • Total voters
    169

bagapath

International Captain
No. Its really not. With respect to bagapath if he is not aware, and to yourself if you were not being facetious, its a terribly fallacious post. That is why Flem and I independently changed the team selection.

Which is what you have done yourself.

If you have a team with five ATG batsmen and four ATG bowlers and one wicket keeper/batsman, who would you take in as your last player, Sobers or Imran?
 

smash84

The Tiger King
If you have a team with five ATG batsmen and four ATG bowlers and one wicket keeper/batsman, who would you take in as your last player, Sobers or Imran?
You can flip this around and say. If you have 6 ATG batsman on wk/bat and 3 ATG bowlers. Who will you pick? That's not a great question.

I think its fair to say that of all batsmen I'd probably think of putting down Sobers's name after Bradman. But that is not the case with Imran in bowling.

So Sobers's stronger suit is arguably stronger than Imran's.

Then you can compare their secondary skill and compare them.

Then fielding etc are summed up.

will probably give you a better answer
 
Last edited:
If you have a team with five ATG batsmen and four ATG bowlers and one wicket keeper/batsman, who would you take in as your last player, Sobers or Imran?
Yeah. This is still fallacious. But I would probably have 6 Sobers 7 Sangakarra 8 Khan 9 Hadlee 10 Marshall 11 Murali,

giving me 7 atg batsmen and 4 atg bowlers. Khan is an atg bowler. Sobers could be replaced by Kallis. Heck given that bowling line up, if Sobers gets injured I may just replace him with a specialist batsman. But Khan, he's far more valuable.

Given a 'salary cap' or some scenario where it was a matter of choosing a marquee player, or betting on those one wicket tournaments that Rice, Hadlee, Khan, Botham and Dev used to play, I'd be betting on Khan to beat Sobers. Of course. He is the better cricketer.
 
Last edited:

bagapath

International Captain
well... I agree with you smali...

if imran had averaged just, say, fifteen runs with the bat, he would still be considered a great cricketer. and ranked on par with ambrose and truman. but lillee and marshall could be rated ahead of him. perhaps, mcgrath and hadlee as well.

if sobers had not taken a single wicket, he would still be considered second only to bradman. he would still have been bigger than sachin, hobbs and viv richards.

so yeah, sobers the batsman is marginally > imran the bowler
sobers the bowler is also marginally > or at worst = imran the batsman (I've personally never been a big fan of imran's batting. always rated both and kapil as better bats - not just potentially, but in terms of actually delivering)
sobers the fielder is a class apart anyways...
so yeah.... sobers will probably be the package any cricket team would dream to possess.
 
Last edited:
well... I agree with you smali...

if imran had averaged just, say, fifteen runs with the bat, he would still be considered a great cricketer. and ranked on par with ambrose and truman. but lillee and marshall could be rated ahead of him. perhaps, mcgrath and hadlee as well.

if sobers had not taken a single wicket, he would still be considered second only to bradman. he would still have been bigger than sachin, hobbs and viv richards.

so yeah, sobers the batsman is marginally > imran the bowler
sobers the bowler is also marginally > or at worst = imran the batsman
sobers the fielder is a class apart anyways...
so yeah.... sobers will probably be the package any cricket team would dream to possess.
Your "analysis" states that Sobers the batsmen is "marginally" greater than Imran the bowler. What is your criteria? Because average runs per wickets suggests you are quite wrong. Sobers the batsmen is behind Imran the bowler.
 
Last edited:

smash84

The Tiger King
Your "analysis" states that Sobers the batsmen is "marginally" greater than Imran the bowler. What is your criteria? Because average runs per wickets suggests you are quite wrong.
Grumpy, I think your approach is a little too stats oriented.

Quite misses out on the context of the runs and wickets.

My point is that if you haven't seen Sobers (and Imran) play and you don't know the context of their wickets and runs, a statistical analysis is not a very meaningful route to take.

I do various stats analyses for a living (from time to time) and the stats are not very useful if they don't fit in with sensible, logical, and coherent theory from the subject area.

Quite like that you can have your foot in boiling water and the other foot in freezing water, and on average you are quite comfortable.
 
Last edited:

bagapath

International Captain
Your "analysis" states that Sobers the batsmen is "marginally" greater than Imran the bowler. What is your criteria? Because average runs per wickets suggests you are quite wrong. Sobers the batsmen is behind Imran the bowler.
don't you ever use " " at me. show some respect for my diabetes.
 
