Trent Woodhill's brave new, data-driven world
"I'm moneyballing," says Trent Woodhill. Find your way to any new-age cricket thinker - and Woodhill is as new-age as cricket will tolerate - and they will, at one stage or another, mention the title of Michael Lewis' book on baseball sabermetrics. It's like cricket's new thinkers are in a book club together, and they only ever discuss one book. If international coaches toy with the ideas behind Moneyball, it is the domestic T20 teams who truly obsess over them.
Woodhill is by trade a batting and assistant coach. He must be pretty good, as his two personal clients are Steve Smith and David Warner. But his methods, and increasingly his willingness to talk openly about his complete disregard for cricket's norms and its hierarchy of former Test players turned coaches (Woodhill never played first-class cricket, the cheeky monkey) mean that he isn't Cricket Australia's (CA) batting coach, but Melbourne Stars'. But it's his job as Stars' list manager that is most interesting.
Moneyball's hero is Billy Beane, a man on the inside who thinks the entire system is wrong, and when faced with a choice, takes a new way over perceived baseball wisdom. Beane is a not a current player or a coach, as heroes often are in most sports narratives. He is the general manager, the man in charge of the list and day-to-day operations. To most athletes that might seem a pretty uncool role, but for people like Woodhill, Beane is about as cool as it gets. Woodhill is a devoted sports nerd, and when he gets excited, he might talk about Martin Guptill's ball-striking or AB de Villiers' hands, but he also gets just as excited about guys like Theo Epstein, the backroom genius who broke the Chicago Cubs' title-drought, and even ESPN baseball writer Keith Law.
Woodhill is not your typical Australian cricket coach. He is not a proselytiser of the baggy green, he never mentions it in our conversation, nor does he talk about Steve Waugh's mental toughness. Recently he spoke out against sledging, and he doesn't bring up the glory days of Australian cricket every seven minutes. Instead, Woodhill quotes, verbatim, from the Big Short - another Michael Lewis book turned movie:
"It's time to call bullshit."
"Bullshit on what?"
Woodhill is calling bullshit on how cricket is played, coached and talked about. He isn't here to listen to former pros tell us how cricket has always been; he doesn't care about your Test record from 1974; he thinks most cricket coaches are living off their playing careers without helping much at all, and he believes cricket's current stats are holding cricket back. His way of putting together the best list for Stars is to use science and hard facts in and around cricket's steaming pile of opinion. "If you've been out of the game for six years, you have no idea what is going on in the modern game".
In a sport that still has a weird relationship with coaches and seems to actively distrust high-performance managers, Woodhill's job will elicit punchlines from the moustachioed set. But in franchise cricket, someone needs to put the list together. The coaches swan from gig to gig, the captains often the same; each team has a GM or CEO, but often they are more in charge of running the franchise, not list curation.
Woodhill's job out of season is to assemble Stars the best 18 players he can find. Some of it is old school: he haggles with managers, talks to players about opportunities, but he is also looking beyond the perceived wisdom on a player, looking for the truth. He wants to know more about them than anyone else, so that when he makes a decision it isn't based on a recommendation, or on how they look, but on as many facts as possible.
The problem is, for all the obsession cricket has with numbers, we haven't yet moved on to a data obsession. Opta, perhaps the first big player in cricket data, still has a person at the ground mapping out where each ball was fielded, where it pitched, and if the batsman was in control of the shot. They don't map field positions, or how quickly fielders move to the ball, and at the moment they don't have their own cameras at grounds. Quite often the data that Opta gets ends up on apps and websites and not with the teams themselves. There are other providers such as Cricviz. But even then we are at the very beginning. Data just isn't something that cricket has invested in greatly yet.
Where this plays out is when, for example, Woodhill wants to know if a bowler is good at the death and he doesn't want to rely on opinions that may be coloured by memory. Let's say he wanted to sign Sean Abbott. The conventional view might be that he isn't good at death bowling because Travis Head went all 4, 6, 4, 6, 6 and 1 in an over against him. But the fact of Abbott's death bowling isn't in one bad over or one good over; and it's not always just in conventional death-bowling stats. Abbott may go at 8.5 over in those overs, but he might also concede 30% of those runs from edges, and there might be dropped catches and fumbles; or his fielders might not be as strong as those of other teams; perhaps he bowls on a smaller ground than others; or he has bowled more against bigger hitters. Woodhill moves to the best stats and data cricket has and then makes his judgement. He may not have all the facts he wants, but he's as far away from lazy misconceptions as he can get.
When Woodhill talks facts, he isn't doing it in some post-truth 2016 way; he's actively trying to find out what is real and how to use it. Other teams have some of the same data, others perhaps have even more, but often those teams are run by people who still believe the old ways are the only ways. Or that they could deliver more in a team environment that backed their talent. They are also looking for a different kind of player, what Woodhill calls a three-tool player, which was something that the pioneering Bob Woolmer always liked. Someone who can help win them the game with batting, bowling or fielding.
