You win a Test by bowling sides out. So I plotted wickets taken by teams over last 20 years. Inspired by this post from Asad Salm, sharing Atif Mian’s charts of long-run Test performance. The idea, the trailing-average framing and the visual style are theirs. The change below is mine.

Mian’s charts score every Test as a result: win = 1, draw = 0, loss = −1, averaged over a trailing window. It’s a clean measure of outcomes. It produces some striking national arcs. Example: Pakistan’s slow slide and India’s climb.

But the result is the consequence. The real mechanism underneath it is wickets. You cannot win a Test without taking twenty wickets. So I swapped the outcome for the driver: opposition wickets taken per match, out of a maximum of twenty, on a trailing five-year average.

Four things fall out of it

New Zealand is the story. From 14.8 wickets a Test in 2014 to 18.1 in 2026. The largest improvement of any side, and the clearest example of a bowling attack turning a team from also-rans into a side that closes matches out. They have successfully churned out world class bowlers like Southee, Wagner, Boult from mid 2010s to now Matt Henry, Will O’Rourke and Jacob Duffy in mid 2020s.

Australia is the only team to never come below median wickets per match in last 20 years. They start the period at 18.3, the best in the world (McGrath, Gillespie, Warne, Lee) and end at 17.6 (Starc, Cummins, Hazelwood, Lyon). Australia didn’t get worse so much as stop being alone. However, their current bowling cartel is getting old (average age ~35 years) and they don’t have any one who can take over from them in next 4-5 years. We will finally see a dip in next 3-4 years.

Bangladesh built a bowling attack. 20 years ago they started at 9.6 wickets a Test and now they are at 15.3. They have built a pace battery that can threaten any team in any conditions. All their bowlers are young and have at least 7-8 years in their prime. Expect to see Bangladesh to continue to improve and be a major test team in next 4-5 years.

India: the curve bends in 2013 from the Kohli era. They were trailing with 15.3 in 2012 being fifth among the established sides, behind Australia, South Africa, Pakistan and England. Nine years later it read 18.2, the best in the world.
I’ve written at length about the quartet (Ishant, Bhuvi, Shami and Umesh) and how those four taught India to be a fast-bowling country before Bumrah bowled a Test ball. The recent drop off is due to Ashin-Jadeja fading and Bumrah being injured which has hurt the wicket taking abilities. It’s time for the new crop of bowlers to step it up.

Method. Ball-by-ball data from Cricsheet, which covers Tests from December 2001 onward — so these charts start in 2005, not 1961. Wickets are credited to the bowling side; retirements are excluded. Matches are restricted to the ten established Test nations playing each other. Afghanistan’s matches are withheld by Cricsheet and Ireland has too few Tests for a five-year window, so both are absent. Every panel shares the same axes so they can be read against each other.