Sachin Tendulkar vs Jacques Kallis

[subshakerz]

RPT is an unreliable way of measuring quality. It's a reliable way of measuring output.

If you want to measure quality, throw this discussion about runs and wickets per test/innings out the window. Both players are better than the numbers suggest.

If you want to be measure workload, we can look at these output numbers. If you want to be consistent, either compare WPT to RPT or WPI to RPI. If you want to be consistent and use the better method, compare WPT to RPT. I prefer to use the better method.

I've given you the maths that shows Kallis ahead, regardless of which method is picked. I know you aren't going to agree with what the answers are telling us. So either show an error in the maths with numbers of your own or attack the validity of the entire concept.

You've given nothing really. Just numbers without any reasonable reference to other cricketers that you call low and somehow converting runs to wickets in a convoluted way to compare bats and bowlers.

Whereas I was comparing Imran with other test standard bats of his era and Kallis with bowlers.

Can you show me where you have ever used RPT to compare bats and WPI to compare bowlers before? If you haven't then bringing it up now is meaningless and simply cherrypicking stats.

You are also flat out contradicting yourself.

You already said RPI is a more meaningful stat than RPT. You are actually measuring the actual times he batted. Now you say, let's throw out all stats to measure quality. Yeah right.

And no, workload is not just output numbers, it's also the average per innings performance a cricketer delivered. That's the actual workload.

So by all means, let's compare RPI with WPI or WPM then. I don't care because Imran still is better on this score than Kallis.

 

[Bolo.]

You've given nothing really. Just numbers without any reasonable reference to other cricketers that you call low and somehow converting runs to wickets in a convoluted way to compare bats and bowlers.

Whereas I was comparing Imran with other test standard bats of his era and Kallis with bowlers.

Can you show me where you have ever used RPT to compare bats and WPI to compare bowlers before? If you haven't then bringing it up now is meaningless and simply cherrypicking stats.

You are also flat out contradicting yourself.

You already said RPI is a more meaningful stat than RPT. You are actually measuring the actual times he batted. Now you say, let's throw out all stats to measure quality. Yeah right.

And no, workload is not just output numbers, it's also the average per innings performance a cricketer delivered. That's the actual workload.

So by all means, let's compare RPI with WPI or WPM then. I don't care because Imran still is better on this score than Kallis.

I will be extremely generous to Imran in the numbers I select in this post to try and get a straight answer out of you.

Imran scores 30 RPI. Kallis takes 1.08 WPI. Imran is scoring 28 runs for every wicket Kallis takes. The batting average for both eras is above 30.

How many runs do you think a wicket is worth?

 

[subshakerz]

I will be extremely generous to Imran in the numbers I select in this post to try and get a straight answer out of you.

Imran scores 30 RPI. Kallis takes 1.08 WPI. Imran is scoring 28 runs for every wicket Kallis takes. The batting average for both eras is above 30.

How many runs do you think a wicket is worth?

First can you answer my question on where you used RPI and WPI before?

Second, yeah it's obvious 30RPI (not including not outs) is better than a wicket an innings in terms of match value. I don't know how this is debatable.

 

[Bolo.]

First can you answer my question on where you used RPI and WPI before?

Second, yeah it's obvious 30RPI (not including not outs) is better than a wicket an innings in terms of match value. I don't know how this is debatable.

I'm not sure 100% sure I understand your question. I don't remember ever considering RPT/RPI in relation to each other for top order bats before. I frequently mention WPM as opposed to WPI for bowlers, but that's basically just because the info is known/accessible. Either way works as a means of quality assessment to me, not sure which is better. I think I have only ever talked about workload per match. ARs very, very rarely get broken by marathon contributions in their secondary. It's both disciplines across a match. Or cumulatively over the course of a career. Kallis had his workload managed by bowling less. Imran had his workload managed through batting lower. Sometimes this meant only batting once, which really managed his workload. Does this answer your question?

