How cricket analytics works

Cricket poses a problem most sports don't have to solve: the same underlying skills — batting, bowling, fielding — are judged completely differently depending on the format being played. A batting approach that is correct in a five-day Test match, where survival and building a long innings matters most, can be far too cautious in a 20-over match, where scoring quickly matters more than staying in. Cricket analytics has developed largely around adjusting for format, rather than assuming one number means the same thing everywhere.

The questions analytics is built to answer

Analysts, coaches and broadcasters use cricket analytics to answer questions that shift meaning by format: is a batter scoring at a rate appropriate to their format and match situation? Is a bowler being genuinely economical, or just bowling to a weak lineup? And, when rain interrupts a limited-overs match, what is a fair revised target given the overs and wickets lost? That last question has its own dedicated, widely used statistical method, because it comes up often enough to need one.

The data it runs on

Cricket analytics is built on ball-by-ball data — for every delivery, the outcome (runs scored, how they were scored, dismissal type if any), the bowler, the batter, and match context (innings, over, required run rate). This level of detail has been recorded for international and top domestic cricket for a long time and is dense enough, because a single match can contain hundreds of individual deliveries, to support fairly granular situational analysis even within one game. Video and, increasingly, ball-tracking systems used for umpiring decisions (line, length, and trajectory) supplement this with a physical layer of data used both for officiating and for bowler analysis.

Measuring batting by format

[[batting-strike-rate|Batting strike rate]] measures runs scored per 100 balls faced: strike rate = (runs / balls faced) × 100. Its usefulness depends entirely on format and match situation. In a Test match, a strike rate is a secondary consideration behind simply not getting out, since the format rewards a long, patient innings over quick scoring. In a T20 match, strike rate is often the primary measure of a batter's value, because the format has a hard limit on deliveries and scoring quickly is usually worth more than preserving a wicket. Reading a batter's strike rate without knowing the format and the match situation — chasing a target, batting first, early in the innings versus the closing overs — is one of the most common ways to misjudge a player's performance.

Measuring bowling

[[economy-rate|Economy rate]] measures runs conceded per over bowled: economy rate = runs conceded / overs bowled. Like strike rate, its meaning shifts by format and situation — a bowler conceding five runs an over looks expensive in a low-scoring Test innings but very economical in a T20 match where ten runs an over is a common par rate. Economy rate also needs context from the overs a bowler was assigned: a bowler used mainly in the tightest, lowest-scoring overs of a limited-overs innings will naturally post a better economy rate than one regularly given the overs batters attack hardest, independent of raw skill.

Adjusting for interruptions

Rain interruptions are common enough in cricket, especially in formats played outdoors over a fixed number of overs, that the sport adopted a dedicated statistical solution: the [[duckworth-lewis-stern-method|Duckworth-Lewis-Stern method]]. It calculates a revised target for a team batting second when overs are lost to weather, based on a statistical model of how a team's run-scoring resources (overs remaining and wickets in hand, combined) decline through an innings, rather than a simple pro-rated adjustment, which would systematically favor whichever team happened to be batting when the rain arrived. It is a good example of analytics built to solve one specific, recurring, format-driven problem rather than to generally describe player quality.

How the work is done in practice

Analysts build situational splits — a batter's strike rate and dismissal patterns broken down by phase of the innings (powerplay, middle overs, death overs in limited-overs cricket), and a bowler's economy rate broken down by the same phases — because average-across-the-match numbers obscure most of what a coach actually needs to make a selection or tactical decision. Opposition analysis typically pairs ball-by-ball data with video, identifying a specific batter's scoring zones and a specific bowler's typical lengths against different types of opponents, then feeding that into a match plan turned around between fixtures. [[elo-rating|Elo-style rating systems]] and [[win-probability-model|win probability models]], adapted from other sports, are increasingly used at the team level to rate international sides and to estimate a team's live chances of winning a match in progress given the current score, overs and wickets.

Common mistakes and misreadings

• Reading a strike rate or economy rate without knowing the format. The same number means something close to opposite things in a Test match and a T20 match.
• Ignoring the phase of the innings a stat was accumulated in. A death-overs economy rate and a powerplay economy rate answer different questions about a bowler, and averaging them together loses the distinction.
• Judging a player's ability from a small number of matches, particularly in shorter formats where a single innings can be a large share of a player's recent sample; short-run numbers are especially prone to regression to the mean.
• Treating Duckworth-Lewis-Stern targets as arbitrary rather than as a deliberately built statistical model — the method is widely used precisely because ad hoc pro-rated targets produced clearly unfair outcomes for whichever team was batting during an interruption.
• Comparing raw statistics across eras or pitch conditions without adjusting for how scoring rates and pitch behavior have changed over time and vary by ground.

Browse every tool in this category for sports-data vendors, and see how sports teams use analytics for how tracking and video tools are used across sports generally.