April 2, 2012

Buchanan puts the science into selection

New Zealand did away with their selection panel last year. Instead they have a selection manager who reads pie charts before drawing up a squad list
20

The Oakland Athletics baseball team did not have the most talented players in the US major leagues, nor the most money. What they did have was a strategy for "the art of winning an unfair game". Unfair because other teams had the cash to buy up the most talented players, which Oakland did not. Oakland, though, had an edge in detailed statistical analysis, which allowed them to make best use of the resources available to them and, in the end, to punch far above their weight. New Zealand Cricket is aiming for something similar.

The country has a population of just over four million, whose overriding sport of choice is rugby. Cricket is small in New Zealand, with a correspondingly small pool of players to choose from. In terms of cricket numbers, though, they are no worse off than some of the game's powerhouses. Across the six provincial teams, with a group of Under-19s, New Zealand have about the same number of domestic players in their system as South Africa and Australia do, around 140.

However, in real terms there is a marked difference. New Zealand's smaller population means that the chances of less talented individuals making it into those 140 are greater. It results in a less competitive cricket environment, the effects of which are felt in the national team. Still, New Zealand compete, fiercely if not always triumphantly, with every other team at the elite level of international cricket. The problem facing John Buchanan when he was appointed director of cricket in New Zealand last May was how to turn that spunk into success. He started by revamping the selection process.

It's a theory somewhat similar to the one applied with Oakland in Moneyball, and zones in on optimal use of the available supply of cricketers to ensure the national team is best stocked. "We are working towards getting the sort of skills that we need to play with in all formats," Buchanan told ESPNcricinfo. "The selection system needed an overhaul and needed to produce something that everybody had confidence in, and we are moving in that direction."

It's a clean-cut and clinical look at cricket, an idea that was born out of Buchanan's eight-year tenure as Australian coach. "I always felt there were way too many people in and around the team environment, whether it is selectors, coaches, captains or team managers," he said. "You hear them talk about a person and then they leave the room and say something completely different about that same person. There were too many confusing messages, particularly when you are dealing with selection issues."

Buchanan explained that selectors' jobs are not only about managing players who already represent the national team, but also about dealing with those who are pushing for places. For that reason, he believes the system that selectors use should be as unbiased as possible, so every candidate who is deserving of a chance gets one. "Selection is not only about those who are in, it's about those who are not in," he said. "It's important to try to streamline the whole system."

Buchanan's approach is based on objectivity. The blueprint he has put in place relies heavily on statistical data to make selection as impartial as possible. New Zealand's Dominion Post newspaper was leaked a copy of the selection policy in practice. Each player has a pie-chart drawn up for him, which is divided into slices of different sizes that stand for different aspects of his game, in order to create a total score.

"The most important selection criteria is now significant performance at 35%, followed by consistent performance at 25%, contribution to the team, 15%, fitness, 10%, fielding, 10%, and least valuable, at 5%, is selectors' intuition," Jonathan Millmow, sports editor of the Post wrote.

The pie-chart attempts to reflect how well and how often players put in notable performances. "Significant performances" are classified as centuries and five-wicket hauls, but they are worth less if scored against a part-time attack or if obtained when blowing away a tail. "Consistent performances" are judged on batting averages and on bowling economy and strike rates. If these numbers are maintained or improved in a season, a player is regarded as consistent. There is also a rating for the role the performances have in team success, with contributions that are key to victories rating higher than those that come in losing causes.

Buchanan said the numbers in Millmow's report were not entirely correct but the principles were. "There's a range of things that go into making [the pie-chart] up. It makes selection far less subjective. Although there will always be subjectivity, hopefully it makes it more objective."

Kim Littlejohn, New Zealand's selection manager, collects all the data from the provincial coaches, analyses it and presents it to John Wright, the national coach. Together they pick a squad and it is up to Wright to choose the XI who will play on any given day.

Buchanan's approach is based on objectivity. The blueprint he has put in place relies heavily on statistical data to make selection a process that is as impartial as possible

Essentially it means if a player has performed well in the domestic season, irrespective of age or personality, such as Mark Gillespie, his chances of playing in the national side are almost guaranteed. It also means if someone has missed out, Littlejohn will be able to tell him exactly which area he needs to work on to increase his chances of being picked. It takes away the uncertainty of not knowing why someone was overlooked, which Buchanan thinks will improve relations between players and administrators hugely. "All the players are looking for is consistency and specific feedback. They need to believe somebody is watching and seeing what they are doing," he said. "We are utilising a tool that is relatively objective and therefore consistent."

