Welcome

Hello.

Welcome to the Bat and Ball Brimborion.
This is a blog about numbers (mostly).
Cricket statistics (mostly).
But they could be any numbers.
Or anything else that I may feel like rambling on about.
Whatever may interest me at the time.
Enjoy.
And, in case you are wondering:
Brimborion – n. Something useless or nonsensical. From ‘The Superior Person’s Second Book of Words’ by Peter Bowler (not the first-class cricketer).

Andrew

Wednesday, October 31, 2007

She's Out!

An amazing thing happened in the Women’s Provincial League match between KwaZulu-Natal and South Western Districts at Chatsworth last weekend: Johmari Logtenberg got out.

This is the first time since the 2003/04 season that she has been dismissed before January in this competition. In 2004/05 she was dismissed (for 110) on 15 January in her 5th innings of the season. She made 377 runs before being dismissed that season. In 2005/06 she was not dismissed until 1 April in her 9th and final innings of the season, having made 464 runs before this dismissal. And, last season she was not dismissed until 25 February in her 4th innings, having scored 227 runs before being dismissed.

Her career record in 29 matches in this competition is 1462 runs at an average of 146.20. She has been not out 17 times in 27 innings and scored 3 centuries and 11 fifties.

Sir Donald would be impressed.

Tuesday, October 30, 2007

Nelson, Nelson and Nelson

Pete le Grange has pointed out that Western Province lost wickets on 111, 222 and 333 in their second innings of their South African Airways Provincial Three-day Challenge match against KwaZulu-Natal at Cape Town last weekend. With the help of CricketArchive, I can reveal that, as far as can be ascertained, this is the first time in first-class cricket that wickets have fallen on each of these scores in the same innings. I wonder what Napoleon would have made of this?

In Test cricket a total of 175 wickets have fallen on 111, but a few more (186) have fallen on 112. Do players relax slightly on getting past 111? Or, as is more likely, is it pure coincidence? If you are Australian you may be interested to know that your team have only lost 23 wickets on 87 while losing 31 on 86 and 29 on 88.

There are two occasions in Test cricket where a team has lost 3 wickets on 111: South Africa v England at Lord’s, Durban in 1913/14 and England v South Africa at Lord’s (the one in London) in 1955. In the first example it was South Africa’s last three wickets of their second innings that fell on 111 to end the game, while in the second case England lost 3 wickets on 111 in their first innings. Somewhat weirdly, this match ended with South Africa losing their last two wickets in the second innings on 111 (these were the 8th and 9th wickets as Jack Cheetham did not return from retiring hurt). This one may be beyond even Napoleonic explanation.

In first-class cricket, Ireland lost 4 wickets on 111 v Derbyshire at Buxton in 1947 and Transvaal B did likewise against Rhodesia at Salisbury in 1970/71. And in 1988 at Nuneaton, Warwickshire lost 5 wickets on 222 against Lancashire. What would Richie have made of that?

Saturday, October 27, 2007

NFL at Wembley

I see the New York Giants are playing the Miami Dolphins in an NFL (American Football) game at Wembley this weekend. Is this the end of civilisation as we know it? Or, the beginning of civilization as we know it? Or both? Or neither?

West Indies Under-19

For those who may not have noticed, West Indies Under-19 lived up to their name by being bowled out for 18 by Barbados in a KFC Cup game last week. This is the lowest List A total ever.

Tuesday, October 23, 2007

The Partnership Breakers

“He is a good partnership breaker, isn’t he?” We have all heard this statement quite often when a bowler like Sourav Ganguly or Paul Collingwood comes on to bowl. Most notably a TV commentator once said it as Collingwood was coming on to bowl. His career stats flashed on the screen. At that stage he had a career record of 0 for about 200. “So which partnership did he break then?”

So, who are the partnership breakers of Test cricket? We can work out the average partnership broken. The clear leader of bowlers with 20 or more wickets is former England captain Norman Yardley. He took 21 wickets at an average of 33.66 in 20 Tests. But more significantly the average partnership that he broke was 92.19. This is well clear of 2nd placed Viv Richards’ average of 67.18 for his 32 wickets. The top five are:

NameYearsWktsTotPartsAvgPart
NWD Yardley (Eng)1938-195021193692.19
IVA Richards (WI)1974-199132215067.18
GA Hick (Eng)1991-200123152466.26
RR Sarwan (WI)2000-200723146563.69
CD McMillan (NZ)1997-200528170760.96


Yardley, incidentally, was also the first captain to be dismissed for 99 in a Test which he did against South Africa at Nottingham in 1947.

