Garry Sobers v Imran Khan,Test Cricket:Poll

[marc71178]

Not to mention that whilst he was doing all that with the bat he was a front line bowler taking a huge share of the overs.

 

[CaptainGrumpy]

Not to mention that whilst he was doing all that with the bat he was a front line bowler taking a huge share of the overs.

Front line bowler? So why he get so few wicket from 90 tests? 5th bowler at best for the majority of his career including when he opened the bowling ahead of 3 spinners in the bowling attack.

 

[OverratedSanity]

Front line bowler? So why he get so few wicket from 90 tests? 5th bowler at best for the majority of his career including when he opened the bowling ahead of 3 spinners in the bowling attack.

So few? He has a far better rate of wickets per match than any other batting allrounder.

 

[Red]

Front line bowler? So why he get so few wicket from 90 tests? 5th bowler at best for the majority of his career including when he opened the bowling ahead of 3 spinners in the bowling attack.

What point do you want to make with all this anyway? Do you just want it acknowledged that your mathematical formula you devised says Imran is better than Sobers or something?

 

[CaptainGrumpy]

So few? He has a far better rate of wickets per match than any other batting allrounder.

But so few for a front line bowler.

 

[cnerd123]

What point do you want to make with all this anyway? Do you just want it acknowledged that your mathematical formula you devised says Imran is better than Sobers or something?

I was wondering how long it would take you all to realise that Grumpy isn't here to discuss, he is here to argue down anyone who disagrees with him in an attempt to 'win'.

 

[CaptainGrumpy]

I was wondering how long it would take you all to realise that Grumpy isn't here to discuss, he is here to argue down anyone who disagrees with him in an attempt to 'win'.

*****, please stay on thread topic. You talk about me far too much. I'm starting to think that you may be in love with me.

 

[The Sean]

My thoughts precisely & I'm someone who'd probably rate Imran as Asia's greatest ever cricketer, but there's only ever been one Sobers.

Lies

 

[marc71178]

Front line bowler? So why he get so few wicket from 90 tests? 5th bowler at best for the majority of his career including when he opened the bowling ahead of 3 spinners in the bowling attack.

So if Imran was such a good batsman how come he got so few runs from his 88 tests?

 

[CaptainGrumpy]

So if Imran was such a good batsman how come he got so few runs from his 88 tests?

Are you calling Imran a "front line" batsman for his whole career now?

I have written and maintained that he was a just above "par" or average batsman for the entirety of his career averaged out. He is just ahead ahead of par. I have never said that he was "such a good batsman" for the entirety of his 88 match career.

Imran was a just above par batsman over his 88 tests, which is brilliant because he was a brilliant bowler as well.

 

[OverratedSanity]

Are you calling Imran a "front line" batsman for his whole career now?

I have written and maintained that he was a just above "par" or average batsman for the entirety of his career averaged out. He is just ahead ahead of par. I have never said that he was "such a good batsman" for the entirety of his 88 match career.

Imran was a just above par batsman over his 88 tests, which is brilliant because he was a brilliant bowler as well.

What would you consider Kallis' bowling as then? He averaged a very good 32 but only got 290 odd wickets in 166 tests. Surely he's above par. So by the same logic, Sobers getting 235 wickets in only 93 tests should be seen as well above par, if you're taking wicket/runs per match as your metric.

 

[CaptainGrumpy]

What would you consider Kallis' bowling as then? He averaged a very good 32 but only got 290 odd wickets in 166 tests. Surely he's above par. So by the same logic, Sobers getting 235 wickets in only 93 tests should be seen as well above par, if you're taking wicket/runs per match as your metric.

Are you actually asking me the question, or are you using rhetoric to state your argument that:

Kallis took 290 wickets at 32 in 166 tests
Kallis is above par as a bowler (which begs the question whether he is above par)
Sobers has an equivalent to better record at first blush (because you have not acknowledged that the average per wicket in Sobers era was less than Kallis's)
Therefore Sobers as a bowler is above par

I cannot give you the answer I think you want. What is the average par performance for a "bowler". I cannot because too many batsman have bowled and will continue to bowl. A bowler may take 5 wickets in a match, but go for 250 runs in collecting those wickets. I hope you understand the reason why. This is a measure of cricketers, their batting and bowling combined.

