A post in which I am the cricketing version of David Hilbert

Blog
About

A post in which I am the cricketing version of David Hilbert

November 17, 2007 at 11:55 am | In Cricket | 8 Comments

So Dave posted on my facebook wall

“I’ve just started running the code that will build me my Test match database – all Test scorecards. The code is hopefully the worst MATLAB/Octave code ever written by someone who passed PHYS3071 – 1100 ghastly lines that take over ten seconds to take a Cricinfo HTML file and strip it down to the scorecard data.

So the point is, if you have any ideas on what I should do with such a dataset, let me know.”

Well here they are. Feel free to suggest some more ideas in the comments.

A bowling average weighted by the batting average of the batsman dismissed.
Is there a statistical sense in which Shane Warne is such a great bowler? His average isn’t particularly good.
Batsman’s averages against spin, pace, left handers …
A list of times that 5 or 7 or 8 runs have been scored off a single ball.
List of longest overs in test match cricket.
Umpiring statistics: Was Dicky Bird really a not outer?
Enforcing the follow on, good idea or not?
What advantage is winning the toss? Overall? Per country? Per decade?
Night Watchmen, do they score more runs than they would batting lower down? Is this good for the team?
Players who have improved their batting over the years. Recently Vettori has averaged 40 batting at number 8. McGrath increased his average from 2.5 to over 5. Who has improved the most?
Batsman who scored the most threes in test match cricket.
Most injury prone cricketer? In terms of unfinished overs? Or proportion of games that they would have been picked for but missed out on?

8 Comments »

RSS feed for comments on this post. TrackBack URI

Unfortunately my database isn’t going to have ball-by-ball data, just the raw details you get in a summary scorecard. So some of these will be impossible for me. Indeed, ball-by-ball data is pretty rare before Cricinfo came along, so even the big statisticians of the world only do analyses on them with matches from the last ten years or so.

Nightwatchmen are bad for the team. See Charles Davis.

Warne was ranked as the #10 bowler all-time by Davis in Wisden Australia 2005/06 – link (with some now partly dated discussion). If he took country-by-country analysis, Warne was fifth.

Spinner’s averages tend to be higher than those for quicks – in my head I always use an adjustment of 5 runs or so.

McGrath actually got his average up to 7.5 by the end of his Test career.

Comment by David Barry — November 17, 2007 #

Reply
More Charles Davis: the toss. It used to be advantageous (57-43 before WWII), but isn’t anymore (51-49 “in recent years”) because captains aren’t choosing to bowl first often enough.

Comment by David Barry — November 17, 2007 #

Reply
Marty

Comment by the real mum — November 18, 2007 #

Reply
I,ve lots of ideas , will discuss them with you when you’ re home. Do you arrive on 22nd Dec in Bris ? What time ? Are you stay ing any time in Bris? Sorry this is probably a misuse of your blog but my E-mail wont send and the phone line is eternally busy Looking forward to seeing you.

Comment by the real mum — November 18, 2007 #

Reply
Hi dad,

I’ll arrive in Brisbane at 8:20am on 22nd of Dec. I’ll probably come home straight away and then spend some time in Brisbane early January.

Have you got me a ticket to come back to the US of A yet?

Our apartment phone must not be working for some reason. I don’t think i’ve heard it ring for weeks. You could try my mobile if you want. I’ve got lots of credit that will run out soon. Number is (plus whatever you need to stick on the front).

Comment by martinleslie — November 18, 2007 #

Reply
Scratch that. Credit has run out.

Comment by martinleslie — November 18, 2007 #

Reply
I would be interested in knowing whether this theory has any merit in cricket: http://www.foxsports.com.au/story/0,8659,22801590-23214,00.html

Obviously there aren’t trades at a national level, but is there anything to it on a domestic level? Or on retirement at an international level?

Comment by Julian — November 22, 2007 #

Reply
I certainly hope the theory has merit in football. Then the Eagles might win the flag again, now they’ve lost Judd and Cousins. I’m not holding my breath.

I don’t think the idea has much merit in cricket. There may be personality clashes, and these have led to players missing out on selection, but the loss of a superstar is almost always bad. I would much rather have Warne than MacGill, for instance. And Warne is a useful example here – it’s clear that he didn’t get along with Steve Waugh all that well, but they put aside their differences on the field.

I think the Bulldogs will be a worse team without Mason. We’ll see how it goes next season.

(In defence of the idea in that article, we only beat the West Indies after Allan Border retired, and only won a series in India after Waugh retired.)

Comment by David Barry — November 22, 2007 #

Reply