Matty Hayden v SUnil Gavaskar - better test opener ?

[archie mac]

Absolutely true...Don't you feel that he doesn't get discussed too much these days on CW? Shouldn't we open a new thread as a tribute to Hammond? What you say?

If you start it, they will come

 

[Ikki]

Naa, being a stats junkie for so many years, i've come to realise that stats are easily the most deceptive thing about players. It is really sad when world class players like Ponting and Tendulkar are reduced to mere numbers, and the actual strokeplay of theirs is hardly appreciated. Where is test cricket as a spectator sport these days? Where are those days when a Trumper or Vishy was regarded in the same vein as Bradman or Gavaskar irrespective of stats?

They can be deceptive, but it's upto you to prove that they are deceptive. You cannot just put your fingers in your ear and deny what is fact.

The most easily abused aspect of judging is subjective reasoning. With stats, you can verify an assertion. You cannot necessarily do that with opinions.

You aren't going to see people making vast and expansive comparisons willy nilly anymore. Because almost everybody now has the capacity to find out whether such statements are true.

 

[SJS]

If you start it, they will come

It is a valid point.

Hammond has suffered immensely due to being from the same era as Bradman. If he was from any other time people would be at least as unambiguous about his position in the ranks of the very short list of greatest of all time as they are, for example of Hobbs the end of his very long career just brushed against the start of the Don's.

 

[Anil]

If you start it, they will come

field of dreams?

 

[archie mac]

field of dreams?

Well spotted

 

[Precambrian]

They can be deceptive, but it's upto you to prove that they are deceptive. You cannot just put your fingers in your ear and deny what is fact.

The most easily abused aspect of judging is subjective reasoning. With stats, you can verify an assertion. You cannot necessarily do that with opinions.

You aren't going to see people making vast and expansive comparisons willy nilly anymore. Because almost everybody now has the capacity to find out whether such statements are true.

Eg- Take comparison of two batsmen. Does it speal anything abt the quality of pitches on they played? Does it talk about the quality of bowling? Does it talk about the conditions? Stats usually give almost as much, if not lesser, evidence as compared to "knowledgable" subjective judgement.

eg- The Murali - Warne debate. Stats, blindly followed easily put Murali miles ahead of Warney.

 

[Ikki]

Eg- Take comparison of two batsmen. Does it speal anything abt the quality of pitches on they played? Does it talk about the quality of bowling? Does it talk about the conditions? Stats usually give almost as much, if not lesser, evidence as compared to "knowledgable" subjective judgement.

All this is easily taken care of once you create a standard. For example X runs made on average by every batsman during this era. Then you see how much players are above the mean. The frequency with which that happens. I mean, many many things. It's very valuable to have facts regarding the type of pitches played and all, but the stats make up a very large portion of that doubt.

eg- The Murali - Warne debate. Stats, blindly followed easily put Murali miles ahead of Warney.

Wrong. It is when you do a statistical analysis of both that you actually realise how close they really are. I won't start this, because this discussion usually takes a life of it's own, but if you are wondering what I am talking about just go see one of the Warne v Murali threads.

People taking stats basically at face value, or overall averages without looking at the performances of said player can be mislead. But you always have a statistical way of verifying further.

This is less fallible than subjective means. There is no element of prejudice or bias there.

 

[SJS]

Lets clarify one thing here.

Anyone like me who says stats are not the best way to rank players is not for a moment saying stats tell nothing. Surely you dont have to look beyond Harbhajan's and Bradman's batting stats to know which is the better batsman

It is when we are trying to compare two very good/great players or players who aren't too far apart that stats need to be handled with care.

Just a small clarification - I am not getting back into this debate

 

[Precambrian]

All this is easily taken care of once you create a standard. For example X runs made on average by every batsman during this era. Then you see how much players are above the mean. The frequency with which that happens. I mean, many many things. It's very valuable to have facts regarding the type of pitches played and all, but the stats make up a very large portion of that doubt.

