Regression–Other Sires and Broodmare Sires

As some readers might remember, I did a post on “sires of sires” among sales foals of 1999-2002 back around when I first started this blog, in the summer of 2010. Basically I identified the 20 most popular “sires of sires” among that group (the ones who showed up most often) and detailed their prices and results. It is about the only study I have ever seen on “sires of sires” (second generation only).

Mr. Prospector topped that list, not the least bit surprisingly, showing up 4,697 times as a “sire of sires” (P1 in the second generation). Valid Appeal was 20th, showing up 549 times.

Northern Dancer was 18th on that list, showing up 592 times. And he was not very good. He had only 22 stakes winners from those 592 foals (3.72%). The base line for this group was 3.78% stakes winners (2,046) from foals (54,174).

And those 22 stakes winners were not very good either, averaging only 507 Performance Points apiece (compared to the overall average of 617). That gave ND a PPI (result) of 0.81, far below his Price Index of 1.52. ND deteriorated a lot as a sire of sires, from 6.5% stakes winners from foals among all North American-bred foals of 1983 to 3.72% among sales foals of 1999-2002.

If you do a composite (add up the individual numbers) of those 20 leading “sires of sires” among sales foals of 1999-2002, you get 1,126 stakes winners from 26,443 foals (4.26%). Not bad, higher than the base line for that group of 3.78%. I thought it might be amusing to see what those same 20 sires did as sires themselves (first generation).

Over the entire course of their careers, those same 20 sires had 2,050 stakes winners from 17,568 foals (11.67%) (statistics from Equineline).

So those same 20 sires went from 11.67% stakes winners from foals in the first generation to 4.26% in the second generation (with a base line of 3% for the former and 3.78% for the latter). 4.26% is an improvement of only 11% over a baseline of 3.78% (second generation). 11.67% is an improvement of 289% over a base line of 3% in the first generation. See how much “influence” is lost over the course of only one generation????? That is regression to the mean.

Of course those 20 sires included some real duds. Nags such as Private Account and Devil’s Bag who at that time were considered decent as “sires of sires” but who really were not very good in that role. So I decided to examine only the top eight sires among those 20, those who had 1,000+ foals among sales foals of 1999-2002.

First let us look at their own sire records (first generation).

Sire                               Foals          Stakes Winners          %            Improvement %

Mr. Prospector           1,195                   182                   15.23                  12.23

Storm Cat                    1,452                   176                   12.12                    9.12

Danzig                          1,099                   198                   18.02                  15.02

Forty Niner                    948                     56                    5.91                      2.91

Deputy Minister         1,162                     90                    7.75                      4.75

Gone West                   1,281                     94                    7.34                      4.34

Seattle Slew                 1,103                   111                   10.06                    7.06

Fappiano                         420                     47                    11.19                     8.19

Totals                           8,660                   954                   11.02                     8.02

That last column (Improvement %) is merely the actual percent of stakes winners from foals minus 3% (the overall norm for the breed). Now let us look at these same eight sires as broodmare sires (second generation).

Sire                               Foals       SWs/%    Actual/Expected Improvement % Regression %

Mr. Prospector           4,629      368/7.95                        4.95/6.12                            81

Storm Cat                    3,312       165/4.98                       1.98/4.56                            43

Danzig                          3,876       188/4.85                        1.85/7.51                            25

Forty Niner                 1,946         80/4.11                        1.11/1.46                            76

Deputy Minister         3,287        185/5.63                       2.63/2.23                          118

Gone West                   2,621         97/3.70                        0.70/2.17                            32

Seattle Slew                3,528        210/5.95                       2.95/3.53                            84

Fappiano                     1,706         102/5.98                       2.98/4.10                           73

Totals                         24,905      1,395/5.60                      2.60/4.01                          65

I should probably explain the chart above using Mr. Prospector as an example. As a sire himself (first generation) Mr. Prospector had 15.23% stakes winners (182) from foals (1,195). Assuming an overall baseline of 3% stakes winners from foals, Mr. Prospector should have had 9.12% stakes winners from foals as a broodmare sire if “influence” declines by ONLY 50% per generation. I arrive at that figure of 9.12% by taking the average of 3% and 15.23%.

