James018's Match Engine Analysis

liquefry
Posts: 432
Joined: Wed May 18, 2016 10:10 am

Re: James018's Match Engine Analysis

Postby liquefry » Fri Jun 21, 2019 4:18 am

Necro (sorry) - James do you still have the raw data?

James018
Posts: 1359
Joined: Sat Apr 23, 2016 12:30 am

Re: James018's Match Engine Analysis

Postby James018 » Sat Jun 22, 2019 7:35 am

I do. Can also collect new raw data if needed.

Stomalomalus
Posts: 376
Joined: Thu Jul 20, 2017 7:20 am

Re: James018's Match Engine Analysis

Postby Stomalomalus » Sat Jun 22, 2019 9:31 am

I don't have access to the data yet, but is there a breakdown of ball quality vs Batsman order? That a Bowler may ball worse quality later, as he's tired, but the Batsman is worse quality so makes more mistakes?

James018
Posts: 1359
Joined: Sat Apr 23, 2016 12:30 am

Re: James018's Match Engine Analysis

Postby James018 » Sun Jun 23, 2019 1:16 pm

Only just realised what all the recent spreadsheet access requests I got lately were about :lol: Sorry. I didn't know what I was doing with Google Drive at the time.

Here's a link to an actual viewable sheet with the data and the analysis I've done so far: https://docs.google.com/spreadsheets/d/1BhAN6_d2S7eileG1prF-zKCfdjpZ6J3lC2dFbnyIUco/edit?usp=sharing (note there are multiple tabs at the bottom and you'll need to switch between them to see everything).

What I've looked at so far is:

  • The runs/wickets/etc. broken up by the time in the innings (blocks of 5 overs), the speed of the ball, ball quality, line, the batsman's shot and the contact level.
  • The association each of these factors have with each other (e.g. time in innings vs ball speed, ball quality vs line)
  • Results broken down by shot type and contact level (e.g. perfect slogs, well hit attacks)
  • Results broken down by time in innings and ball quality (e.g. 10/10 balls in the first 5 overs, 4/10 balls from overs 31-35)

Happy to take requests on others.

To answer your question, Stomalomalus, I haven't analysed ball quality by batsman order. It's hard to figure out whether a batsman comes in at number 1 or number 10 simply through the raw commentary data. Like I said, though, I have looked at ball quality vs time in innings, and it definitely seems like ball quality gets worse later on in the innings.

Stomalomalus
Posts: 376
Joined: Thu Jul 20, 2017 7:20 am

Re: James018's Match Engine Analysis

Postby Stomalomalus » Sun Jun 23, 2019 1:44 pm

James018 wrote:Only just realised what all the recent spreadsheet access requests I got lately were about :lol: Sorry. I didn't know what I was doing with Google Drive at the time.

Here's a link to an actual viewable sheet with the data and the analysis I've done so far: https://docs.google.com/spreadsheets/d/1BhAN6_d2S7eileG1prF-zKCfdjpZ6J3lC2dFbnyIUco/edit?usp=sharing (note there are multiple tabs at the bottom and you'll need to switch between them to see everything).

What I've looked at so far is:

  • The runs/wickets/etc. broken up by the time in the innings (blocks of 5 overs), the speed of the ball, ball quality, line, the batsman's shot and the contact level.
  • The association each of these factors have with each other (e.g. time in innings vs ball speed, ball quality vs line)
  • Results broken down by shot type and contact level (e.g. perfect slogs, well hit attacks)
  • Results broken down by time in innings and ball quality (e.g. 10/10 balls in the first 5 overs, 4/10 balls from overs 31-35)

Happy to take requests on others.

To answer your question, Stomalomalus, I haven't analysed ball quality by batsman order. It's hard to figure out whether a batsman comes in at number 1 or number 10 simply through the raw commentary data. Like I said, though, I have looked at ball quality vs time in innings, and it definitely seems like ball quality gets worse later on in the innings.


And the number of wickets with lower quality balls also goes up...

Which has given me an idea... :)

bumpuss
Posts: 409
Joined: Sat Apr 23, 2016 12:30 am

Re: James018's Match Engine Analysis

Postby bumpuss » Sun Jul 14, 2019 8:11 am

fasil Fartland
Top Order: 24,128 56,244
Middle Order: 18,302 53,746
Tail Batting: 15,018 24,864
Powerplay Bowling: 26,526 51,448
Mid-Innings Bowling: 23,793 57,917
Death Bowling: 22,391 53,899
Fielding: 31,328 51,643
Overall: 23,069 49,966


Bhavik Arasaratnam is out for 4 from 2 balls, Tanaka Yeye will replace him.
0.4 W Imam to Arasaratnam 99 kph 8/10 off stump
push
skewed tough infield CATCH for Connor.
Arasaratnam is very unlucky getting out there.
0.3 4 Imam to Arasaratnam 100 kph 8/10 off stump
block
Sweet Boundary! through a gap in the infield
They are consolidating now
Nyika Zaca is out for 0 from 2 balls, Bhavik Arasaratnam will replace him.
0.2 W Imam to Zaca 102 kph 8/10 off stump
block
total miss LBW!
0.1 . Imam to Zaca 99 kph 7/10 middle stump
block
well hit


So what does 8 ball rating mean here? If the average is 6 for the whole match, given that my bowlers are twice as good as theirs by ratings, does this mean I am more likely to get twice as many "8" ball ratings as their bowlers?

OR does this mean, an 8 ball rating by one of their bowlers, is equivalent to 8 ball rating by one of my bowlers even though my bowlers are twice as good?

James018
Posts: 1359
Joined: Sat Apr 23, 2016 12:30 am

Re: James018's Match Engine Analysis

Postby James018 » Sun Jul 14, 2019 11:47 am

"Twice as good" doesn't necessarily correspond to double the amount of good balls, or double the average ball quality, or anything like that.

But in essentials, the first option you gave is correct. The ball quality of all bowlers in the match is standardised to the average of the bowlers on both sides, so your bowlers will be getting far better ball qualities than theirs, on average. You'll be able to see this in the average ball qualities of each bowler, once the match is finished.


Return to “Match Engine”

Who is online

Users browsing this forum: No registered users and 1 guest