This is like "Who will bell the cat", the problem is that it is very difficult to develop a "standard" for every era as you say, without assuming some subjective element. For eg- In Tendulkar's case, his career has spanned 3 decades during which the conditions have changed quite drastically. While a batting avg of 50 in 1990 put you among the top 3 in the whole world, today it doesnt even help you to be in top 10 or 15. Stats are just numbers and can never be conclusive proof of anything. It can be used as a guide thats all.

Wrong. It is when you do a statistical analysis of both that you actually realise how close they really are. I won't start this, because this discussion usually takes a life of it's own, but if you are wondering what I am talking about just go see one of the Warne v Murali threads.

People taking stats basically at face value, or overall averages without looking at the performances of said player can be mislead. But you always have a statistical way of verifying further.

This is less fallible than subjective means. There is no element of prejudice or bias there.

I've seen some threads in CW, and still remain hugely unconvinced. The problem with stats is that the more you go into it, the more unsure you become.

 

[Ikki]

Lets clarify one thing here.

Anyone like me who says stats are not the best way to rank players is not for a moment saying stats tell nothing. Surely you dont have to look beyond Harbhajan's and Bradman's batting stats to know which is the better batsman

It is when we are trying to compare two very good/great players or players who aren't too far apart that stats need to be handled with care.

Just a small clarification - I am not getting back into this debate

The fact that two batsmen are so close statistically is probably the biggest indicator that they are comparable as players. The refusal to accept this is what seems to be the problem here.

This is like "Who will bell the cat", the problem is that it is very difficult to develop a "standard" for every era as you say, without assuming some subjective element. For eg- In Tendulkar's case, his career has spanned 3 decades during which the conditions have changed quite drastically. While a batting avg of 50 in 1990 put you among the top 3 in the whole world, today it doesnt even help you to be in top 10 or 15. Stats are just numbers and can never be conclusive proof of anything. It can be used as a guide thats all.

Why is it difficult to develop a standard? It can be as easy as rounding up all the players' batting records in that era and getting an average for all their batting. And then see how many batsmen deviate above or below it.

Tendulkar's career has really only seen one major change and that is at the start of the new millennium where pitches have been largely flat.

Whilst the discussion deserves more dissecting, let's just say there are more batsmen that average 50+. That's fine, but the problem is your standard is simply: great batsmen average 50+. That's as bad as just looking at overall career figures and not delving further. What should be discussed is how much each batsmen tended to average in any given era and see how much that has changed era by era. For example, this current era has the highest averages and runs scored, so averages are inflated somewhat. But by how much? Really, only a few points. It should also be noted that there are more players in more teams now that are capable of averaging 50+ so that can be attributed as well.

Everything else being equal, but pitches:

Era A: batsmen average 33 runs per dismissal
Era B: batsmen average 35 runs per dismissal

Compare played A from Era A and played B from Era B:

player A: averages 51 runs per dismissal
player B: averages 53 runs per dismissal

Where would the averages mislead about their quality? Only if you look at it with face value. If you take into account the era, you can conclude that they are likely of equal ability.

This is a very simple measure, but I just wish to give you an idea.

And frankly, easier batting conditions do not guarantee more runs scored. Since about the time pitches have been getting flatted Tendulkar averages about 51 (45 I think without minnows). So he is scoring less. Whilst injuries and age come into it, it's a massive drop considering in the same time someone like Ponting is averaging in the 60s, even disregarding minnows.

I've seen some threads in CW, and still remain hugely unconvinced. The problem with stats is that the more you go into it, the more unsure you become.

LOL, what? No, the more you go into it, the more sure you should become. I guess it depends what you are looking at. I mean, I'll admit, some of the stuff in the Warne v Murali thread is ridiculous. Huge analysis just to split hairs. But digging enough and properly you'll see a truer comparison.

 

[Precambrian]

How do you determine flatly a cutoff when the pitches started becoming roads?

 

[Ikki]

How do you determine flatly a cutoff when the pitches started becoming roads?