If Mr. Prospector had 9.12% stakes winners from foals as a broodmare sire (second generation), he achieved 100% of his expectation in that role based on his own record as a sire (first generation, 15.23%). He actually achieved 7.95% as a broodmare sire, which is 81% of his expectation (4.95%, his actual improvement over 3%, divided by 6.12%, his expected improvement over 3%).

The second-last column shows 4.95% actual improvement for Mr. Prospector and 6.12% expected improvement. The regression % is simply the former divided by the latter, or 81%, meaning that Mr. Prospector achieved 81% of his expectations as a broodmare sire based on his own record as a sire.

That 81% is actually pretty good compared to the other seven sires listed above. Their composite averaged 65%. Only Deputy Minister scored better, at 118%, and that was only because Deputy Minister achieved only 7.75% stakes winners from foals as a sire himself (first generation). Mr. Prospector was leading broodmare sire nine times (all in a ten-year period from 1997-2006). His actual 7.95% stakes winners from foals is by far the best of the eight broodmare sires listed above (Fappiano is next at 5.98%).

Some of these other seven sires were not nearly as good. Danzig scored only 25% of his expectation, Gone West only 32%, and Storm Cat only 43%.

Now let us do the same for these same eight sires as “sires of sires” among sales foals of 1999-2002 (second generation).

Sire                               Foals       SWs/%    Actual/Expected Improvement % Regression %

Mr. Prospector           4,697       224/4.77                         1.77/6.12                            29

Storm Cat                    2,986      134/4.49                         1.49/4.56                            33

Danzig                          2,285        94/4.11                           1.11/7.51                            15

Forty Niner                 2,111         77/3.65                          0.65/1.46                           45

Deputy Minister         1,961       104/5.30                         2.30/2.23                         103

Gone West                   1,591         73/4.59                          1.59/2.17                           73

Seattle Slew                1,480          65/4.39                         1.39/3.53                          39

Fappiano                     1,411           43/3.05                         0.05/4.10                           1

Totals                         26,443    1,126/4.26                          1.26/4.01                           31

The base line for sales foals of 1999-2002 was 3.78% stakes winners for foals. Just for the sake of consistency I have used 3% in the chart above (to make it similar to the other two). That in effect gives these eight sires a huge benefit of the doubt in this chart. Forty Niner (3.65%) and Fappiano (3.05%) were actually below the correct base line of 3.78%.

Without belaboring the point, it is obvious that the scores for these eight sires  as “sires of sires” (second generation) are much lower than for the same eight sires as broodmare sires (also second generation). The composite above has a score of 31%, well below the comparable figure of 65% for these same eight sires as broodmare sires. And if I used 3.78% instead of 3%, it would be even lower (12%).

This should not be too surprising at all. The more of a “reputation” a “sire of sires” acquires, the worse his actual results are going to be. Storm Cat is a classic example. Zillions of lower-class sons of Storm Cat have gone to stud and polluted the breed.

That is another reason why “influence” regresses so quickly to the mean. Because breeders exercise little or no judgment at all in breeding to lower-class sons of Storm Cat (or whoever the “hot sire of sires” is at the time). That inexorably diminishes the actual observed “influence” of Storm Cat, et al., not only in the second generation but also carrying farther back into the third, fourth, fifth generations, etc.

It happens to all popular sires to a greater or lesser degree as they recede back and become more remote “names in pedigrees.” Their actual observed “influence” is greatly diminished by the indiscriminate use of all their sons, regardless of their actual worth (or lack thereof, I should say). That score of 31% for these eight sires as “sires of sires” reflects a reality in which too many of their lower-class sons are used and overused.

