Do Current Conditioning Methods Impact the Breeding Shed?


Under-conditioning of our thoroughbred starts is not only producing ever-slowing Belmont Stakes times, it is also slowing the progress of the breed as a whole. Exercise affecting genetics? Impossible, you say? Current research is full of examples where intensity of exercise programs has been shown to influence genetic expression. I’m not qualified to expand on that point, but I can certainly draw comparisons between AEI and the conditioning programs of any given racing era.

The stallion stars of the 21st century, even the immortal Storm Cat, are not passing down their racing genes to their respective progeny as often as was the case several decades ago. At least that is the story told by the Average Earnings Index.

If you know full well the definition of AEI, please skip the next 2 paragraphs and get to the ‘good’ stuff, well maybe not so good if you are a breeder in today’s marketplace.

What is AEI?

Developed by long time BloodHorse editor Joe Estes, the Average Earnings Index is a statistical method to compare stallions across different eras. One can’t simply compute earnings, as the value of the dollar changes over time. Nor should crop size distort the picture. Mr. Estes simply computes the average racetrack earnings by age group each year, terms this figure 1.0, and compares how a stallion’s offspring performed vs. that average. If the average earnings for Storm Cat 3 year olds in 2000 was $20,000, and the breed average for 3yo that season was $10,000 – the AEI would be 2.0, for example.

The primary complaint is that AEI can be skewed by one big runner, but isn’t that part of the joy of breeding – hitting the Home Run Horse? Furthermore, when we compare stallion careers in this manner – the sample size becomes quite large and this ‘fault’ is mostly remedied.


In his time, Mr. Estes considered an AEI of 4.0 or greater as an indicative of an outstanding sire. Sadly, today’s breeding stars are only able to post figures just over 2.5.

Leading Stallions by AEI from 2013-2001:

2013: 2.74 War Front
2012: 2.68 War Front
2011: 3.03 Storm Cat
2010: 3:09 Storm Cat
2009: 3.26 Street Cry
2008: 3.15 AP Indy
2007: 3.38 Storm Cat
2006: 3.50 Storm Cat
2005: 3.57 Storm Cat
2004: 3.66 Storm Cat
2003: 4.32 Danzig
2002: 4.42 Danzig
2001: 4.53 Danzig

The past decade plus has shown us the end of the line for both Storm Cat and Danzig, and although both were breed leaders in their respective years, the AEI distinguishes one from another quite distinctly. AEI also seems to decline with age, which I think can be expected due to nature. However, Storm Cat finishes at just over 3.0 and Danzig in the vicinity of 4.3 – a huge level of statistical significance. I cannot find lifetime AEI data on these 2 since they are no longer active at stud. But, I did find a mention of Danzig’s AEI in 1999 at a robust 5.17, so it can be safe to assume his lifetime AEI performance to be vastly superior to Storm Cat. (EDIT Found it: Danzig 4.53, Storm Cat 4.11, not as big a difference as suspected, but still remarkable). What is also somewhat curious as I look at the data is that Storm Cat’s AEI fell off more drastically than that of Danzig.

Danzig’s not the only stallion who can claim AEI superiority over Storm Cat, not by far, but I’ll save that for a bit later in the post. The suspense is whether or not Storm Cat will make the All-time Top 20.

Here is a quick window into other stats indicative of breeding success for these two taken from a 2005 summary that briefly backs up the concept of the AEI indicative of superior breeding success:

% of stakes winners: Danzig 18%, Storm Cat 13%
% of grades stakes winners: Danzig 10%, Storm Cat 6%
% of G1SW: Danzig 5%, Storm Cat 3%

And here is a graphical representation of the top 5 in AEI for each of those years which further illustrates the increasing decline of stallion performance measured by offspring race earnings compared to peers:


In 2013 you can hit the Top 5 list with an AEI of 2.2, but back in 2001 it took a 3.27. The last 2 years (2013 and 2012) no stud (of several hundred) broke the 3.0 barrier, yet in 2001 the average of the top 5 was 3.9. Remember, Estes talked of superstars in the shed at 4.0 back in his day. Talk about lowering the bar. That graph looks familiar, where have I seen that trend before? Oh yes, recent Belmont Stakes winning times:


My apologies for this terribly primitive chart from an earlier post. The Y-axis is the winning time in seconds, averaged for each decade to control for equine gods such as Secretariat and also for varying track conditions at Big Sandy as well as different pace scenarios which can effect final times. The X-axis represents each decade, with the 1 corresponding to the limited data set of the 2010’s – 4 years, and the 9 representing the 1930’s – the first full 10 year span where the race was run at its current 12 furlong distance.