New Zealand do not have a selection panel, unlike other countries, because Littlejohn runs the department himself. Buchanan said they would gradually look at appointing a group of scouts who will assist Littlejohn in putting the information together, but for the moment all the work sits on his plate.

"It is about placing a lot of responsibility on the provincial coaches to provide very good information on not just their players but other players that they see. The coaches are much closer to individual players than anybody else, so we are trying to utilise their information along with statistical information," Buchanan said.

After its first season the system is still a work in progress, Buchanan said, but he is pleased it is being applied at all levels, from the national team and women's team to age-group levels. One of the glaring weaknesses it has highlighted is the need for greater depth in all aspects of the game, particularly batting, because not enough players are scoring as highly on the consistency index, so replacements for the current internationals are not easily available. "I wouldn't say that if we lose two or three players we can immediately replace them," Buchanan said, "but we are beginning to develop and increase that depth,"

Kane Williamson was the team's only centurion in the recently completed Test series against South Africa; the rest of the top five struggled. On the bowling side, when Tim Southee was dropped, New Zealand called up Gillespie and Brent Arnel. Only Gillespie was successful, leaving them with the same problem they had when Southee played, in that one of their four seamers was the weak link.

Even without the pie-chart to assist them, New Zealand should have been able to isolate the shallow playing pool as their biggest concern. It is how he addresses that problem that will be a measure of Buchanan's tenure. He insists it will take some time before he can be judged on results, especially after this summer ended with losses in all three series against South Africa. "It depends how you measure success. If you measure it by results in the field, you couldn't regard it as a success," Buchanan said, although he was able to take some positives as well. "If we look at systems and processes, then we are going along nicely."

Firdose Moonda is ESPNcricinfo's South Africa correspondent

Comments have now been closed for this article

  • Lermy on April 4, 2012, 3:15 GMT

    So now we know why we got such bizarre selections as Rob Nicol and Brent Arnel for test cricket. The fact is you have to be subjective when selecting players, because test cricket is different, even the pitches may vary markedly from domestic pitches. Results at domestic level don't always reflect relative ability. A guy that scored 100 may have been dropped three times before reaching 10. Someone in better form may have knicked the first good ball he got. One guy might score a power of runs on slower and lower pitches, yet be far inferior at test level. Using Buchanans system, we have picked a medium pace trundler like Arnel over Andy Mackay, but I know who I'd rather have in the team. Mackay clearly showed in the limited overs games that he could hurry the SA batsmen, probably one of the few NZ bowlers capable of that. So of course lets leave him out of the test team. John, you're a genius!

  • Knightriders_suck on April 3, 2012, 9:32 GMT

    Says the only guy who could read a pitch off a laptop

  • g_golu on April 3, 2012, 3:43 GMT

    Great Planning with little results... Buchanan really hasnt made much progress with NZ and they are still at the bottom rung... Constant Changing has not helped either and there is no clear direction.. May be it is high time NZC took notice before it is too late... Cmon we need 11 players on the field, not some bizzare team based on mathematics... !! Surely NZ has more talent...

  • MattBlake on April 2, 2012, 21:37 GMT

    And this wonderful approach gave us Rob Nicol opening despite a first class average as an opener of about 32, and Brent Arnel...

  • on April 2, 2012, 21:21 GMT

    I love a scientific approach, but John Buchanan often seems to me the best argument against such an approach. Certainly the to claim that he made Australia great is a shaky claim at best - he had at his disposal some of the greatest players in history. He certainly was spectacularly unsuccessful at Calcutta. Maybe we are not ready for him, but till then, everybody who is so hung up on Moneyball should also read Scorecasting. This book, along with debunking concepts like "momentum", also points out that people are more readily forgiven for failure if they follow convention. One has to be very sure of one's place to do something different.

  • nzcricket174 on April 2, 2012, 20:37 GMT

    If he was in charge 7 years ago Israel Dagg would be playing cricket not rugby!

