Fifteen years ago John Randall and I embarked on the compilation of a book which was published in the following December as A Century of Champions.

It was an ambitious and fascinating project. We sought to identify the best horses, worldwide, and to rate them, and, among many other things, we listed the names of those individual horses and humans who, we felt, had been most influential during the 20th century. Of course, there could be nothing definitive about our conclusions, which simply represented the considered opinions of two blokes who presumed to believe that they were entitled to make such judgements.

One of the tasks we set ourselves was to determine who had been the outstanding broodmare of the century. Having completed our ratings of racehorses, we had already decided the identity of the best female runner, awarding that accolade to the 1904 Fillies’ Triple Crown heroine Pretty Polly, whom we made 1lb superior to Sun Chariot and 2lb superior to Sceptre and Pebbles among British- and Irish-trained performers. Naming the best broodmare of the century proved to be an altogether more demanding task, as there were so many worthy candidates with similar accomplishments.

How on earth could we separate them and settle upon a plausible number one? In the end, I suppose we chickened out, and by doing so laid ourselves open to ridicule. We awarded the title to a mare who died at the age of 30, and whose obituary in the esteemed Bloodstock Breeders’ Review noted that she ‘did not greatly distinguish herself as a producer of winners.’

Naming the best broodmare of the century proved to be an altogether more demanding task, as there were so many worthy candidates with similar accomplishments

So by what criterion could we justify promoting the cause of the aforementioned Pretty Polly, five of whose ten foals failed to win, as the outstanding broodmare of the century? We decided that she deserved the second accolade because, whereas many others had been more conspicuous as dams of important winners, none of them had founded a family that had attained such enduring success. There was no female line, descending from a mare foaled in the 20th century, which had flourished to anything like the same degree. Pretty Polly could be separated from – and placed above – all those other worthy mares by that criterion.

Whether or not that was the most sensible criterion we might have used, I long ago decided that one day, when I had nothing better to do, I would take a closer look at Pretty Polly’s descendants in tail-female and monitor its progress. There had always seemed to be something significant happening in every generation; was that still the case, I wondered? Most families have peaks and troughs, so why wouldn’t that be the case with Pretty Polly’s? And it had to be acknowledged that Pretty Polly herself could hardly be considered influential, a realistic source of merit, in the pedigrees of descendants who came along many decades and many generations after her.

As it turned out, I needed more than a day with nothing better to do to complete my research, and I don’t claim that my study was exhaustive. I needed data that were readily accessible, and that meant I would confine my investigations to winners of European Pattern races, whose five-generation pedigrees I have recorded since the programme was introduced in 1971. There has obviously been a lot going on elsewhere in the world, but to my mind that didn’t make a solely European study invalid; it provided a legitimate representative area of study.

Not a quick job
After the work of several days when I had nothing better to do I had a new set of data, still to be put into some sort of order to enable me to make any kind of sense of it. This was going to be a rather longer job than I had envisaged.

Pretty Polly’s brood included only four daughters. They were Molly Desmond (1914, by Desmond), who won the Cheveley Park Stakes; Dutch Mary (1915, by William the Third), who ran five times without winning; Polly Flinders (1918, by Polymelus), who won the National Breeders’ Produce Stakes; and Baby Polly (1926, by Spearmint), who ran five times without winning. They had all established branches that had provided an abundance of success before the first year included in my study; 1971 was 70 years after Pretty Polly’s birth and 40 years after her death. The great mare was arguably already an inconsequential figure in the pedigrees of horses running then.

The tail-female descendants of Pretty Polly continued to show up prominently throughout the 43 European Pattern seasons under review

Before 1971 the Molly Desmond branch had been responsible for a host of celebrities, including Guersant, Premonition, Nearctic, St Paddy and Luthier; the Dutch Mary branch had peaked early with Donatello and had delivered a Derby hero in Psidium; the Polly Flinders branch had provided such as Supreme Court, Court Harwell, Only For Life and Huntercombe; the Baby Polly branch, represented rather less numerously than the others, had given Colorado Kid (the mare’s second foal and first son) and, two generations on, Vaguely Noble’s sire, Vienna.

So what did I learn over something like three weeks with nothing better to do? The bare statistics provided the most obvious fact: the tail-female descendants of Pretty Polly continued to show up prominently throughout the 43 European Pattern seasons under review. There have been 239 individual winners of 435 races, comprising 112 successes in Group 1, 89 in Group 2, and 234 in Group 3. And the rate of success was not slackening; in 2013 there were 11 individual winners of 16 Pattern events, including Group 1 winners Style Vendome and Vorda, while Dank, a winner at Group 2 and Group 3 level here, surpassed those achievements with success at the Breeders’ Cup.

Brigadier the best
The only winner as close as five generations distant from Pretty Polly was her best-ever descendant, Brigadier Gerard, whose exploits in 1971 and the following year brought him 13 Pattern victories, including six in both Group 1 and Group 2. (If the Pattern had come along a year earlier, his Middle Park victory would have given him another top-level score.) Brigadier Gerard represented the Molly Desmond branch, easily the most successful of the four, responsible for 130 individual winners.

The others from the branch with Group 1 wins to their credit included Alydaress, Artaius, Artan, Borderlescott, Cape Cross, Capricciosa, Daring Display, Desert Lord, Desirable, Diktat, Emmson, Fantastic Light, Inkerman, Kutub, Legal Case, Love Divine, Malevic, Markab, Masarika, Music Show, Northern Taste, Opale, Park Appeal, Pitasia, Rosanara, Russian Rhythm, Shadayid, Shake The Yoke, Sixties Icon, Style Vendome, Sulk, Swain, To-Agori-Mou, Val d’Erica, Vintage Crop, Vintage Tipple, Vorda and Workforce. Phew.

Sixteen individual winners represented the Dutch Mary branch, with just four at the highest level. They were Deauville, Polaris Flight, Stone and Stouci, none of whom could reasonably be considered in the same league as once-defeated champion Donatello, foaled as long ago as 1934 and out of a grand-daughter of Dutch Mary. There has clearly been some decline in this branch.

In contrast, the Polly Flinders branch has been holding up pretty well, delivering 91 individual winners. There were 21 successful in Group 1 company, namely Belle et Celebre, Be My Chief, Bianca Nera, Caerwent, Double Form, Eva Luna (the Irish-bred one, not the daughter of Alleged who traces to Molly Desmond), Fedora Grey, Loch Garman, Marling, Marwell, Paean, Revoque, Sarab, Shavian, Sholokhov, Sigy, Simply Perfect, Soldier Of Fortune, Sunspangled, Tenby and Unite.

For a family branch to thrive there must be daughters who produce daughters who produce daughters ad infinitum, preferably in significant numbers. Of Baby Polly’s four, only two amounted to much, the first being Rusk (1935, by Manna), who became Vienna’s grand-dam. The other was Precious Polly (1937, by Hyperion), and she features in the background of the only two European Pattern winners descending from Baby Polly, as fourth dam of Arc hero Carroll House and fifth dam of the estimable international Group 1 scorer Phoenix Reach. Both, of course, are males, perhaps an ominous sign for this branch.

If you fancy doing some research of your own into Pretty Polly’s descendants, you’re welcome. I’ve now found something better to do.