Does Winning the Toss Win Tests? Research

Across 2,537 Tests, the side that won the toss won 932 (36.7%), lost 815 (32.1%) and drew 790 (31.1%). Put the draws aside and look only at the 1,747 Tests somebody won: the toss winner took 53.3% of them.

That is an edge, but a modest one. If the coin made no difference at all, the figure would sit at 50%. Winning the toss moves a Test result by roughly three points in a hundred, which over a five-Test series is worth, on average, a fraction of one match. It is enough that no captain should be indifferent to the call, and nowhere near enough to explain why a team wins a series.

A coin toss is random, which makes this one of the cleanest natural experiments in sport: the teams that win it are, as a group, no better than the teams that lose it. Any difference in results comes from the toss itself, not from who tends to win it.

The remaining question is whether 53.3% could be noise. Over 1,747 decisive Tests, a plausible range for the true figure is 51% to 55.7%, and a split this uneven would turn up by pure chance about once in 196 tries. The all-time edge is real.

The modern picture is less clear. Since 2000 toss winners have taken 52.9% of 836 decisive Tests, a range of 49.5% to 56.3%. That still leans the same way, but on its own it cannot be told apart from a fair coin. Whatever the toss is worth today, it is worth less than the all-time figure suggests.

Playing at home is worth far more. Across the Tests with a home side (2,481 of them), the hosts won 42.2% and the visitors 26.6%: a gap of about 16 points, several times the size of the toss effect.

The two stack. A home side that won the toss went on to win 44.5% of those Tests; one that lost it won 39.6%. For visitors the same call moved their win rate from 25.3% to 28.0%. The toss helps the home side a little more than the away side, which fits the idea that a home captain knows best what a pitch will do on day four and so gets more out of the choice. But the order of magnitude is clear: the toss is a nudge, home is a shove.

The edge has never been steady. It peaked in the 1930s, when toss winners took 60.4% of decisive Tests, and was weakest in the 2020s at 47.4%. Early Test cricket was played on uncovered pitches that rain could turn from a batting paradise into something close to unplayable overnight, so batting first on the best of the surface was a large advantage. Covered pitches, better preparation and five-day scheduling have steadily taken the extremes out of the surface, and the toss has mattered less as a result.

The draw complicates the chart. In the 1960s, 47.8% of Tests were drawn, so a toss edge had fewer finishes to show itself in. By the 2010s draws were down to 19.4%, and toss winners won 57.3% of the decided ones, the highest since the 1960s. The dip since 2020 rests on only 160 Tests, too few to call a trend.

For most of Test history, winning the toss meant batting. Before 1960, about one captain in ten chose to bowl. The 1970s brought that to 22.7%, and in the 1980s, with four-pronged pace attacks at their height, it jumped to 41%. It has settled around a third ever since. Across all 2,537 Tests, 1,841 toss winners batted and 696 bowled.

Did the bowlers get it right? Before 1960, the few who chose to bowl did badly: they won 12 and lost 19 of their decisive Tests. In 1980 to 1999 the choice paid off, with bowl-first toss winners taking 55.5% of decisive Tests against 48.9% for those who batted. Since 2000 the two are almost level, 53.4% for batting first and 51.8% for bowling, which suggests captains have become good at reading which pitches reward which choice.

The host country changes the story more than the era does. In South Africa, the home side won 52.4% of Tests after winning the toss and 41.3% after losing it, the biggest gap among the major hosts. In Australia the same comparison is 59.8% against 56.3%: Australian sides at home win at a high rate whoever calls correctly.

New Zealand is the outlier. It is the only major host where most captains choose to bowl (only 42.2% batted), and it is the one place where the toss is worth nothing measurable: toss winners took 49.3% of decisive Tests, and New Zealand won 32.2% at home after winning the toss against 32.5% after losing it. Green, seaming pitches give the bowling side its best conditions on the first morning, and whatever that is worth appears to be cancelled out by batting last on a surface that holds together. India sits at the other extreme of the choice: toss winners there batted 92.4% of the time, the highest of any host, because a turning pitch on days four and five is the last thing a captain wants to bat on.

Small samples need care. Bangladesh and Zimbabwe have hosted far fewer Tests than the others, and their home records are low enough that a handful of results moves the percentages a long way.

The heaviest defeats for a side that won the toss and chose to bowl have one thing in common: they were mismatches. In most of them a weaker side bowled first hoping for early help from the pitch, did not get it, and batted twice without reaching the first-innings total.

That pattern is the most useful lesson in the list. Bowling first is a bet on conditions, and it is a bet that a strong attack can cash more often than a weak one. For a side that is already the underdog, handing the stronger team first use of a good pitch can turn a likely defeat into a heavy one.

Three things come out of the numbers. First, the toss is worth having but not worth fearing: a side that loses it still wins close to half of the Tests that are decided. Second, the choice matters more than the call, and the right choice depends on the country: batting is still the default almost everywhere, while in New Zealand bowling first is the norm and costs nothing. Third, home advantage is the far bigger force, which is why touring sides lose so often even when the coin falls their way.

You face the same decision in The Follow-On, CricArcade's Test game, where you call the toss and choose with the pitch report in front of you. Our guide to how a Test pitch plays covers what changes over five days, and the captaincy guide covers the calls that follow.

• The archive covers men's Tests to March 2024; Tests played since then are not included.
• "Home" is decided by the host country's name, so a team's home Tests at a neutral venue, such as Pakistan in the United Arab Emirates, count as neutral rather than home.
• The toss is random, but the choice after it is not: captains bowl first when they expect the pitch to help, so the bat-or-bowl comparison shows how well captains chose, not what either choice is worth on an average pitch.
• Draws hide part of the effect. A toss edge that turns a likely defeat into a draw does not show up in the decisive share.

Figures computed from CricArcade's match archive with our own analysis scripts. Drafted with AI assistance, checked against the scorecards and edited by Vishal Jaiswal. Found a number that looks wrong? Tell us and we will check it against the scorecard.

• CricArcade Test match archive: 2,537 Tests, 1877 to March 2024
• MCC Laws of Cricket, Law 13: Innings (the toss)
• Figures computed with scripts/research/does-winning-the-toss-win-tests.ts