.png){kind=spacer}
Between 2018 and 2021, Marnus Labuschagne averaged 60.4 in Test cricket. Steve Smith, across the same window, averaged 52.91.
Same team. Same era. Same dressing room. Nearly identical sample sizes: 35 innings to 33. Labuschagne ahead by more than seven runs an innings.
And yet ask any bowling unit who they’d rather see at the crease, and the answer hasn’t changed since Labuschagne debuted. Smith is the wicket. Labuschagne is a really good player.
The cricket world’s intuition isn’t wrong. But the averages, on the face of it, are saying something else. Somewhere between “Labuschagne averages more” and “Smith is the better player” sits everything an average can’t tell you.
This is a piece about that gap.
The raw data says Labuschagne was the better player in this period. Dig in, and the picture changes.
Labuschagne’s 60.4 leans hard on twelve innings against New Zealand and Pakistan at home, where he averaged above 80. Strip those out and he averages around 52 across the four years, almost identical to Smith’s headline number.
Smith’s 52.91 hides the opposite distortion. In the 2019 away Ashes, on English pitches that have broken most touring batters for a generation, he scored 774 runs in 7 innings at 110.57. Australia’s next-best in that series was Labuschagne, with less than half the runs at less than half the average. The rest of the Australia top six averaged under 35.
Two players, almost the same average over the same window. One number was inflated by easier opposition at home. The other was dragged down by everything that wasn’t a generational series in England. The same headline number can describe two very different players.
That exercise revealed something real about two top batters. But the direction of the research came from a hunch. We already suspected NZ and Pakistan at home would be softer for Labuschagne. We already knew Smith had a generational series in 2019. We told the data what to confess.
It worked because Smith and Labuschagne have a lot in common. Same team, same era, same position. The dimensions that needed unpacking were obvious.
Try comparing two players from different teams, in different positions, across different eras. The factors multiply. The hunches stop working.
To compare any two players fairly, you’d need to know how hard their runs were to score. And plenty of things make a run hard to score: opposition strength, conditions, batting position, era. These are all things we can account for, and cricket has tried. The ICC has its ratings. ESPN once ran the MVPI. Various adjusted-average projects have come and gone.
None has become the default. The default is still comparing the raw averages of two players with very different careers.
The strip-out moves we did above shouldn’t be a one-off exercise for a Smith-Labuschagne argument. They should be baked into how anyone reads a Test record. An adjusted number wouldn’t replace the raw average. It would sit next to it, and stop the misreadings before they start. Right now, no mainstream tool does this. There should be one.
Fine Margins is that tool. It takes every Test innings from 2001 onward, weights it by opposition strength, conditions, and era, and produces an adjusted view of a player’s record alongside the raw one. You can compare Labuschagne to Smith without manually stripping out the soft windows. You can compare batters from different teams and different eras without losing your mind in caveats.
It launches August 12. Follow along between now and then.