Menow Versus Bull Lea

“What is the overall scope of influence of Menow on modern pedigrees?”

If you ask that question of the average devotee of pedigrees, the answer would be something like the following:

“Well, Menow sired Tom Fool. And Tom Fool sired Buckpasser. Buckpasser sired the second dam of A.P. Indy. And that is about the total extent of Menow’s influence.”

That answer would be only partially correct. Menow shows up through a bunch of different avenues as well.

Menow also sired Spring Run, dam of Red God, sire of Blushing Groom.

Menow also sired Flaring Top, dam of Flaming Page, dam of Nijinsky II.

Menow also sired First Rose, the fourth dam of Storm Cat.

Menow also figures in the pedigrees of sires such as Majestic Light, Conquistador Cielo, Royal Academy, The Minstrel, Irish Castle, etc.

Considering that Menow figures in the pedigrees of Blushing Groom, Nijinsky II, and Storm Cat as well as Tom Fool, Buckpasser, and A.P. Indy, you can see that he is more pervasive than you might have thought and hence worthy of some statistical consideration.

So I examined the 45,562 sales foals of 2008-2011 and noted down how many times the name of Menow appears. I categorized those appearances into the sources listed below. I categorized Tom Fool, Buckpasser, and A.P. Indy independently as separate entities. In other words, Buckpasser does NOT contain A.P. Indy. Tom Fool does NOT contain Buckpasser nor A.P. Indy.

Just as a reminder, I also counted every time Menow’s name showed up (every presence) as a foal. Some foals had zero Menow. Some had only one Menow. Some had two Menows. Some had three or more Menows.

Prices for the various sources of Menow are listed in the chart below.

Source                      Foals          Average          Maverage          Price Index

Tom Fool               27,798     $45,447       154.19             1.00

Buckpasser           27,778      $44,634       151.04             0.98

A.P. Indy                5,351       $74,919        200.61            1.30

Blushing Groom    9,207       $43,546       149.62             0.97

Nijinsky II           14,280       $50,454       167.04             1.08

Storm Cat             13,102      $52,412       164.59             1.07

All Others              7,534       $37,072        136.63            0.89

Totals                 105,050      $47,515        157.10             1.02

The first thing I notice is that Menow showed up 105,050 times in these 45,562 foals, or an average of 2.31 times per foal. That compares to 95,976 times for Bull Lea, or 2.11 times per foal. So Menow appears about 10% more often than Bull Lea among these sales foals of 2008-2011, which is somewhat surprising.

And there is no chronological bias at work. Both Menow and Bull Lea were foals of 1935 (and both ran in the 1938 Kentucky Derby, the former finishing fourth, the latter eighth of ten, both behind winner Lawrin).

The biggest difference between Menow and Bull Lea is in the all others category. Only about 7% of Menow’s foals (7,534 of 105,050) fall into this miscellaneous category. For Bull Lea it was the majority (53,139 of 95,976). So it is safe to say that Bull Lea was much more widely dispersed throughout pedigrees than Menow was. “Widely dispersed” could also mean “contains a lot more dross.”

The miscellaneous Bull Leas were cheap. So are the miscellaneous Menows (average of $37,072, maverage of 136.63, Price Index of 0.89). Not surprisingly, A.P. Indy was by far the most expensive of the Menows (average of $74,919, maverage of 200.61, Price Index of 1.30). The other sources were not too far from the norms (average of $46,418 and maverage of 154.0).

Listed below are the racetrack results for the various sources of Menow. APPPSW in the chart below stands for average Performance Points per stakes winner, the benchmark now being 701.

Source                      Foals          Stakes Winners          %          APPPSW          PPI (Result)

Tom Fool               27,798            965               3.47         722               1.05

Buckpasser           27,778             923               3.32         661               0.92

A.P. Indy                5,351              254               4.75        699               1.39

Blushing Groom    9,207              392               4.26        739                1.32

Nijinsky II           14,280              589               4.12        752                1.30

Storm Cat             13,102             444               3.39        702                1.00

All Others              7,534              247               3.28         610                0.84

Totals                 105,050           3,814              3.63         702               1.07

As you can see, A.P. Indy (PPI of 1.39), Blushing Groom (1.32), and Nijinsky II (1.30) were by far the most successful of the Menows. Buckpasser was a drag at 0.92. All others were even worse at 0.84.

The chart below summarizes the all-important relationship between prices and results.

Source                      Foals          Price Index          PPI (Result)          Difference

Tom Fool               27,798         1.00                  1.05                +0.05

Buckpasser           27,778          0.98                  0.92                –0.06

A.P. Indy                5,351           1.30                 1.39                 +0.09

Blushing Groom    9,207           0.97                  1.32                 +0.35

Nijinsky II           14,280           1.08                 1.30                 +0.22

Storm Cat             13,102          1.07                  1.00                 –0.07

All Others              7,534           0.89                  0.84                –0.05

Totals                 105,050          1.02                  1.07                +0.05

Blushing Groom was the big winner at +0.35. Nijinsky II was also very good at +0.22. Storm Cat (–0.07), Buckpasser (–0.06), and all others (–0.05) were the worst. A.P. Indy (+0.09) and Tom Fool (+0.05) were slightly positive.

Stir it all together, and the 105,050 foals of Menow posted a price of 1.02 and a result of 1.07. They sold for prices about 2% above average and achieved results about 7% above average. Not bad at all for that volume of foals. Better, in fact, than Bull Lea, whose descendants sold for a price of 1.00 and achieved a result of 1.01.

