• Welcome to the Cricket Web forums, one of the biggest forums in the world dedicated to cricket.

    You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, respond to polls, upload content and access many other special features. Registration is fast, simple and absolutely free so please, join the Cricket Web community today!

    If you have any problems with the registration process or your account login, please contact us.

The Barnes Standard

watson

Banned
In recent weeks on CW we've seen the landmark of 50 (or 52) Tests matches used to calculate the peak averages of great batsman so that we can compare them to Don Bradman. In the following article the author proposes a similar standard using the career of Sydney Barnes as the benchmark.

The Barnes standard

By Kartikeya Date

If Don Bradman is the best batsman in Test history, Sydney Barnes, it is frequently argued, is the best bowler. In 27 Tests, Barnes took 189 wickets at 16.43, one every 41.6 balls. Barnes' claim is not as strong as Bradman's, but until 1990, Barnes was unmatched by any other bowler. He has the highest rating for a bowler in the ICC's ratings for Test bowlers, just as Bradman has the highest rating for batsmen. Therefore, I propose a Barnes standard for Test bowlers to complement the Bradman standard for Test batsmen.

The most successful bowler is one who takes the most wickets, while conceding the fewest runs, in the fewest deliveries. Put another way, the most successful bowler takes the most wickets, most cheaply and most frequently. Using such a rule, we get a Barnes standard measure: wickets taken divided by the product of bowling average and strike rate. Barnes' measure is 0.276 over the length of his career.

Among bowlers with a career haul of at least 100 Test wickets, only George Lohmann, a 19th century Englishman, and Dale Steyn have strike rates better than Barnes. Steyn's wickets cost about six runs more than each of Barnes' did. No bowler, not even Lohmann, has matched Barnes' average of seven wickets per Test.

It turns out that only one player in Test history has had a 27-match stretch - the length of Barnes' Test career - in which he achieved a higher Barnes standard measure. In 27 consecutive Tests between 1990 and 1994, Waqar Younis took 177 wickets, at the rate of one every 33 deliveries. This remains the single-most lethal wicket-devouring span over 27 consecutive Tests. Barnes, over the entirety of his Test career, produced a record comparable to Waqar's between 1990 and 1994.

Bowling in strong line-ups is not necessarily a drawback, since the Barnes standard measures the frequency and speed of wicket-taking in Tests. Malcolm Marshall was at his peak between 1984 and 1988 while bowling in a strong bowling attack. Amazingly, Wasim Akram's most successful period coincided with Waqar's. From 1990 to 1994, Pakistan had arguably the most lethal new-ball attack in Test history - two bowlers at their peak at the same time. Their successor, Shoaib Akhtar, also features high on the list but perhaps the most telling fact is that his best stretch of 27 Tests came over a six-year period - between November 26, 1999 and November 29, 2005 - when Pakistan played 55 Tests. Javagal Srinath too has a similar record. His best spell of 27 Tests came in a four-year period between November 20, 1996 and November 18, 2000, a spell in which India played 36 Tests. Playing three out of four Tests is not bad for a fast bowler, but, tragically for India, five of the nine Tests that Srinath missed in that phase were during the 1997 tour of West Indies.

Sir Richard Hadlee and Muttiah Muralitharan expectedly feature high on the list. Their dominant phases are not surprising either. Shane Warne's most successful phase was in the early 2000s. This is somewhat surprising as this was after Warne had been through a couple of rough tours of India and West Indies, and had a competitor for the legspinner's slot - Stuart MacGill.

Apart from telling us when the best bowlers in Test history were at their best, the Barnes standard also reveals when bowlers combined and peaked. With Wasim and Waqar at their best from January 12, 1990 to September 28, 1994, Pakistan won 15 of their 30 Tests and lost only six. Similarly, between 1968 and 1973, John Snow and Derek Underwood, who both peaked around the same time, led England to the top of Test heap. England beat West Indies in the West Indies, and Australia in Australia during this period. Their one major reverse was against India at home in 1971, when Bhagwath Chandrasekhar destroyed them at The Oval.

The Barnes standard record also shows peculiarities about individual bowlers. Dennis Lillee and Fred Trueman, both known to be genuinely quick bowlers, had their most productive phases of their Test careers after each had lost his tearaway pace to injury or age. In the early '60s, Trueman famously reduced the length of his run-up and gave up some pace for guile. Lillee suffered a serious back injury that interrupted his career, first in the early '70s and then in 1977, and remodelled his action. He was most productive between 1977 and 1981.

The fabled contest between Lillee and Viv Richards, often billed as one between the world's fastest bowler and the world's fastest batsman, probably held true in the 1975-76 series in Australia, but later on in the '70s, it was more a contest between a highly skilful fast bowler and a marauding batsman. In contrast to Lillee and Trueman, Waqar and Allan Donald were less productive once they lost some of their tearaway pace. They were still very good bowlers, just not the wicket-taking machines they had been at their respective peaks.

