What the Sires' Premiership Actually Measures
Ninety-seven point nine per cent.
That's the share of St Jean's Australian progeny earnings that came from a single horse last season. He had eleven runners here. One of them had a career year, and it carried him from 304th on the sires' earnings table to 20th.
By the premiership, St Jean is now a top-20 Australian sire. His connections did everything right. One very good horse had one very good season, and every dollar was fairly won. The interesting question isn't about the horse. It's about what a ranking should do with that fact, because as published, the table can't distinguish an eleven-runner book with one star from a two-hundred-runner production line. It was never designed to.
I rebuilt the sires' premiership from scratch for the past two seasons (every race, every runner, every dollar actually paid) and compared them. 363 stallions had runners in both years, and the movement in and out of the top 30 is where the table shows you what it really measures.
The movers
Seven sires entered the top 30 this season. Five of them arrived on a concentration jump. Not a broad crop improving, but one horse having one big year. St Jean is one. Shamexpress is another: 221st to 28th on fourteen runners, with 96.5% of his total earned by Ka Ying Rising, a sprinter doing his earning in Hong Kong. Seven sires exited, and four of them left the same way they'd arrived. The single horse carrying them stopped, and the ranking moved on.
The contrast is worth looking at. The Autumn Sun rose from 32nd to 3rd on the other model entirely: his runner count grew from 146 to 183, his strike rate held, and his best horse accounts for 44% of his total. High, but attached to a genuinely deep book. Two sires, two very different kinds of season, rendered identically by the same column of figures.
Last season's data showed the champion-sire-by-winners table correlates with runner count at 0.98. It substantially ranks book size. The movement data shows the earnings version has the opposite sensitivity: a single exceptional horse can outweigh a hundred ordinary ones. Neither number, alone, answers the question most of the table's readers are actually asking: what does the typical foal by this horse do?
The Fastnet Rock case
Fastnet Rock illustrates why that matters. A year ago Via Sistina was 62.6% of his season. This year his runner count fell 31% and his total earnings fell 23%. A fading line, by the table. But his earnings per runner rose 12.4%, to $155,028, and Via Sistina's share eased to 52.4. His book shrank faster than his money did. The headline says decline; the per-head figure says his remaining stock performed better. Both are true. Only one gets published.
The structural note
There's also something the table never carries. The Everest's $20 million, partly funded by twelve slot-holders at $700,000 a year, moved $12.4 million into three horses in one October afternoon. Giga Kick's $2.9 million from that single race was 76% of his sire Scissor Kick's season. The earnings table treats that dollar identically to one ground out across forty midweek maidens, which is arithmetically correct and analytically silent.
None of this makes the premiership wrong. It adds up. I've verified it does, to the cent. It's simply incomplete for the decisions people use it to inform. And the fix is modest: publish the median foal's earnings and the top horse's share alongside the total. Three columns instead of one, and the table starts answering the question its readers are asking of it.
Next in this series: putting stallion service fees beside what the median foal actually earns. The gap between those two numbers turns out to be the most practical finding in the dataset.
Notes
Both seasons rebuilt race-by-race from complete public results (2024-25: 18,939 races; 2025-26: 18,752), prizemoney as paid rather than advertised. 363 sires with runners in both seasons, identical methodology, top-30 movement computed from per-season rankings. Season aggregates reconcile with Racing Australia's published figures on the checks available (race counts, field sizes). Concentration is the top horse's share of a sire's total progeny earnings for the season. The Everest is included in all totals because the money was genuinely paid; its funding structure is covered in the note on the first piece in this series. All analysis is my own, built on a database I maintain of every Australian race result; queries and figures available to anyone who wants to check them.