Paul Collingwood now has 6 Test wickets with quite a decent Average Partnership Broken of 57.16. Ganguly’s figure is a lower, but still respectable 44.53 for 28 wickets.

Saturday, October 20, 2007

Strike-rates in declared 3rd innings

When teams are looking to declare in the 3rd innings and set a target it obviously helps if you can score your runs quickly. We have already seen in the earlier post which players have good averages. Now, let’s look at players for whom we have full details of balls faced in their career with over 500 runs in declared 3rd innings in Test cricket. Kumar Sangakkara has the highest strike-rate here of 64.95, followed by Inzamam-ul-Haq (63.70) and Sanath Jayasuriya (63.69). Perhaps of more interest though is the degree to which the batsman can increase his strike-rate in this situation. So, ranking these players by the difference in their 3rd innings strike-rate and their overall strike-rates the leader is Graham Thorpe, who scored 959 runs in 16 declared 3rd innings at an average of 106.55 and a strike-rate of 58.76 compared to his career strike-rate of 45.89. The following are the top five:

NameDecl 3rd innsCareerDifference
GP Thorpe (Eng)58.7645.8912.87
SC Ganguly (Ind)62.4349.6812.75
WJ Cronje (SA)55.9444.5011.44
Inzamam-ul-Haq (Pak)63.7054.009.70
Younis Khan (Pak)63.3053.759.55


At the other end of the scale Justin Langer has a strike-rate of 46.48 in declared 3rd innings, which is 7.75 lower than his career 54.23 and Herschelle Gibbs is 45.49 compared to a career 50.07. Not surprisingly, it seems that batting lower down the order when the declaration is imminent helps here rather than opening the batting when the game may still be in the balance.

If we decrease the qualification to 300 runs an interesting name appears at the top: Freshly retired Craig McMillan scored 305 runs in 5 declared 3rd innings at a strike-rate of 79.63 which is 24.71 ahead of his career strike-rate of 54.92.

And Jacques Kallis? His strike –rate is 44.08, slightly higher than his career figure of 43.13.

Friday, October 19, 2007

Bradman celebration

On 6 August next year Sir Donald Bradman would have been 99 years, 345 days old had he been alive. What, you may well ask, is the significance of this? Well, 345 is 0.94 of 366 (next year being a leap year). So, Bradman would have been 99.94 years old on this day.

Now, his centenary is on 27 August 2008. But, I submit that 6 August 2008 is a far more appropriate day to celebrate the great man.

So, where’s the party?

Thursday, October 18, 2007

A bit about the blog

Here are some random notes on the blog:

Clicking on the ‘View my complete profile’ link doesn’t provide anything more useful than links to other ‘Statisticians’ and ‘South Africans’ with blogs on Google’s Blogspot. Which mostly won’t be very helpful. I am not sure how I can get rid of this without deleting the profile bit.

I am planning to try to post something at least every 2 to 3 days and maybe more often.

You can post comments, but I have to ‘moderate’ them and they will only appear after I have done this. I may or may not respond to comments.

In reply to some of the comments:

I am planning on doing a follow-up on the 'Declaration Innings' with reference to strike-rates.

‘Zscore’ has neatly replied to the query from ‘Statcat’ about Inzamam-ul-Haq and run outs in the comments under ‘Milestone for first-class cricket’. I can just add that you can link to this to get more on the subject:

http://content-www.cricinfo.com/ci/content/story/224522.html

Basically, Imzamam is only mid-table in this department. Steve Waugh’s record of being run out only 4 times himself while his partner is run out 23 times is a big difference and a sufficiently large enough sample that one can conclude that this appears to more of a skill than pure chance, especially given that he also extends this form into ODIs.

One of my plans with the blog is to try and introduce new ways of measuring various things, or creating ways of measuring things that are not currently measured. The bit on wicket-keepers yesterday is an example. These measures may, or may not, add anything to what we know and will probably also be refined over time. Some of this stuff might be quite heavy going, but fortunately, I won’t do it too often if for no other reason than that it is usually quite a lot of work.

The good news is that tomorrow’s post is going to be very short and sweet and spectacularly meaningless - a true brimborion.

Wednesday, October 17, 2007

Is Mark Boucher the new Jack Blackham?

As we know, Mark Boucher recently broke the world record for wicket-keeping dismissals and became the first keeper to take 400 dismissals. This invariably starts people wondering about the question of who the best wicket-keeper ever is.

I don’t plan on answering that question here, but I am going to make a suggestion as to how we can create measurements that allows us to more accurately compare keepers over time.