The bigger part of what Stars are doing is looking at cricket in a more forensic way. When John Hastings went down injured, they weren't just interested in replacing him as a bowler who batted a bit; they delved deeper than the allrounder tag. While Hastings is known as a big hitter, for Stars he only got two hits last year, and only one of them came off. In his entire career for the team, he has batted in about half their matches, never scored more than 22, and has scored, in total, 139 runs. Hastings is for them a specialist death bowler. "While cricket might see him as an allrounder, or rate his batting," says Woodhill, "we have to deal with what he did for us, and replace it."
Every time they need to fill a place they need to look at it in far more detail than looking at batsmen, bowlers, wicketkeepers and allrounders. How many six-hitters? How many strike-rotaters? How many slow-pitch tamers? How many new-ball seam bowlers? How many new-ball spinners? How many death bowlers? How various should the bowling options be? How many different wicketkeeping options? How many batsmen who are specialists against spin, or pace? How many boundary riders? How many players who spin it away from a right-hand batsman? How many batsmen who spin it away from a left-hand batsman?
And these are just playing types, excluding the different personalities to consider. Players who have experience in multiple locations. Players with previous captaincy experience to help bulk out your strategic thinking. You might want a few leaders, or you might want a group where the leaders evolve naturally. You might want a diverse group of people, as diverse groups of people are better at problem-solving, or you might want a similar group of people, as they tend to work more smoothly. And all this before your team owner or chairman comes in and says, "We need to you to sign a big name for marketing this year."
It's a tough job to put a quality list together. And that means the gut in cricket selection will become more of a science. Many international sides still pick potential over performance. Franchise cricket is doing that less and less. That means picking older players, for whom teams can look at a data profile beforehand and know what they should expect.
While the old-school baseball front office no longer exists, many cricket teams have barely started flirting with new ways of looking at the sport. It is often the T20 franchises that are doing the most adventurous work, because as a new franchise you need to win to guarantee your survival. New franchises, even the ones run by cricket boards, can't float for decades on legacy money and nationalist supporters. They are fighting for the best players, the best coaches, and all of cricket's new fans. That means instant results, and teams will take them any way they can get them.
While the Australian team will talk of picking Matt Renshaw and blooding him for the future, despite the fact he might not be seen as the best player now, in franchise teams they need the best here and now. Rob Quiney, who has been passed over by Victoria this year as they try new players, is a perfect fit for a franchise. They have data on what he can do, they know he is a seasoned pro who has played cricket overseas, and they have decided he will fit well into their side. T20 was first seen as a young man's game, but quickly it has become a game for older, more experienced cricketers whom traditional cricket often moves on from. It's also a common theme of sabermetrics: a young player simply doesn't have the data to mine; they are unknown where the old players have history.
But it isn't all milk and honey for those older players. Franchise cricket can't afford to live on reputation. They need to win, so giving a player a romantic ride at the end won't happen. The players have to adapt to stay relevant. Often that is perfecting a shot they have never mastered, or learning a new delivery. Sometimes it's by adding a new skill altogether. A 37-year-old former international who bowled offspin in his youth, but didn't need it at international level, might need it again if he wishes to cash in on the last few years of his career. His offspin might only be useful on three pitches a season, against certain types of left-handers who struggle to score quickly against spin, but even if he only bowls six overs, that adds to his versatility and makes him more desirable to a franchise. Six overs of mediocre offspin to a list manager can be the advantage that wins a game.
Every list in the BBL has to produce a squad of 18 that has enough variation in batting and bowling, for the entire series. Because once the 18 names are signed, there can be no trades, and unless there is an injury, no replacements. So, unlike American sports, where you can add a player for a two-week stint at a whim, or trade when your team has deficiencies mid-season, the BBL is a static, closed league.
If your team finds itself struggling against fast bowling coming up to a week where three of your opposition have the best fast-bowling attacks, instead of hiring a fast-bowling specialist for the week, like an American sport might, you will just have to wing it with your available list. Likewise, if after three weeks of the tournament, Perth Scorchers think they have too much batting and not enough bowling, and Hobart Hurricanes think the opposite, they can't make a trade to rectify that. And that is one of the ways that the Big Bash is still not fully formed.
There are also no long-term contracts, perhaps in part because the players are still yet to sign what they hope will be a very lucrative new MOU. Perhaps more importantly there are no drafts or auctions. The original BBL teams were made up primarily of players from that state, and even now it's first in, first served, and once served, you can't do a damn thing.