Well, the pitches in general started changing in 2000/2001. But here's the thing, not all pitches after this time have been flat or dull, and pitches prior to this time had not always been lively either...it's not like they always batted on mine fields.

There are also things to take into account like batting/bowling approaches or approaches towards matches (less draws nowadays), strength of opposition, rule changes etc. But not just one person benefits or otherwise from this...they all do. That's really why it is important to keep a standard for certain eras and judge accordingly. All these variables really answer why simple subjective analysis is not enough.

 

[Precambrian]

Well, the pitches in general started changing in 2000/2001. But here's the thing, not all pitches after this time have been flat or dull, and not pitches prior to this time had been flat and dull as well...it's not like they always batted on mine fields.

Isnt that subjective judgement? And pitches are not the only thing why batting has improved. Also the quality of the bats used is a factor.

There are also things to take into account like batting/bowling approaches or approaches towards matches (less draws nowadays), strength of opposition, rule changes etc. But not just one person benefits or otherwise from this...they all do. That's really why it is important to keep a standard for certain eras and judge accordingly. All these variables really answer why simple subjective analysis is not enough.

Yeah, all these things have to be taken to account. So how exactly are you going to do that? Take ball-by-ball stats? Or match-by-match analysis? Even then don't you think subjective bias would creep in? What about the fatigue factor in the subcontinent due to oppressive heat?

 

[Top_Cat]

Well, the pitches in general started changing in 2000/2001. But here's the thing, not all pitches after this time have been flat or dull, and pitches prior to this time had not always been lively either...it's not like they always batted on mine fields.

There are also things to take into account like batting/bowling approaches or approaches towards matches (less draws nowadays), strength of opposition, rule changes etc. But not just one person benefits or otherwise from this...they all do. That's really why it is important to keep a standard for certain eras and judge accordingly. All these variables really answer why simple subjective analysis is not enough.

Conversely they also explain why the numerical measures we have available aren't adequate either.

Let's take the most basic unit, number of runs scored. What's a run? Batsman runs from one end of the pitch to the other, credited with a point. Sounds simple right? Okay well, what influences how the runs are scored or whether they're scored at all? Here's the short list of things external to the batsman him/herself; bowling type/quality (this has its own set of complex criteria), pitch type/condition, weather, ball condition, time of day, etc. You have to take every factor into account or you're assuming all runs scored by a batsman are equal which, patently, they are not.

This is why I treat all sports stats collected as a guide and engaging in paralysis by analysis on those which are pretty blunt performance measures seems a little pointless. They're the best measures we have, this is true, but they come with a lot of caveats.

 

[Ikki]

Isnt that subjective judgement? And pitches are not the only thing why batting has improved. Also the quality of the bats used is a factor.

It's about as subjective as the quality of bats improving.

Yeah, all these things have to be taken to account. So how exactly are you going to do that? Take ball-by-ball stats? Or match-by-match analysis? Even then don't you think subjective bias would creep in? What about the fatigue factor in the subcontinent due to oppressive heat?

It's not a problem to be subjective, the problem is being so subjective that bias will come into it and not adequately being able to defend any assertions. That's why you see certain members here go in and out the discussion, seemingly trolling without adding anything. They're too stuck in their own opinion and have been countered by verifiable stats that prove the contrary to a lot of their assumptions.

It's like religion. I have no problem with people who want to believe in religion, but to take pot-shots at science for not being "exact" is kinda funny and ironic.

You see, stats may not be perfect. They may not go down to the accuracy of several decimal places, but they are still a much better gauge and a much more impartial way to judge than merely giving "I saw player X bat, and he is WAY better than player Y" - especially if the statistics surrounding it are proving the contrary.

 

[Ikki]

Conversely they also explain why the numerical measures we have available aren't adequate either.

