High Impact - Assessing the Great Batting Performances

[hang on]

ikki,

have read.

but this is why i don't agree:

Yeah that's the obvious issue with this (and other similarly-justified) measures, it's pretty tough to filter batsmen who come in at the top of the collapse and then survive it.

Ah, actually this is where this analysis fails. If a batsman comes in at 1/100 chasing 350, suddenly sees four partners depart in the space of 10 runs, and then puts on a 240 run partnership with the No. 7 batsman, he won't get much credit at all compared to the No. 7 batsman (if I'm understanding the method correctly). Langer probably got screwed in this analysis in that Hobart Test match against Pakistan where they chased 360.

 

[Spark]

I've seen similar measures used in rating batsmen "in pressure situations". The central problem is that you can't statistically define a "pressure situation" because beyond the obvious "wickets can fall in clumps just after you got in", there's insane amounts of unquantifiable context involved in cricket which is impossible to ignore. A batsman coming in at 3/200 can be under more pressure in certain circumstances than one coming in at 5/50, if the context dictates.

 

[Agent Nationaux]

I've seen similar measures used in rating batsmen "in pressure situations". The central problem is that you can't statistically define a "pressure situation" because beyond the obvious "wickets can fall in clumps just after you got in", there's insane amounts of unquantifiable context involved in cricket which is impossible to ignore. A batsman coming in at 3/200 can be under more pressure in certain circumstances than one coming in at 5/50, if the context dictates.

Agree with this statement.

The 'match importance' criteria should be considered in this analysis. In a 5 match test series where the score is 2-2, with one more match needed to be played, the pressure in the 4th innings in much greater if the score is 2-1 with 2 matches remaining. In this instance we appreciate a match winning performance much more. However I understand that this would be possible if the person doing the analysis had more time on his hand.

 

[Spark]

Yeah but how do you quantify "match importance"? Probability of winning, ok, you can put a number to that in various ways, and then you can do a solid analysis from there.

 

[Agent Nationaux]

You could give it a score or a percentage I guess. I'm not sure, maybe somebody else would know better.

 

[Ikki]

ikki,

have read.

but this is why i don't agree:

That's not what you said though mate. You said you didn't agree merely because Sehwag and Sachin scored more in the same match.

I've seen similar measures used in rating batsmen "in pressure situations". The central problem is that you can't statistically define a "pressure situation" because beyond the obvious "wickets can fall in clumps just after you got in", there's insane amounts of unquantifiable context involved in cricket which is impossible to ignore. A batsman coming in at 3/200 can be under more pressure in certain circumstances than one coming in at 5/50, if the context dictates.

It's certainly not the norm though and as far as analysis goes these things always have a few holes here and there but the foundations are sound IMO. Your gripe is akin to saying a 50 in one match is better than an 80 in another. That may be the case but when you look at career averages you'd not say an average of 80 is worse than an average of 50. This is not to belittle those not on the list or champion those that are; but you'd think eventually there would be enough tough 4th innings opportunities in a career where a batsman will have to step up for him to get on this list.

I wonder if there was a way of taking these stats and creating a ratio of these kinds of tough runs accumulated instead of single innings here and there. Because, for example, a batsman may be in a very strong team and rarely need to face tough situations but may encounter a very difficult one and succeed - enough so to make this list - and yet you may have a batsman that is regularly in tough chases, and may have succeeded in them but not high enough %-wise to make this list.

 

[hang on]

That's not what you said though mate. You said you didn't agree merely because Sehwag and Sachin scored more in the same match.

i didn't mention sehwag. so, please avoid ascribing opinions to me that are not the case. also, i never mentioned that the tendulkar knock was more valuable or of higher impact purely by virtue of being a bigger score. so, again, do not put words into my mouth. here is what i said:

the singh innings of 80 odd that ranks ahead of the 103 of tendulkar's (in the same innings) surely gives the lie to this analysis.

regarding the lack of explanation: that's because i didn't have the time, or the wit and the verve at that time, to write it as well as those two. but i certainly had that in mind, especially the second one that i quoted. in fact i was coming back on to try and explain it after going off. but it was already done....and better than i could have managed, i'm sure. the lack of an explanation doesn't necessarily imply the lack of thought. mate.

 

[Ikki]

So how does the Singh innings show that it is a lie then? That innings was one where the match was won. You quoted Tendulkar's score (which was higher)? I am not sure what else you could have been referring to. VCS' point about lost matches is valid, but it has nothing to do with the innings you had a problem with because India won.