So, the number 4 on the horizontal axis is the fastest decade (1980’s) when you average the 10 winners’ times and come to a value of 147.94, or 2:27.94.

I don’t have enough data on AEI to precisely overlay these two charts. But we can make two general statements:

  1. 2:30+ winning times over fast Belmont tracks have become the new normal in the 2010-2013 timeframe, performances not seen since the 1930’s.
  2. AEI numbers of our ‘leading’ stallions are approaching 2.5, versus a historical 4.0 or better.

Now of course these two trendlines don’t overlap precisely. Belmont winning times were on the improve after Bold Ruler’s day, but the sudden reversal in the early 90’s is apparent, coinciding with the legalization of Lasix in New York State around 1995. The Kentucky Derby and Preakness Stakes have stopped improving over the last several decades as well; but the longer Belmont has actually reversed the trend in winning times. Meanwhile, AEI declines consistently over the entire interval.

Danzig is perhaps the lynchpin in this AEI analysis, before him big AEI numbers, afterwards – not so much. Conditioned by Woody Stephens, although only raced 3 times due to injury, Danzig never won the Belmont – but Stephens did with Danzig Connection in 1986, the last of his 5 consecutive victories beginning in 1982.

Missing from this post is any reference to mares. Many believe, as do I, the majority of good racing genes comes from the female side of the mating. However, billions are tied up in stallions due to the numbers game foisted upon them by nature. I think we can agree that during each racing era; 90% of all elite runners are conditioned similarly. From the mid 1980’s onwards it’s all Lukas and speedwork every 6/7 days, racing with 5-6 weeks rest. Starting with Stephens and running back towards Preston Burch and Max Hirsch in the 40’s – it was 2-3x weekly speed sessions and racing every week or two. Remember the Derby Trial was run on a Tuesday preceding the Derby on Saturday. Nowadays that race should be called the Preakness Trial, because no elite horse is prepared to run twice in a month, much less a week.

So, although AEI only looks at half of the equation, the mares being bred too were also conditioned/raced more aggressively in the days of Bold Ruler, etc. The Comparable Index somewhat addresses this, but I’ll have to leave that for another day, or maybe someone smarter than myself.

Minimal conditioning practices decrease the rate of passing on athletic genes to offspring. When horses were worked long and often, and raced more frequently, they became better athletes and maximized genetic expression which carried over to the breeding shed. Do I have ‘proof’ of this? Nope, but I don’t need to as I am merely a blogger who is biased towards ‘old school’ conditioning practices.

Bottom line: Going to a top stud in Danzig’s day gave you a decent chance at earning 4x the breed average and covering your bills; today the ROI is half that. The importance of choosing a top sire is lessening.

Here is a 99% complete list of all time studs ranked by AEI, I may have missed a few from 1990-2010, but based on recent trends it’s not likely they would have made this list anyway:

Bold Ruler                           7.73
Alydar                                   5.21
Nasrullah                             5.16
Northern Dancer              5.14
St. Simon                             4.75
Nijinsky                                4.74
Danzig                                   4.53
Tom Fool                             4.51
Hail to Reason                   4.47
Fappianoo                           4.46
Bull Lea                                 4.37
Mill Reef                              4.36
Blandford                            4.29
Count Fleet                        4.28
Mr. Prospector                4.25
Seattle Slew                       4.12

17. Storm Cat                     4.11

Ribot                                     4.09
Round Table                       4.01
Blushing Groom                3.97
Buckpasser                         3.94
Hoist the Flag                     3.88
Equipoise                            3.86

*I stopped with Equipoise because I always chuckle at the fact this is also the name of a very popular steroid. Surely not known back in the 1930’s, but used today quite often, not only in horses, but also in humans.


Could I get past the Jockey Club today naming a horse Testosterone? Furosemide? Cobra Venom? Clenbuterol?

Ok, now for the arguments against my primary claim. Most of them will involve other nuances of the AEI, and rightfully so. However, the crux of the Index lies in comparing racehorses to their peers in each individual year. So, the fact that horses started much more often in Bold Ruler’s day than in Storm Cat’s shouldn’t matter. Nor should the fact that purse sizes now are significantly larger due to inflation and casino money. Now, the predilection of today’s stallions to get larger books of unraced mares? Absolutely a valid claim and surely a factor in declining AEI in the modern era.