Tom Fool was the only son of Menow to have survived into modern pedigrees. I thought it might be interesting to divide these results by sex of the source. The descendants of Menow through a male source all come through Tom Fool (including Buckpasser and A.P. Indy). All other descendants of Menow come through a female source (Blushing Groom, Nijinsky II, Storm Cat, and miscellaneous). Prices are summarized below.

Source                           Foals          Average          Maverage          Price Index

Male (Tom Fool)        60,927        $47,665              156.83                   1.02

Female (all others)    44,123        $47,309              157.49                    1.02

As you can see, prices for the two groups are virtually identical.

Source                           Foals          Stakes Winners          %          APPPSW          PPI (Result)

Male (Tom Fool)        60,927                  2,142                3.52             693                    1.02

Female (all others)    44,123                  1,672                 3.79              715                    1.13

As you can see, the female group is significantly better than the male group. Since both groups had the same Price Index of 1.02, it is easy to see that the male group was right on the money (price of 1.02, result of 1.02). The female group outperformed (price of 1.02, result of 1.13, difference of +0.11).

Blushing Groom and Nijinsky II were the reasons the female group outperformed its prices. Storm Cat and all others were a drag on this group.

To recapitulate, Menow shows up about 10% more often than Bull Lea among these sales foals of 2008-2111. Menow is also better than Bull Lea. The former has a difference of +0.05, the latter +0.01 (due to rounding; actually closer to +0.0024).

The descendants of Menow through a female source (especially through Blushing Groom and Nijinsky II) are better than the descendants of Menow through a male source. This is somewhat amusing because most people are not even aware of all the descendants of Menow through a female source. Their familiarity extends only to the male sources.

I must admit that I undertook these two projects on Bull Lea and Menow more for the intellectual challenge of it than for any other reason.

All of this has little to do with breeding/buying/developing/racing a better horse (and that is the whole point of the exercise). Sires and dams (first generation) are much more important than any other names in pedigrees, especially names many generations back.

Having said that, however, in order to have some fun with pedigrees, sometimes you have to go back many generations to make your point. Both Bull Lea and Menow illustrate that point (the former better than the latter). What inevitably happens to names in pedigrees is that sooner or later (sometimes a long time later, sometimes a short time later) both their overall prices and overall results converge on the average of 1.00.

Posted in phalaris | Tagged , , , , , , , | Leave a comment

The Bull Lea Quartet (Storm Cat, Alydar, Nijinsky II, and Roberto)

Two posts ago I remarked:

“So the four principal sources of Bull Lea in modern pedigrees are Storm Cat, Nijinsky II, Roberto, and Alydar. And there are other sources as well.

“With those four sources, I feel pretty certain that Bull Lea would hold up well to a statistical analysis. . . .”

I decided to investigate this matter, to see if Bull Lea holds up at all under statistical analysis and how much of that success (or lack of it) is attributable to the quartet of Storm Cat, Alydar, Nijinsky II, and Roberto.  So I examined the 45,562 sales foals of 2008-2011 and noted down how many times Bull Lea shows up through this quartet, as well as through other sources, and their prices and results.

First a word of explanation. I use the term “foals” in the charts below, but “presences” might be a more accurate word. I count each presence as a foal (and multiple presences as multiple foals).

For example, Alydar has a double dose of Bull Lea (his dam is 3×3 to Bull Lea). Therefore, each presence of Alydar is counted as two foals (ditto for stakes winners).

A more extreme example is Afleet Alex, who carries five crosses of Bull Lea. A hypothetical foal by Afleet Alex out of a mare with Alydar in its pedigree would count as seven foals. Or possibly more if the mare had even more Bull Lea in its pedigree (ditto for stakes winners).

Prices for the quartet and for all other sources of Bull Lea are summarized below.

Source                    Foals          Average          Maverage           Price Index

Storm Cat        13,102     $52,412         164.59             1.07

Alydar              8,798       $46,691         156.22             1.01

Nijinsky II        14,280      $50,454         167.04             1.08

Roberto           6,657        $54,245         171.32             1.11

Totals Above   42,837       $50,869        164.73             1.07

All Others        53,139       $41,332        146.58             0.95

Grand Totals   95,976       $45,589        154.68             1.00

Bull Lea showed up 95,976 times in 45,562 foals (about 2.11 times per foal). Of course many foals had no Bull Lea at all, and many foals had three or more presences of Bull Lea.

I was expecting Storm Cat to be the leader of this quartet in terms of accounting for the most presences of Bull Lea. Somewhat surprisingly, he was actually behind Nijinsky II (14,280 to 13,102) in this respect. They were followed by Alydar (8,798) and Roberto (6,657), although keep in mind that the Alydars are doubled, and therefore only 4,399 foals produced his 8,798 presences.

The quartet above accounted for 42,837 presences among 45,562 foals (0.94 per foal). So they did indeed account for a lot of Bull Lea.

A lot, but not the majority, which surprised me. The remaining 53,139 presences (1.17 per foal) came from all other sources of Bull Lea. I was surprised that there was so much Bull Lea out there not through the quartet above.

Some other prominent sires carrying Bull Lea were Dixieland Band (though Delta Judge), Cox’s Ridge (through Best Turn), and Gone West and Conquistador Cielo (both through Tim Tam).

