New feature - Batting impact, with a word on Trumper

[chasingthedon]

Dave looks at batting impact as compared to traditional batting averages, and attempts to set the record straight as regards Victor Trumper's somewhat average average of 39.04.

A study in impact - batting part one, with a final word on Trumper | Cricket Web

 

[Prince EWS]

Interesting stuff. Trumper benefited quite heavily when I used a completely different method to sort of standardise performances across time as well (his equivalent average went from 39 to 49 -- just higher than Ricky Ponting and Steve Waugh).

It's often said that Trumper's Test batting average didn't reflect his talent, but when people say this there tends to be an implication that he under-performed relative to that talent. In reality I think his batting average without context not only belies his ability, but his actual Test performances across his career as well.

 

[chasingthedon]

Cheers mate. Haven’t done Ponting and Waugh yet, but that’s interesting.

Did you publish your work?

 

[Prince EWS]

Cheers mate. Haven’t done Ponting and Waugh yet, but that’s interesting.

Did you publish your work?

Just on the forum. I've been contemplating writing an article for it but I still consider the methodology to be a bit of a work in progress (it's gradually got better and better since I started the thing eight years ago). I want to save that for the more finished product.

Thread is here:
http://www.cricketweb.net/forum/cri...test-averages-2018-update-post-697-698-a.html

And here's a Google Doc with a list of rankings using the 'alternate' longevity calculation method that people other than me seem to like more:
https://docs.google.com/spreadsheets/d/1Dulfao4Zy_gwmHePVRvBD6MsGAw2m4C4R7Eddnuw3Uk/edit#gid=0

This is a bit of a derail from your work though, haha. I've got mine automated and it looks like a lot of your work is still manually calculated, so let me know if you'd like any help automating it for you -- depending on the specifics it might be easily doable.

 

[Victor Ian]

Great stuff CTD. Always love your analysis (and any pews input too) . Nice to see Trumper in better company on a list than he normally gets.

 

[chasingthedon]

Just on the forum. I've been contemplating writing an article for it but I still consider the methodology to be a bit of a work in progress (it's gradually got better and better since I started the thing eight years ago). I want to save that for the more finished product.

Thread is here:
http://www.cricketweb.net/forum/cri...test-averages-2018-update-post-697-698-a.html

And here's a Google Doc with a list of rankings using the 'alternate' longevity calculation method that people other than me seem to like more:
https://docs.google.com/spreadsheets/d/1Dulfao4Zy_gwmHePVRvBD6MsGAw2m4C4R7Eddnuw3Uk/edit#gid=0

This is a bit of a derail from your work though, haha. I've got mine automated and it looks like a lot of your work is still manually calculated, so let me know if you'd like any help automating it for you -- depending on the specifics it might be easily doable.

oh yes, of course I was aware of your standardised averages, been a while since I last looked though, will get up to speed.

 

[chasingthedon]

This is a bit of a derail from your work though, haha. I've got mine automated and it looks like a lot of your work is still manually calculated, so let me know if you'd like any help automating it for you -- depending on the specifics it might be easily doable.

Don't worry about promoting your work, I'm all about getting to the bottom of things regardless of who gets there.

As regards automating my work, thanks for the offer. The difficulty was in converting the specific match status into a win probability. I have a database with all of the FOWs, and I was using an Access query to find all of the matches which enjoyed that status, and converting that to a win-draw-loss probability. So for example, if it's 252/7 in the second innings and 102 behind, how many matches has that situation been duplicated? That is, how many matches has the team batting second been 102 behind with three wickets to give in their first innings? Note that isn't just an FOW point, but e.g. a match where the team batting second went from 151 behind with seven first innings wickets down 47 behind with eight wickets down, passed through that match situation, i.e. 102 behind with seven wickets down.

There were other difficulties, such as how to incorporate the impact of time remaining which for most Tests is not easily doable (all we have is stumps scores). Then there are the complications of retired hurts coming back in etc. Also when did the extras get scored, for most Tests for which there are no ball-by-ball?

So much of what is taking my time now is double-checking the automated output is not effed up by necessary corrections.