Grumpy, I think your approach is a little too stats oriented.

Quite misses out on the context of the runs and wickets.
The context is averaged out. Day 1 and Day 5 conditions, good bowlers, bad bowlers, wins, losses, draws. The longer the career or statistical pool, the more accurate it becomes. That is statistics. Its the players worth above average in the eras that they played in.

People always question statistics, math, logic when it upsets what they thought was "common sense" or a "truism.
 
Last edited:
don't you ever use " " at me. show some respect for my diabetes.
I had no idea that you had diabetes. But your analysis was rather weak. You sort of said well Imran was not as good as a bowler as "lillee and marshall could be rated ahead of him perhaps, mcgrath and hadlee as well" where it is arguable and minute. Then said Sobers is the second best batsmen to Bradman (that premise alone is arguable) and then concluded from that that Sobers as a batsman > Imran the bowler which is an illogical conclusion to make, unless you meant only in terms of your questionable ranking. But its only a ranking. Its not a measure like runs per wicket average is.
 
Last edited:

smash84

The Tiger King
The context is averaged out. Day 1 and Day 5 conditions, good bowlers, bad bowlers, wins, losses, draws. The longer the career or statistical pool, the more accurate it becomes. That is statistics. Its the players worth above average in the eras that they played in.

People always question statistics, math, logic when it upsets what they thought was "common sense" or a "truism.
sorry, i edited that post.

My point is that if you haven't seen Sobers (and Imran) play and you don't know the context of their wickets and runs, a statistical analysis is not a very meaningful route to take.

I do various stats analyses for a living (from time to time) and the stats are not very useful if they don't fit in with sensible, logical, and coherent theory from the subject area.

Quite like that you can have your foot in boiling water and the other foot in freezing water, and on average you are quite comfortable.
 
sorry, i edited that post.

My point is that if you haven't seen Sobers (and Imran) play and you don't know the context of their wickets and runs, a statistical analysis is not a very meaningful route to take.

I do various stats analyses for a living (from time to time) and the stats are not very useful if they don't fit in with sensible, logical, and coherent theory from the subject area.

Quite like that you can have your foot in boiling water and the other foot in freezing water, and on average you are quite comfortable.
With all due respect, you are illogical and your sophistry is crude and incredibly fallacious with regard to your feet.

The stats may favorably support a theory, or disfavorably suggest that the theory is in fact incoherent or not a likely explanation of what the theory purports to explain. You don't ignore the stats because they don't prove your "accepted theory", unless the sample is too small. You might rethink and examine your formulation of the statistic, because there could be human error. Then you need to question the theory, and see what theory best explains the stats, because that is the theory that the stats are supporting. You have this all wrong. I am not sure who you work for or what you do for them, but your approach seems not right.

A statistical analysis of runs per wickets on average for their eras does ignore having seen the players play. But it 100% demonstrates their below or above par performances for the eras that they played in over their career. Yes, it will say that over Botham's career he was behind Khan, when in 1981 he was special. But we know in 1981 that he was. That is not the issue. The issue is a player's worth above "par" of all players over a long career in their respective eras. Otherwise you would assess players solely on their peak one off performance, or series performance. But we know that that is not the best indicator, and that is why we use statistics and averages to compare players. To demonstrate their consistency and value over a period of time, not their peak potential. A side might play weaker sides more than the stronger teams, but that is less of a concern for Imran and Sobers, than it for comparing current Zimbabwe players to Australian players. But the negative impacts of this should be averaged out to a certain degree.

A long career will face good bowlers and bad, bowl good batsmen and bad, lucky runs and hard runs, lucky wickets and unfortunate missed wickets. Statistics elucidate. They becomes more accurate with a greater sample - that is a longer career. The differences between eras is standardized out for a lot of it. Runs per wickets is sensible. It makes far more sense than the difference between batting average and bowling average, which is commonly used by many cricket journalists. Or do you agree with that?

On a runs per wicket average for their actual runs and innings, and their actual wickets for their actual runs conceded then standardised against the era averages, Bradman is scoring at 95 runs ahead in terms of value in a match than a batsman who averages 36 runs per innings in the 1980's. Seems right to me. Seems coherent. Makes sense. I did not watch Bradman bat.