The important point I wish to make is that these eight sires are pretty much par for the course. Yes, you can find a Deputy Minister who exceeds his expectations from the first generation in the second generation. If he starts off low enough in the first generation, that is, as Deputy Minister did.

But much more often you will see that sires in the second generation fail to reach 100% of their expectations based on their own performances in the first generation. And those expectations are based on “influence” diminishing by ONLY 50% per generation.

So a score of 100% means that a sire diminished by ONLY 50% from the first generation to the second generation. The vast majority of sires diminish by MORE THAN 50% from the first generation to the second generation. They score below 100% (which is actually only 50%, accounting for the difference in generations) by this measure.


This entry was posted in Uncategorized. Bookmark the permalink.

5 Responses to Regression–Other Sires and Broodmare Sires

  1. vineyridge says:

    So what are we supposed to take away from this series. That all breeding theories that aren’t first generation based are voodoo? That one shouldn’t breed to a sire until he has the numbers to prove his quality? That breeding the best to the best is a huge gamble no matter how much money you throw at the sire and dam?

    • ddink55 says:

      “Voodoo” is your choice of word, not mine. But I would say that any breeding theory that relies on sires (and dams) in the third, fourth, fifth, or further generations to transmit a POSITIVE and RELIABLE “influence” is bound to fail any true test of objectivity.

      • Ned Williams says:


        Does this make any sense? If you look at an extended pedigree, you will likely see sires and dams that are productive matches. You will also see matches that are far less productive. Chasing the less productive lines because they were productive several generations back is not wise. Breeding, to race (I say “to race” because the sales game is unfortunately a totally different beast) to non proven sires is not wise. Breeding an unproven mare that has a bad race record is not wise (unless you are familiar with her and know of some good reason as to why her true potential was not fulfilled). However, none of these unwise decisions guarantee that you will not produce a good horse…this is the brutal truth of genetics. However, your chances are improved, if you avoid some or all of these unwise decisions. Good families are good because they are currently good. Families that are currently bad can become good. . .through the vagaries of genetics. The question is: do you want to be the person who spends his/her money and time trying to bring a family back? Racing ability in the first generation is your best hope of giving yourself some edge.Moving to the second generation has some value, but less so than the first. After that you are just reading interesting names.

        When I first started reading Boojum I was hoping for the silver bullet. However, I have learned that there is no easy way to predict a foals future racing ability. Boojum, in my humble opinion, has punched holes in many of the faulty breeding and pedigree theories that pervade the breeding industry today. He has done so in a systematic, scientific, and fair way.

        So if you are like me you might be thinking, “What am I waisting my time for? Is there any way to improve my chances?” YES. Breed the best racing quality mare that you can get your hands on. If you can get a proven mare, all the better. Breed to the best quality proven stallions that you can afford. Read this blog often and use Boojum to help guide you as to quality and what “best” means. The more I understand the more the old saying becomes truer: Breed the best, to the best, and hope for the best. Just make sure “best” means “best producer or best racing class” not best nick somewhere back in the pedigree.

  2. Steve Lichti says:

    Aren’t you forgetting 1/2 of the equation? Northern Dancer, at (eventually) a stud fee of $1M, was
    certainly seeing the finest producing and racing broodmares in the world. Most of his sons, at
    much lesser stud fees, were seeing a pool of less productive broodmares. Less potent sons with less
    productive broodmares points to a regression of at least 75% (1/2 x 1/2 = 1/4).
    Assuming a 50% regression, only gives weight to the sire of a horse and ignores the fact that great
    sires certainly see a better broodmare than those seen by his sons (in totality).
    A.P. Indy must have seen a better class of broodmare than the average broodmare seen by his
    Don’t forget mom (the better half?) as a big factor in explaining why the regression must be
    greater than 50%.

Leave a Reply

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

You are commenting using your 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 )

Google+ photo

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

Connecting to %s