That's 28 runs per innings, not 30. Kallis is 1.08 WPI, not 1. The distinction is meaningful because I (and many other people) rule of thumb the value of a wicket as 30. The bowling average in the Kallis era was about 33-34. For Imran its around 32. Do you stand by your position of 28 runs > than 1 wicket in output?

 

[subshakerz]

I'm not sure 100% sure I understand your question. I don't remember ever considering RPT/RPI in relation to each other for top order bats before. I frequently mention WPM as opposed to WPI for bowlers, but that's basically just because the info is known/accessible.

Right so just like I thought. You are using metrics you never have before just for these two cricketers because you think they show some type of parity between them.

That's 28 runs per innings, not 30. Kallis is 1.08 WPI, not 1. The distinction is meaningful because I (and many other people) rule of thumb the value of a wicket as 30. The bowling average in the Kallis era was about 33-34. For Imran its around 32. Do you stand by your position of 28 runs > than 1 wicket in output?

Sorry but Imran is literally 30 RPI.

I don't know where you came up with some universal value of 30 runs = 1 wicket across all eras but no I don't buy it at all.

Think of how crazy that is. You are saying a bowler with 2WPI (many pacers) is equivalent to a bat scoring 60 RPI, literally second to Bradman? Sobers had a RPI of 50 lol.

In batting terms, Imran is right up there with the mean for a lower order bat, Kallis isn't close for a bowler.

I think it's clear by now you are just playing with numbers and making false parities between bats and bowlers without the faintest idea what you are doing. You shouldn't do that and waste people's time.

 

[sayon basak]

Imran scores 30 RPI. Kallis takes 1.08 WPI. Imran is scoring 28 runs for every wicket Kallis takes. The batting average for both eras is above 30.

That's a huge oversimplification imo. It's not taking into account their actual era adjusted values, quality of opposition (Kallis would drop in every stats possible if BAN and ZIM are excluded) etc.

And this method also gives advantage to the bowlers. Murali takes 3.478 WPI. So, that would make it equivalent to 104.347 RPI, surely that's not right is it?

 

[subshakerz]

That's a huge oversimplification imo. It's not taking into account their actual era adjusted values, quality of opposition (Kallis would drop in every stats possible if BAN and ZIM are excluded) etc.

And this method also gives advantage to the bowlers. Murali takes 3.478 WPI. So, that would make it equivalent to 104.347 RPI, surely that's not right is it?

He is just pulling these numbers from his rear. He admits he never used these numbers before.

 

[_00_deathscar]

Sooooooo
Sachin or Kallis eh?

 

[Bolo.]

Right so just like I thought. You are using metrics you never have before just for these two cricketers because you think they show some type of parity between them.


Sorry but Imran is literally 30 RPI.

I don't know where you came up with some universal value of 30 runs = 1 wicket across all eras but no I don't buy it at all.

Think of how crazy that is. You are saying a bowler with 2WPI (many pacers) is equivalent to a bat scoring 60 RPI, literally second to Bradman? Sobers had a RPI of 50 lol.

In batting terms, Imran is right up there with the mean for a lower order bat, Kallis isn't close for a bowler.

I think it's clear by now you are just playing with numbers and making false parities between bats and bowlers without the faintest idea what you are doing. You shouldn't do that and waste people's time.

I think I need to remind you that I assess neither batting nor bowling ARs quality simply on the basis of workload. My position has always been that both need to be judged on their impact on the team vis-a-vis stuff like partnership runs, team selection, and workload management for other players. Workload is your criticism of Kallis' bowling. I'm responding in kind. I assess the quality of neither Imran nor Kallis in terms of stuff like RPT/I or WPT/I. The badis of this conversation is you criticizing Kallis for workload in secondary but not applying the same standards to Imran.

Now, to use reductive terms that you seem attatched to, please answer my question. How many runs is a wicket worth in terms of output?

 

[sayon basak]

Now, to use reductive terms that you seem attatched to, please answer my question. How many runs is a wicket worth in terms of output?

Yeah I get your point. Kallis vs Imran is debatable in secondary discipline if we are only going by workload. I rate Imran higher in secondary discipline based on opposition and value to the team.