Let’s think about this yet another way. Assume we have two twin girls who become mothers at age 25. Let’s also assume they married twin brothers. Identical genes all around, or as close as we can get. Mother A and Father A both read books obsessively pretty much from school age on, while Mother B and Father B never read more than a cereal box, spending the rest of their time watching television.

Baby A and Baby B are born. Who is more likely to have, or even develop, a strong inclination to read books? Have the reading habits of Parents A altered their genetic expressions to influence the likelihood of passing on the ‘reading gene’ to their offspring. Have Parents B done the opposite? Surely possible, albeit unlikely to ever ‘prove’.


About bpressey

Equine Exercise Physiologist

Posted on January 24, 2014, in Uncategorized. Bookmark the permalink. 21 Comments.

  1. Ralph Livingston

    Hello Bill:

    I love it when folks read books I recommend. You got more out of Mr. Estes’ slim, but profound, volume than I ever did!

    As for your question “Have the reading habits of Parents A altered their genetic expressions to influence the likelihood of passing on the ‘reading gene’ to their offspring.” -the answer, that science is discovering, is a very definite “affirmative” !!

    Science is learning that genetic composition is not fixed for life, the stresses of existence cause ‘mini mutations’ in an individual’s lifetime. These subtle changes in the DNA helix can cause dramatic changes in the phenotype in just one generation.

    They don’t teach Math in schools like they used to, they have “lowered the bar”; your research is proving that the bar has been plummeting in TB performance.

    Will we ever see another Round Table, 66 starts-56 in the money; or John Henry 83 sarts, 63 in the money, any time soon?

    Keep up your great work!!

    Best regards,

    Ralph in Kelowna, BC

    Science is learning that genetic structures are not fixed for a life time but rather undergo mutations due to stresses inthe individuals lifetim

    • Yes, Ralph – thanks very much for the Estes book recommendation; would have NEVER found it without you. I also had some help from pedigree guru Frank Mitchell on finding some old-time sires’ lifetime AEI numbers. A big thank you to him as well.

      Sarah here in the comments terms the concept of phenotype influencing genotype as Epigenetics. Fascinating.

  2. Epigenetics is the study of how environment influences the how the inherited genetics are expressed. It is most clearly demonstrated in cloning, where the exact same genes can result in astoundingly different individuals. Different aspects of different genes will be ‘turned on’ and expressed depending on the experiences of the individuals. The traits needed to for a horse to survive growing up in a 12 x 12 box eating molasses is very different form what is needed by a horse grows up eating grass and moving15-20 miles a day. So yeah, there is science out there to back up your blog!

    • Thanks Sara, a name to put to the concept, I never even thought of the cloning phenomenon. Has anyone cloned identical cows perhaps; and measured environmental influences on the milk production of offspring? or done the same with respect to muscle mass development in steers?

      • There is a fair amount of cloning in meat and milk production, mostly in conjunction with AI, keeping the one stud going, and gross measurements seem to transfer fairly predictably. Probably because their environment stresses those traits and there is not a lot of subtlety in the butcher business! The meat/milk producers are not concerned with genetic diversity either, which could be seriously bad news, but that is a whole other subject.
        Cloned cats on the other hand are totally unpredictable, even down to coat color. There have been a few clones among performance horses- polo and jumping individuals mostly. There has not been a lot of fanfare about it probably because the results are so disappointing. Apparently what makes a great horse can’t be put in a test-tube. It depends on environment, conditioning, and the horse’s connection to the people on their team.
        I am not up on the most recent research and results, but I am sure there is some interesting stuff out there if you search for it.

      • sounds like they have found what equine genetecists have found; no matter how much you improve the breeding process only 25-30% of athletic performance is hereditary. A few industry types have told me: horses with a dearth of identified ‘good’ genes stay poor, but only some of the ones who have an abundance of ‘good’ genes become good. Many of the ‘good’ ones fail because they cannot seem to adapt to the conditioning process – so the search is now on worldwide to identify a set of ‘trainability’ genes.

