Horse running through field

The Next Little Thing

The ad agencies that want you to think every new gadget represents a “paradigm shift” have rendered that term even less meaningful (although possibly more valuable) than it was when Thomas Kuhn invented it in the early 1960s. To those who actually do science, as opposed to those who sell it, the name Karl Popper is much more important. He was the philosopher of science who showed how knowledge advances by accumulations of the incremental, piecemeal efforts of uncelebrated researchers asking small questions and getting small, but important, answers.

I’m reminded of Popper’s empiricism–and in an ironic way–when I read about the crazy economic policy coming out of both the White House and Paul Ryan’s House budget committee (because a new paradigm in economic policy has no use for the lessons of the past) and, for a different reason, when I’m asked a question about pedigree. Everywhere, people are looking for a paradigm, and it seems like the shiftier it is, the better it sells. The typical pedigree question is right out of the Thomas Kuhn playbook, but it’s still only the little Karl Popper moves that really matter.

My problem with the typical pedigree question is that it’s paradigmatic in scope, which means it has to be deconstructed before it can be answered. For example, how are the Gone West-line sires doing with mares in descent of Mr. Prospector? I might answer that one in the following way, starting with my usual, “Well, it depends.”

All of the Gone West-line sires are managing returns of Mr. Prospector differently from one another. But here’s a way to look at the big picture if that’s what you really want to do. Gone West is by Mr. Prospector, so the cross always involves inbreeding. Breeders are generally sold on that, so it gets lots of opportunity. It happens, though, that every single one of the proven Gone West-line sires has a lower frequency of superior runners from those that are inbred at 4×4 or closer than from those that are not. Insofar as inbreeding to Mr. Prospector figures in that, the picture doesn’t look good, but, again, it depends.

The best way to deal justly with a broad question like that is to give it enough rope to hang itself. It happens, incidentally, that almost all of the A.P. Indy-line sires, most of which have Mr. Prospector in their ancestries, have a higher frequency of superior runners from those that are inbred 4×4 or closer than from those that are not–just the opposite of the Gone West-line sires.

Interesting facts, but facts that have dubious practical import. You’re snake-bitten if you breed the mare to one of the two or three fashionable A.P. Indy-line sires that don’t cope very well with mares in descent of Mr. Prospector.

So, what’s my view of close inbreeding? Well, it depends. . . .

Dissident Ancestors

by Roger Lyons

The point of my last post was, in part, to suggest that certain dissident ancestors of broodmares pose special problems for stallion selection. They’re not carrying signs and throwing rocks, but the troublemakers assert themselves so forcefully that some seemingly well bred mares will have severely limited opportunity for a good match–or none at all.

I’ll just name three ancestors on my watch list: Graustark, Halo, and Nureyev. I’m not really sure about Nureyev. Let’s call him an ancestor of interest, but one can’t afford to be squeamish when rounding up suspected dissidents. You never know where a seemingly innocent association might lead. What these three have in common is that an unusually large proportion of stallions don’t like them very much. That’s enough to warrant indefinite detention on pedigree security grounds. If they’re innocent, let them prove it.

Never mind that in the right pedigree context those ancestors can have powerfully beneficial effects. Individual merit doesn’t count for much when a pedigree security interest is at stake.

According to eCompuSire, the online pedigree intelligence asset, six stakes winners worldwide are out of mares in descent of Graustark, Halo, and Nureyev within five generations. I’m afraid that’s not enough to establish their innocence. Even the most exculpatory evidence, such as a record of stakes production, readily turns against the perpetrators. Just watch.

The dam of Treasure (Anabaa ex Treasure Queen, by Kingmambo), winner of the Prix de la Vallee d’Auge in France, has Halo as the sire of her second dam, and she has Nureyev and Graustark by virtue of Kingmambo’s dam, Miesque.

The dam of Bottega (Mineshaft ex Sun Is Up, by Sunday Silence), winner of the Criterium du Languedoc in France, has Halo as the sire of Sunday Silence, and she has Nureyev and Graustark by virtue of her second dam, Miesque.

The dam of Sunday Sunrise (Lemon Drop Kid ex Sun Is Up [same dam as Bottega], by Sunday Silence), winner of the Veteran S. in New Mexico (must impose a travel ban), has Halo as the sire of Sunday Silence and has Nureyev and Graustark by virtue of her second dam, Miesque. Furthermore, Sunday Sunrise is by a son of Kingmambo, out of Miesque.

