First thread - Who will be the top run scorer and wicket taker?

[Jono]

But (some) batsman do tend to be more likely to have a poor innings after a string of good ones....

Unless you clearly define what is a string of good ones, this is kind of meaningless isn't it?

 

[Valer]

Unless you clearly define what is a string of good ones, this is kind of meaningless isn't it?

I haven't done my own analysis but he used arithmetic mean (average) over the last 5 outs*.


*The last innings of players finishing n.o were ignored.

 

[Prince EWS]

But that's not the point PEWS is saying. All he is saying is that people saying someone is due a poor series is a stupid statement with no genuine argument or even remote evidence behind it. Its a figment of everyone's imagination.

In fact there was nothing stats monger about that post you quoted of PEWS. He may have used stats terminology (i.e mean) but really, what he said was actually just common sense.

I actually thought honestbharani was agreeing with me - that trying to guess what a player would score based on what he'd scored in his last five innings or whatever completely ignores the fact that it's actually a contest between bat and ball out there; not a programmed simulation.

I've actually completely rejected any sort of statistical analysis on this.

 

[Prince EWS]

But (some) batsman do tend to be more likely to have a poor innings after a string of good ones...

And some don't. It's a coincidence. No-one has had a career long enough to make strings like this statistically significant. I'm sure if you generated a series of random numbers then you'd find what appeared to be repeating patterns and quirks amongst those too but it'd just be mental masturbation.

Batting scores aren't random, but that just strengthens the case for that "analysis" to be ridiculous, really. They're a product of your physical performances out there in the middle with the bat against the bowlers you're facing and the fielding side; they're not a product a statistical sequence or randomly generated, and they're certainly not a product of the cricket gods deciding your recent record is flattering/unflattering and that you're due for a bad/good score.

So, all in all, I actually agree with honestbharani here. As you all know I love my stats but the sports world of late has been infected by what I like to call "voodoo stats" - statistics without a logical explanation that are clearly products of uncontrolled variables, poor sample sizes or just good old fashioned coincidences. Certain batsmen being more or less likely to score big after a string of five (totally arbitrary from what I can see) big innings is just another one of those.

 

[uvelocity]

should change your location to 'voodoo stats'

i'd change mine but im fond of seamy road

 

[honestbharani]

But that's not the point PEWS is saying. All he is saying is that people saying someone is due a poor series is a stupid statement with no genuine argument or even remote evidence behind it. Its a figment of everyone's imagination.

In fact there was nothing stats monger about that post you quoted of PEWS. He may have used stats terminology (i.e mean) but really, what he said was actually just common sense.

I was talking about the people who come up with those articles reg. form and being due to score and all that.. Not necessarily PEWS.. should have clarified that

 

[honestbharani]

I actually thought honestbharani was agreeing with me - that trying to guess what a player would score based on what he'd scored in his last five innings or whatever completely ignores the fact that it's actually a contest between bat and ball out there; not a programmed simulation.

I've actually completely rejected any sort of statistical analysis on this.

Yep.. I am having a go at the guys who think all of that actually makes sense.. Predicting a game between human beings with stats.. Sure, that will end well...

 

[Jono]

I actually thought honestbharani was agreeing with me - that trying to guess what a player would score based on what he'd scored in his last five innings or whatever completely ignores the fact that it's actually a contest between bat and ball out there; not a programmed simulation.

I've actually completely rejected any sort of statistical analysis on this.

I was talking about the people who come up with those articles reg. form and being due to score and all that.. Not necessarily PEWS.. should have clarified that

Yep.. I am having a go at the guys who think all of that actually makes sense.. Predicting a game between human beings with stats.. Sure, that will end well...

my bad

 

[Valer]

And some don't. It's a coincidence. No-one has had a career long enough to make strings like this statistically significant. I'm sure if you generated a series of random numbers then you'd find what appeared to be repeating patterns and quirks amongst those too but it'd just be mental masturbation.

You can easily get statistical significance on a difference in slope t-test. If I had his data I'd show you but I cbf setting up something to pull it.

Heres some methods if you feel like doing it yourself.
https://www.ets.org/Media/Research/pdf/RR-09-32.pdf

Batting scores aren't random, but that just strengthens the case for that "analysis" to be ridiculous, really. They're a product of your physical performances out there in the middle with the bat against the bowlers you're facing and the fielding side; they're not a product a statistical sequence or randomly generated, and they're certainly not a product of the cricket gods deciding your recent record is flattering/unflattering and that you're due for a bad/good score.

How is any of this a problem, you're describing a typical data set. Cricket is not a special flower.

So, all in all, I actually agree with honestbharani here. As you all know I love my stats but the sports world of late has been infected by what I like to call "voodoo stats" - statistics without a logical explanation that are clearly products of uncontrolled variables, poor sample sizes or just good old fashioned coincidences. Certain batsmen being more or less likely to score big after a string of five (totally arbitrary from what I can see) big innings is just another one of those.

Spend 5 minutes to read what's been written there. You don't understand what he's written.

How is the existence (h0 form does not exist) form not a explanation? As for anti-form don't bowlers tend to have better plans against in form batsmen, also see playing for your place.


As an aside are you suggesting form doesn't exist at all or just that some players don't slump/improve after a series of good/bad innings?

 

[Adders]

I cbf setting up something to pull it.

This is the trouble with nerds.........just lay back and let nature take its course man!!

Heres some methods if you feel like doing it yourself.

Oh my goodness.......so many possibilities. But I think I'll stop now.

 

[Furball]

Didn't Pickup do some sort of analysis years ago which basically concluded that there was no such thing as form?

 

[Valer]

Didn't Pickup do some sort of analysis years ago which basically concluded that there was no such thing as form?

I haven't see it but...
Note the changing sign. If he did an anaylsis on a group of players rather than an individual you'd likely get less conclusive results.

Adders

 

[honestbharani]

my bad

and as off topic as it is, loving Ross's batting right now

 

[LongHopCassidy]

But (some) batsman do tend to be more likely to have a poor innings after a string of good ones....

Ah, but how long is a piece of string?

 

[luckyeddie]

Trott to score the most runs

Broad to take most wickets

 

[luckyeddie]

Didn't Pickup do some sort of analysis years ago which basically concluded that there was no such thing as form?

I believe that Salman Butt and Mohammad Asif had form

 

[Jono]

luckyeddie, what a huge appearance!

 

[GIMH]

Yes yes yes

 

[stevewebster]

Cook and Anderson are the safe bets for England but for those who like a flutter go for Trott and Swann.

As for Australia the bookies have Clarke (by a mile) and Pattinson as favourites. Should the pressure of captaincy and expectation be too much for Clarke chuck some money on Haddin at 10-1. The bowling?...go with the odds and say Pattinson.

 

[grecian]

Cook and Anderson are the safe bets for England but for those who like a flutter go for Trott and Swann.

As for Australia the bookies have Clarke (by a mile) and Pattinson as favourites. Should the pressure of captaincy and expectation be too much for Clarke chuck some money on Haddin at 10-1. The bowling?...go with the odds and say Pattinson.

Talking of odds, I've got money on you to win the BMW this week. Welcome anyway