Finally, the standard illustrates, like few other statistics do, the weakness of India's Test bowling over the years. India's highest-placed bowler in the Barnes standard table is Anil Kumble, who places 51st. Kapil Dev, Srinath, Harbhajan Singh and Zaheer Khan all follow, between 55 and 71.

Blogs: Kartikeya Date: The Barnes standard | Cricket Blogs | ESPN Cricinfo
Even if you don't like the authors proposed standard, his table (please click on link) - 'The Best 27 Match Spell for Test Bowlers' - makes for some interesting reading. The top 10 bowlers are;

01. Waqar Younis
02. Sydney Barnes
03. Muttiah Muralitheran
04. Imran Khan
05. Malcolm Marshall
06. Dale Steyn
07. Richard Hadlee
08. Shane Warne
09. Alan Donald
10. Wasim Akram

.....which now begs the question: Why not use the Waqar Standard instead, since he out-Barnes, Barnes?
 
Last edited:

watson

Banned
If this means we accept that Waqar's the greatest ever, I approve :ph34r:
I guess so, as long as 'longevity' is not taken into account;

In contrast to Lillee and Trueman, Waqar and Allan Donald were less productive once they lost some of their tearaway pace. They were still very good bowlers, just not the wicket-taking machines they had been at their respective peaks.
 

OverratedSanity

Request Your Custom Title Now!
I guess so, as long as 'longevity' is not taken into account;
Yeah I was only half serious. But 177 wickets in 27 matches is ridiculous for a fast bowler... Anything over 5-5.5 wpm is unheard of and here he was going at 7+ with an SR of 33. Unbelievable.

And really, how underrated is Donald? Even in this article he's marked down for losing his pace and not being the same force he was before, but apart from those two ill-fated matches against Australia, he was amazing throughout his career. His record is up there with the very best. It annoys me to no end when from his era Ambrose, Mcgrath, Wasim, all get talked about in a discussion about the greatest fast bowler, and he barely ever gets a mention... :censored: He was my favorite bowler growing up
 

Goughy

Hall of Fame Member
I dont know who Kartikeya Date is and I have only skim read the article because the premise appears dumb.

Strike rate is already incorporated in to bowling average. This is counting strike rate twice. It is completely illogical to double up strike rate. Someone who has actually read the article to point out where I am wrong but one a quick view and this is one of the most flawed Cricinfo articles I have 'read.'

You go for 3 an over and take a wicket every 36 balls and you have an average of 18. You go for 3 an over and take a wicket every 60 balls and you have an average of 30. There is no fathomable reason to then further make a product of average and strike rate.
 
Last edited:

Neil Pickup

Cricket Web Moderator
The formula basically boils down to (wickets cubed) / (runs x balls) - so every extra wicket massively inflates the formula upwards.

Really not sure what this is meant to demonstrate beyond an entirely arbitrary collection of figures.
 

watson

Banned
I dont know who Kartikeya Date is and I have only skim read the article because the premise appears dumb.

Strike rate is already incorporated in to bowling average. This is counting strike rate twice. It is completely illogical to double up strike rate. Someone who has actually read the article to point out where I am wrong but one a quick view and this is one of the most flawed Cricinfo articles I have 'read.'

You go for 3 an over and take a wicket every 36 balls and you have an average of 18. You go for 3 an over and take a wicket every 60 balls and you have an average of 30. There is no fathomable reason to then further make a product of average and strike rate.
I kind of understand what you saying, but the fact remains - if you divide the total number of wickets by just the Average, or just the SR, then the order of the players' ranking would be completely different on each occasion. And there is no logical reason to prefer Average over SR, or vice versa. What do you do to decide between the two, flick a coin? Therefore, you still need some equation that can successfully 'compare' the Total Wickets to the Average and the SR.

Any suggestions for a suitable equation/algorithm?
 
Last edited:

The Battlers Prince

International Vice-Captain
It's interesting, but I also notice that Barnes played just 10 of his 27 tests in England (home tracks). How many '27 test peaks' would have so few home tests. May havehappened, I haven't checked.
 

Hurricane

Hall of Fame Member
The formula basically boils down to (wickets cubed) / (runs x balls) - so every extra wicket massively inflates the formula upwards.

Really not sure what this is meant to demonstrate beyond an entirely arbitrary collection of figures.
What is a better formula for combining avgs and srs? Avg + SR and the lower the number the better? I tried running those numbers and Waqar is still near the top.
 

kyear2

Cricketer Of The Year
There is no perfect formula, but we can at least use basic statistical standards to create groups. If we were to select the players since 1900 that over their careers has managed to meet the minimum requirements to average at least under 26, allowed less than 3.5 runs per over, struck at under 58 balls per wicket and taken at least 4 wickets per match one would end up with only 18 players. Much less than I would have imagined and leaving some of the great fast bowlers in history out in the cold. The players who made this initial cut would be.