As for the prices themselves, the overall average for all 45,562 foals was $46,418, and the overall maverage was 154.0 (corresponding to a price index of 1.00). The quartet were all above those benchmarks, with Roberto being highest (average of $54,245, maverage of 171.32, and Price Index of 1.11).

Not surprisingly, the 53,139 foals from all other sources of Bull Lea were below those benchmarks (average of $41,332, maverage of 146.58, and Price Index of 0.95). Put all the groups together, and the 95,976 foals from all sources of Bull Lea had prices very close to the norms (average of $45,589, maverage of 154.68, and a Price Index of 1.00, rounded down from the actual number of 1.0044).

Listed below are the racetrack results for the same groups. APPPSW in the chart below stands for average Performance Points per stakes winner, the benchmark now being 701.

Source                    Foals          Stakes Winners          %          APPPSW           PPI (Result)

Storm Cat        13,102           444               3.39          702              1.00

Alydar              8,798             292               3.32          639              0.89

Nijinsky II        14,280            589               4.12          752              1.30

Roberto           6,657             283               4.25          834              1.48

Totals Above   42,837          1,608             3.75           732              1.15

All Others        53,139          1,658             3.12           682              0.89

Grand Totals   95,976          3,266             3.40           707              1.01

Storm Cat (PPI of 1.00) neither helped nor hindered Bull Lea. Alydar (PPI of 0.89) was a drag on Bull Lea. Nijinsky II and Roberto helped Bull Lea quite a bit, particularly the latter, with PPIs of 1.30 and 1.48 respectively). The quartet collectively helped Bull Lea (PPI of 1.15). All other Bull Lea was a drag (PPI of 0.89).

Stir it all together, and the PPI of all 95,976 foals was 1.01 (rounded up from 1.0068).

The chart below details the relationship between prices and results for the various groups.

Source                    Foals          Price Index          PPI (Result)          Difference

Storm Cat        13,102          1.07                 1.00                 –0.07

Alydar              8,798           1.01                  0.89                 –0.12

Nijinsky II        14,280          1.08                 1.30                  +0.22

Roberto           6,657            1.11                 1.48                  +0.37

Totals Above   42,837           1.07                 1.15                  +0.08

All Others        53,139          0.95                  0.89                 –0.06

Grand Totals   95,976          1.00                  1.01                  +0.01

Storm Cat (difference of –0.07) and Alydar (difference of –0.12) were a drag on Bull Lea. Nijinsky II (difference of +0.22) and Roberto (difference +0.37) were very helpful to Bull Lea, particularly the latter. The quartet overall was helpful to Bull Lea (difference of +0.08). All others were a drag (difference of –0.06).

Stir it all together, and the 95,976 foals with Bull Lea in their pedigrees sold for a price of 1.00 and achieved a result of 1.01 for a difference of +0.01. Actually, that figure of +0.01 is a bit misleading. The actual number is +0.0024 (1.0068 minus 1.0044). Rounding the former up and the latter down yields the listed figure of +0.01.

Bull Lea was a foal of 1935. The average Thoroughbred generation is 10-11 years. The foals in this survey were born from 2006 to 2011. Therefore, the prime position for Bull Lea in foals in this survey is the seventh generation.

I decided it might be fun to see how Bull Lea fared by generations. The three charts below summarize simplistically by dividing the sample into two groups: those with Bull Lea seven generations or closer and those with Bull Lea eight generations or farther back. The former (52,344 foals) has a slight majority over the latter (43,632 foals).

Generations      Foals          Average         Maverage         Price Index

4-7                 52,344        $47,589          156.00             1.01

8-11              43,632         $43,458          153.09            0.99

As you can see above, the 4-7 group has prices ever so slightly above the 8-11 group.

Generations      Foals          Stakes Winners          %          APPPSW        PPI (Result)

4-7                 52,344              1,644                3.14           726              0.95

8-11              43,632               1,622                3.72           687             1.07

As you can see above, the 8-11 group has results more than slightly better than the 4-7 group.

Generations      Foals          Price Index          PPI (Result)           Difference

4-7                 52,344            1.01                   0.95                    –0.06

8-11              43,632             0.99                  1.07                     +0.08

In terms of prices versus results, the 8-11 group is significantly better than the 4-7 group (+0.08 to –0.06).

I must admit that I undertook this project more for the intellectual challenge of it than any other reason. I think it demonstrates how “influence” really works.

Bull Lea is not such a “failed sire” as most people think. In terms of prices versus results, he is very slightly positive (+0.01 or +0.0024; take your pick; I prefer the latter number).

I prefer the latter number because it shows what inevitably happens to names in pedigrees. Sooner or later (sometimes a long time later, sometimes a short time later) both their overall prices and overall results converge on the average of 1.00.

Some Bull Lea is better than others. Roberto and Nijinsky II are very nice. Alydar, Storm Cat, and all others are not so nice.

It seemingly defies logic, but closer generations are not always better than farther generations. The reasons for this were discussed two posts ago, in The Seven Corners Conundrum.

All of this has little to do with breeding/buying/developing/racing a better horse (and that is the whole point of the exercise). Sires and dams (first generation) are much more important than any other names in pedigrees, especially names many generations back.

Forgive me if I do not say it often enough. Sires and dams (first generation) are much more important than any other names in pedigrees, especially names many generations back.

