Numbers in the Dressing Room: How Analytics Changed Selection

Picture the team hotel on the eve of a big match. Twenty years ago the conversation over dinner might have run on memory and instinct: he looked good in the nets, he got a hundred here last time, the pitch will turn. Today there is often a laptop open at the same table. It shows how the opposition's number four scores against left-arm spin in the middle overs, which end suits a particular seamer, and how many balls a new batter typically needs before he starts hitting boundaries. Cricket has always loved statistics, but the way those statistics are gathered, sliced and used in selection has changed almost beyond recognition.

From scorebook to ball-tracking

For most of the game's history, selectors worked from averages. A batter's runs per innings, a bowler's wickets at a given cost: these were the currency, and they were crude. An average of forty could hide the fact that a player scored heavily on flat pitches against weak attacks and struggled whenever the ball moved. Scorers had long kept detailed records, but turning them into something a coach could use quickly was a laborious job.

Two developments changed that. First, ball-tracking and detailed ball-by-ball databases meant that every delivery could be logged with its speed, line, length, shot and outcome. Second, cheap computing power meant that analysts could query those databases in minutes. The wagon wheel, which shows where a batter scores his runs, and the pitch map, which shows where a bowler lands the ball, became standard tools rather than television novelties.

What the numbers are actually good at

The clearest gains have come in match-ups and roles. Instead of asking whether a bowler is good, analysts ask whether he is good against this batter, in this phase, on this type of surface. In limited-overs cricket, where the game is divided into powerplay, middle and death overs, such questions have obvious answers: some batters accelerate against spin, some stall; some bowlers are at their best with the new ball, others when the field is spread.

Data has also changed how teams think about the value of runs. A fifty in twenty-five balls and a fifty in fifty balls look similar in a scorebook but can have very different effects on a T20 match. Modern analysis attempts to measure the impact of an innings relative to the situation, rather than just its total. That has pushed teams towards picking players for specific jobs, such as the batter who can attack from the first ball or the bowler who can be trusted in the final overs, rather than simply choosing the eleven best individuals.

Fielding and fitness have also become measurable. Teams can track how much ground a fielder saves, how many chances are put down and where, and how much bowling load a quick has taken on. That last point matters enormously for selection, because the best bowler available is not much use if he breaks down in the second Test.

Where franchises led the way

Franchise tournaments accelerated the trend. A side that assembles a squad at an auction, with a limited budget and a short window to play together, has every reason to value information. Support staff in these competitions often include full-time analysts, and some franchises employ people drawn from other data-heavy industries. International teams, watching their players come back from such tournaments with new ideas, have steadily built up their own analytical departments.

England's white-ball revival after the 2015 World Cup is often cited as an example of a team that blended a clear tactical philosophy with careful use of information, choosing players who suited an aggressive style and then supporting them when it occasionally backfired. The principle, rather than any single statistic, is the lesson: data works best when it serves a clear idea of how the team wants to play.

The limits of the spreadsheet

Cricket is a stubborn sport for analysts. Sample sizes are small: a Test batter might face a given bowler for only a handful of overs in a career. Conditions change, with a pitch that behaved one way in the morning behaving differently after lunch. And a player's form is not simply a number; it is confidence, rhythm and a dozen small habits that do not appear on a chart.

There is also the matter of temperament. A selector deciding whether to give a debut to a twenty-one-year-old is making a bet on character as much as technique. How does he react when he is hit for consecutive boundaries? Does he recover from a first-ball dismissal? Numbers from domestic cricket tell you something, but not everything. The same applies to leadership and dressing-room chemistry, which can make a side greater than the sum of its parts.

• Small samples can make ordinary players look brilliant, or brilliant players look ordinary.
• Context matters: a low score on a difficult pitch can be better than a century on a flat one.
• Over-reliance on match-ups can lead captains to predictable decisions that opponents prepare for.
• Players who feel reduced to statistics may lose the freedom that makes them effective.

Getting the balance right

The most effective teams use analysts as translators. A good analyst does not hand a captain forty slides; he offers three clear points before the toss and answers questions during the game. A good captain, in turn, knows when to follow the data and when to trust the feel of a match. Some of cricket's most celebrated decisions, such as bowling a particular player at a crucial moment, have worked precisely because they went against the percentages.

Equally, data can protect players from the unconscious biases of selectors. A quiet performer who consistently delivers value may be overlooked by those who prize flair, and a spreadsheet can make the case for him. In that sense analytics does not just add information; it challenges assumptions.

What comes next

As the tools improve, teams will probably get better at modelling fatigue, predicting injuries and tailoring training to the individual. The risk is that every side ends up with the same information, which makes the human edge more important rather than less. The teams that win will be those whose coaches ask the sharper questions of their data and whose players are secure enough to play with freedom within a plan.

So the next time a selector explains a surprise omission, remember that there is probably a database behind the decision, and a conversation, and a gut feeling that no database could replace. Cricket's great charm has always been that it blends the measurable with the unmeasurable. Analytics has sharpened the first half of that equation, but the second is still where matches are won.