Guess I'm now gonna prepare the average of RPI and WPl of top 50 batters and bowlers respectively to answer your question. That should work imo.

 

[Bolo.]

That's a huge oversimplification imo. It's not taking into account their actual era adjusted values, quality of opposition (Kallis would drop in every stats possible if BAN and ZIM are excluded) etc.

And this method also gives advantage to the bowlers. Murali takes 3.478 WPI. So, that would make it equivalent to 104.347 RPI, surely that's not right is it?

Haha, as if I needed more reason to rate Murali. I already rated him as the GOAT bowler by an appreciatiable distance without knowing this stat. Now I'm going to be insufferable in my level of praise for him.

More seriously though, this is just a measure of output. It's tangentially related to quality, but not the same thing. As a measure of output, if Murali is taking 40 or 45% of wickets per innings (whatever the actual number is), he does have a similar output to a bat going at 100 RPI.

 

[sayon basak]

@subshakerz @Bolo. Here's an attempt to answer how much is a wicket worth in tests.

I went with the list of 84 batters with the most runs in the history of test cricket and averaged their RPI.
Average RPI for those 84 batsmen came out to be 42.768.

Did the same thing with bowlers (top 83 bowlers with most wickets+SF Barnes)
Average WPI for those 84 bowlers came out to be 2.2643.

So a wicket could be regarded as being equivalent to 42.768/2.2643= 18.89 runs.

So, Muralitharan's 3.478 WPI can be interpreted as being equivalent to an output of 65.69 RPI.

And 1.0735 WPI for Kallis could be interpreted as an output of 20.27 RPI, which is inferior to Imran's 30.21 RPI output. Sounds about right to me.

This would also be consistent with the idea that a 5'fer is equivalent to a century.

 

[Qlder]

All this discussion on how many runs = 1 wkt, it is and always has been 20 as a rule of thumb

There's a reason bowlers celebrate a 5fer same as bats celebrate a hundred. Also, look how many bowlers average 4 wpm and batsmen averaging 80 runs per match gives a good criteria for finding the elite in each.

Then for the Super elite you have 5 wpm and 100 rpm, that would include Lillee, Hadlee & Murali and Bradman (Headley 99.55 and G. Pollock 98.09 oh so close)

 

[subshakerz]

I think I need to remind you that I assess neither batting nor bowling ARs quality simply on the basis of workload. My position has always been that both need to be judged on their impact on the team vis-a-vis stuff like partnership runs, team selection, and workload management for other players. Workload is your criticism of Kallis' bowling. I'm responding in kind. I assess the quality of neither Imran nor Kallis in terms of stuff like RPT/I or WPT/I. The badis of this conversation is you criticizing Kallis for workload in secondary but not applying the same standards to Imran.

Dude you were the one introducing the RPI vs WPI measure not me. Can you at least admit you were off base with your numbering since it makes it easy for your average pacer to be better than a worldclass bat.

Now, to use reductive terms that you seem attatched to, please answer my question. How many runs is a wicket worth in terms of output?

I'm not sure, I haven't crunched the numbers but I would assume below 20.

 

[subshakerz]

@subshakerz @Bolo. Here's an attempt to answer how much is a wicket worth in tests.

I went with the list of 84 batters with the most runs in the history of test cricket and averaged their RPI.
Average RPI for those 84 batsmen came out to be 42.768.

Did the same thing with bowlers (top 83 bowlers with most wickets+SF Barnes)
Average WPI for those 84 bowlers came out to be 2.2643.

So a wicket could be regarded as being equivalent to 42.768/2.2643= 18.89 runs.

So, Muralitharan's 3.478 WPI can be interpreted as being equivalent to an output of 65.69 RPI.

And 1.0735 WPI for Kallis could be interpreted as an output of 20.27 RPI, which is inferior to Imran's 30.21 RPI output. Sounds about right to me.

This would also be consistent with the idea that a 5'fer is equivalent to a century.

Yes this makes sense. So Imran's output essentially is around 50% better than Kallis on average.