I can tell you that Sobers bowling is ahead of average for a cricketer. All cricketers bat. Not all cricketers bowl. Some batsman bowl bowl very badly on occasion. That all affects the "average" runs per wicket that batsmen plunder from. I can not tell you what the average 'bowler' contributes to par. All bowlers are different. But I can tell you whether a cricketer, batsman or bowler as well, is head or below par on the method of runs per wicket average for their era.

If Sobers never scored a run, and had a batting average of 0.0, and using a uniform 1.6 innings per match batted - his cricketing contribution which is effectively his bowling is negative 58.8 below par per match as a cricketer because of insufficient wickets. Now think about it, would you expect a bowler averaging 0.0 with the bat, and 32.2 with the ball who gets a little over 2 wickets per match at a very low economy rate to help you beat the opposition or hinder you? No. You wouldn't think he will help you win too many matches. Could he have bowled more overs? Irrelevant. This measures only the overs they bowled per match. That is, he was actually a less than par bowler with his bowling record, his SR is very high and his captain preferred others to bowl than him. He was not a "front line" bowler. He was a 5th bowler. I would imagine that most, if not all '5th bowlers" have below par records. this isn't about "peak potential", its about what the players did over the entire course of their careers (although selected time periods can be used).

Imran's bowling is 53.79 runs ahead of "average" with 4.1 wickets per match at 22's. If Imran had never scored a single run, with a batting average of 0.0, he would have been 5 runs below par per match as a cricketer for his era. That is including 1.6 ducks on average per match. He was very much a "front line bowler". This is because the method is to directly correlate batsman and bowlers on the same metric which is average runs per wickets. With a batting average of 3 he would be at par. A batting average of 5 and he is winning you more games than he's losing for you.

This has been calculated on a 1.6 innings per match batting (which is subject to further refinement). I also plan on changing runs above and below par, to be divided by the era's runs per wicket to give a ratio score. It makes little difference for Sobers and Khan's period as it 35-36 runs per wicket average consistently, but would allow for fairer comparison to Grace's period because while it will give an accurate score for those cricketers of that era in a runs and wickets average, runs were more scarce in each match then.

 

[Red]

If Sobers never scored a run, and had a batting average of 0.0, and using a uniform 1.6 innings per match batted - his cricketing contribution which is effectively his bowling is negative 58.8 below par per match as a cricketer because of insufficient wickets. Now think about it, would you expect a bowler averaging 0.0 with the bat, and 32.2 with the ball who gets a little over 2 wickets per match at a very low economy rate to help you beat the opposition or hinder you? No. You wouldn't think he will help you win too many matches. Could he have bowled more overs? Irrelevant. This measures only the overs they bowled per match. That is, he was actually a less than par bowler with his bowling record, his SR is very high and his captain preferred others to bowl than him. He was not a "front line" bowler. He was a 5th bowler. I would imagine that most, if not all '5th bowlers" have below par records. this isn't about "peak potential", its about what the players did over the entire course of their careers (although selected time periods can be used).

Imran's bowling is 53.79 runs ahead of "average" with 4.1 wickets per match at 22's. If Imran had never scored a single run, with a batting average of 0.0, he would have been 5 runs below par per match as a cricketer for his era. That is including 1.6 ducks on average per match. He was very much a "front line bowler". This is because the method is to directly correlate batsman and bowlers on the same metric which is average runs per wickets. With a batting average of 3 he would be at par. A batting average of 5 and he is winning you more games than he's losing for you.

Can you see any flaws in your logic?

 

[watson]

Whether you prefer Imran or Sobers is purely contextual.

If your team needs a star bowler because its bowling strength is weak then you would select Imran.

If your team needs a star batsman because its batting strength is weak then you would select Sobers.

If your team is all weak from top to bottom then nothing is going to rescue your team anyway.

If your team is strong from top to bottom then merely toss a coin and be happy with whatever way it falls.

 

[CaptainGrumpy]

Can you see any flaws in your logic?

