A Lancastrian born in Bolton, Tony Lewis forms one half of the duo credited with creating the best model for target revision in rain-interrupted matches. Or, as his partner Frank Duckworth once put it, it is all about getting a ‘Fair Result in Foul Weather’. Arunabha Sengupta, Cricketcountry’s chief cricket writer, caught up with Lewis at the Lord’s Tavern after the first One-Day International (ODI) between the England and Australian women’s teams.
The hour started with discussions about the rain rule, the difficulties of speaking mathematics to the cricketing community, the criticism faced frommembers of the media largely ignorant of the technique and later the conversation diverged to horses and pop groups named after the duo, hilarious episodes of being mistaken for the former England captain, being compared to Osama Bin Laden and much more… And every answer was accompanied by a heart-warming jovial bout of laughter.
A brief look at the men and their method
Limited-overs cricket had for years been tormented by interruptions. Teams batting second sometimes walked away as victors having scored 126 for eight in 25 overs after chasing 250 to win. The rules were tweaked, but the modifications resulted in South Africa trying to figure out how to get 22 runs in one ball on the huge stage of the1992 World Cup semi-final. The cricket world was looking for a solution that could reduce overs and still keep the situational balance intact. It was also very evident that this was not a simple problem with solutions offered by elementary arithmetic.
Tony Lewis had for long been intrigued by the problem, mostly while watching matches at the WACA ground in Perth where he was working as a lecturer at the Business School of Western Australian College of Advanced Education. When he returned to England with a lecturing job in Bristol Polytechnic, later University of the West of England, he witnessed the notorious semi-final between England and South Africa in 1992. While he thought about the problem, he came to know of another mathematician, and another Lancastrian to boot, Dr Frank Duckworth, of the Berkeley Nuclear Laboratories, who was passionately trying to devise a more realistic rain rule. The two communicated from 1993 onwards and started working extensively on the concepts, using old scorecards of the Wisden to test their evolving method.
In 1996-97, the Duckworth-Lewis method was used for the first time in an international match — during the second ODI game of the series between Zimbabwe and England. An elegant mathematical solution based on exponential decay functions, the technique has been refined often, and still remains the best available rain-rule for interrupted matches.
It has continued to be a labour of love. Contrary to popular perceptions, the two gentlemen have not made huge amounts of money from their sterling service to cricket. They did get occasional payments and expense reimbursements, and they do have a contract with ICC for maintaining the method, and all this has brought in a moderate income to make their retirements a little more comfortable. But, one cannot copyright or patent a method. It is neither a document nor a material invention. They were not commissioned by the cricket authorities to produce the method. They responded to a need and offered it for use. They did not get royalties when it was used. But, they don’t really regret it.
Here is Tony Lewis in a candid interview:
CricketCountry (CC): Could you tell us about explaining sophisticated mathematical techniques to non-mathematicians, convincing administrators about the method, explaining it to the media and the fans…
Tony Lewis (TL): I think the key is to keep it simple. There is an acronym KISS.
CC: Keep It Simple Stupid?
TL: Or ‘Keep it Short and Sweet’ which is a variation of the same. One tries to avoid technical details and shows how it works in practice. Lots of examples of the past can be used: the 1992 World Cup semi-final for instance and many before and since then. We need to show that the system we are using is fair and works in most circumstances. The driver or key to getting things accepted is to create scenarios, to show how it works in the wide variety of scenarios.
We had good fortune in many ways that the people we were talking to were quantitatively acute; they did figure out what we were doing. The chief executive of ICC (International Cricket Council), David Richards, had an economics background. So he did have certain mathematical expertise. So, we did explain the general principles without going into the background mathematics of the model. His technical background ensured his interest and helped him in understanding what we were getting at.
But the best way was to put scenarios on paper and try them out. Keeping the technical side as limited as possible, mentioning whatever is necessary and keeping it as simple as possible.
CC: But, you have been hounded by some members of the media, who have made no effort to understand the system. Jonathan Agnew and David Lloyd to name two of them.