Having said that, however, in order to have some fun with pedigrees, sometimes you have to go back many generations to make your point. And the point here is that Bull Lea is actually pretty typical of sires from that long ago (he was a foal of 1935). He has a collective price of 1.0044. He has a collective result of 1.0068.

What inevitably happens to names in pedigrees is that sooner or later (sometimes a long time later, sometimes a short time later) both their overall prices and overall results converge on the average of 1.00.

Posted in Seattle Slew | Tagged , , , , , , , , | Leave a comment

By a Safety in Overtime

“I’ve always maintained that inbreeding should be practiced only through superior performing animals in an effort to filter out negative recessives. That’s why, despite a lack of evidence, I’ve half-heartedly clung to the notion that stakes-winning broodmares bearing the Rasmussen Factor might proved to be somewhat better producers.”

I received the comment above from Allison about six months ago on my post “You Learn Something New Every Day.” Per his request, I investigated stakes-winning mares bearing the Rasmussen Factor (RF) as producers. Let us start with prices.

Category                    Foals          Average          Maverage          Price Index

RF SW Dams              404           $64,521             202.22                   1.31

I identified 404 foals among sales foals of 2008-2111 whose dams were both stakes winners and possessed RF. Their prices were slightly above the norms, as you can see above, which was to be expected from stakes-winning mares. Their prices were slightly lower, however, than the prices for all 6,659 foals out of stakes-winning mares (average of $85,050, maverage of 222.66, Price Index of 1.45).

The chart below shows results for these 404 foals. APPPSW stands for average Performance Points per stakes winner, the benchmark now being 701.

Category                    Foals          SWs          %          APPPSW          PPI (Result)

RF SW Dams               404            20         4.95             780                    1.62

The RF group had very nice results, a PPI of 1.62, well above their price of 1.31, as demonstrated in the chart below.

Category                    Foals          Price Index          PPI (Result)          Difference

RF SW Dams              404                 1.31                       1.62                        +0.31

This appears to be a clear-cut victory for the RF group. But the 6,659 foals out of all stakes winners also had very similar prices and results, as summarized below.

All SW Dams             6,659               1.45                       1.74                        +0.29

The 404 foals had a difference of +0.31. All 6,659 foals out of stakes-winning dams had a difference of +0.29. The latter group had a price 0.14 higher than the former group (1.45 to 1.31). The latter group had a result 0.12 higher than the former group (1.74 to 1.62).

If this were a football game, the score would be 31-29 in favor of the RF group. They won by a safety in overtime.

So it does appear that stakes-winning dams with the RF factor were slightly better producers than all stakes-winning dams, but the emphasis has to be on the “slightly.”

Posted in La Troienne | Tagged | Leave a comment

The Seven Corners Conundrum

There is no particular reason you should be familiar with the sire Seven Corners (1956 by Roman–Miss Traffic, by Boxthorn). Seven Corners was a decent racehorse, with a record of 71-10-16-11 for earnings of $79,679. He placed in eight stakes but was not a stakes winner. Seven Corners was no great shakes as a sire: 106 foals, one stakes winner, AEI of 1.00 (exactly average, at least theoretically).

Yet Seven Corners still shows up in modern pedigrees primarily through only two sources: Gulch and Silver Deputy. He sired the second dam of the former and the third dam of the latter.

Asbury Mary was a winning daughter of Seven Corners. She produced Jameela (by Rambunctious), who posted a sterling record of 58-27-15-6 for earnings of $1,038,704, Jameela won 16 stakes, including three G1s.

Jameela produced only two foals, the full brothers Gulch and Big Mukora (both by Mr. Prospector). Gulch was 1988 champion sprinter and a pretty decent sire. But having produced no fillies, the female line of Jameela expired there.

Seven Corners also sired Proud Pied, whose race record reads 3-0-0-0 for earnings of $0. Proud Pied produced six foals but no stakes winners. She did become the second dam of stakes winner Silver Valley (by Mr. Prospector), the dam of Silver Deputy (by Deputy Minister).

Silver Valley produced a dozen foals, eight colts and four fillies (Silver Deputy was her first foal). But the female family of Silver Valley has not exactly flourished, aside from Silver Deputy.

So you still see Seven Corners in modern pedigrees, almost always in connection with either Gulch or Silver Deputy. I look at a lot of pedigrees, but I do not recall ever seeing one with Seven Corners NOT through either Asbury Mary or Proud Pied (Gulch or Silver Deputy).

Therefore, if you were examining sires in the fourth generation and including all of Gulch’s foals, Seven Corners looks pretty good. His record is virtually the same as Gulch’s. And that is pretty damn good for a sire in the fourth generation.

Similarly, if you were examining sires in the fifth generation and including all of Silver Deputy’s foals, Seven Corners looks pretty good. His record is virtually the same as Silver Deputy’s. And that is pretty damn good for a sire in the fifth generation.

What we have here is an example of statistics telling us one thing and common sense another thing. Statistics might claim that Seven Corners was actually a pretty good sire. He has survived into the present through two pretty good sires, Gulch and Silver Deputy.

So do you believe those statistics???? In this case, no. Or rather, you go back to the original statistics, those of Seven Corners himself as a sire: 106 foals, one stakes winner, AEI of 1.00. The latter statistics are much more credible than the former statistics.

The truth of the matter is that Seven Corners was an ordinary sire who happened to be in the right place at the right time. Yes, he has survived into the present through two pretty good sires, but he did so through sheer luck, not through any particular merit.