The dam of Indigo Cat (Storm Cat ex Bluemambo, by Kingmambo), winner of the Hampton Court S. in the UK, has Halo as the sire of her second dam and Nureyev and Graustark by virtue of Kingmambo’s dam, Miesque.

The dam of Air Zipangu (El Condor Pasa ex Air Passion, by Halo), winner of the Stayer S. in Japan, has Halo as her sire, Nureyev as her broodmare sire, and Graustark as the sire line of her second dam. Not incidentally, Air Zipangu is by a son of Kingmambo, whose dam is Miesque.

The dam of Link Man (Torreador ex Western Smoke, by Among Men), winner of the Gold Medallion (G1) and other group stakes in South Africa, is the only one among the dams of these six SWs whose contribution of Halo, Graustark, and Nureyev is not controlled in some fashion by Miesque–and in one case (Sunday Sunrise) inbreeding to Miesque. If you disqualify the five because they’re all mediated by one freakishly good broodmare, then what you have left is only one SW worldwide–a G1 winner, to be sure–to testify in defense of mares descending from all three of these ancestors.

By the way, eCompuSire is what you need when you haul an ancestor in for questioning. It’s the waterboard of enhanced pedigree interrogation techniques. See subscription details at the eNicks website–Products tab. Did I not mention that I have a personal stake in that product? Wouldn’t want to breach any ethical constraints.

The Politics of Pedigree

by Roger Lyons

The recent series of posts relating to the table of ancestor preferences suggests that different ancestors have different roles and relations with respect to the stallion population. Maybe it’s time for a more schematic rendering of that variety.

Franco Varola famously placed the variety of touchstone sires (chefs de race) on a typological spectrum, using the analogy of the left and right wings (the dominant liberal and conservative ideological commitments) of the English-style parliament. His dosage analysis focused on the functional relations of the different types. However, his typology didn’t address the question of compatibility between individual ancestors. Could Varola’s analogy lead to a way of characterizing individual ancestors on that basis? Let’s give it a whirl.

Two structural changes are required.

First, our shift of emphasis means that we’re no longer subject to parliamentary rules. Out here in the street there’s a broader ideological spectrum, including radicals and anarchists, and, as the autocrat of your thoroughbred breeding operation, you ignore them at your peril. As we’ll see, it’s important that we sequence the four major ideological commitments in this way: radical, liberal, conservative, anarchist.

Second, the linear structure used by Varola won’t do. We need a structure that reflects the way in which the four ideological commitments relate to one another. So, the solution is to place them on a clock, with the radical at twelve o’clock, the liberal at three o’clock, the conservative at six o’clock, and the anarchist at nine o’clock. The question of compatibility is addressed by placement of ancestors on the clock, relative to one another and respective of their “ideological” commitments.

The radical, at twelve o’clock, is the easiest type, thanks to Varola, who familiarized us with “the Phalaris revolution.” Clearly, the radical is an incipient figure in Varola’s analysis and a pivotal one. As change-makers go, Native Dancer would be a prime representative of the radical commitment, but not the extremist that Phalaris was. The advantage of our compatibility clock is that we can place Native Dancer at around one o’clock if that seems right–the radical commitment shading into the liberal. Appropriately, the radical is diametrically opposed to the conservative, at six o’clock.

Accordingly, the liberal is opposed to the anarchist. Why? One might associate Mr. Prospector and Northern Dancer with the liberal commitment, whose distinction is that it defines the mainstream of the population at any given time and in any given place. The liberal knows how to conform and expects the same, gets along best with other liberals. For that reason, the liberal has a tendency to be complacent about the company it seeks and needs to be revitalized continually by radical and conservative associations. The liberal has no use for anarchy and will give it no quarter.

The conservative can get along with the liberal just fine if often on contrasting terms and might tea party on the anarchic side, but the conservative can’t abide a radical. I’m inclined to think Varola’s pure types, such as Bold Ruler and Double Jay on the left and Vaguely Noble and Alleged on the right, could all be considered conservative in their way. Unlike the radical, who wants to start something new, the conservative wants to preserve something long established, regardless of Varolan type. Remember, our clock doesn’t stop with what an ancestor contributes. Its hours mark where an ancestor is likely to fit successfully in the population.

The anarchist, at nine o’clock, opposes the liberal mainstream any way it must, just wants its distinctiveness to be respected. The survival of the anarchist is always under threat. Largely because anarchists don’t play well together, their best hope lies in alliances with either the radical or the conservative elements, at twelve o’clock and six o’clock, respectively. Graustark is almost certainly an anarchist, and I suspect Halo of leaning libertarian.

There, I think that winds up the compatibility clock enough to get its wheels turning.