      • And as you emphasize. take a profound look at the way horses are housed and conditioned…I found this old text quite interesting:

  3. Bill – I remember the days of Bold Ruler and some of the others. Back then a stallion ‘s book was limited to around 60 mares. Therefore, if you wanted to breed to Bold Ruler you had to have a mare that would “qualify” as worthy in the stallion managers eyes (or else know someone on the inside). I used to call up and submit a mare and then be called in return to see if I had got in. Today’s stallions are covering hundreds of mares and getting one in to a particular stallion is more about your bank balance than the quality of mare you have. I’ve always felt that this is part of the decline in our racing stock today.

    • But does AEI already reflect that, Sue? If more ‘poor’ racemares are sending offspring to the races, they would presumably earn less – dragging down the average earnings for everyone? Leaving the AEI relatively unscathed? I am not enough of a mathematician to eyeball that one…

    • Absolutely, I agree with the bank balance aspect of your comment. No breeders gives a damn about AEI, but rather the hammer prices of yearlings at auctions, that is what the market values.

  4. I don’t know either. It’s just a change I’ve noticed venturing back into breeding these days vrs what used to be in the early 70’s. All business evolves but I can’t help but think we’ve strayed from what has worked in the past.

    • I’ve always thought that horse racing and horse breeding/selling are two separate and distinct sports.

    • It’s my understanding that AEI refers to the stallions progeny performance vrs the standard of that year, therefore a big performer would skew the number to the high side. Is that correct?

      • Correct, but we are looking at hundreds, if not thousands, of starts which should blunt the effect somewhat. or, the other way to look at declining AEI over the decades is that you are far less to hit a ‘home run horse’ in the current environment – which is not good news, either.

  5. Bill – what do high speed treadmills run these days? Whose model do you recommend?

  6. Sue’s comment that in the seventies a popular stud would limit the book to sixty mares is accurate and. in the fifties and forties a popular stud had a book that was limited at forty mares.
    The smaller books statistically change the impact of a high earning offspring. The high performer has a higher impact because the pool of runners is smaller.
    This can be seen currently with the excellent young stud, Into Mischief,. Currently he has a 3.30 with a 10% grade stakes winners. However, This has been done with only 84 foals and 8 stakes winners and mares with a Combined Index of a pedestrian 1.12 CI.
    If his books increase to close to two hundred per season as is happening with the commercially popular studs today, his AEI will drop even if he maintains of his current rate of
    10% rate of stakes winners,
    It is a function of larger numbers.
    Of course, that has no effect on slowing times. That is a function of conditioning.

    • Agreed for the most part, but even our ‘small book’ studs like Into Mischief can only post AEI figures of 3.3 – perhaps 50 years ago when book sizes were all like his, AEI leaders were nearly double.

      However, I do like your AEI to CI comparison. As I understand, only roughly a third of stop studs actually have larger AEI compared to their books collective CI. Wonder if that is a new phenomenon?

  7. I can’t tell you if the AEI to CI comparison ratio has changed over the years. Don’t have the stats to answer that question.
    You could ask the Blood Horse. They might have the numbers.

    I would not be surprised if the ratio was better 50 yrs ago when the books were smaller. The function of the volume change might also affect the AEI to CI ratio also.

    Even though the mares have exactly the same proportion of DNA passed on to the foal, superior mares can have have 20% or 30% stakes winners of their progeny. The best stallions cannot. A 10% stakes winners to foals is the mark of an excellent stallion.

    Perhaps, the nurture time with the mare might be the difference. The mare might instill behavior characteristics that make the variance.

  8. I suspect the decline of the AEI of top stallions is at least partially attributable to the systematic early retirement of their top offspring. This seems particularly true in the case of Storm Cat and A.P. Indy, whose top colts were rushed off to the breeding shed before the close of their 3-year-old seasons. Imagine how much more money those horses could have made if they were allowed to race as 4- and 5-year olds, which would led to a higher AEI for their sires.

    • Certainly a factor, thanks for the insight. It’s the steady decline that seems odd to me. We’ve been watching early retirements to stud for several years. Book size has been big for years, etc. It almost seems like something ‘systemic’ is at work. Same with the Belmont Stakes winning times. They decline steadily for 30+ years. No change in track surface can effect that type of downward curve, unless one posits that the BEL crew adds a half inch of cushion every 5 years like clockwork.

      I hope to make one giant graph of things like AEI, classic race winning times, starts per year, etc. and see how similar all these declines are in graphic form.

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 )

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: