How much has the T20 format changed while you've been a top coach?
When I first started, T20 was just about going out and expressing yourself, and if a couple of guys had a good day, you ended up on the winning side. Now I think T20 is the most analysed game, because things happen so quickly. If you can gain a small advantage from a tactical perspective, that makes a big difference to the outcome of the game. In the 2012 World T20, we played Super Overs against each of the finalists, so we knew that skill-wise, we were pretty close to two good sides. So we tried to find other ways of trying to gain that small advantage. Hence there is a lot of work done away from the ground by our analyst and support coaches. A lot of senior coaches add to it as well.

Is data key to getting that advantage?
Only if it is real. We can all gather loads of data, but I guess the key is to break it down to what's actually useful. What data is relevant is actually to the conditions you play in? What about the opposition? Is it a day game or a night game? What position are you likely to be in at that stage? You're trying to look at trends rather than chuck a whole heap of numbers at players.

Is T20 the format where data is most effective at yielding that advantage?
I think as long as you have enough of a sample. If you don't have enough of a sample and you guess based on a small amount of data, you can be exposed. But there is a tipping point where you've decided this is a trend, and that you need to pay attention to it. For example, if you look at someone like Shikhar Dhawan, against offspin, he's struggled. But you've only really got a nine or ten-ball sample - so you've got to make a decision on whether it's too small to be a pattern, or do you take a punt on it.

That's what we looked to do in that first game in Nagpur, but it wasn't something we were sure of. Similarly, Glenn Maxwell really struggles against left-arm spin and becomes a bit one-dimensional. On a turning wicket he became a bit of a one-trick pony, really, the way he looked to take Mitchell Santner on. The data sort of suggested that, and it played out during the day as well.

"When the selectors talk, we speak about adding context to players' performances, rather than lining them up and picking who made the most runs"

Who gathers that data?
We have relationships with different countries - the ECB and Sri Lanka Cricket - where we share information. The matches are played in different time zones and through different TV networks, so we gather that. Some of it is more reliable than others.

You have data on your own players. How much do you pass on to them? What do you hold on to?
Depends who the player is. Some players respond really well to information. Other players, you want to keep it as simple as possible, or you could just cloud them by throwing information at them. Grant Elliott loves stats and data. He's a very thoughtful cricketer. The way he bowls, he has to stay one step ahead of the game, because he doesn't run in and bowl 145. He has to be clever in how he operates. And also because he bats in the death, he has to have some cues about what they're likely to bowl, so that he can hit. Kane [Williamson], meanwhile, has great intuition that data doesn't see, so there's a great strength there.

How much more potential is there for data to be used in cricket?
I don't think we do it as well as we know we can. For example, in baseball you can determine strike rates in little pockets or zones where the ball arrives. That's really useful. Once you get to that point, then the game will keep evolving. Batsmen will identify where bowlers are going to target them. What we've found with all the different shots people are playing is that bowlers have caught up, and batsmen will have to keep evolving.

What were you looking for when you picked the spinners for this tournament?
We felt that we needed two guys who could bowl in the front six. Nathan McCullum is definitely one. Mitchell Santner is someone who had done it a couple of times. And we wanted someone through the middle who could take wickets for us. Through that period in the middle where sides are trying to build, if you're able to keep taking wickets, then you give yourselves opportunities to reduce the length of the death. That's where Ish [Sodhi] has really stepped up. Just impressed with the way he's bowled in the last 12 months. He started in Test cricket. He was in and out a little bit. He had some things that he's gone away and worked on, and he's certainly come back a far better bowler.

You've said that you pick players to play specific roles. If someone is performing really well in domestic cricket, do you try to find a role for them in the team?
It's all about context. Where are they scoring their runs? What role are they playing? How are they going against the best bowlers? How about on the tougher surfaces? Or when there's spin? Has he put the team first? When the selectors talk, we speak about adding context to players' performances, rather than lining them up and picking who made the most runs.

Santner is a guy who - his domestic stats don't suggest that you'd pick him. I think it's more about what we saw in terms of the skill set. New Zealand wickets don't spin. For a spinner, if you average 40 domestically, you're going okay. Also, players take a little while to develop. You're trying to select the player for what they are now, rather than what their stats suggest over a two or three-year period.

Is pitch reading an art or a science?
I think it's a combination. You know information about the ground and previous surfaces, but we also know that every pitch is different. You need to use your intuition and also gather information from other people around. I think there are some pretty simple characteristics, which if we understand, we know how the pitches are going to play. In terms of the clay, the feel of it. Firstly, you want to work out if there's going to be any pace in it. You can work that out pretty quickly by looking at the grass cover. Very rarely do you see a pitch with no grass that will have pace in it.

Sometimes you will also have a sheen on the wicket, and the new ball can kiss through. As the sheen disappears, it will grip on the surface. And then the feel of it. Are the cracks flaky? Does it move when you touch them? You can gather information on the dew factor as well. There's a whole heap of information you can process, and you don't always get it right.

"Gone are the days when you try and bowl six yorkers, because if you get two wrong, you go for 14 an over"

I guess one of your information sources is the curator, but that's not always a reliable source, is it?
Of the four group games, three of the curators said it was going to be hard, fast and bouncy. In Mohali he was spot on. He didn't actually make a judgement, he just gave us information. The other three said hard, fast and bouncy. That's when you have to make your own assessment. It's not as if people are being tricky. I think what's hard, fast and bouncy for us is possibly a little different for others.

Are regional variations gaining more emphasis in the years you've been involved?
I think so. I've been really surprised how slow the pitches have been for this time of the year. I know that towards the back end of the IPL the pitches slow up, because they've been used. But at this end of the season it has been slow, and I'm not sure of the reason for that.

As pitches become more and more different, do you need to have a wider skill set in your team?
I think your best players adapt everywhere. As your other players mature, they do as well. Some players are better suited to certain conditions. That's something that we're becoming better at, as a support staff group. We know that our best XI is not always our best XI in those conditions. There are certain parts of the world, like South Africa or India, where you have to pick your best team for those conditions.

Is T20 more a horses-for-courses format?
In New Zealand, for one-day cricket, the pitches are very similar. We produce very flat wickets with good pace - your best team is your best team. In India, sometimes we thought three spinners was the way to go - in Nagpur and Kolkata, for example. Mohali, we thought had more pace and carry. And with Dharamsala, we thought with Mitch McClenaghan bowling hard into the wicket, it would create some variation in bounce.

Is it tough to leave out performing players when they are not needed in certain conditions?
I think every player reacts differently. Most players want to play all the time. But once players look at the bigger picture, or when we tell them we don't want to keep playing them till they break, they come around. Without a doubt, the management in that is challenging, but we've got an outstanding group which thinks about the team. If it's in the best interests of the team, sure there might be five or ten minutes when they're frustrated, but really quickly they turn that around. They say, "That's what the team needs", put a smile on their face, and get on with it. We've got a really selfless group, and that's evolved over a period of time. Selection has a little bit to do with that. Sometimes if you feel like you're this close to being dropped, you can be a little selfish and try to get a score. But we try and make our judgements based on what you add to the team, rather than necessarily what your figures suggest.

How important is communication when you're batting first and assessing a par score?
We get information from the guys that are out there - they are in the best position to assess. We think we know how it's going to play, but we never truly know. There's a lot of sharing going on in the group, a lot of talk between the guys going out to bat.

We played a one-day game in Dunedin against Sri Lanka where it was nibbling around, and we were 80-odd for 5. The guys coming back said it was tough to bat on, so if we scrap our way to 200, it could be a good score. And then Luke Ronchi and Grant Elliott got going, and said: "It's flattened out, we need to get close to 300 here." It was a big change from 200 to 300, but I think we ended up getting 360. That day there was a continual passing back of information, so that we could change our sights.

And there are times when you have to revise your score downwards. In our first World T20 match, in Nagpur, we knew it would slow up, but we didn't know it would slow up as much as it did. Then it actually spun as well. I was thinking 140-150, but with Corey Anderson's innings, he said, "If we can get ourselves to 120, we're in the game here." And in the end it was sort of an ugly 126. I thought we'd had about par there. We had to lower our sights there.

You scout and analyse opposition players as well. Have you come across players who are tough or impervious to that kind of analysis?
Someone like Chris Gayle - his tempo is very different from one day to the next. The data you have is often irrelevant. You know if he flicks the switch, the game can change so quickly. Other days, you can actually create some dots, but you know the game is not over. You always know that as long as he's out there, the game can change very quickly, so you need to get him out. You're asking for trouble if you try to contain him.

"Some players respond really well to information. Other players, you want to keep it as simple as possible, or you could just cloud them by throwing information at them"

AB de Villiers is also tough because he can score 360 degrees. He's very difficult to bowl a dot ball to, because he's able to manoeuvre the ball anywhere. He's very hard to stop in full flow. You need to try and identify an area that you want him to hit.

Any bowlers who are similarly difficult?
With Mustafizur Rahman, we know what he does, but he's got a good wrist, so he's able to get good dip on the ball. It's all very well picking him, but it's tough playing him, because of the dip he gets and the grip. We haven't seen a lot of him. We've seen a little bit but not on surfaces that grip so much as the one at Kolkata. He's impressive.

Jasprit Bumrah is also very difficult to pick up if you haven't faced him before. If you've had the opportunity to even line him up a couple of times, then it's not so difficult, but the first time is a real challenge.

How much has scouting changed in the time you've been coaching?
We are more accepting that some players don't want it. For some, one little gem is all they need. With Ish - if you keep it really simple with him - he knows how to bowl. You might see a hole in the swing of a batsman, and you tell him the wrong'un could be a good option. His plan's really, really simple. He will set the batsman up to bowl the wrong'un.

Can you give an example of when scouting gave you a very significant advantage?
Against Bangladesh in Kolkata, a lot of really good information came out from the bowlers' meeting, with help from analyst Paul Warren and bowling coach Shane Jurgensen. We worked out the lines and change of pace that we needed to bowl. You're used to bowling offspinners to left-handers, but maybe sometimes you'd bowl an offspinner to a right-hander. Basically we married up the data with the players we were facing, and the scouting plans were excellent.

Has the yorker become less of a weapon now?
There are certain players and grounds where the yorker is great, and if you're able to execute it, it's a really good ball. But if guys lap and reverse lap, and you put your field back straight, then that challenges your yorker. Gone are the days when you try and bowl six of them, because if you get two wrong, you go for 14 an over. But there's certainly a place for good yorker bowlers in the game.