High Upside, High Downside Risk

by Roger Lyons

What do Lyphard, Fappiano, and Nureyev have in common? As ancestors of dams, each of these sires passes on respective traits that rub a lot of stallions the wrong way, but that contribute in favorable ways to the offspring of a lot of other stallions. We know this because of the high approval and high disapproval rates shown in the table of ancestor preferences for these three ancestors. In that respect, they represent a distinctive category of ancestors–a category for which stallion selection is faced with very high upside and very high downside risk.

Based on that table, there’s a 29% chance that a stallion whose name you pick out of a hat will have a higher-than-average strike rate with mares in descent of Lyphard; but there’s a 27% chance that a stallion chosen at random will have a lower-than-average strike rate with those mares.

Just a cursory look at the table suggests what a distinctive profile that is for an ancestor. Fappiano is similar, with a 27.5% approval rate and a 31% disapproval rate, and Nureyev has a 28% approval rate and a 26% disapproval rate. You can find other stallions with a similar profile. It’s not particularly rare, but it’s rarely this extreme.

In order for the stallion population to respond the way it does, the distinctive traits these three ancestors contribute–along with other ancestors of this statistical profile–would have to pass through to foals on a very consistent basis. Furthermore, in order to complement the contributions of a lot of stallions, but clash with the contributions of many others, those traits would have to be distinctive in character, but without being particularly idiosyncratic, as in the case of Graustark.

That much can be inferred from the numbers. As a numbers cruncher, I have no idea what traits are at issue in the contributions of Lyphard, Fappiano, and Nureyev and wouldn’t have the eye to discern them if they were pointed out to me. Still, the numbers say it’s something to consider.

Native Dancer at the Vanishing Point

by Roger Lyons

Native Dancer has reached a genetic vanishing point in the sense that his influence is so pervasive that his contribution can no longer be contrasted in any significant way with the mainstream of the population. This is evident from the table accompanying my last post, which shows that in no instance do the strike rates of US stallions with mares in some descent of Native Dancer exceed their overall records. That is, no “contemporary” US stallion has a statistically measurable preference for mares in descent of Native Dancer. Keep in mind, though, that it’s because the broodmare population is so largely defined by Native Dancer. It’s like the air they breathe.

Moreover, only two stallions–Diesis and Dixieland Band–(that the table rather loosely defines as contemporary) fall below their normal strike rates when crossed with mares in some descent of Native Dancer, and they are both deceased. They were bred quite differently from one another with respect to Native Dancer–Diesis, by Sharpen Up, by Atan, by Native Dancer, and Dixieland Band, by Northern Dancer, out of Natalma, by Native Dancer.

Reasons can no doubt be given for why they had trouble with mares in descent of Native Dancer, but the fact that they did so has always had more practical import than the reasons why. More interesting here is that their aversion to mares in descent of Native Dancer and their shared birth year of 1980 renders their stud records especially useful as measures of the proliferation of Native Dancer’s influence.

Consider what it could mean that both of these stallions had significantly higher strike rates with mares born prior to 1985 than with mares born after that year.

Dixieland Band got superior runners from 50 of 435 individual mares born prior to 1985, for 11.5%; but, of the 644 mares born after 1984, 51 produced superior runners by him, for only 7.9%. Furthermore, only 23 of the 396 mares born after 1989 produced superior runners by him, for only 5.8%. As Native Dancer’s influence spread throughout the later population cohorts, Dixieland Band’s strike rate with those mares declined. Of the 219 mares with Northern Dancer in their ancestries, only 10 produced a superior runner, and he was barely in range of his average with mares in descent of Mr. Prospector.

Diesis’ stud record reflects the same broad pattern. While 48 of 343 mares born prior to 1985 produced superior runners by him, or 14%, only 28 of 453 (6.2%) born after 1984 did so. Of the 277 mares born after 1989, only 10 (3.6%) produced superior runners by him. Diesis got a superior runner from only one of 64 mares with Mr. Prospector in their ancestries, and he was barely in range of his average with mares in descent of Northern Dancer.

With each successive broodmare population cohort, in which crosses of Northern Dancer and Mr. Prospector assured a rapid accumulation of Native Dancer influence, the strike rates of Dixieland Band and Diesis declined. It seems warrantable to conclude that Native Dancer’s proliferation in the broodmare population had a more profound effect on their stud records than could possibly be attributed to a decline in viability due to increasing age.