Whereas Sobers with 28RPI is roundabout Imran level.

 

[Bolo.]

That's a huge oversimplification imo. It's not taking into account their actual era adjusted values, quality of opposition (Kallis would drop in every stats possible if BAN and ZIM are excluded) etc.

And this method also gives advantage to the bowlers. Murali takes 3.478 WPI. So, that would make it equivalent to 104.347 RPI, surely that's not right is it?

People tend to overestimate the impact of minnows on records for most players. The majority tend to hardly play against minnows. And underestimate the impact of who/when they were playing in relation.

Look at this breakdown of wicket value that @ankitj posted.

How valuable is that wicket!

Long time back I came across a comprehensive analysis of bowlers based on the "value"of the wickets they take. The approach was wonderfully elegant and very logical. The article is not online anymore. Nevertheless I decided to do the exercise again. I am happy to share the results of my hard...

www.cricketweb.net

The value of Kallis' wickets is ridiculously high, whatever type of minnow bashing he did. He also mainly bowled at set bats to break partnerships or hold the fort, with an old ball. Which probably had a significantly bigger impact. If his home country was anywhere other than RSA, I would probably upgrade his bowling average from meh to very good.

 

[Bolo.]

@subshakerz @Bolo. Here's an attempt to answer how much is a wicket worth in tests.

I went with the list of 84 batters with the most runs in the history of test cricket and averaged their RPI.
Average RPI for those 84 batsmen came out to be 42.768.

Did the same thing with bowlers (top 83 bowlers with most wickets+SF Barnes)
Average WPI for those 84 bowlers came out to be 2.2643.

So a wicket could be regarded as being equivalent to 42.768/2.2643= 18.89 runs.

So, Muralitharan's 3.478 WPI can be interpreted as being equivalent to an output of 65.69 RPI.

And 1.0735 WPI for Kallis could be interpreted as an output of 20.27 RPI, which is inferior to Imran's 30.21 RPI output. Sounds about right to me.

This would also be consistent with the idea that a 5'fer is equivalent to a century.

Interesting, thanks.

If you dont mind putting in the effort, what does this look like in reverse? What's the number of wickets for the top run scorers and the number of runs for the top wicket takers?

 

[ankitj]

The value of Kallis' wickets is ridiculously high, whatever type of minnow bashing he did. He also mainly bowled at set bats to break partnerships or hold the fort, with an old ball. Which probably had a significantly bigger impact. If his home country was anywhere other than RSA, I would probably upgrade his bowling average from meh to very good.

True that. Look at this list of batsmen dismissed by Kallis. There are hardly any tail-enders at the top (only Brett Lee). For most premier bowlers you would often see some meme tailenders appearing high up.

 

[sayon basak]

People tend to overestimate the impact of minnows on records for most players. The majority tend to hardly play against minnows. And underestimate the impact of who/when they were playing in relation.

Look at this breakdown of wicket value that @ankitj posted.

How valuable is that wicket!

Long time back I came across a comprehensive analysis of bowlers based on the "value"of the wickets they take. The approach was wonderfully elegant and very logical. The article is not online anymore. Nevertheless I decided to do the exercise again. I am happy to share the results of my hard...

www.cricketweb.net

The value of Kallis' wickets is ridiculously high, whatever type of minnow bashing he did. He also mainly bowled at set bats to break partnerships or hold the fort, with an old ball. Which probably had a significantly bigger impact. If his home country was anywhere other than RSA, I would probably upgrade his bowling average from meh to very good.

That was a great effort by @ankitj.

Seems like Kallis and Sobers had pretty similar values of wickets and both took those wickets at a similar average.

But still think Kallis lacked the workload with the ball, and I'd still place Sobers and Imran ahead of him as they're workload is superior.

 

[sayon basak]

Interesting, thanks.

If you dont mind putting in the effort, what does this look like in reverse? What's the number of wickets for the top run scorers and the number of runs for the top wicket takers?

That's a good idea. I'll definitely look into it.
(Pretty sure it'd spit out some bizzare number though)