Anyway, I was looking for positive feedback on the improvement of the statistical formulation and critical assessment without massive pouring in of additional player statistics and data from those who understand and accept statistics, not an essay on subjectivity of how runs were scored or wickets taken. As if that does not average out over the sample. Nothing is perfect with statistics. We all know and accept that. This is not about perfection. The eras alone is to further average out imperfections that can never be made perfect, but merely improved. It is about debunking a myth that Sobers was the greatest or most valuable all round cricketer. He's not. Bradman is on this analysis. Khan is also a more valuable cricketer than Sobers on this analysis. Which is my point.
 
Last edited:

smash84

The Tiger King
With all due respect, you are illogical and your sophistry is crude with regard to your feet.

The stats may favorably support a theory, or disfavorably suggest that the theory is in fact incoherent or not a likely explanation of what the theory purports to explain. You don't ignore the stats because they don't prove your "accepted theory", unless the sample is too small. You might rethink and examine your formulation of the statistic, because there could be human error. Then you need to question the theory, and see what theory explains the stats. You have this all wrong. I am not sure who you work for or what you do for them, but your approach seems not right.

A statistical analysis of runs per wickets on average for their eras does ignore having seen the players play. But it 100% demonstrates their below or above par performances for the eras that they played in over their career. Yes, it will say that over Botham's career he was behind Khan, when in 1981 he was special. But we know in 1981 that he was. That is not the issue. The issue is worth more over a long career. Otherwise you would assess players solely on their peak one off performance, or series performance. But we know that that is not the best indicator, and that is why we use statistics and averages to compare players. To demonstrate their consistency and value over a period of time, not their peak potential. A side might play weaker sides more than the stronger teams, but that is less of a concern for Imran and Sobers, than it for comparing current Zimbabwe players to Australian players. But the negative impacts of this should be averaged out to a certain degree.

A long career will face good bowlers and bad, bowl good batsmen and bad, lucky runs and hard runs, lucky wickets and unfortunate missed wickets. Statistics elucidate. They becomes more accurate with a greater sample - that is a longer career. The differences between eras is standardized out for a lot of it. Runs per wickets is sensible. It makes far more sense than the difference between batting average and bowling average, which is commonly used by many cricket journalists. Or do you agree with that?

On a runs per wicket average for their actual runs and innings, and their actual wickets for their actual runs conceded then standardised against the era averages, Bradman is scoring at 95 runs ahead in terms of value in a match than a batsman who averages 36 runs per innings in the 1980's. Seems right to me. Seems coherent. Makes sense. I did not watch Bradman bat.

Anyway, I was looking for positive feedback on the improvement of the statistical formulation and critical assessment without massive pouring in of additional player statistics and data from those who understand and accept statistics, not an essay on subjectivity of how runs were scored or wickets taken. As if that does not average out over the sample. Nothing is perfect with statistics. We all know and accept that. This is not about perfection. It is about debunking a myth that Sobers was the greatest or most valuable all round cricketer. He's not. Bradman is on this analysis. Khan is also a more valuable cricketer than Sobers on this analysis. Which is my point.
With all due respect. Either you are deliberately trolling or are extremely stupid. With that out of the way I'll now use your strategy of editing my post over the next one hour and posting a complete reply :p
 
Last edited:
With all due respect. Either you are deliberately trolling or are extremely stupid. With that out of the way I'll now use your strategy of editing my post over the next one hour and posting a complete reply :p
With all due respect, again you are wrong. I could be extremely stupid and deliberately trolling. The concepts are not mutually exclusive, nor are they binary as you have also limited your options incorrectly.

It could be just that you do not understand it, that I am not stupid and not trolling. If something needs further explanation, I will provide you with it if you ask. Or many further possibilities.
 
Last edited:

smash84

The Tiger King
With all due respect, again you are wrong. I could be extremely stupid and deliberately trolling.
You're right. This is quite plausible :p

The problem with you is that you let the data do the thinking whereas you should be the one doing the thinking.

Cricket is not something "new" or "creative" that you are trying to do in order to come up with novel ideas. You are not doing an exploratory data analysis. A lot of cricket is a matter of historical record that quite a few people have been around to witness. Those people have witnessed the pitch conditions, the weather conditions, the quality of batsmanship/bowling at play and have arrived at certain conclusions.