I've invited any suggested improvements to the math. So far all you have said is that you dont like averages, maths and stats but you read and watched some games, and rank Sobers batting higher than Imran's bowling without an expressed objective or logical measure that I or anyone else can use. I like a more objective measure than that so as to compare batting to bowling. So I use the runs per wicket average of their eras to find how far above or below par that any given cricketer was. Sobers' bowling improves his par score by 5 runs per match, he is about 40 runs ahead of par per match on average, that is brilliant which it should as he was better than average bowler and an outstanding batsman and a true match winner, but he was no "front line" or "strike" bowler. He was a very good or excellent 5th bowler. He was a better than average bowler overall. That's what the math confirms. Could he have bowled more? Possibly, but the West Indies captains and Sobers chose not to for whatever reason. I doubt it was to help the opposition make more runs. So we are limited to what actually did happen in a career in terms of wickets taken. I find that measure quite objective. If Sobers was not such a master batsman, he wouldn't have played much as a specialist bowler at 2 odd wickets per test. That is not a difficult conclusion for anyone to accept. So sticking to what he actually did on the field, he is 35 odd runs ahead of par for batting which is improves by another 5 runs for his bowling. The precise numbers are on the preceding page. I think you may be missing the point that to improve on par, a player actually has to score enough runs, or take enough wickets below the average, to have a better than par score. So a batsman must average 36 or so (depending on the era) before he starts improving your teams fortunes and carries the specialist bowler's runs per wicket. That makes sense because specialist batsman below that score are obviously a hinderance. A batsman averaging over 40 over a decent length career is typically referred to as a good batsman.

I think you should think about the runs per wicket average and understand how it works, and if you have suggestion for improvement, I am open to suggestions for improvement. But if you merely misunderstand it, please read around on it. It is an objective measure, that is used for teams regularly and widely by many cricketing writers and analysts. A team is merely the sum of its parts, so I am using it for the players themselves to see how individually they fared.

I accept maths and stats as a means of sports analysis. I understand the limitations thereof. Fielding is excluded. So is captaincy. Running between wickets. Just a record of batsman vs bowler and a comparison who is above and below par as a batsman, bowler and ultimately as a cricketer. Bad and good pitches, batting under pressure, batting under little pressure, bowling under pressure, bowling under little pressure, easy runs and hard runs, lucky wickets and unlucky wickets/non wickets will average out, not perfectly but appreciably.

What the stats reveal is interesting. The West Indies 81-89, who made them a great team? Their batting or their bowling? Well despite Richards (who I think was declining towards the end of his career), Lloyd (for half the period), Haynes and Greenege, the West Indies were only above merely average with the bat, but they had Gomes and Hooper and some average batsmen play a lot of matches in sum. But they were ridiculously ahead of par with the ball. Cricketers always talk highly of Viv and Greenidge, and the entire West Indies attack, so it makes sense. Their batting was average at 36.27 (35.86 being the exact average) with Marshall batting 8 in the absence of a Hadlee, Imran, Botham or Dev. But their bowling was exceptionally good at 26.13. That's like a 100 runs lead on average at the conclusion of the first innings. What does it lead to in real terms? A win loss ratio, 5.71 wins for every loss with 58% of test played won.

The Australian team of 2000-2007 - well their batting was way ahead of par at 44.39 per wicket when the average is 38, but their bowling was at 27.25, or similarly as dominant as the West Indian bowling when standardizing eras. So while both facets are match winning, the bowling contributed just a little bit more. That is utter destruction of opposition. That is like a 170 run lead on the average at the conclusion of the first innings. What does it lead to in real terms? A win loss ratio, 7.2 wins for every loss with 77.4% won. That is just ridiculous. Now not every player contributes the same to this winning advantage - surely that you can accept. So the player's individual performances can also be measured.

There is great objectivity in the wickets per runs average. It can be used for players and teams alike. By analogy, and I am not saying that you would, you could tell me and argue till your blue in the face that the 1980's WI dominant because of its batting. I would respond that the numbers tell me that they were merely just above average overall for batting, but they were amazing for bowling, which makes sense because they batted Marshall at 8 and played the likes of Logie and Gomes as well as well as someone like Viv Richards (who was in decline towards the end of his career).

