Purveyors of Pedigree Bullshit

I take great pleasure reading your articles; I appreciate the research and the insight that you put into them. Many of your recent posts have elaborated on the subject of inbreeding and I thought this article might interest you http://fmitchell07.wordpress.com/2012/12/25/have-a-very-inbred-christmas/ It would be interesting to hear your thoughts on it!


I received the comment above a few days ago. I read the inbreeding treatise at Bloodstock in the Bluegrass from the link above and responded thusly:


Thank you. I will comment briefly.

“Unfortunately, most of the inbreeding that actually takes place today is either accidental (fortuitous or inadvertent) or planned by those that do not have access to top-class stallions and mares, and they inbreed using less than first-class animals, with less than first-class results. . . . ”

I agree 100% with this statement. It probably explains why if you examine ALL inbreeding (or ALL inbreeding to a certain sire or mare), the results are rarely more than a little above average (and frequently well BELOW average).

“For today, inbreeding to Mr. Prospector through Woodman, Conquistador Cielo, Forty Niner, and Majestic Light would provide an excellent possibility to concentrate speed, acceleration, and class while minimizing the problems of unsoundness and front leg confirmation defects.”

Inbreeding to Mr. Prospector does seem to work better than to most other sires. Majestic Light was by Majestic Prince out of a Ribot mare and thus had no Mr. Prospector at all. This is a serious factual mistake. Serious factual mistakes such as this tend to undermine the overall validity of a treatise such as this. If you can’t get the FACTS right, your opinions lose all credibility. I am far from overwhelmingly convinced by this treatise.

I would agree that if you inbreed ONLY to DOMINANT individuals and exercise selectivity in doing so, you will probably get better results. I simply question how many DOMINANT individuals there really are and WHO they really are. Take Northern Dancer, considered by almost everyone to be a DOMINANT sire, for example. The net result of ALL inbreeding to Northern Dancer among sales foals of 2003-2007 was a Price Index of 1.087 and a PPI (result) of 1.085. I rest my case.


The original commentator wrote again thusly:

Thank you for your informative reply! I was interested in getting your point of view because I do think the statistics you use in your articles offer a more objective analysis of the subject.

I thanked her again and told her that I had been thinking about this subject matter some more and could feel another post coming on. So here goes.

“Doc,” the author of the inbreeding treatise in question, did get another thing right, and I did not give him sufficient credit for it. He said something along the lines that when you look at a five-cross pedigree on a piece of paper or your computer screen, what you see is a collection of names. Nothing more, nothing less.

Most people look at that collection of names and automatically assume that each name makes some positive contribution to the foal in question. As “Doc” pointed out, this is far from the truth. I have been told that after five generations, the foal in question may have received ZERO genes from a particular ancestor in the fifth generation. There are 32 ancestors in the fifth generation. The foal in question may have received 100% of its genes from only 30 of those ancestors in the fifth generation (just to pick a number). All 32 ancestors are not automatically represented.

Even if all 32 ancestors in the fifth generation were automatically represented, that does NOT mean that the genes they have passed on make a POSITIVE contribution to the foal in question. The contribution they do make may have nothing to do with racing class or ability. They may have passed on only the gene for coat color, for example, or for some other attribute that has nothing to do with racing class or ability. The same holds true for generations even closer than the fifth.

“Doc” was correct is pointing all this out, and I give him credit for it. In this regard “Doc” is way far ahead of most of the PPBSers (purveyors of pedigree bullshit) who pass themselves off as pedigree “experts.”

“Doc” seems to think that that inbreeding to “dominant” ancestors using only the best sires and best dams produces good results. Well, DUH?????!!!!!!! Of course it produces good results!!!!!! If you use only the best sires and best dams, of course you will get good results. Regardless of whether the foal in question is inbred.

“Doc” seems to think that “selective” inbreeding as described above produces results that are even BETTER than the results of the same quality of matings without the inbreeding. That is his opinion, and he is perfectly entitled to it. The only problem is that he offered no scientific PROOF of that opinion. And it is virtually impossible to prove (or disprove) the validity of that opinion.