Measuring Ancestor Appeal

by Roger Lyons

Recent posts about Dixieland Band and Graustark preceded their broader context, which arrives as a table, titled “Ancestor Preferences of Major US Sires.” The table covers 164 ancestors and reflects, among other things, how well or poorly individual ancestors of broodmares are playing to the contemporary stallion population. Links to three different versions of the table are provided below.

The table is very simple once you catch on to the concept. To over-simplify, there’s about a 31% chance that a stallion whose name you pick out of a hat is going to react well with Storm Cat in the ancestry of a mare and only about a 9% chance of an unfavorable reaction. That leaves about 60% that do okay with Storm Cat, but not many breeders dream of results that are okay.

If your mare is by Storm Cat, the odds are not bad that you’ll draw a suitable stallion by chance, at least in that respect. However, if her dam happens to be by Halo, with which only 13% of the stallion population react favorably against the 22% that react unfavorably, it’s more complicated. Thus do questions of compatibility arise from the layers of a mare’s ancestry.

Clearly, the breeding of some mares renders them far more flexible as to the selection of a mate than other mares. Ideally, you would want a stallion to have high stike rates with a mare’s entire ancestry. Some mares may have a range of such options, but for other mares there’s no such stallion.

Anyway, for each ancestor, I surveyed 71 proven sires to determine, first, how many of them sired foals out of at least 10 mares in some descent of the subject ancestor. Second, the qualifying sires were divided into two groups: 1) those that, from mares in some descent of the subject ancestor, had strike rates significantly above their overall records and 2) those that had significantly lower strike rates with those mares.

Here’s the legend for the resulting table:

Ancestor–the subject ancestor as represented by broodmares.
Sires–the number of stallions (from among the 71) that sired 10 or more foals with mares representing the ancestor.
Approve–the number from the “Sires” group with significantly higher-than-average strike rates.
Approve%–the percentage of “Sires” with significantly higher-than-average strike rates.
Disapprove–the number from the “Sires” group with significantly lower-than-average strike rates.
Disapprove%–the percentage of “Sires” with significantly lower-than-average strike rates.

There are three versions of the table, one listing 164 ancestors by “Approve%” rank to show them in order of the frequency of high strike rates by the stallion sample; another listing them by “Disapprove%” rank to reflect the downside risk; and then an alphabetical listing so you can have fun looking them up individually.

Bear in mind, the survey includes only US sires. Danzig’s approval rating of only 14% and disapproval rating of 23% would be vastly different based on a survey of European or Australasian sires. Nevertheless, Danzig is a problem for a lot of US stallions.

I’ll comment further on this table in future posts, beginning with the reason why Native Dancer’s approval rating is the lowest in the list.

Graustark’s Hard Edge

by Roger Lyons

A rough measure of an ancestor’s relevance in the contemporary broodmare population can be taken from how well proven stallions, as a group, are doing with mares that descend from that ancestor. I mentioned this approach a couple of posts back in regard to the influence of Dixieland Band, and I’m driven back to it by the intriguing case of Warrior’s Reward (Medaglia D’Oro-For All You Do, by Seeking the Gold), winner of the 2010 Carter H. (G1) and now standing at Spendthrift Farm in Kentucky for $15,000, first foals eagerly anticipated in 2012.

Note that Warrior’s Reward’s third dam is by Graustark (1963 Ribot-Flower Bowl, by Alibhai), an ancestor whose influence was quite extreme a decade ago. We know this because back then an unusually large proportion of the stallion population had serious trouble with mares in some descent of Graustark. Since then, his effects have been tempered by generational distance. That’s evident in that only around 18% have serious problems now, as against the 16% whose records actually benefit from his influence.

He’s already gravitating toward the vanishing point that awaits him at the margin of effective pedigree space. Nevertheless, Graustark still has profound effects in many cases, and Warrior’s Reward is probably one of them. His dam is one of five with a superior runner by Medaglia D’oro from the 25 that produced a foal by him and had Graustark in their ancestries. Graustark is hard to process, but Medaglia D’Oro has the knack.

What’s interesting is that, although about 30% of Medaglia D’Oro’s SWs win stakes at two, three of the five whose dams have Graustark in their ancestries, including Warrior’s Reward, didn’t win a stakes until age four. That could have something to do with the size and bone mass associated with Graustark.

Despite the stallion population’s problem with processing Graustark, his full brother, His Majesty (1968), continues to enjoy the approval of about 30% of the stallion population, one of the highest among major ancestors, with only about 14% disapproval. That’s the reverse of where Graustark was a decade ago, and in that time His Majesty’s effect has shown no sign of abating, maybe because His Majesty’s stud career started six years after that of Graustark.