If you were to try and tell me, and I'm not suggesting that you are, that the West Indies batting line up top to bottom was as strong for their era as the Australian team 1999-2007, top to bottom, I will respond that the statistics strongly suggest otherwise and that its probably not the case that the 1980's West Indies batting line up top to bottom was as strong as the Australian batting line up top to bottom for their respective eras. And then you think well Australia had Gilchrist at 7 with 6 brilliant batsmen ahead of him, and a strong tail bar McGrath, it makes sense, even with Warne at 8.


Batsman WI Ave Aus Ave
1&2 43.10 48.52
No.3 44.85 62.77
No.4 37.50 45.12
No.5 41.90 46.37
No.6 39.26 39.75
No.7 32.87 51.52

Era comparison is 35.86 average for the 1980s. 38.37 odd for the Australian era.

Just look at that "Gilchrist factor" in the Australian team. You must accept that statistic surely?

 

[Red]

What the stats reveal is interesting. The West Indies 81-89, who made them a great team? Their batting or their bowling? Well despite Richards (who I think was declining towards the end of his career), Lloyd (for half the period), Haynes and Greenege, the West Indies were only above merely average with the bat,

I

West Indies from 81-89 with the bat. Greenidge (avg 47). Haynes (41). Richards (46). Richardson (48). Gomes (41). Lloyd (61). Dujon (36). Marshall (20)

Just quietly, they were ticking along ok with the bat in the 80s.

Apart from that, I have absolutely no idea how what you posted above relates to Imran and Sobers.

 

[RossTaylorsBox]

I've only skimmed over this thread but I think there are some problems with the "runs per wicket" statistic. Firstly, runs per wicket is essentially the average partnership including not outs. So it's comparing a partnership to that of an individual batsman which I think is slightly flawed.

Secondly, even if this were OK I don't think using the mean average is appropriate in this case. The distribution of individual innings and partnerships are highly skewed, the greatest frequency of runs here is actually 0 so the mean average will turn out to be higher, possibly penalising outstanding batsmen. I'm wondering what would happen if medians were compared.

Finally, I feel trying to generalise for different roles highly problematic. Even given that runs and wickets are universal measures, it's difficult to really quantify and compare different types of all-rounders and generating an objective answer will end up futile.

 

[CaptainGrumpy]

West Indies from 81-89 with the bat. Greenidge (avg 47). Haynes (41). Richards (46). Richardson (48). Gomes (41). Lloyd (61). Dujon (36). Marshall (20)

Just quietly, they were ticking along ok with the bat in the 80s.

Absolutely. The West Indies overall were ticking along at just above average thanks to those fine and great batsmen mentioned above performing well. Its a shame Lloyd didn't play the whole decade.

Apart from that, I have absolutely no idea how what you posted above relates to Imran and Sobers.

With all due respect, I suspected you possibly wouldn't. But all the necessary information is in the previous post. Please read it again. Cricket is a game of runs and wickets. The runs per wicket average combines batsmen and bowlers, because bowlers bowl to the batsmen against each other. It also combines players into teams because teams bat and bowl against each other. It has been used for team analysis widely and for a long time. It takes a lot of work to do it for the individual players as well but it can be done because not all players contribute the same to team success.

If your batsman score more than the average number of runs for their own wicket, and the bowlers are taking wickets and the cost of the wickets is less than the average, he is ahead of par and causing you to be ahead of the game meaning more wins than losses on average. If an individual player does the same, he is above par and causing you to be ahead of the game meaning more wins than losses on average.

 

[CaptainGrumpy]

I've only skimmed over this thread but I think there are some problems with the "runs per wicket" statistic. Firstly, runs per wicket is essentially the average partnership including not outs. So it's comparing a partnership to that of an individual batsman which I think is slightly flawed.

Please clarify exactly what you're trying to say about a "flaw" here. Because ten wickets is an innings. If ten batsmen lose their wicket, its innings over. Someone will be always be left "not out" in every innings. A batsman who bats lower will statistically be left "not out" more often and yet still have a lower batting average than the batsmen ahead of him. Its a skill to not lose your wicket cheaply. Its a measure of how many runs the batsman contributed for his own wicket. Individuals contribute the runs to the partnerships in differing proportion of contribution. Surely you accept and acknowledge this fact.