And that is the Seven Corners conundrum in a nutshell. When looking at statistics on sires four or five or more generations back, it behooves you to remember that you are looking at what has SURVIVED of those sires. And what has survived of those sires is usually the best of those sires. A lot of their dross has been winnowed out over the years and generations.

In the case of Seven Corners, what survived is Gulch and Silver Deputy and very little else. Seven Corners sired 106 foals. Two of them (Asbury Mary and Proud Pied) survived into modern pedigrees. The other 104 foals were dross and did not survive.

I make this elaborate point in connection to my previous post, “As Many Grains of Salt (Phalaris Revisited).” Look at the final comment to that post (by Pete) and my response to it. I have been beating my brains out trying to understand how Phalaris could have better results than his paternal grandson Nearco and whether that is indeed a legitimate result.

One of the reasons that happened is the Seven Corners condundrum. Phalaris is two generations further removed from modern pedigrees than Nearco is. That actually works to the advantage of Phalaris. It gives him two more generations with which to winnow out all his dross. So the numbers I presented in that post might need to be taken with an entire cake of salt, not just a few grains.

I am beginning to think that the whole approach of counting “presences” of a given sire (or dam) in modern pedigrees (no matter how far back) and trying to determine if “more is better” is simply fraught with too many unaccountable variables to yield legitimate results. Chronology is one variable I have discussed and tried to minimize already. Another is the topic of today’s discussion, the Seven Corners conundrum.

To restate that conundrum, you have to remember that only the best survive into modern pedigrees. A lot of dross has been lost (good riddance) along the way. It would be nice to know exactly how much dross a given sire (Phalaris or Nearco, for example) has shed along the way. The more it has shed, the less impressive the current results. The less it has shed, the more impressive the current results. But the precise amount shed appears to be unquantifiable at this point.

I sometimes think that you could pick just about any name that has survived into modern pedigrees and “prove” statistically that “more is better” with respect to that name. Because the deck is stacked in favor of a verdict of “more is better,” both through chronology and through the process of the dross being weeded out and only the best surviving.

Take Bull Lea, for example. Bull Lea was the leading sire five times and leading broodmare sire four times. So he was obviously a pretty good sire in his own time. But his male line failed miserably, and so most people today think of Bull Lea as a long-term “failure.”

This is short sighted for several reasons. The obsession with male lines leads to this kind of fuzzy (to put it charitably) thinking. Just because a horse fails miserably in the male line (almost all of them do) does not mean that horse is a “failure” as a sire.

Let us consider the sources of Bull Lea in modern pedigrees. Let us start with Storm Bird (Northern Dancer–South Ocean, by New Providence). New Providence was by Bull Page, by Bull Lea. Storm Bird is the sire of Storm Cat. So right there is a sizable chunk of modern pedigrees bearing Bull Lea through Storm Cat.

Bull Page was also the broodmare sire of Nijinsky II (and several other sire sons of Northern Dancer). Yet another sizable chunk of the population.

Roberto was by Hail to Reason–Bramalea, by Nashua. The second dam was Rarelea, by Bull Lea. Yet another sizable chunk of the population.

Alydar was by Raise a Native–Sweet Tooth, by On-and-On. Sweet Tooth was inbred 3×3 to Bull Lea. So Alydar has a double dose of Bull Lea. Yet another sizable chunk of the population.

So the four principal sources of Bull Lea in modern pedigrees are Storm Cat, Nijinsky II, Roberto, and Alydar. And there are other sources as well.

With those four sources, I feel pretty certain that Bull Lea would hold up well to a statistical analysis of the “more is better” variety. Because the deck is stacked in favor of “more is better.” And in Bull Lea’s case, because he lost a whole lot of dross in the process of surviving into modern pedigrees. Probably a lot more dross than Phalaris and/or Nearco lost, for example.

Many, many moons (more than 25 years) ago I remember working on the entire crop of foals of 1983. Bull Lea was one of 20+ sires I was examining in the fourth generation. The average for the entire crop was just above 3% stakes winners from foals. Bull Lea was around 2.5%. Pretty bad, in other words.

I finished the fourth generation and started on the fifth generation. Bull Lea was one of the 20+ sires I was examining in the fifth generation. I was thinking to myself, “Bull Lea was horrible in the fourth generation. He must be even worse in the fifth generation.”

Actually, Bull Lea was around 3.5% stakes winners from foals in the fifth generation. I was very surprised, to say the least. At the time I just chalked it up to chronology.

In retrospect, I see the Seven Corners conundrum at work. That extra generation gave Bull Lea more opportunity to get rid of his dross. His surviving results were much better in the fifth generation than in the fourth generation.

That seemed rather illogical at that time, but I have observed it over and over again as I have continued to examine populations of pedigrees over the years. X+1 generations are not always worse than X generations. Sometimes they are. Sometime they are not. All depends on the sire and the chronology and the amount of dross a sire has managed to shed.

I believe that Bull Lea is actually fairly typical of how “influence” actually waxes and wanes (mainly the latter). “Influence” actually decreases a lot more quickly than most people think. Sometimes it drops below 1.00 (the theoretical average) a lot more quickly than most people think.

But having dropped below 1.00, it can also rebound back up above 1.00, as Bull Lea did in the fourth and fifth generations for foals of 1983. But generally speaking, once it hits or drops below 1.00, it remains around 1.00 (with some fluctuations) no matter how many generations farther back you go.

