An analyst with a cricket team told me that the question he is most commonly asked is, "If you were working for the opposition, how would you get me out?"
This presents a dilemma. The player will almost invariably have a technical weakness, but how much do you want to tell them what their greatest flaw is? Handle the situation incorrectly and you turn a minor technical issue into a major mental one.
This creates an environment where players and data can butt heads. On the one hand, you're a one-of-a-kind, free-spirited maverick who needs to be given the chance to fail and the room to fly. On the other, you're a left-hand, middle-order bat who strikes at 127, matches up well against pace, doesn't face many dot balls, but should never be allowed to start against a leggie. It's overwhelming.
However, data is just information. It's just that rather than coming to light via a keen eye from a coach, a feeling that something was off in your set-up, or an off-the-cuff comment from a team-mate, it's in graphs and charts shown to you on a screen. And numbers can be confronting, especially ones we don't like.
If you want data used properly, you need to remember that you're still dealing with people, and that there's a relationship that has to be managed.
Different players process data differently
Benny Howell has one of the best domestic T20 records in world cricket. Operating primarily as a bowler capable of providing late-order hitting, he has over 100 wickets at an average of 19 and an economy below 7. It is a record that recently earned him a deal with the Melbourne Renegades, and with two T20 World Cups on the horizon, could earn him a lot more.
Howell is explaining to me all the different bits of data he goes through to prepare for a match. Probably the most in-depth of these is what the opposition bowlers tend to bowl the ball after they get hit for a boundary.
Does Howell ever think about the fact that others will be looking at similar data about him?
"No. They don't think like that."
And he's right. His Gloucestershire team-mate Ryan Higgins credits Howell with being able to take on data and implement it in his game in a way that he hasn't seen anyone else do.
"If I did what he did, I think that I would really struggle to go away from that and use my natural instinct, whereas he is so adaptable and probably has 15 different game plans in his head."
Higgins uses Howell's example of patterns in bowlers' behaviour to illustrate. "So he may know that a bowler is likely to bowl a full ball after he bowls a short one, but if the bowler still bowls a short one, he can adapt, whereas I'd be thinking, 'He's going to bowl full, he's going to bowl full, he's going to bowl full.'"
For Howell and Higgins the same information represents completely different things. For Howell, knowledge is power; for Higgins, ignorance is bliss.
There's something larger here. Howell has ADHD (attention deficit hyperactivity disorder). He told Jarrod Kimber his impulsive nature has played a big part in him creating upwards of 50 different types of slower balls.
This fits with the picture society at large has of ADHD: the kid in the classroom who's unable to sit still. What doesn't fit, however, is the thorough, methodical approach Howell uses to prepare for each game.
Howell's success is an example of "neurodiversity", a term coined in 1998 by sociologist Judy Singer, who suggested that mental disorders such as autism, Asperger's or ADHD were differences rather than disorders. That's not to downplay the real suffering that can be caused by a neurodevelopmental condition but to understand that difference can be a source of strength. So much so that the idea of neurodiversity is finding its way into the corporate world and it's gradually being understood that the more you have people who think differently, the more organic ideas you're going to create, and the more solutions you're going to find.
The relevance of this to data is, again, in how you best harness it. Data has undoubted value and as databases grow, the quality of that data is only going to improve. The challenge for teams is tailoring that black-and-white information to a sea of individuals who come in grey.
If a data insight doesn't work once, players might be turned off data
Ahead of a match against Kent last year, Middlesex hatched a plan for legspinner Nathan Sowter. "As a match-up they said Bell-Drummond struggles against legspin early," Sowter says. "So I bowled the second over of the day. And Bell-Drummond didn't struggle that day, he whacked me for 17."
The next time the two sides met, the plan was shelved. "One of the worst things you can do is turn a player or a coach off analysis by pushing something that is a theory of yours and doesn't have enough data to back it up," says Max Backhouse, Lancashire CCC and Manchester Originals analyst. "Because it only takes a few things to go wrong for them to go, well, I won't trust him again."
This, I imagine, has analysts across the land screaming into their pillows that correlation doesn't equal causation. It wasn't that the data was wrong, it's just that it didn't happen that time.
But if you're a player and you're presented with an idea from a laptop on one hand and it runs counter to an experience that you've lived and breathed yourself on the other, which are you going to side with? Data sets are built up over years of leagues and competitions, but as a player you're never going to experience the overarching trends as a whole but only as seemingly unrelated events. You're forever at the sharp end of a data point.
That is, in essence, the argument for using data. Your experience as an individual, while tangible, pales into insignificance beside the mass of evidence that analysts across the globe have at their disposal.
However, professional athletes are known for their unwavering self-belief, and that is often part of what sets them apart from the rest of us. For them to place their trust in something that is outside of themselves is to make a mental switch, and the difficulty of that shouldn't be understated.
Lack of data is a limitation but it also tells you how much potential there is
There's also a disconnect between the international and domestic game to be borne in mind. It may appear that if England, the best white-ball side in the world, are big on data, so must every team be. However, both England and their opponents play year-round in televised fixtures, which means high-quality, up-to-date data is available from all their games. The majority of domestic players play half the year and are only occasionally televised. This means that rather than multiple Hawk-Eye cameras, those lower-level games are tracked by one analyst and their clicky finger.
A major sticking point appears also to be the feeling that the period between seasons gives players the chance to work on weaknesses in their game for months on end, meaning you start each season back at square one. This view implies that data is always going to be playing catch-up and only able to tell you what has happened as opposed to giving you any indication about what might.
But there is little evidence that players are capable of turning a weakness into a strength in such a short period of time. More often than not, a player is only capable of turning a weakness into a mild competency that is still worth targeting.