Secondly, even if this were OK I don't think using the mean average is appropriate in this case. The distribution of individual innings and partnerships are highly skewed, the greatest frequency of runs here is actually 0 so the mean average will turn out to be higher, possibly penalising outstanding batsmen. I'm wondering what would happen if medians were compared.

Medians won't work. The only way batsman are penalised are those that bat a longer time to collect more extras. So Trevor Franklin's value will be possibly slightly underestimated because he only scored 23 runs per innings and gave more time to collect extras. This is offset by batsman who score runs quickly, like David Warner, but actually do less for the team in collecting in extras. But the extras are collected by all batsmen who bat. Some would say that fast scoring batsmen do more good for the team in breaking the other team's spirit or setting the team up for victory, so it averages out to some extent. Warner and Franklin are deliberately chosen as being at either end of the quick scoring opener scale.

Some bowlers get a free ride with byes and leg byes for comparison among other bowlers who give up less byes and leg byes. But they always have and do on the averages. That is a flaw with the scorecard if you think some bowlers are worse offenders of giving away leg byes (leg stump bowling) or byes which are not the wicket keepers fault than other bowlers, but its minor compared to runs off the bat. Because it is consistent, it doesn't effect the comparison of batsman to bowlers. They are still operating within the same dynamic and framework. There is an assumption of 1.6 innings per match used for comparison that needs further scrutiny. Khan really batted 1.4 times per match, and Sobers 1.7, Khan still wins on the actual, but it seemed unfair to compare Viv at 1.5 to Youseff at 1.7 times per match. Youseff was getting an advantage of playing for a weaker team in an indivdual comparison. So the same metric had to be used.

Finally, I feel trying to generalise for different roles highly problematic. Even given that runs and wickets are universal measures, it's difficult to really quantify and compare different types of all-rounders and generating an objective answer will end up futile.

No, its really easy on a wickets per runs average. Its just a lot of work to collect and process all the information.

There is an issue that I am using 1.6 as the average innings batted, when that is an approximisation. I can have every players actual innings batting, but it will take some time to average what the actual number will be for eras, and overall. I suspect it will be around 1.5 -1.7. But the player's actual number is always available for analysis. Just makes comparing Youseff to Viv Richard's fairer.

 

[longranger]

I like cricketing stats, I really do. Often I crunch numbers, to show trends and patterns that often miss the eye. For example, wickets per match (WPM), where Murali leads with 6.02 and Wasim Akram lags with 3.98. Does this mean Murali is 1.5 times a better bowler than Wasim? I don't think many would agree with that. Numbers in isolation make sense if cricket is like a computer simulation. But it's not - there are intangibles like the pitch, the umpiring, the crowd support, the animosity between the teams, the importance of the match, the weather and most importantly the match situation. Saying it 'averages out over a career' is quite optimistic.

You cannot just compare players by equating a certain number of runs to wickets as a fixed ratio and then saying they are apples against apples and can be prepared. A 300 is magnificent, a 100 less so. But what if the 300 had led to a tame draw, and the 100 was a fourth innings classic to chase down a target to win a match, that eventually won an away series. How do you account for that? By just saying 'it averages out'?

That's part of the charm. There are cheap runs and cheap wickets. Openers often have a tough time negotiating the swing of a new ball and sometimes the number 6 or 7 can come in at 300/4 with tired bowlers and smash a few boundaries around. While those runs are runs, do we all not see that they are cheap runs? Or imagine a bowler who cleans up the tail by taking the wickets of Ishant Sharma, Mohammed Shami and Karn Sharma. Are his 3 wickets 1.5 times better than a bowler who took the wickets of Murali Vijay and Virat Kohli?

I repeat, statistics are very important. They give context to opinions. But they're just one of the many tools available to judge quality. Don't underestimate the human brain which can judge after taking in inputs from various sources.

As for Imran vs Sobers, no clue. I'd vote for Imran, but maybe that's because a subcontinent fast bowling captain who could bat well appeals to me more than another great yesteryears West Indian player.