  • popcorn on April 2, 2012, 13:44 GMT

    Not for nothing is John Buchanan rated the BEST COACH EVER N THE WORLD. His radical approach to batting - scorng 400 in a day, raising the bar on EVERYTHING the Aussies did, mental disintegration of the opponents, teaching them The Art of War, self - belief, and a will to win, a never-say -die spirit,"it's not over till it's over" that won Australia the Test against India when Michael Clarke spun them out in Sydney 2008,3 World Cups,2 Champions' trophies,16 consecutive Test wins, not once, but twice - and there you are - a Coach that is beyond compare. Good luck, Kiwis. Listen to him.

  • Anand_Nandakumar on April 2, 2012, 12:31 GMT

    I had done a similar analysis here http://3piecesuitwithterrycot.blogspot.com/2012/01/sowho-should-be-dropped.html and here http://3piecesuitwithterrycot.blogspot.com/2011/11/on-merit-pleasepicking-next-indian.html for the Indian cricket team. Very interesting results...

  • Busie1979 on April 2, 2012, 12:31 GMT

    Great article, great approach. Buchanan was seen as a mad man as coach of Australia, but he developed Australia into a world cricketing super power, where the team picked itself. Yes he had a lot of talent to play with, but from that talent he developed a team. I like the idea that older players can still get into the team on merit. If this approach was used by Australia, David Hussey would be playing test cricket, and Pete Forrest would have never been picked. One query I have is the reliance on first class coaches to rate player performance. It would be better to keep people rating them independent to remove the conflict of interest - but then the coaches are better placed to rate other criteria. Significant performance is also dubious - I'm not sure about the criteria of 100s and 5 wicket hauls. Does this mean a match turning 99 is not significant while a painstaking 100 is? Also, is 4/15 not significant while 5/150 is?

  • on April 2, 2012, 12:24 GMT

    Great Article! Great planning, and I cant wait for New Zealand to start winning..they always play great cricket, but the ending is always the same..

  • Lermy on April 4, 2012, 3:15 GMT

    So now we know why we got such bizarre selections as Rob Nicol and Brent Arnel for test cricket. The fact is you have to be subjective when selecting players, because test cricket is different, even the pitches may vary markedly from domestic pitches. Results at domestic level don't always reflect relative ability. A guy that scored 100 may have been dropped three times before reaching 10. Someone in better form may have knicked the first good ball he got. One guy might score a power of runs on slower and lower pitches, yet be far inferior at test level. Using Buchanans system, we have picked a medium pace trundler like Arnel over Andy Mackay, but I know who I'd rather have in the team. Mackay clearly showed in the limited overs games that he could hurry the SA batsmen, probably one of the few NZ bowlers capable of that. So of course lets leave him out of the test team. John, you're a genius!

  • Knightriders_suck on April 3, 2012, 9:32 GMT

    Says the only guy who could read a pitch off a laptop

  • g_golu on April 3, 2012, 3:43 GMT

    Great Planning with little results... Buchanan really hasnt made much progress with NZ and they are still at the bottom rung... Constant Changing has not helped either and there is no clear direction.. May be it is high time NZC took notice before it is too late... Cmon we need 11 players on the field, not some bizzare team based on mathematics... !! Surely NZ has more talent...

  • MattBlake on April 2, 2012, 21:37 GMT

    And this wonderful approach gave us Rob Nicol opening despite a first class average as an opener of about 32, and Brent Arnel...

  • on April 2, 2012, 21:21 GMT

    I love a scientific approach, but John Buchanan often seems to me the best argument against such an approach. Certainly the to claim that he made Australia great is a shaky claim at best - he had at his disposal some of the greatest players in history. He certainly was spectacularly unsuccessful at Calcutta. Maybe we are not ready for him, but till then, everybody who is so hung up on Moneyball should also read Scorecasting. This book, along with debunking concepts like "momentum", also points out that people are more readily forgiven for failure if they follow convention. One has to be very sure of one's place to do something different.

  • nzcricket174 on April 2, 2012, 20:37 GMT

    If he was in charge 7 years ago Israel Dagg would be playing cricket not rugby!