I think Menow might be another good example. Both Menow and Bull Lea were foals of 1935. I think it might be interesting to compare the two statistically. But that is a project for another day and time.

The important thing to remember is that when you are looking at names in pedigrees several generations back or more, you are looking at what has survived of that name, and usually only the best survives. The dross is almost always winnowed out with time.

So the population of pedigrees containing that name is not the same now as it was originally. Which makes evaluating names in pedigrees more difficult than it should be, especially the farther back you go in pedigrees, and especially if you are testing the “more is better” hypothesis. As Heraclitus put it, “You could not step twice into the same river; for other waters are ever flowing on to you.”

I have been testing the “more is better” hypothesis for just about a year now. I feel like I have been barking up a tree without a coon. I feel like I have been hunting a snark. I feel like I have fallen down a rabbit hole.

I apologize to readers for having subjected them to this misadventure into Wonderland. But I think I did learn something from it. I learned how to explain the Seven Corners conundrum. I hope I have done so sufficiently (albeit not very succinctly).

Posted in phalaris | Tagged , , , , , , , , , , , , , , | 4 Comments

As Many Grains of Salt (Phalaris Revisited)

I promised I would go back to Phalaris and repeat the process using the restricted sample of 10,670 foals by sires born in 1998-2002 and out of mares born in 1998-2002. I will try to keep it short and sweet.

There are 32 ancestors in the fifth generation of a foal’s pedigree. I categorized them by how many of those ancestors traced to Phalaris in the male line. Since Phalaris was much more a “sire of sires” than a broodmare sire, that effectively captures the vast majority of his overall presences.

The average number of presences of Phalaris in this group was about 14. I did not find anything extraordinary at the upper or lower extremes of the distribution. So I divided the sample into only two groups, 0-15 (low) and 16+ (high).  Prices are listed below.

# of Presences           Foals           Average           Maverage              Price Index

0-15                             6,722          $33,862               132.14                     0.86

16+                              3,948           $38,093              141.54                      0.92

Not surprisingly, 16+ had slightly higher prices than 0-15 ($38,093 to $33,862 by average, 141.54 to 132.14 by maverage, and 0.92 to 0.86 by Price Index).

Listed next are the racetrack results for the two groups. APPPSW srands for average Performance Points per stakes winner, with the benchmark now being 696.

# of Presences            Foals          Stakes Winners          %          APPPSW          PPI (Result)

0-15                             6,722                    223                   3.32             786                    1.10

16+                               3,948                    183                   4.64             739                    1.45

The 16+ group was decidedly better by percentage of stakes winners from foals (4.64% to 3.32%). The 0-15 group was slightly better by APPPSW (786 to 739). The net result was that 16+ was much better overall by PPI (result, 1.45 to 1.10).

The table below summarizes the differences between prices and results.

# of Presences          Foals             Price Index          PPI (Result)            Difference

0-15                             6,722                 0.86                      1.10                        +0.24

16+                               3,948                 0.92                      1.45                        +0.53

Bear in mind that the overall difference for all 10,670 foals was +0.35 (0.88 to 1.23). The 0-15 group was slightly lower at +0.24. The 16+ group was higher at +0.53.

In my last post on Nearco I divided the sample into three groups. For purposes of comparison in the chart below I combined the lower two groups. The chart below (listing only two groups, 0-9 and 10+) for Nearco can now be compared to the chart above for Phalaris.

# of Presences          Foals           Price Index          PPI (Result)             Difference

0-9                              5,838                  0.87                      1.31                        +0.44

10+                             4,832                  0.89                      1.14                        +0.25

As you can see, Nearco had the opposite result from Phalaris. The lower group for Nearco (0-9) was actually better than the higher group (10+) +0.44 to +0.25.

Phalaris was the opposite. The higher group for Phalaris (16+) was better than the lower group for Phalaris (0-15) +0.53 to +0.24.

Just as a reminder, I used the smaller sample of 10,670 foals to minimize the effect of chronology on the results. Adjusted for chronology, it becomes clear that it is not so simple as the more Nearco, the better. A moderate amount of Nearco (7-9) was better than a low amount of Nearco (0-6) in my previous post, but a moderate amount of Nearco (7-9) was also better than a high amount of Nearco (10+).

So in the case of Nearco, too much of a good thing was not necessarily desirable. To put it somewhat more bluntly, Nearco fails the test for being a thoroughly positive influence on modern pedigrees when the sample is adjusted for chronology.

Not so with Phalaris. He still passes the test (adjusted for chronology) with flying colors.

How to interpret all this???? That is the rub. To do so I will have to indulge in some speculation. So take the following with as many grains of salt as you desire.

The most obvious and prevalent source of Nearco in modern pedigrees is Northern Dancer. The next most obvious and prevalent source is Nasrullah (many branches), Bold Ruler, . . . Seattle Slew, A.P. Indy, etc. I think what the numbers are trying to tell us is that those sources of Nearco are NOT the best ones in the long run.

Northern Dancer was a foal of 1961. By foals of 2017, it becomes easier and easier to build up a lot of Northern Dancer in the nether regions of pedigrees. If you are designing matings to do so (build up a lot of Nearco through Northern Dancer in distant regions of pedigrees), you might want to think again.