If you examine any inbreeding scheme in its entirety, you will find that the qualifiers possess a wide range of pedigrees. Some are good pedigrees, sell for high prices, and produce good results. Some are bad pedigrees, sell for low prices, and produce poor results. The majority are average pedigrees, sell for average prices, and produce average results.

That is one reason why I listed all those foals in my series on close inbreeding. By listing all those foals, I was trying to make that very point. They all had a certain inbreeding scheme in common. Some were good pedigrees, some were bad pedigrees, and many were in between. I was hoping that people would actually look at those lists in order to grasp this concept for themselves, but I am not optimistic that many people gave those lists more than a cursory glance.

My point is that the only effective way to examine any inbreeding scheme is to do so in its entirety. Evaluate ALL of the qualifiers, which will include a mixture of good, bad, and average pedigrees. The prices for ALL the qualifiers will tell you whether that particular collection of pedigrees was good overall, bad overall, or close to average overall.

If the collection of pedigrees was good overall (prices above average), than it SHOULD have good results. If the collection of pedigrees was bad overall (prices below average), than it SHOULD have poor results. I then compare the prices to the results and can see whether a particular collection of pedigrees outperformed its opportunities or underperformed them (or performed pretty close to expectations).

In the case of closely inbred foals, they had prices ABOVE average and results BELOW average. Clearly not a desirable combination. Clearly not a successful inbreeding scheme.

In the case of of all 56,000+ inbred foals, they had prices very slightly above average and results very slightly above those prices. All in all, a very slightly successful breeding scheme.

I have already commented on the results for all inbreeding to Northern Dancer. That scheme had a Price Index of 1.087 and a PPI (result) of 1.085. The two numbers being so close means that all in all, this inbreeding scheme was very close to expectations based on its prices (or slightly negative, if you prefer to interpret it that way).

To repeat a quote from my commentator, “I do think the statistics you use in your articles offer a more objective analysis of the subject.” Thank you again. A more objective analysis of the subject is exactly my goal.

And that is the crux of my objection to the treatise of “Doc.” It is his own opinion, and he is perfectly entitled to it. But do not mistake an OPINION for a FACT. There was very little “objective analysis” in his treatise. It was all OPINION. And some of the FACTS he threw out proved to be WRONG, which undermines the validity of his OPINION.

And that is also the crux of my objection to all the assorted purveyors of pedigree bullshit. (I do not count “Doc” as one of them. He has a little more sense than that.) All the assorted purveyors of pedigree bullshit have to offer is OPINIONS. They are extremely LONG on OPINIONS and extremely SHORT on “objective analysis” (also known as the scientific method). They purvey no SCIENCE at all.

This entry was posted in Uncategorized. Bookmark the permalink.

10 Responses to Purveyors of Pedigree Bullshit

  1. Stéphanie says:

    Once again, a well thought of and researched article! Thank you!

  2. Michael Martin says:

    By analyzing all the foals with Northern Dancer inbreeding, you neglect to take notice of the functional aspects of “Doc’s” piece. Doing so (analyzing all foals) makes his point that breeders use pedigree only, rather than performance and pedigree, in selecting sires, especially when budgets are limited. Perhaps an additional selection criteria is needed, such as inbreeding among only graded stakes winners. Comparing these to graded-stakes-breedings of unrelated individuals could reveal an advantage. This might reintroduce the science into analyzing Doc’s effort, and avoid the broad brush evaluation. In fact, pigeon breeders might raise better racehorses with the closely-related matings mentioned.

    • ddink55 says:

      Yes, you could examine only the foals with the “best” pedigrees. Those foals would certainly sell for higher prices. As such, they would be EXPECTED to produce BETTER results. Would those results be better than indicated by the prices????? “Doc” thinks so, but that is only his (unsubstantiated) opinion. Personally, I am NOT gonna waste my time on this. I have other fish to fry. If YOU think such an approach would be productive, why don’t YOU do it????????????????????? It is still a FREE country (more or less and probably not for very much longer).