Let's take the most basic unit, number of runs scored. What's a run? Batsman runs from one end of the pitch to the other, credited with a point. Sounds simple right? Okay well, what influences how the runs are scored or whether they're scored at all? Here's the short list of things external to the batsman him/herself; bowling type/quality (this has its own set of complex criteria), pitch type/condition, weather, ball condition, time of day, etc. You have to take every factor into account or you're assuming all runs scored by a batsman are equal which, patently, they are not.

This is why I treat all sports stats collected as a guide and engaging in paralysis by analysis on those which are pretty blunt performance measures seems a little pointless. They're the best measures we have, this is true, but they come with a lot of caveats.

I agree. They are the best measures we have. Are they adequate? Yes, to a large extent they are. And if they aren't, then it's nice to hear why they aren't. For example, someone looking plainly at stats may not appreciate that Sobers bowled different types or know why his stats may be affected more negatively than usual. It's good to know certain facts like these and they help in gauging talent/ability even more. But to say, for example, that Sobers was good enough to open the bowling for the Windies does not necessarily say enough about how he actually did and how strong that attack was at the time to need him to open, etc.

 

[Top_Cat]

I agree. They are the best measures we have. Are they adequate? Yes, to a large extent they are. And if they aren't, then it's nice to hear why they aren't. For example, someone looking plainly at stats may not appreciate that Sobers bowled different types or know why his stats may be affected more negatively than usual. It's good to know certain facts like these and they help in gauging talent/ability even more. But to say, for example, that Sobers was good enough to open the bowling for the Windies does not necessarily say enough about how he actually did and how strong that attack was at the time to need him to open, etc.

The point is, excessive reliance on sports stats or subjective viewing results in substantial inaccuracy.

 

[Precambrian]

It's about as subjective as the quality of bats improving.

Who said it is not? The question is how to figure that in the statistical model that you seem to be suggesting?

It's not a problem to be subjective, the problem is being so subjective that bias will come into it and not adequately being able to defend any assertions.

Again, how do you define "subjectivity without bias"? Wouldn't that then become "objective"?

You see, stats may not be perfect.

Exactly the point I was making.

They may not go down to the accuracy of several decimal places, but they are still a much better gauge and a much more impartial way to judge than merely giving "I saw played X bat, and he is WAY better than player Y" - especially if the statistics surrounding it are proving the contrary.

Stats are not perfect even to get to a round figure, let alone decimals. "I saw played X bat, and he is WAY better than player Y" statements are as reliable as statistics. Because both are not free from subjective-bias.

Very simple, you were arguing that Ponting-Tendulkar warrants a comparison. Why?

Because you think Ponting's better. Isnt that subjective bias? Then you proceed to process statistics supporting your view. So how can they not be subjective?

I think Tendulkar is better than Ponting. And I proceed to process another set of numbers that support the same. So how is my analysis free from subjective bias?

You say that Ponting's bad average in India is ignorable and he's much rounded player than Tendulkar. I say he's not since his averages are terrible in India. Now, how do we define "Roundedness" of a player? That again is subjective.

So, statistics are merely tools in the hands of a person to arrive at some conclusion that he wants to get. To say that they are end of everything is sadly an argument that has no basis.

 

[Uppercut]

The point is, excessive reliance on sports stats or subjective viewing results in substantial inaccuracy.

Quite. Tbh i try never to compare players across eras. Players i can see now are often, i feel, better than their stats would suggest- Daniel Vettori for instance. But even though i find cricketing history rather interesting, there's no way i could compare Sunil Gavaskar, based on a series of writings and an analysis of his stats, to Matthew Hayden who i've seen play countless times. Stats, writings and even performances are all relevant to the time at which they happen, so it can't really be done by anyone who hasn't seen both in the flesh IMO.

 

[Uppercut]

So, statistics are merely tools in the hands of a person to arrive at some conclusion that he wants to get. To say that they are end of everything is sadly an argument that has no basis.

To be fair, some work done by posters like Days of Grace (and others elsewhere) has simply analysed stats for all players using a system into which you put each player's numbers and just see who comes out on top. There were of course still rankings that most would consider highly questionable, but you couldn't say it wasn't a fair, unbiased analysis.