The most obvious and prevalent source of Phalaris NOT through Nearco is Native Dancer, Raise a Native, Mr. Prospector, etc. A distant second in that regard might be Tom Fool and Buckpasser. The rest of Phalaris (NOT through Nearco) is widely scattered and mainly European or South American (Fairway, Fair Trial, Petition, etc.).

I think what the numbers might be trying to tell us is that the sources of Phalaris NOT through Nearco are the best ones to use if you are trying to build up a lot of Phalaris in distant portions of pedigrees. And Mr. Prospector is by far the most prominent name in that group.

Take this speculation with as many grains of salt as you desire.


See the next post, The Seven Corners Conundrum, to see why a cake of salt might be more appropriate.

Posted in phalaris, Seattle Slew | Tagged , , , , , , , , , , , , | 5 Comments

Too Much of a Good Thing (Nearco)

I received a comment on a post I did on Phalaris a couple months ago. That comment was to the effect of now do the same thing with Nearco (paternal grandson of Phalaris and a superb sire in his own right). Not a bad idea, I responded.

I started to do so with sales foals of 2008-2111, first with 2008, then with 2111. Along the way I noticed that the percentage of foals with a surfeit of Nearco increased dramatically from 2008 to 2111. That got me to thinking that a chronological bias was at work that might be affecting the results on both Phalaris and Nearco.

Not many people dispute the fact that younger mares make better producers. As my last two posts attest, the same holds true for younger sires, at least with the particular population of sales foals of 2008-2111.

Younger mares and younger sires both have higher numbers of both Phalaris and Nearco in their pedigrees due to the simple fact that both these sires are constantly expanding their number of presences in modern pedigrees. That is the chronological bias of which I speak. Even if both Phalaris and Nearco have little effect on racetrack results in modern pedigrees, it will appear as if they do because of this constant expansion.

The best racehorses in this population generally are by younger sires and out of younger mares. Younger sires and younger mares generally have more presences of Phalaris and Nearco than the general population. Therefore, if you analyze the population by the number of presences of Phalaris and/or Nearco, it will appear that those two sires have a positive influence on racetrack results. Which may or may not be true.

I started thinking about ways to compensate for this chronological bias. My first inclination was to restrict the sample pool to foals by sires born in 2000 and out of mares born in 2000. I quickly realized that would yield a very small sample (less than a thousand). So I decided to expand the sample group to foals by sires born in 1998-2002 and out of mares born in 1998-2002. That helps to minimize the variation in number of presences of Nearco and/or Phalaris by age of parents.

Along the way I noticed that this sample group was a pretty good one in its own right (price of 0.88, result of 1.23), which reinforced my perception that younger sires were also good (along with younger mares) and led to my last two posts.

Today I will examine that 1998-2002 sample group with respect to number of presences of Nearco. I did so similarly to the way I did Phalaris. There are 32 ancestors in the fifth generation of a foal’s pedigree. I categorized them by how many of those ancestors traced to Nearco in the male line. Since Nearco (like Phalaris) was much more a “sire of sires” than a broodmare sire, that effectively captures the vast majority of his overall presences.

The average number of presences of Nearco in these 10,670 foals of the 1998-2002 sample group was about 9.5. The median and mode (number that appears most often) were both nine. Though I list the categories below as 0-6, 7-9, and 10+, I did not find any foals with zero Nearco. I think two was the lowest I found. Listed below are the prices for these categories.

Presences          Foals          Average          Maverage          Price Index

0-6                     1,177          $40,691              137.59                    0.89

7-9                     4,661          $35,834              133.15                    0.86

10+                    4,832          $35,443              137.51                    0.89

Totals              10,670          $36,194              135.62                    0.88

The overall average for all 10,670 foals was $36,194. The 0-6 group was highest at $40,691, but truthfully the differences by average among the three groups was very small. Ditto for maverages and Price Indexes (0.89 for 0-6 and 10+, 0.86 for 7-9, 0.88 overall).

Listed below are the racetrack results for the three categories. APPPSW stands for average Performance Points per stakes winner, the benchmark now being 696.

Presences          Foals          Stakes Winners          %          APPPSW          PPI (Result)

0-6                     1,177                     32                     2.72             766                    0.88

7-9                     4,661                   182                     3.91             857                    1.42

10+                    4,832                   192                     3.98             678                    1.14

Totals              10,670                  406                     3.81             765                    1.23

The 7-9 group was decidedly best at 1.42, followed by the 10+ group at 1.14 and the 0-6 group at 0.88. The relationship between prices and results is summarized in the chart below.

Presences          Foals          Price Index          PPI (Result)          Difference

0-6                      1,177               0.89                       0.88                     –0.01

7-9                       4,661              0.86                       1.42                      +0.56

10+                      4,832             0.89                        1.14                      +0.25

Totals                 10,670            0.88                        1.23                     +0.35

Remember, the 1998-2002 group was pretty awesome in its own right (difference of +0.35). The 7-9 group was even better at +0.56. The 10+ group was not quite as good as the overall group at +0.25. The 0-6 group was –0.01, not good at all.

If Nearco was truly still a positive influence on modern pedigrees, 10+ would be better than 7-9, and 7-9 would be better than 0-6 (the more Nearco, the better).

The 7-9 group was decidedly better than 0-6, but 7-9 was also better than 10+. So this is a mixed result. About all you can say is that a lack of Nearco (0-6) is not good. A surfeit of Nearco (10+) is better than a lack of Nearco. But a moderate amount of Nearco (7-9) is by far the best of all.

