Is Total Shots Ratio Still Useful in the Expected Goals Era? What the Simplest Shot Metric Can and Cannot Tell You

Total shots ratio, usually shortened to TSR, is a team's share of all the shots in its matches: shots taken divided by shots taken plus shots conceded. A value of 0.50 means a side takes as many shots as it allows. Shot counts are among the most widely recorded figures on RubiScore, which makes TSR one of the few analytical measures anyone can calculate for almost any league.

This study asks whether that simple ratio still earns its place now that expected goals (xG) is everywhere. It looks at what TSR captures, why analysts adopted it, what tends to happen when it is compared with xG-based measures and where it still adds something. Shot and match statistics for the leagues covered are available at https://rubiscore.com, so the calculation can be repeated for any team.

The Question

Before xG became common, football analysts needed a measure that described how well a team controlled matches without relying on goals, which are scarce and noisy. TSR filled that role. The question now is whether a count that treats a speculative long-range effort and an open goal from two yards as equal still tells us anything that xG does not.

There are two parts to the question:

• Description: does TSR describe how a team has played?
• Prediction: does TSR help forecast how a team will do next?

Where the Idea Came From

Shot-share metrics were borrowed from ice hockey, where measures such as Corsi and Fenwick count shot attempts for and against to estimate puck possession and territorial control. In the early 2010s, football analytics writers began applying the same logic, and TSR quickly became a standard way to rank teams by process rather than results.

Its appeal was practical. Shots are recorded in almost every competition, the formula is transparent and the result sits on an intuitive scale. A team above 0.50 is out-shooting its opponents; a team well above it is usually dominating territory.

How to Calculate It Properly

The formula is simple, but a few choices change the answer:

• Aggregate, then divide. Add up all shots for and against across the period first, then calculate the ratio. Averaging single-match ratios gives one-sided matches too much weight.
• Decide on blocked shots. Some analysts include all attempts, others exclude blocked shots to reduce the influence of defences that throw bodies in front of the ball. Either is defensible, but the choice must be consistent.
• Consider removing penalties. Penalties are rare and random in timing, and they distort small samples.
• Split by game state if possible. TSR while the score is level is a cleaner measure of balance than TSR across all minutes.

A season TSR built this way is easy to reproduce and to compare between teams in the same league.

What TSR Captures

TSR measures volume of attempts, and through volume it captures several things indirectly:

• Territorial control: teams that spend more time near the opposition box usually shoot more.
• Defensive suppression: a low count of shots conceded shows an ability to keep opponents away from goal.
• Repeatability: shot volume is driven by style and squad quality, both of which persist, so TSR tends to be fairly stable from one part of a season to the next.

That stability is the main reason TSR became useful. Goal totals swing with finishing and goalkeeping, while shot shares move more slowly and reflect the underlying balance of play.

What TSR Misses

The obvious weakness is that every shot counts the same. A team that takes many low-quality shots from distance will post a strong TSR without creating much danger, while a counter-attacking team that takes fewer, better chances will look worse than it is. Expected goals was developed precisely to weight each attempt by its likelihood of being scored.

TSR also ignores blocked-shot context, the difference between headers and shots with the foot, and the gap between an open-play chance and a set-piece scramble. Each of these is part of what an xG model tries to value.

What the Comparison Tends to Show

When analysts compare TSR with xG share, the share of total expected goals a team accounts for in its matches, a fairly consistent picture emerges:

• Both measures relate strongly to league position over a full season, far more strongly than single-match results.
• xG share usually predicts future results better than TSR, because it adds information about quality that shot counts ignore.
• TSR remains surprisingly competitive in short samples, because shot counts settle quickly, while xG values per shot take longer to stabilise.
• Shots on target ratio, which counts only attempts on target, adds a little quality information but introduces more noise, since whether a shot hits the target partly depends on finishing.

The honest summary is that xG-based measures are better, but not so much better that TSR becomes worthless. The gap is largest for teams with extreme shooting styles and smallest for teams whose shot quality is close to the league average.

The Confounders

Several factors distort TSR, and most of them also affect xG to some degree:

• Score effects. Teams that take an early lead often sit deeper and concede more shots, while trailing teams push forward and shoot more. A strong side that leads often can see its TSR understate its quality.
• Shot selection. Coaching philosophies differ on long-range shooting. Teams that discourage it will have lower shot volumes but better average quality.
• Schedule. A run of fixtures against weak opponents inflates TSR; a run against strong ones deflates it.
• League environment. Shot volumes differ between leagues and eras, so TSR is best compared within a single competition and season.
• Recording standards. What counts as a shot, especially a blocked one, can vary between data providers.

Game-state splits and opponent adjustments reduce these problems but rarely remove them.

Where TSR Still Earns Its Place

TSR remains useful in at least four situations.

Where xG is not available. Many lower divisions, youth competitions and some women's and non-European leagues have shot counts but no public xG. TSR offers a consistent process measure in those settings.

Historical analysis. Shot counts exist for seasons long before modern xG models, which makes TSR one of the few ways to compare team dominance across eras within a competition.

Early-season reading. After a handful of matches, TSR is often a steadier signal than goal difference and a useful cross-check on xG.

Disagreement as information. The most valuable use may be comparing TSR with xG share. When the two disagree, the gap describes a style:

• High TSR, lower xG share: a team that shoots often from poor positions, or keeps opponents to many low-quality attempts.
• Lower TSR, higher xG share: a selective or counter-attacking team that takes fewer but better shots, or concedes many speculative efforts.

On RubiScore, reading total shots, shots on target and xG side by side makes these style signatures easy to spot.

A Worked Example

Consider two hypothetical teams halfway through a season. Team A has a TSR of 0.60 and an xG share of 0.52. Team B has a TSR of 0.48 and an xG share of 0.55.

On TSR alone, Team A looks clearly stronger. The xG figures suggest something different: Team A out-shoots opponents but mostly from distance, while Team B takes fewer shots of much higher quality and restricts opponents to poor chances. If both teams keep the same styles, the xG view would expect Team B to collect points at a slightly higher rate in the second half of the season, though both are above average.

How TSR Fits With PDO

TSR is often paired with PDO, the sum of a team's shooting percentage and save percentage. The logic is simple: results depend on how many shots a team takes relative to its opponents and on how efficiently those shots are converted and stopped. TSR describes the first part and PDO the second. Because PDO tends to drift back towards average over time, a team with a strong TSR and a low PDO is a common candidate for improvement, while a weak TSR propped up by a high PDO is a common candidate for decline.

Verdict

TSR is no longer the best single process measure in football; xG share usually is. But TSR is cheap, transparent, quick to stabilise and available almost everywhere. Its best use today is as a companion to xG: a check on small samples, a tool for leagues and eras without xG and, when the two measures disagree, a way to describe how a team plays rather than only how well.