 

[hang on]

my argument is not about matches lost. and the vcs quote is not about matches lost either. i was referring to the fact that tendulkar stayed for longer, saw wickets go down (with the pressure that it entails), and was instrumental, with singh, in ensuring the target was reached. arguably, he was the fulcrum of the whole indian innings, while sehwag's was the initial impetus and singh's the final one. the greatest pressure, arguably, was on tendulkar. thus singh's innings being assigned a greater impact value makes me question the methodology underpinning the analysis. btw, i quoted tendulkar's score not to indicate its greater magnitude necessarily but as a reference or index, if u will, pretty much in the same way one refers to lara's 277 or tendulkar's 136 etc.

 

[Dissector]

Ah, actually this is where this analysis fails. If a batsman comes in at 1/100 chasing 350, suddenly sees four partners depart in the space of 10 runs, and then puts on a 240 run partnership with the No. 7 batsman, he won't get much credit at all compared to the No. 7 batsman (if I'm understanding the method correctly). Langer probably got screwed in this analysis in that Hobart Test match against Pakistan where they chased 360.

Your example is pretty much spot on about why this method is seriously flawed. It looks at the pressure when the batsman comes in but not what happens afterwards. And I agree that Tendulkar's century was clearly superior to Yuvraj's knock and it's only the flaw in the methodology which says otherwise. And Laxman's 73 should rank even higher; it was a ridiculously improbable innings which probably deserves a place in the top 5.

 

[Dissector]

Let me add that while I think this method is flawed, I appreciate the effort that's been put into it and I think with a few tweaks it could be a very useful tool.

 

[Ikki]

my argument is not about matches lost. and the vcs quote is not about matches lost either. i was referring to the fact that tendulkar stayed for longer, saw wickets go down (with the pressure that it entails), and was instrumental, with singh, in ensuring the target was reached. arguably, he was the fulcrum of the whole indian innings, while sehwag's was the initial impetus and singh's the final one. the greatest pressure, arguably, was on tendulkar. thus singh's innings being assigned a greater impact value makes me question the methodology underpinning the analysis. btw, i quoted tendulkar's score not to indicate its greater magnitude necessarily but as a reference or index, if u will, pretty much in the same way one refers to lara's 277 or tendulkar's 136 etc.

I see, I missed that point. Can certainly see your criticism now.

 

[hang on]

no worries. i should have explained it at the outset.

 

[vcs]

Your example is pretty much spot on about why this method is seriously flawed. It looks at the pressure when the batsman comes in but not what happens afterwards. And I agree that Tendulkar's century was clearly superior to Yuvraj's knock and it's only the flaw in the methodology which says otherwise. And Laxman's 73 should rank even higher; it was a ridiculously improbable innings which probably deserves a place in the top 5.

AWTA, especially when you take the injury into account. Easily one of the best 4-5 knocks of the decade, IMO.

 

[Sanz]

Well this is just another analysis, Like any other analysis, it can't be perfect but it definitely was interesting.

 

[hang on]

would also like to say, "good analysis" and thanks for doing it. didn't mean to come across as churlish yesterday when i criticized it.

 

[Isura]

huh ? use some logic will you ? so sehwag's innings was also over-rated as even with it 2 other 50s and a hundred was required ?

its insane in how many places people just can't use plain common sense ; common sense is not so common after all !

"common sense" related to sports analysis by casual fans is very often wrong

 

[Spark]

"common sense" related to sports analysis by casual fans is very often wrong

This. Cherry picking and selective memory, inherent bias, lack of perceptiveness of crucial subtleties and context, all reasons why "common sense" can often fail.

 

[chasingthedon]

Thanks for all the comments

Both complimentary and otherwise - I do appreciate everyone taking the time to read. It did, as someone mentioned, take a very long time as features go (several months getting the database up to scratch).

It certainly is not a perfect system, merely a proposal at this stage - what I always attempt to do with this kind of thing is to see if our perceptions are correct when we look at things more objectively.

I would like to take other factors into consideration in the future, e.g. importance in the series (Butcher would lose out in that regard) and also extend it to other innings. The latter is always difficult as there is often a lot of cricket still to be played, and certainly a first innings knock would suffer most in this regard.

Always welcome suggestions for improvement, now that I have the tools to check it. Also, please bear in mind I'm not saying innings A is "better" than innings B - this is merely one facet of the greatness of an innings.

I'm looking at the same thing right now as applied to bowlers, and again there are issues, but it's interesting to see the results and then enjoy a full and frank discussion.

 

[robelinda]

Surprised the two tons in WI chase of 418 didnt make the grade?