You armed with just the final output i.e. the scorecard have absolutely 0 basis of awarding points for quality. Perhaps you just take Samaraweer/Mohd Youuf's averages and compare them to Viv Richards and reach the conclusion that "Oh, Viv with that average of 50 doesn't match these blokes." Or that Javed Miandad (as good as he was) was a better batsman than Viv in that same generation since his final average is higher than Viv's. Those who witnessed Viv play (and the contemporaries who played against him) can put those averages in context. Excel warriors can capture ALL the numbers but NOTHING that actually fed into those numbers. Which is why most of your analyses are not just stupidly simplistic they are an insult to people who actually take the time to follow the game and appreciate the nuances it has to offer.

You talk of samples sizes yet are ignorant of the basic facts around those samples. The grounds, the wickets, the opposition, the match situation were all very different for Imran and for Sobers which go a long way in determining each character's part in their respective teams' successes. Your jargon doesn't control for those factors and neither does your thinking (which definitely needs a thorough workout).
 
Last edited:

Red

The normal awards that everyone else has
If you have a team with five ATG batsmen and four ATG bowlers and one wicket keeper/batsman, who would you take in as your last player, Sobers or Imran?
Taco girl gif

Hutton
Gavaskar
Bradman
Tendulkar
Richards
Sobers
Gilchrist
Imran
Hadlee
Marshall
Warne
 

Prince EWS

Global Moderator
Hahaha. Ok.

If there is one last spot in a team, who would you take, niazi or soby?
Well it obviously depends on the composition of the rest of the team; that's the point we've been trying to make to you.

I think if you took a cross section of every team in the history of cricket, more of them would benefit more from replacing a player with Imran than replacing a player with Sobers, which is why I rate Imran higher. However, stacking the deck with a certain team composition with an obvious requirement as you did with your ten player team doesn't really help to answer the question.
 
You're right. This is quite plausible :p

The problem with you is that you let the data do the thinking whereas you should be the one doing the thinking.
No. That is not the problem. The ideal would be to put as much data in that allows for incredibly minor nuances, but who has the the resources or the inclination?

Cricket is not something "new" or "creative" that you are trying to do in order to come up with novel ideas. You are not doing an exploratory data analysis.
I seriously hope this analysis is not novel. So what is your point that I am not the first explorer? I have determined this method independently, but I freely accept that someone else may have done it first.

A lot of cricket is a matter of historical record that quite a few people have been around to witness. Those people have witnessed the pitch conditions, the weather conditions, the quality of batsmanship/bowling at play and have arrived at certain conclusions.
Without making an objective data output, this remains as subjective information. It is not surprising, that a batsman like Bradman averages 99.94 from those that watched him play regularly. Or that Sachin Tednulkar averaged over 50 whereas Chris Martin averaged 2.5. I am not asking for a statistical formula to tell me how beautiful a rose looks or how sweet it smells. Please try and get this into your head. This is results based averages. Things average out in careers. Good balls, bad balls, lucky dismissals, unlucky dismissals, good pitches, bad pitches. They will never average out perfectly. But where there is a gulf between the outputs, that is a useful comparison tool. There is a gulf between Tendulkar and Stepehn Flemming as batsmen's averages. Well subjectively there was for their batting techinque and success. Flemming may have had slightly less, the same or more talent than Tendulkar, but he did not realise it as often. That is what sports statistic measure. Not talent. Results.

You armed with just the final output i.e. the scorecard have absolutely 0 basis of awarding points for quality. Perhaps you just take Samaraweer/Mohd Youuf's averages and compare them to Viv Richards and reach the conclusion that "Oh, Viv with that average of 50 doesn't match these blokes." Or that Javed Miandad (as good as he was) was a better batsman than Viv in that same generation since his final average is higher than Viv's. Those who witnessed Viv play (and the contemporaries who played against him) can put those averages in context. Excel warriors can capture ALL the numbers but NOTHING that actually fed into those numbers. Which is why most of your analyses are not just stupidly simplistic they are an insult to people who actually take the time to follow the game and appreciate the nuances it has to offer.
With all due respect, you havn't read my posts or you have not understood them. The batsmen are standardised against the averages in their own eras meaning Youseff will come down in value as against Viv, "55 is the new 50". Statistics are not perfect, but things average out. Statistics are a valuable tool of analysis, you can reject them as much as you like. But for the enlightened, people accept their value.

