Charles of the Yukon
Charles Davis (left) takes a break from statistics, on a recent visit to the Yukon

There are cricket historians. There are those who dabble in cricket data. Few, however, have managed to combine the two, that too at a level matched by none. Charles Davis had taken up the painstaking (and intimidating) task of cataloguing and retelling cricket history through numbers. He has delved deeper into cricket numbers than anyone in history, and has managed to recreate ball-by-ball breakdown of every batsman’s innings, pointing out many an error in existing scorecards in the process. His chronicles, Z-Score’s Cricket Blog, is a go-to point for cricket statisticians and quizzers across the world. Abhishek Mukherjee interviews the man who proved that Don Bradman had actually scored those four runs that would have given him a Test batting average of 100.

 CricketCountry (CC): The world got to know of you when you discovered the incident of Don Bradman’s missing four runs, assigned erroneously to Jack Ryder. How did you manage to stumble across it?

 Back-story: During his enormous project, Davis came across the score-sheet of the 1928-29 Ashes Test at MCG. While the team totals matched when he compared the batsmen’s and bowler’s sections, there was an anomaly when it came to individual scores: four runs that should have been Bradman’s were awarded to Ryder. Had they been (rightly) given to Bradman he would have finished with a career average of 100.

 Charles Davis (CD): I should say firstly that the error I found is by no means clear cut. What I found is that the only surviving official score has a four in the wrong place; this is an anomaly, but assigning the runs to Bradman is just one of several possible ways to resolve it. Ultimately, giving the four to Bradman is no more than a tantalising possibility.

I came across it as part of a long-standing effort to digitise all surviving old Test match scores into ball-by-ball form. When I entered the detailed bowling figures from that Test into my spreadsheet, the computer assigned the four runs to Bradman, so I looked into it.

 CC: Once you discovered the anomaly, did you contact the authorities of cricket data, like ACS and Wisden? We know the databases have not been updated…

 CD: I did not contact them myself, but they were aware of it. I have not called for changes to be made to official records, given that there is a high level of uncertainty around it all.

One surprising reaction I did encounter was that some people just did not want to know about it. The idea that there may be any uncertainty over old records shocks them, and they reject it out of hand.

 CC: What prompted you to undertake the creation of the ball-by-ball database, a project so humongous that it would have intimidated most?

 CD: In studying cricket statistics, I found it frustrating that the known detail about old matches fell so far short of modern standards, and decided to do something about it. I started with Australian matches and expanded from there. If you do these things bit by bit, in spare time, it is not quite so intimidating.

 CC: As in the Bradman-Ryder incident, you have almost certainly come across discrepancies between scoring shots of batsmen and scoring shots scored bowlers. Who do you trust in such cases?

 CD: I much prefer to go with the ‘official’ scores unless there is (1) overwhelming evidence of error, and (2) a high level of certainty about the correction. This is actually very unusual in the case of ‘core scores’ such as batsmen’s runs. Far more often, study of the old scores just exposes anomalies that cannot be satisfactorily resolved.

One example out of many would be Victor Trumper’s 185 not out in 1903-04. The surviving score does give his score as 185, but his scoring strokes add up to 187. Most problems like this are better regarded as uncertainties rather than errors.

It is more common to find errors in secondary statistics such as balls faced (pre-2000), batting order, or daily close of play scores. For instance, the online sources name the batsmen out at each fall of wicket in all Tests. I have found about four hundred errors in that data prior to 2000.

 CC: You had spent two decades as a scientist in research laboratories, in the mining industry as well as the food industry. What made you switch to cricket statistics? Was there a specific incident?

CD: If there was an ‘incident’, it was losing my job during a company downsizing: this happens a lot in the cyclical mining industry. It happened just at a time when I was formulating ideas about the potential of cricket statistics (something that had always interested me), so I took the opportunity to write a book on the subject, The Best of the Best, which I was very fortunate to have published by ABC Books.

 CC: How do you access scorebooks and newspaper archives across the world? Not every country has been meticulous in maintaining archives. As for the newspaper archives, not all of them have online archives.

CD: I have obtained the majority by personal visits, but some people have been of great help. Andrew Samson has scoured the archives in South Africa, and Shahzad Khan has found many scores in Pakistan that might otherwise have been lost, and sent me copies.

The best archives are at Lord’s, at Cricket NSW and the Melbourne Cricket Club. By a stroke of fortune, the legendary Australian scorer Bill Ferguson bequeathed many of his scores to Cricket NSW rather than Cricket Australia (I say fortunate because Cricket Australia has lost all its scores from before 1980). The people at these places are very helpful, and I must mention Colin Clowes at Cricket NSW, a great help to my research.

I have visited most of the major Australian, New Zealand and England grounds, and obtained whatever records I could. I have made five visits to England that have included cricket research, with many hours spent at the British Library newspaper section (now closed). I have searched newspapers at Australian libraries in every state. I have even done research at the Library of Congress in Washington, much more user-friendly than the British Library.

Some of the material that I have extracted has since become available online, but most of it has not. I have perhaps five thousand pages of Test match newspaper reports in digital or hardcopy, and notes made from a great many other reports that I accessed but could not copy.

I should also mention my friend Roger Page, who has an almost unrivaled private collection of cricket books and magazines, and who has helped me greatly.

 CC: Z-score’s Cricket Stats Blog is a point of reference for a lot of cricket statisticians. Do you have any go-to places yourself on the internet?

 CD: Just the usual online cricket databases, especially Cricket Archive. I often haunt the Ask Steven Facebook page where cricketing questions are asked.

 CC: Your father was an umpire, standing in places as exotic as Malaysia (in the match where Garry Sobers took 5 wickets in 5 balls) and Mexico. Do you have memories of cricket in other unusual locations from your earlier days?

 CD: I cannot really help you there. Perhaps the oddest experience was seeing the 1970 England World Cup football team playing cricket in Mexico City. I got all their autographs.

I did once play in a match against a team of blind cricketers, and they beat us! They used a ball with a bell in it — you had to bowl underarm along the ground — and things like that.

 CC: What is the biggest hurdle you have faced during your research?

CD: There has been widespread indifference to preserving the past among the powers that be, especially at national level, and in some countries no procedures have ever been in place to preserve records of past matches. Even if attitudes change, what is lost is lost and cannot be recovered.

 CC: Have cricket boards expressed interest in your research? Do they seem keen on history?

 CD: In general, no and no. There are individual exceptions, perhaps.

 CC: You have acquired data nobody else in the world does. Do you have plans to use this data for statistical analysis?

 CD: I still plan to write a magnum opus on cricket stats, collating all the directions of my work and trying to use the data in new and interesting ways, as opposed to long lists of names and numbers. Unfortunately, it is enormously difficult to get a book like this published any more. Most publishers do not even reply to proposals these days, even to authors they have previously published.

 CC: Do cricket writers and the cricketers themselves approach you for insights?

 CD: I have had correspondence with quite a few writers of books or articles, and tried to be helpful when I can. As for cricketers themselves, not really.

 CC: What keeps you going through your research?

CD: Cricket history and statistics is a small pond, but it is heartening to be regarded as an important contributor to one’s field. There is a limited but committed circle of people, many of them enthusiastic amateurs or volunteers, whom I respect and who encourage me in my work. I hope also that my work can be used by future generations of enthusiasts.

My partner, Ann, has always encouraged me, in good times and difficult times.

(Abhishek Mukherjee is the Chief Editor at CricketCountry and CricLife. He blogs here and can be followed on Twitter here.)