  • popcorn on April 2, 2012, 13:44 GMT

    Not for nothing is John Buchanan rated the BEST COACH EVER N THE WORLD. His radical approach to batting - scorng 400 in a day, raising the bar on EVERYTHING the Aussies did, mental disintegration of the opponents, teaching them The Art of War, self - belief, and a will to win, a never-say -die spirit,"it's not over till it's over" that won Australia the Test against India when Michael Clarke spun them out in Sydney 2008,3 World Cups,2 Champions' trophies,16 consecutive Test wins, not once, but twice - and there you are - a Coach that is beyond compare. Good luck, Kiwis. Listen to him.

  • Anand_Nandakumar on April 2, 2012, 12:31 GMT

    I had done a similar analysis here http://3piecesuitwithterrycot.blogspot.com/2012/01/sowho-should-be-dropped.html and here http://3piecesuitwithterrycot.blogspot.com/2011/11/on-merit-pleasepicking-next-indian.html for the Indian cricket team. Very interesting results...

  • Busie1979 on April 2, 2012, 12:31 GMT

    Great article, great approach. Buchanan was seen as a mad man as coach of Australia, but he developed Australia into a world cricketing super power, where the team picked itself. Yes he had a lot of talent to play with, but from that talent he developed a team. I like the idea that older players can still get into the team on merit. If this approach was used by Australia, David Hussey would be playing test cricket, and Pete Forrest would have never been picked. One query I have is the reliance on first class coaches to rate player performance. It would be better to keep people rating them independent to remove the conflict of interest - but then the coaches are better placed to rate other criteria. Significant performance is also dubious - I'm not sure about the criteria of 100s and 5 wicket hauls. Does this mean a match turning 99 is not significant while a painstaking 100 is? Also, is 4/15 not significant while 5/150 is?

  • on April 2, 2012, 12:24 GMT

    Great Article! Great planning, and I cant wait for New Zealand to start winning..they always play great cricket, but the ending is always the same..

  • pitzicada on April 2, 2012, 12:13 GMT

    John Buchanan has to be the most overrated person involved in cricket in the last 20 years. The Australian team he is deluded enough to take credit for contained some of the best players the game has seen. Warne, McGrath, Ponting et al. The trouble is he believes his own hype.

  • on April 2, 2012, 12:11 GMT

    Interesting experiment. Combining cricket selection with heuristics

  • ElPhenomeno on April 2, 2012, 11:54 GMT

    I know buchanan is going to get bashed around which is almost a given thesedays, but I think he is onto something here. A good mathematical approach to anything is going to give you consistent results, it has been proven time and again in all parts of life including but not limited to other sports. I know soccer has used something similar for establishing a players true worth. Baseball uses what this article refers to and is called "SABERMETRICS" to establish the same. It works if you know what you are doing.

  • i_witnessed_2011 on April 2, 2012, 11:24 GMT

    In cricket, Condition or Pitch are also matters most. Teams like India,Eng are very good in one condition and poor in other. Teams like SA (and AUS up to some extent) play well in all conditions. It would be interesting to know how these parameteres are fit into statistical observation. For ex: Playing Ajmal in subcontinent condtion is difficult but he would not be as effective in Eng. So how the batsmen who faces ajmal (against spin) are rated? But one thing is sure, It makes selection process more objective but at this moment it is work in progress.

  • Gizza on April 2, 2012, 9:19 GMT

    Geez, an article on an interesting topic which people would have very different opinions on and already 2 of the first 5 comments are just country bashing. Actually, such a scientific approach may as well be useful for India, England, Australia as well as every other teams since every team in world cricket has at least a few apparent weaknesses at the moment and no team is the finished article. So @Unnithan, no need to react like that. It won't actually be a bad idea if the BCCI conducted something similar. But let's be honest. Out of the 10 Test-playing nations, India is the LEAST likely to do this. Two reasons. 1) Resistant to change 2) Some of their individual superstars might be embarrassed by the resutls which focus on contributing to the team (most obviously Tendulkar but probably others too).

  • thefountain on April 2, 2012, 7:55 GMT

    I liked this story. NZ have to be radical to compete. I hope this works out for them and they are able to grow their depth. Another alternative would be to let rugby players who are cricket talented have part time contracts so they can switch over in the summer.