You talk of samples sizes yet are ignorant of the basic facts around those samples. The grounds, the wickets, the opposition, the match situation were all very different for Imran and for Sobers which go a long way in determining each character's part in their respective teams' successes. Your jargon doesn't control for those factors and neither does your thinking (which definitely needs a thorough workout).
Again, with all due respect, things average out over a long career, and the players are standardised against their eras. Sports statistics are not perfect, but grounds, wickets, match situations average out. Both players would have had to bat time on a 5th day to save a test in their career on a turner. Or bowl on the 5th day to bowl the opposition out. Or played in a boring draw heading into day 5. Again, we all know statistics are not perfect, but things average out for comparitive tool of how above or below par players were for their eras, and then to compare players of different eras.

At the end of his career, Imran was worth 55 runs ahead of par per match. Sobers was 40 odd. Bradman 95. Runs and wickets is what determines who wins in cricket. You win by runs, or you win by wickets in hand. That is a hard notion to reject. Now you can get into specific match conditions and context are ignored, sure. But things do average out over a long career. Not perfectly. But enough that makes comparison of players between eras and in different matches, feasible.

If all you want to do is reject the idea of statistical analysis in sports, you can do so. But it does not improve on statistical analysis of sport - which is here to stay.
 
Last edited:
I think the formula needs tweaking and actual innings may not be best to use. But an arbitrary figure is set of between 1.5 and 1.7 innings per match for fairer comparison.

Imran, 4.11*22.81; – 1970’s 36.72; 4.11*13.09; 53.79 runs under average; Batting average 37.69; 1.43 * ~1 run = match value ~55 positive.
Sobers, 2.52*34.03 - 1960’s 36.36; 2.52*2.33 = 5.8; 57.78; 21.42 * 1.72 = match value 36.85 + 5.8 = ~42.65 match value positive;

then becomes

Imran, 4.11*22.81; – 1970’s 36.72; 4.11*13.09; 53.79 runs under average; Batting average 37.69; 1.6 * ~1.6 run = match value ~55.5 positive.
Sobers, 2.52*34.03 - 1960’s 36.36; 2.52*2.33 = 5.8; 57.78; 21.42 * 1.6 = match value 34.272 + 5.8 = ~40.07 match value positive;

I will look into some statistical information for the best innings per match average.

This is because I plugged Youseff v Viv into the formula and found a possible human error. Viv was ahead on runs per wicket average, but was then harsly penalised for playing far less innings per match as a comparison tool, because I was using his actual runs for his career, and not standardised his number of innings per match. See smalishah84, you're helping. Even if its in a manner you would never have thought of.

I am not sure sure whether it is because Viv played in such an anomalous successful team, and was such a great batsman himself that his first innings and second innings averages are closer together. But I think the arbitrary figure based on statistical average has credit.

So

Youseff 156/90; 52.29; 38.37, 24.128 (actual as 1.7 innings per match)
1.6 * 13.92 = 22.27 (standarised to 1.6 innings)

Viv 182/121; 50.23; 35.86, 21.6 actual (actual 1.5 innings per match - very low, great bowlers and strong batting team played in)
1.6* 14.37, = 22.9

Obviously that increases if it 1.7 innings per match.

I will look for the average number of innings per match for players and welcome constructive thoughts to improve the formula, that includes arbitrarily standardizing the number of innings batted per match or keep it as actual innings completed. It may be that the original formula is fine, and that teams such as the West Indies team will cause an anomaly for being incredibly successful.

I think it may be best to use a comparing player's innings per match, but that does not permit a ranking list. Bradman was also 80/52=~1.5. At 1.6 innings per match he is then

1.6 * (99.94-37.38) (62.56) = 100.096 runs ahead of par in a match standardized whereas his actual was 95 runs.

Wickets per match does not penalize a team for winnings by an innings. But it does penalize a batting team for winning by an innings in terms of restricting the batsman's opportunity to bat. A great bat in a weak team, suddenly gets far more batting opportunities to score more runs per match. I imagine Andy Flower would get a great advantage here, and Viv Richards the greatest disadvantage. You can extrapolate the multitude of reasons for yourselves.

I welcome thoughts on actual innings as against a standardised or averaged innings per match (even if from each player's era)
 
Last edited:

Top