TL: That’s been a difficult area. We have taken the view that we are happy to explain things to them. We have offered many sessions that were suitably pitched. I am a lecturer by trade, and can tweak the material to suit the audience as necessary. Unfortunately the media in general has not shown much interest in taking up the offerings.
We have given lots of seminars, initially to scorers and umpires. They need the training, the scorers certainly do. The umpires too attended at first before they realised that the scorers were doing the job. The cricketers — we haven’t really talked to too many of them. Their situation is that they realise it works, and they don’t need to know why it works. But the media has been a little difficult.
[In their excellent book Duckworth Lewis: The Method and the Men behind it, it is written: "Initially we had produced a simple explanation of the methods with worked examples, and ECB had circulated this to the media. But few, if any, ever gave the indication that they had read it, let alone gone through the examples"]
CC: Do you get irritated when people, of the media and otherwise, who cannot quite explain an elementary concept like, say, the net run-rate, go on to lambast a far more sophisticated method that they obviously cannot understand?
TL: Mathematics is a conceptual and written language, when the journalists try to convert it into a spoken language, it often sounds silly — that’s what irritates me sometimes. They have said quite a few things down the years. If the journalists ask us to comment on something, yes, we talk to them. But then they have often gone on to publish pieces which are rubbish, nonsense, without understanding what is going on — making fun of the mathematics. It sort of denigrates the system. We decided early not to respond in print or even directly to adverse comments. We generally ignore such comments and that’s a policy that has worked very well.
CC: There have been some occasions when D/L has been in focus. Like the time Sri Lanka and South Africa tied the match at Durban during the 2003 World Cup.
TL: The only issue [during that match] was a misreading of the par-score sheet. The par-score sheet is an over-by-over, ball-by-ball table of what the par score should be. And the sheet of paper says this is the par-score; it is very clear. What South Africa did was that they considered the par score of 229 as their target (they finished on 229 for five from 45 overs chasing Sri Lanka’s 268 from 50), whereas it was not their target, it was the par score.They needed one more to win. So it was a total misreading by the South African camp that led to that misunderstanding. [Sanath] Jayasuriya the Sri Lankan captain knew exactly what the situation was, he knew it was a tie and it was okay for the Sri Lankans.
CC: But, it got you compared to Osama Bin Laden?
TL: (Laughs heartily) Yes, an anonymous correspondent did write that. ['How does it feel to be the most hated person in the world, more hated even than Osama Bin Laden?’] We did not respond.
CC: But you did respond to Tony Grogan the cartoonist. [The cartoon appeared in a South African newspaper depicting Duckworth and Lewis being hanged from the gallows]
TL: (Laughs heartily again)The cartoon appeared on BBC television, Channel 4. It was a fantastic cartoon. We rang up Channel 4 and were referred to Tony Grogan himself in South Africa. I rang him up and said who I was, and he was alarmed and asked, ‘have we offended you?’ I said, ‘No, it’s fantastic. Can we use it?’ And he said we could use it wherever we liked. [The cartoon is reproduced in the book written by the duo].
CC: What about the time John Dyson, the coach of West Indies, who misread the table and lost the match as a result [against England at Guyana in 2008]
TL: John Dyson, yes. I am not entirely sure what happened there. They seemed to have the par-score sheet, and on the top of the sheet we have the wickets as the column headings from zero to nine. Dyson’s sheet did not have the wickets and he labelled them from one to 10. (Dyson read the par score as 242 and asked the batsmen to take the offered light at 244 for seven after 44.2 overs, when the par score was actually 245. As a result the West Indies lost by one run)
CC: The method would have been severely criticised had India won against Australia in the 2003 World Cup Final. (In the 2003 final, India were 145 for three from 23 overs in response to 359 for two of Australia. They were just four runs behind par. If the game had stopped after two overs and India had scored five runs on par without losing a wicket, they would have won)
TL: Yes. That was something we knew about. We were then using what is now known as the Standard Edition. We knew that under normal circumstances it worked well, but we had also known for a while that in case of very big totals the approach was not really that good. We did have in place a computerised version, but it meant that the transparency was lost. You couldn’t do it manually by looking up the tables. Up until that point the ICC were very happy with the manual version and the transparency that came with it. But Australia got 359 and that showed up the flaws and straightaway the next edition was introduced which handled high scores much better. The par score for India is likely to be much higher now.