We do not have any reliable data on dropped catches and missed stumpings, so we cannot create a ‘fielding percentage’ similar to baseball. There are two measurements that we do have of wicket-keepers: dismissals and byes conceded. For the purposes of measurement, I am going to look at dismissals as a percentage of team dismissals and byes conceded per 100 balls kept. And, although wicket-keeping purists may not like it, we can also consider batting ability. So, with apologies to the Keith Andrews and Richie Ryalls of the world, I am going to include batting average.

Now, the role of the keeper, and with it the measurements have changed over time. A lot of this is to do with the fact that keepers used to stand up to the stumps a lot more in the past than they do today. In modern cricket keepers have a higher percentage of dismissals, lower rates of conceding byes, and are expected to score more runs than in the past. The table below shows how these numbers have changed over time:


PeriodBat Avg%Dis%Byes
1877-191416.4311.761.28
1919-193923.5614.230.93
1946-197022.4517.390.70
1971-199025.1019.380.63
1990-200729.7520.430.62


So, let us consider Mark Boucher and Jack Blackham. Blackham played in the first ever Test in 1877 and represented Australia in 35 Tests between then and 1894. Boucher has a dismissal percentage of 23.09, byes conceded rate of 0.65 and a batting average of 30.21. Blackham’s figures are 11.36, 1.38 and 15.63, respectively. Boucher’s figures are basically twice as good as Blackham in all departments.

My suggestion is that we create an era adjusted measure by dividing the keepers stat by the overall average of all other keepers who played in the years that they played and multiplying it by the current average (i.e.20.43 for dismissals, 0.62 for byes and 29.75 for batting from the table above).

This produces a very interesting result if we compare Boucher to Blackham. Boucher’s adjusted figures are 23.80 for dismissals, 0.61 for byes and 29.40 for batting. No great change as there, as he is currently playing. Blackham has 23.59 for dismissals, 0.79 for byes and 29.46. Which are remarkably similar, especially in dismissals and byes.

So, here are the tables for the leading 5 keepers in each of the adjusted measures (with a minimum of 20 Tests as wicket-keeper):

Percentage dismissals

NameAdjustedActual
GRA Langley (Aus)29.0221.92
AFA Lilley (Eng)28.5315.54
CO Browne (WI)28.1327.36
H Strudwick (Eng)27.0116.47
ATW Grout (Aus)25.0121.61



Byes conceded per 100 balls


NameAdjustedActual
H Carter (Aus)0.270.74
PR Downton (Eng)0.270.28
DJ Richardson (SA)0.340.33
WAS Oldfield (Aus)0.360.64
Khaled Mashud (Ban)0.360.40



Batting average


NameAdjustedActual
LEG Ames (Eng)63.4643.40
A Flower (Zim)56.8053.70
AC Gilchrist (Aus)49.4448.66
Imtiaz Ahmed (Pak)44.5630.45
H Carter (Aus)44.3722.97


Note that only matches where the player was the designated keeper are included. If they played as a specialist batsman this is excluded. We do not have full details for when keepers were replaced during a match, so it is assumed that the keeper kept for the whole game. This could affect the byes calculation marginally.

Note also that this does not suggest that Les Ames, for example, would average 63.46 if he were playing today. It is simply a measure compared to contemporaries, i.e. if Ames were playing today and was as far ahead of current keepers in terms of batting average as he was in his day then he would average 63.46.

Of course, we still do not have a measure for that other great skill required from modern wicket-keepers: Encouraging the bowlers. You know, the incessant “Great bowling … “ comments after each ball. Sometimes emitted mere nanoseconds before the half-volley is eased to the mid-wicket boundary. But, the TV technology whizzes may be working on this as we speak. Soon ‘Keeper-o-Meter’ will give us measurements of decibel levels, volume and quality of encouragements. And then statisticians will be able to correlate this to wickets falling.

Monday, October 15, 2007

Milestone for first-class cricket

First-class cricket reached the milestone of 50 000 matches this weekend, according to the Association of Cricket Statisticians and Historians list of first-class matches. The first match on this list was played in 1801. The Pura Cup match between Western Australia and New South Wales in Perth is the match that has the honour of being number 50 000 chronologically, starting a few hours after the South Australia v Victoria match on Sunday 14 October.

By comparison, Major League Baseball, in case you are interested, has seen about 193 000 games since 1871.

The 10 000th first-class match was played in 1922, 20 000 was reached in 1953, 30 000 in 1974 and 40 000 in 1992.

The 50 000 matches have taken place over about 160 000 days and, at a very rough guess, 750 000 hours.

The County Championship accounts for 20 402 (over 40%) of the matches.