  • FastTrack_Bullies on April 2, 2012, 7:51 GMT

    @Soythpaw, how about englan team , 4-0 so far and counting...oops i forgot about the numerous (read some 14) continous defeat to indians in ODI

  • soumyas on April 2, 2012, 7:27 GMT

    Nice methods by Buchnann, But looks like they are manipulating too much with players, asking them to improve in different dimensions and putting pressure can spoil gifted talents. It may work with nation which has small pool of talent like NZ, but may not with abundant talent available countries like India.

  • SouthPaw on April 2, 2012, 5:25 GMT

    Is it possible to do the same analysis (say Stephen or Rajesh or Binoy) for a country like India, for the 8 tests that the Indian team played in England & Australia (2011-12) & see what that turns up?

  • Sushrut-Cricketcrazy on April 2, 2012, 5:23 GMT

    Sounds all greek and latin to me. I thought the formula to win matches was to score big and take 20 wickets. Buchanan is the creature of the modern day theorists who are intent on complicating on what is essentially a very simple game.

  • No featured comments at the moment.

  • Sushrut-Cricketcrazy on April 2, 2012, 5:23 GMT

    Sounds all greek and latin to me. I thought the formula to win matches was to score big and take 20 wickets. Buchanan is the creature of the modern day theorists who are intent on complicating on what is essentially a very simple game.

  • SouthPaw on April 2, 2012, 5:25 GMT

    Is it possible to do the same analysis (say Stephen or Rajesh or Binoy) for a country like India, for the 8 tests that the Indian team played in England & Australia (2011-12) & see what that turns up?

  • soumyas on April 2, 2012, 7:27 GMT

    Nice methods by Buchnann, But looks like they are manipulating too much with players, asking them to improve in different dimensions and putting pressure can spoil gifted talents. It may work with nation which has small pool of talent like NZ, but may not with abundant talent available countries like India.

  • FastTrack_Bullies on April 2, 2012, 7:51 GMT

    @Soythpaw, how about englan team , 4-0 so far and counting...oops i forgot about the numerous (read some 14) continous defeat to indians in ODI

  • thefountain on April 2, 2012, 7:55 GMT

    I liked this story. NZ have to be radical to compete. I hope this works out for them and they are able to grow their depth. Another alternative would be to let rugby players who are cricket talented have part time contracts so they can switch over in the summer.

  • Gizza on April 2, 2012, 9:19 GMT

    Geez, an article on an interesting topic which people would have very different opinions on and already 2 of the first 5 comments are just country bashing. Actually, such a scientific approach may as well be useful for India, England, Australia as well as every other teams since every team in world cricket has at least a few apparent weaknesses at the moment and no team is the finished article. So @Unnithan, no need to react like that. It won't actually be a bad idea if the BCCI conducted something similar. But let's be honest. Out of the 10 Test-playing nations, India is the LEAST likely to do this. Two reasons. 1) Resistant to change 2) Some of their individual superstars might be embarrassed by the resutls which focus on contributing to the team (most obviously Tendulkar but probably others too).

  • i_witnessed_2011 on April 2, 2012, 11:24 GMT

    In cricket, Condition or Pitch are also matters most. Teams like India,Eng are very good in one condition and poor in other. Teams like SA (and AUS up to some extent) play well in all conditions. It would be interesting to know how these parameteres are fit into statistical observation. For ex: Playing Ajmal in subcontinent condtion is difficult but he would not be as effective in Eng. So how the batsmen who faces ajmal (against spin) are rated? But one thing is sure, It makes selection process more objective but at this moment it is work in progress.

  • ElPhenomeno on April 2, 2012, 11:54 GMT

    I know buchanan is going to get bashed around which is almost a given thesedays, but I think he is onto something here. A good mathematical approach to anything is going to give you consistent results, it has been proven time and again in all parts of life including but not limited to other sports. I know soccer has used something similar for establishing a players true worth. Baseball uses what this article refers to and is called "SABERMETRICS" to establish the same. It works if you know what you are doing.

  • on April 2, 2012, 12:11 GMT

    Interesting experiment. Combining cricket selection with heuristics

  • pitzicada on April 2, 2012, 12:13 GMT

    John Buchanan has to be the most overrated person involved in cricket in the last 20 years. The Australian team he is deluded enough to take credit for contained some of the best players the game has seen. Warne, McGrath, Ponting et al. The trouble is he believes his own hype.