Woodhill calls bullshit on this, and you can see why. He is trying to build the best list for his captain and coach to work with each week, and once the 18th contract is signed, he has no control. How does that ensure the best 88 players in the country are in action each game? How does that ensure that teams that had a bad start can fill holes in their rosters that might be caused by injuries, or national selections that are out of their hands? That makes it harder to Moneyball, or even use an old-school approach to list management.
But perhaps the biggest hole in Woodhill's plan isn't the access to open player markets but to the data. Baseball and basketball's evolution has come from sabermetrics, big data, and analysts taking over. Woodhill is not a data analyst; he's a batting coach. While he craves more data to factor in, cricket has barely touched the surface. What all forms of cricket need, perhaps none more than T20, is a wins-above-replacement stat, so we know good a player is compared to a standard replacement player for that position; or a plus/minus runs stat that allows us to know what real effect a player has on the game.
For every player that Woodhill picks, or wants to pick, what he desperately wants to know is what impact they have on games. He can ignore batting averages and cling on to strike rates, but he needs far more than that. He needs balls per boundary, per types of bowler, perhaps per types of pitch, and whether their runs come in overs 1-6, 7-10, 11-15 or 16-20. He needs to know what teams win T20 matches on different kinds of pitches. Is the number of times contact is made with the ball more important? Or who plays fewer dot balls? Or the number of singles/boundaries? Should every batsman be a boundary-hitting monster, or should you be picking batsmen based on their ability to rotate strike? Woodhill says, "All I want from some batsmen is to run twos so he can get the big hitter back on strike."
A player like Sachin Baby may look out of place in a batting line-up like Royal Challengers Bangalore, but if his job is to ensure that Chris Gayle, Shane Watson, Travis Head, de Villiers and Virat Kohli face more balls, his value is greater than his runs. When Kohli smashed Australia in the World T20 this year, MS Dhoni wanted credit for all the twos he helped steal to add to Kohli's total and keep him on strike. At the moment, cricket is still trying to work out how many extra runs non-striking batsmen are worth, meaning that Woodhill is guessing, rather than knowing.
And it isn't all we don't know.
Which player in the Big Bash runs the quickest three? Which bowler has the most successful slower ball? Who has the fastest hand speed? Which player fumbles the least? Which fast bowler saves the most runs in the field? What is the catch/drop percentage of each player? Which players only hit boundaries when the ball is new? How many balls at the start of an innings does each player need to hit their first boundary? A lot of these are simple things that American sports, and even domestic sports like Australian Rules and Rugby League would have access to.
And that's before you get to fun stuff like the probability of a player playing a good shot compared to a bad one. What kind of things can we learn about a batsman or captain using spatio-temporal pattern recognition? These are things that draft camps might tell you, that you would ideally want to learn before picking players, that cricket is still trying to find answers to.
Woodhill doesn't have all the answers. Some he wishes he does, others he wishes he had the staff to prepare answers for him. CA still don't have a great big analytic lab. Most of cricket's data comes from a few companies that have limited funding, and without open-sourcing of data, cricket can't have the big leap forward that baseball and basketball have had. MIT have the Sloan Sports Conference to move sports forward using science. Something like that would massively help cricket, but since the sport won't pay for its umpiring technology system, we're a way off a yearly cricket science conference. For now, Woodhill relies on the tiny bits of data he gets and hopes that he finds things that other teams aren't yet looking for.
Without access to MIT scientists, Woodhill improvised by crowdsourcing via social media. He uses the new cricket media, from professional writers to amateur tweeters, to question his beliefs. Without a front office to bounce ideas off, he might do that with a random tweeter, or he might read a blog by someone that makes him rethink a belief he had. For his way to work, he needs not only to call bullshit on the old ways, but also on new ways of doing it that may just be wrong.
Being that so many people in cricket are claiming to be moneyballing, and so many teams aren't very good, it is easy for many critics to laugh at how often the term comes up. Most critics just think it's all nonsense. There are bound to be missteps when you are mainly trying to use another sport's evolution to force your own. Not every team has the intelligence of Andy Flower and Nathan Leamon. The two teams Woodhill works for, Bangalore and Stars, both came second last season. And his list for Stars was so good, they signed the same list again this year. And it is that list that has them as favourites for this year's tournament.
Right now cricket's moneyballing is more Mick Lewis than Michael Lewis. Woodhill is one of a very small crew of list managers in Australia, and world cricket. He is just an amateur, playing with the beginnings of a professional structure, and he knows it. He has no idea if what he is doing is correct, but he is using the best current thinking to make his decisions. Which is at least a step towards a new age.
Whether Woodhill's methods will be proved right or wrong we don't know, but what he is doing has been working, and the more it works, the more list managers, hands-on general managers and data we'll see in cricket. The lazy ex-players might call bullshit on the new ways, but that is their opinion. Woodhill's more interested in facts.
Jarrod Kimber is a writer for ESPNcricinfo. @ajarrodkimber