Around 5 000 (10%) have been played in London alone, 2 709 of them at Lord’s.

There have been approximately 39 million runs scored and 1.5 million wickets taken in the equivalent of about 13.5 million 6-ball overs in first-class cricket to date.

Over 3 500 batsmen have been out hit wicket (14 by Arthur Milton alone, and 6 by Gavin Cowley in just 87 dismissals) and nearly 180 000 lbw. How many appeals there have been is beyond reasonable estimation.

And over 10 000 catches have been taken by subs.

Saturday, October 13, 2007

Do Test pitches deteriorate over 5 days?

Conventional wisdom is that Test pitches should become more difficult to bat on over the course of 5 days. Pakistan saved the Test in Lahore against South Africa relatively easily. There also appear to have been a number of successful high 4th innings run-chases in recent times.

So, the question is: Do modern Test pitches still deteriorate over 5 days as they presumably did in the past?

The short answer is yes. The table below gives the average runs per wicket by match day for Test matches that lasted exactly 5 days since the Second World War. Shorter games, and the few that have lasted longer than 5 days, in this period are excluded.

Day1946-19791980-19992000-2007
140.4141.0548.53
232.5233.8440.37
334.7236.6436.75
430.9033.4233.67
528.4929.8928.18

So, if anything modern Tests have a greater trend of runs per wicket steadily decreasing over the 5 days than in the past. Which suggests that pitches still behave as we would expect them to over the course of 5 days.

Friday, October 12, 2007

Jacques Kallis: Master of the declaration innings

Jacques Kallis likes batting in the 3rd innings of a Test when South Africa are setting up a declaration.

His 107* in South Africa’s 2nd innings against Pakistan at Lahore yesterday is the 5th century that he has made in the 3rd innings of a match in which South Africa has declared. In all, he has batted 19 times in these situations and scored 1264 runs with 5 centuries and 6 fifties. With the help of 10 not outs, he averages 140.44 in these innings.

Only Matthew Hayden has more runs and centuries in 3rd innings that have been declared. He has scored 1352 runs and 8 centuries in 18 innings in this situation. As an opening batsman, he has been less adept at collecting asterisks when the declaration is imminent than Kallis and his average is a mere 79.52 by comparison.

Another master of 3rd innings declaration innings was Garry Sobers. In 10 innings he scored 906 runs at a useful average of 181.20. Like Kallis he has 5 centuries.

And, in case you are wondering what Don Bradman did in these situations: 3 innings, 241 runs, average 120.50, 1 century and 1 fifty.

Thursday, October 11, 2007

Peter Bowler's brimborion

I found the word ‘brimborion’ in a book by Peter Bowler, an Australian lexicographer. Not to be confused with Peter Bowler the first-class cricketer.

But this reminded me of the main contribution that Peter Bowler (the first-class cricketer, not the Australian lexicographer) made to cricket’s reservoir of trivia. Or, cricket’s reservoir of brimborion if you prefer. Bowler was the first player to score a century on first-class debut for two different counties. He scored 100* for Leicestershire v Hampshire at Leicester in 1986 and then 155* for Derbyshire v Cambridge University at Cambridge in 1988. In 1995 he made his first-class debut for his third county, Somerset, and made a bold effort to add them to the list before running out of partners on 84* in the 2nd innings v Glamorgan at Taunton.

Three players, Jonathan Lewis (Essex and Durham), Neil Taylor (Kent and Sussex) and Andrew Symonds (Gloucestershire and Lancashire) have subsequently joined Bowler in this rare feat.

In South African first-class cricket there are three players who have made a century on debut for two different provincial teams (excluding ‘B’ teams). They are Ken Viljoen (Orange Free State and Transvaal), Terence Lazard (Western Province and Boland) and Arno Jacobs (North West and Eastern Province). Lazard is the most interesting here, as he also scored a century on his overall first-class debut which was for Western Province B. His 307* on debut for Boland v Western Province at Worcester in a pre-season friendly in 1993/94 single-handedly put an end to first-class status for pre-season friendlies.

One final point on Bowler: In an interview with Wisden Cricketer Monthly on his retirement from first-class cricket he said that he regarded an innings of 14 that he made for Derbyshire v Warwickshire in Birmingham in 1992 as the best of his career. Yes, 14! He opened the batting with an hour to go on the first day on a green-top against a rampant Allan Donald. He was 8* overnight out of 21-5 and Donald finally knocked him over for 14 the next morning. The man scored nearly 20 000 first-class runs with 45 centuries and yet regards an innings of 14 as his best. Such are the real joys of cricket.