CC: How is the model kept in synch with the changing approach to run scoring?
TL: What we do is we monitor how the game changes. With Twenty-20 thinking coming along, they score a lot quicker early in the innings. Then we need to tweak the parameters. We did that in 2009. The basic model stays the same, but the parametersare adjusted. The database is updated every two years.
CC: There have been some sterling helping hands like David Kendix and Dr Steven Stern.
TL: Yes. David Kendix, the MCC scorer, is a very good sounding board. Dr Steven Stern of Australia has shown a lot of interest in the method. He published apaper with a suggestion about how to evaluate the par scores differently. He has done some good work in the rating of teams and cricketers as well. He has also written the Windows Software for our method (WINCODA). It was in a basic format and needed to be written for Windows. We did not have the expertise, but he did. The software is brilliant and highly thought about around the world.
CC: You have received academic acclaim for your method. Duckworth Lewis method has also received awards in the academic world. And later you have been honoured in the cricketing world as well. Which is more satisfying?
TL: Both are extremely satisfying; there is no difference. You see something you do actually used across the world on such a scale. The academic world appreciates the mathematics behind it, recognising the contribution in our academic profile. In cricket, we have not made a fortune, but we have met great personalities. It was a great pleasure, for example, meeting Clyde Walcott. As I said, both are satisfying.
CC: What do you make of Sunil Gavaskar’s comment that it does not matter if another method is used in Indian domestic cricket because never rains in India?
TL: (laughs) It does rain sometimes … but the difference is that the need for a rain rule is not that important as it is here.
CC: You have had two pop groups named after the two of you, as also a race horse.
TL:Yes. The first Duckworth-Lewis group disappeared years ago, but the second one is still producing music. The horse didn’t do too well, though. He has retired.
(Coley Racing, a syndicate based in Cheltenham, included several well-known cricket personalities. They gave their horses names such as Caught in the Slips, Nudge and Nurdle and Silly Mid-off. One horse was named Duckworth-Lewis and the two mathematicians were invited to its first outing at Uttoxeter in November 2006. The horse started as joint favourite but finished tenth. The best it managed in a two-year career was a third position.)
CC: Once an alternative rain rule from India was sent to the Prime Minister’s office…
TL: Yes. Tony Blair was the Prime Minister at that time. And his office received the suggested method and sent it to ECB. ECB directed it to us for our comments.
CC:What about being mistaken for the former England skipper Tony Lewis. Does that still happen? (One such invitation meant for Tony Lewis, the cricketer, to attend a garden party at Hertfordshire estate was received from Lord MacLaurin of Knebworth, then the president of the ECB)
TL: Not very often, but it still sometimes happens. Sometimes in a seminar people are expecting the other Tony Lewis, and I sometimes end up apologising. I was giving a seminar in Cardiff University a few years ago. And Cardiff is the homeland of Tony Lewis. I was introduced and it felt embarrassing after coming to Cardiff and Glamorgan. And I said, ‘if you are expecting the other Tony Lewis I would be happy if you want to go.’
I used to receive invitation to parties meant for him, with his my wife’s name written as Joan instead of Loris, which I used to think was a mistake. But that doesn’t happen anymore.
CC: A last question from very close to the heart. Often we try to produce statistical articles to settle various age-old arguments about who the better player was and so on. Quite often, the responses read ‘numbers don’t mean anything’. Any advice from someone who has managed to get a mathematical system successfully across to the cricketing world?
TL: (laughs) When we suggested our method, cricket was looking for a solution to a long standing problem. So, it was accepted. What you are trying to do is settle cricketing arguments. In this case mathematics defeats the purpose. People don’t want solutions, they want to argue about cricket. So, it is very different.
(Arunabha Sengupta is a cricket historian and Chief Cricket Writer at CricketCountry. He writes about the history and the romance of the game, punctuated often by opinions about modern day cricket, while his post-graduate degree in statistics peeps through in occasional analytical pieces. The author of three novels, he can be followed on Twitter at http://twiter.com/senantix)