Furthermore, while it is true that there is less data available in the domestic game, analysts argue that far from the data being unreliable, it is actually a fount of untapped potential. Dan Weston, analyst for Leicestershire CCC and the Birmingham Phoenix, says that cricket's use of data is currently operating at "two out of ten in terms of saturation".
"I had a player [from another team] come to me independently and ask, 'What's a good strike rate at the death?' I said, 'It blows my mind that your team doesn't tell you this.' The player replied that his team were too old school. And the fact that a player can't go to his coach or his analyst and find that out and has to come to someone independent is just nuts."
Cricket is currently a robber in a bank who's taking cash from behind the counter but leaving the vault untouched. The value is there, but through a combination of a lack of buy-in from some teams and hesitancy from some players, it's being left behind.
Data and instinct: they don't need to be incompatible
Zubin Bharucha and AR Srikkanth are the performance director at the Rajasthan Royals and analyst at the Kolkata Knight Riders respectively. Both have been involved in some of the most high-profile, outside-the-box decisions taken in the IPL. In 2013, Bharucha was among those behind the Royals' decision to sign 41-year-old amateur cricketer Pravin Tambe. Srikkanth played a major role in Sunil Narine's move to the top of the batting order.
When I asked how they managed scepticism towards data from players, I expected exasperated sighs, frustrated off-the-record comments, or flat-out denial. What I didn't expect was how much empathy they showed for players' concerns. The glitz and glamour of the IPL may be thousands of miles from the English county game but the concerns, worries and doubts of the players as people are the same.
Bharucha says that the current generation of cricketers is one that has developed in an era of hands-off coaching: express yourself, find your own way, do what works for you, and pass us the textbook on your way out, I've been meaning to throw it away for years.
However, with the advent of data, there is suddenly a wealth of information and recommendations to give to these players so they can improve. It's not that either method is wrong, it's just that they contradict each other, Bharucha points out. You can't tell a player to go their own way and then sit in the back seat and tell them they should have taken the last left.
Srikkanth goes further. Having worked as an analyst for 14 years he says he used to harbour frustrations against those who were sceptical about data, but now believes players wouldn't be human if they didn't have at least a few doubts. Why? Because he concluded that cricket is, and always will be, primarily a game of instincts.
There's a video where Glenn Maxwell talks through a battle he had with Jhye Richardson during a one-day game between Victoria and Western Australia. Maxwell hit Richardson for three consecutive sixes after Richardson had been bowling, by Maxwell's admission, "rockets". Maxwell figured the best way to get him out of the attack was to launch a rocket himself.
"I saw [the first one] really early and hit it for six. The next one, I thought: pretty sure he's going to bowl me a slower ball here, he's done me a couple of times with a slower ball. So I stood dead still and waited for it, and sure enough it was a slower ball, back of a length. And the next one I thought he won't go short here, he knows I'm expecting it, so he'll go full again - and, sure enough - full."
In this situation, Maxwell could feasibly say it was all instinct and that no computer could teach him that inherent feel for the game. And equally, an analyst could say, "But what informed those instincts?"
Maxwell succeeded through his knowledge of Richardson's behaviours and patterns, and because was able to compute the likelihoods of certain events as he played. He had that information because, a) he is friends with Richardson, and (b) he has played against him a lot.
All data is trying to do is enhance your instincts and make every bowler in the world your best mate.
You need captains and coaches to buy in
Analysts say part of the challenge is that, because of the creed of individualism that is common among many top players, they often feel better if they think they have achieved something by themselves and not with the aid of data.
This means analysts will often aim to win over the captain or coaching group, as opposed to individual players; that way the effects of data influence the wider group without the players even noticing. A bit like how your mother used to feed you vegetables by chopping them up really small and hiding them in your spag bol.
Srikkanth emphasises that you can't expect players to be data-literate, meaning teams have to be creative to make sure that the information provided is as accessible as possible. One way KKR do this is through a team app, where players can scroll through pdf reports, wagon-wheel charts, beehive bowling graphics, or even tailored clips from matches. It's all there and the player can choose which format they want the inputs in, whether that is a written report, a graphic, or video footage.
While this all feels like it is inherently sensible in the short term, it also feels like a holding pattern. If data is to truly reach its potential, both parties have to understand how it is used and the way it operates.
Bharucha argues that for data to really take the next step, it won't be through technological advancement but the use of data at youth- and academy-level cricket. That way, players of the next generation will grow up with an understanding that data need not be the back-seat driver telling them what to do, but the sat-nav suggesting the best route to take.
Players need to adjust - but so do analysts and coaches
Each player interviewed for this article was asked whether, given the option, they'd like to know their greatest weakness in cricket. They all laughed and said no, before changing their answer and saying they wanted to know how to improve. In a nutshell, this is where the balance lies between making effective use of data and getting player psychology right.
Professional athletes are a jumble of hard work, talent, adaptability and boneheadedness. They have devoted their entire lives to their careers and it requires a lot for them to cede some control to others. But as much as players may need to compromise, there also has to be a recognition of the players' lifetime commitment from analysts and other proponents of data. Overall, while a player's apprehension about data may be irrational, that's not to say it's invalid. It is perhaps a collision of best intentions that the individualism of players now stands somewhat at odds with cricket harnessing the true power of data.
The rise of data in cricket is often likened to the sport shifting on its axis from art towards science. This is a misconception. Data doesn't provide answers, it provides suggestions. Much in the same way that card-counting in blackjack increases your odds of success rather than guaranteeing victory. However, it's important to remember that it only works if you trust in and stick to the process. For players wishing to get the most out of data, it may be the case that the best bet is to go all in.