      • Michael Martin says:

        It will be a free country–at least as much as it is now–for a while longer. The sky is not yet falling. Of course the prices for such matings would be better, but personally, lacking the resources and time, it was only a suggestion. I will simply assert that, since intense inbreeding has worked in other species to stabilize type and performance, it must work in Thoroughbreds. Without a functional performance criterion, though, you can’t tell if it did increase performance above the average for other high performing matings. Lighten up! It’s a FREE country. Happy New Year, and thanks for your work.

      • ddink55 says:

        OK, I am actually thinking about doing this now. Let me make sure I know what you are looking for. Of the 71,000+ sales foals of 2003-2007, a little over 2,000 were out of mares who were graded stakes winners. I would expect close to 500 of those 2,000 to be inbred to ND. If I understand you correctly, you think I should compare the 500 or so inbred to ND to the 1,500 or so not inbred to ND (all 2,000 or so out of graded stakes winners). Let me know if that is NOT correct.

  3. Michael Martin says:

    Almost; the difficulty would be that the other 1,500 are likely to have linebreeding to another dominant stallion, but that is essentially the point, not that my assertion was solely about ND. Maybe I am naive in my assertion, as so many of the graded stakes winners are the product of linebreeding. Does the linebreeding produce better performing individuals, above the average for offspring of other graded stakes winners which are not linebred, not just better sale prices? Such an analysis ought to clarify our discussion, and thank you for that consideration.

  4. Stacy says:

    So are the more successful one’s those mares who were physically similar or dissimilar to the stallions they were bred to? It would make sense that the majority of graded stakes winners would be physically superior in some ways to those who are not and would therefore be more likely, statistically speaking to pass on those good physical traits to their offspring.

    • ddink55 says:

      That sounds like a question I can not really answer. At least not statistically. That is a question more properly addressed to my colleague at Bloodstock in the Bluegrass. Most people who make a living measuring horses would say that it is better to mate “like” to “like” than “like” to “unlike” or vice versa. At least that is my understanding.

      That does bring up another question, however. That question is when to realize that a question can be answered statistically (at least with the tools you have available) and when it can not. Which reminds me of a story. So here goes.

      Back wen I was still slaving at Thoroughbred Times I got called into the office of the editor one day. I thoroughly despised that stupid SOB, he knew that I thoroughly despised him, and I knew that he knew it.

      He gave me a list of about a dozen “principles” of Tesio. He told me that I should spend the rest of my professional life proving or disproving those “principles” statistically. I glanced over them, and it immediately became apparent to me that all of them were TOTALLY incapable of being proved or disproved statistically.

      Many of them had to do with conformation and matings. In order to prove or disprove them statistically, I would have had to have access to precise measurements of all those sires and dams (most of them dead for at least 50 years). Not to mention a large control group and the sires and dams of all the members of that control group and their sires and dams.

      So it was clearly an IMPOSSIBLE task. To this day I do not know if he was so STUPID that he did NOT realize how IMPOSSIBLE that task was. Or if he was just blowing me shit about statistics in his own PERVERSE way.

      At that point I wanted to collect another paycheck or two before departing. So I did not say too much of anything to him in response. I did depart several weeks later, sticking around long enough to see that some of my stuff was published appropriately (and I did submit a RESIGNATION before I was FIRED).

      The point is that it is important to know what can be proved statistically and what can not (given whatever tools you have at your disposal). I try to concentrate on the former and avoid the latter.

      Some recent comments have spurred me to do some research on a subject that I originally thought belonged in the latter category. But I decided to give it a try and found that it actually belonged in the former category. I would not publish the results if I thought they did not really “prove” anything. I am going to publish them next week precisely because I think they DO “prove” something. And that is all I am going to say right now. Stay tuned.

  5. Pingback: Full of Sound and Fury | Boojum's Bonanza

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s