The discovery threatened one of Australia’s iconic truths.
No number (to two decimal places) is better known in Australia than 99.94: Donald Bradman’s average.
That’s four runs shy of an average of 100.
So when historian and author Charles Davis spotted something odd about the scoring for the 1928-29 test at the MCG, he knew it had the potential to change history books.
A boundary that was given to another batsman — but that, according to bowling records, may have been Bradman’s.
The Bradman discovery
Davis has entered every ball from hundreds of matches into his computer, in total more than 5 million deliveries. When he saw the Bradman anomaly in 2008, he knew what it could mean.
“I spoke to a friend of mine who was an editor at the Sydney Morning Herald. His eyes sort of lit up when I mentioned it to him.”
Davis’s discrepancy doesn’t guarantee Bradman the runs, “because it wasn’t noted at the time and there are various other ways of shifting the data around so that Bradman doesn’t get the four”.
After the initial buzz, the interest in the story faded. Ultimately, nobody could prove or disprove Davis’s discovery.
But the saga highlights the problem when statistics lose their certainty.
Davis found another anomaly for Bradman for a test in 1930.
“But what do you say? The scorer made a mistake … do you change Bradman’s score?”
A scorer’s legacy
Bill Ferguson recorded every ball that Don Bradman faced in a test match.
In those days, the scorer doubled as a baggage handler on overseas tours. While little is known about his ability with luggage, Ferguson’s reputation as a scorer is exemplary.
But Davis notes Ferguson was just one man.
“He used to record his scores in a linear form, and then re-copy them into a traditional-type scoresheet in the evenings,” Davis says.
“The trouble with that is the traditional scores are the ones that have survived, but they’re actually re-copies.
“He’s copying his data out in evenings into something else and I’m pretty sure he made quite a few errors.”
Adam Morehouse is a statistician and historian for ACT Cricket and was scorer for the 2019 Manuka Oval test match.
He says all scorers make mistakes, but Ferguson’s reputation should not needlessly be brought into question.
“He was the best scorer, or the most experienced scorer, he scored over 1,000 first=class games and tour matches.
“We’re all getting hung up on four runs he might not have scored?
But it’s not just about Ferguson.
Davis has found similar anomalies in other scorebooks.
Earlier this year, he came across another curiosity that means one more batsman could have been robbed of their own milestone.
In a test in 1910, South African Aubrey Faulkner was — according to the official records — dismissed by England for 99.
Davis’s process of re-scoring earned Faulkner one extra run and the possibility was supported by newspaper reports from the time uncovered by Davis.
But, as Davis has noted, it’s not a sure-fire century as the run in question might also be attributed to another batsman.
“So, uncertainty must remain,” Davis says.
“I would say, though, that when I tried a few possible changes to the bowling (which itself creates new anomalies), Faulkner still gets his 100.”
Dealing with anomalies
Nobody will ever know for sure whether those missing runs were Bradman’s. But it’s clear cricket scores — like much of history — are not rigid.
Davis is now part of a community of historians that re-scores old matches.
Part of the goal is to generate scoring rates before ball-by-ball scoring was common. In 2018 he discovered Englishman Maurice Tate scored more rapidly in the 1920s and 1930s than David Warner has done in tests of the past decade.
But he continues to be intrigued by anomalies. He pitched an article to cricketing bible Wisden about what he has found so far.
“I suggested to Wisden I could write an article on my work and the question marks that have come up.
“They said we don’t want to have that next to our record section because people will think there’s something wrong with the records, which there is.”
Harriet Monkhouse, the statistical editor at Wisden Cricketers’ Almanack says “an uncertain scoresheet is not the basis for altering the records”.
On the Faulkner anomaly — perhaps one with a stronger argument than Bradman’s — she says she would require evidence “a lot more conclusive than this to make a change”.
“We do, of course, issue errata when there is clear evidence of a mistake in our version of a scorecard, but this doesn’t seem to be a mistake on our part, more a matter of conflicting sources.”
To Wisden’s credit they do regularly issue errata, even to Bradman’s career.
In 2000, the publication updated Bradman’s individual records against Essex and Sussex that were originally printed in 1949.
For Morehouse, the iconic character of Bradman’s average means there is even less motivation to revisit old scorebooks.
“Going back and changing things in history, and going through every scorebook … people make mistakes.”
As a lifelong cricket fan, Davis’s passion for the past is simply one way to revisit the greats.
“It doesn’t worry me if Victor Trumper is one or two runs out.
“But it certainly worries some people.”