Syd Barnes
Malcolm Marshall
Dale Steyn
Fred Trueman
Richard Hadlee
Dennis Lillee
Curtly Ambrose
Muttiah Muralitharan
Joel Garner
Michael Holding
Allan Donald
Shane Warne
Imran Khan
Waqar Younis
Colin Croft
Peter Pollock
Andy Roberts

When the standard is raised to an average below 23, economy rate below 3, a strike rate below 52 and a minimum of 4.5 WPM we are left with

Syd Barnes
Malcolm Marshall
Glenn McGrath
Fred Trueman
Richard Hadlee
Allan Donald

Slightly raise the average requirement to below 24 and that would add to the list;

Dennis Lillee
Colin Croft

So I would suggest that a statistical search for the greatest bowler should start with those 6/8 gentlemen. (yes Croft probably doesn't belong in that group and would probably be a situation similar to Graeme Pollock, strong team and short career in terms of years)

Just out of curiosity if we raise the requirements to an average < 23, and strike rate <50, that would leave us with

Sydney Barnes
Malcolm Marshall
Fred Trueman
Allan Donald

Even to lower the average requirement to <22 that would still only eliminate Donald. Don't see too many ATG XI's with an attack of Marshall, Trueman, Donald and Barnes. Probably we should.
 

watson

Banned
Assuming that all your stat's are correct kyear - that's an awesome post (although it does ignore 'peak' averages, and that sort of thing).

However, an attack of Marshall-Trueman-Donald-Barnes probably wouldn't please the purists because it includes a 'strange' finger spinner who bowled at about medium pace. I think that most people would prefer a more traditional spinner like Warne/Murali who flighted the ball a bit.
 
Last edited:

ankitj

Hall of Fame Member
Not sure if economy rate is necessary in those stats, when there is average and SR already in there. ER = average/SR
 

watson

Banned
Not sure if economy rate is necessary in those stats, when there is average and SR already in there. ER = average/SR
I've always wondered way some of my cricket books only quote the number of Tests, total Wickets, bowlers Average, and not much else.

Now I know it's because the SR and ER are mathmatically redundant, and near useless pieces of information. Although I still like having them there in front of me for 'comfort value'.
 

OverratedSanity

Request Your Custom Title Now!
There is no perfect formula, but we can at least use basic statistical standards to create groups. If we were to select the players since 1900 that over their careers has managed to meet the minimum requirements to average at least under 26, allowed less than 3.5 runs per over, struck at under 58 balls per wicket and taken at least 4 wickets per match one would end up with only 18 players. Much less than I would have imagined and leaving some of the great fast bowlers in history out in the cold. The players who made this initial cut would be.

Syd Barnes
Malcolm Marshall
Dale Steyn
Fred Trueman
Richard Hadlee
Dennis Lillee
Curtly Ambrose
Muttiah Muralitharan
Joel Garner
Michael Holding
Allan Donald
Shane Warne
Imran Khan
Waqar Younis
Colin Croft
Peter Pollock
Andy Roberts

When the standard is raised to an average below 23, economy rate below 3, a strike rate below 52 and a minimum of 4.5 WPM we are left with

Syd Barnes
Malcolm Marshall
Glenn McGrath
Fred Trueman
Richard Hadlee
Allan Donald

Slightly raise the average requirement to below 24 and that would add to the list;

Dennis Lillee
Colin Croft

So I would suggest that a statistical search for the greatest bowler should start with those 6/8 gentlemen. (yes Croft probably doesn't belong in that group and would probably be a situation similar to Graeme Pollock, strong team and short career in terms of years)

Just out of curiosity if we raise the requirements to an average < 23, and strike rate <50, that would leave us with

Sydney Barnes
Malcolm Marshall
Fred Trueman
Allan Donald

Even to lower the average requirement to <22 that would still only eliminate Donald. Don't see too many ATG XI's with an attack of Marshall, Trueman, Donald and Barnes. Probably we should.
Just one problem with that post. You can't have the same parameters for quick and spinners. Spinners have a different role and generally have a far higher strike rate. Unfair on Warne and Murali.
Apart from that, it just backs up my point how shockingly overlooked Donald is in ATG teams. He should always be a part of the discussion
 

kyear2

Cricketer Of The Year
Assuming that all your stat's are correct kyear - that's an awesome post (although it does ignore 'peak' aages, and that sort of thing).

However, an attack of Marshall-Trueman-Donald-Barnes probably wouldn't please the purists because it includes a 'strange' finger spinner who bowled at about medium pace. I think that most people would prefer a more traditional spinner like Warne/Murali who flighted the ball a bit.
As I said not supposed to be perfect, just a different perspective.

Frank Tyson would have made the first cut, but only played in 17 Tests and my cut off was 20.

Not sure if economy rate is necessary in those stats, when there is average and SR already in there. ER = average/SR
Economy rate was only a filter to eliminate the bowlers who were out right too expensive. Basically was about Average, WPM and Strike Rate.
 

Top