I will not go so far as to say this proves that Nearco is NOT a good influence on modern pedigrees, but evidently too much of a good thing does have its drawbacks.

Needless to say, now I will have to go back and do the same thing all over again with Phalaris (using the chronology-adjusted sample group of 10,670) to ascertain whether his results were really as good as they appeared to be in my previous two posts on that subject. As usual, I will keep you posted.

Posted in phalaris | Tagged , | Leave a comment

Attrition (Age of Sires)

I promised some details on prices and results by age of sires for sales foals of 2008-2o11. Some theoretical observations first though. In a theoretical world, youngest sires should have the lowest prices, and prices should increase as age increases.

Why???? Pretty simple actually. Because no matter how good a racehorse a sire was, he should not be considered a good sire until he proves it. Or rather, until his progeny prove it by winning races (especially the best races, stakes races).

Unfortunately, the real world is quite different from the theoretical one. The chart below summarizes prices by age of sires.

Age          Foals          Average          Maverage          Price Index

4-5           1,963         $63,369             193.88                   1.26

6-7           11,775        $35,359             140.69                   0.91

8-9           10,131        $30,963             120.95                  0.79

10-11         7,493        $42,915              145.73                  0.95

12-14         7,113         $55,681              167.51                  1.09

15-19         5,313        $64,284              179.29                  1.16

20+            1,774         $80,875             186.90                  1.21

If you ignore the two youngest groups (4-5 and 6-7), the prices do follow the theoretical model. The 8-9 group is the cheapest (average of $30,963, maverage of 120.95, and Price Index of 0.79). Prices increase uniformly from there, with the oldest group (20+) being by far the most expensive with an average of $80,875, a maverage of 186.90, and a Price Index of 1.21.

The first fly in the ointment is that the youngest group (4-5) is actually quite expensive, with an average of $63,369, a maverage of 193.88, and a Price Index of 1.26. The Price Index is actually higher than that of the 20+ group (1.21), though the average is lower ($63,369 compared to $80,875).

That might seem contradictory, but what it really indicates is that the 20+ group hits a lot of home runs (prices of $500,000+) but also strikes out a lot (prices of less than $10,000). The 4-5 group does not hit many home runs (lower average) but does not strike out much at all (higher maverage and Price Index).

The vast majority of foals in the 4-5 group were by first-crop sires. Many sires race through age three, go to stud at age four, and have their first foals at age five. A very few sires go to stud at age three and have their first foals at age four and their second crop of foals at age five. Those are the exceptions to the rule. For purposes of this survey, let us stipulate that the vast majority of the foals in the 4-5 group were by first-crop sires.

Why should first-crop sires have such high prices???? They are all unproven as sires. I am tempted to say that it makes no sense at all. It also has been suggested to me that pinhookers have a great deal of effect on these prices.

But before I make that statement it is a good idea to examine racetrack results as well, which are summarized in the chart below. APPPSW in the chart below stands for average Performance Points per stakes winner, with the benchmark now being 696.

Age          Foals          Stakes Winners          %          APPPSW          PPI (Result)

4-5           1,963                   95                     4.84             801                    1.65

6-7           11,775                 388                    3.30             676                    0.95

8-9           10,131                297                     2.93             719                    0.90

10-11         7,493                289                     3.86            698                    1.15

12-14         7,113                262                     3.68             655                   1.03

15-19         5,313               168                      3.16             715                   0.96

20+            1,774                  48                      2.71             663                   0.76

It is a good thing that I did not make that statement, because as it turns out, the 4-5 group more than justified its high prices with racetrack results. It had a Price Index of 1.26 and a PPI (result) of 1.65 (by far the best of any group in the chart above). Those 1,963 foals sold for prices about 26% above average and produced results about 65% above average.

This is very puzzling. Perhaps the only explanation is “fashion.” For some reason first-crop sires are considered the height of fashion in the crazy world of selling young, untested horses. First-crop sires receive the best mares they will ever receive (unless they really hit it big) in their first year at stud.

The second fly in the ointment is the 6-7 group. Prices drop drastically from the 4-5 group to the 6-7 group (average from $63,369 to $35,359, maverage from 193.88 to 140.69, and Price Index from 1.26 to 0.91). They drop even more from 6-7 to 8-9 (average from $35,359 to $30,963, maverage from 140.69 to 120.95, and Price Index from 0.91 to 0.79). Prices then increase uniformly from 8-9 through 20+, as previously noted.

What the heck is going on here???? The only explanation I can offer is the aforementioned “fashion.” First-crop sires are considered the height of fashion and receive the best mares they will ever receive (unless they really hit it big) in their first year at stud.

But when the next year comes along, all that is forgotten and the next crop of first-year sires is the height of fashion, and the original group of first-crop sires slides slowly (or maybe not so slowly) into market oblivion.

A lot of sires do have excellent first crops, then not much else afterward, for whatever reasons. A striking recent example is Uncle Mo. His first crop was two-year-olds of 2015 and included seven two-year-old stakes winners (and 20 overall). Included among them was 2015 champion two-year-old and 2016 Kentucky Derby winner Nyquist. Uncle Mo got off to a great start at stud.

His second crop (two-year-olds of 2016) has been an entirely different story, however. So far it has produced ZERO stakes winners. That pattern has been entirely uncommon for a lot of sires.

An opposite example was Secretariat. His first crop was no great shakes. But his second crop included General Assembly and Terlingua. That pattern is not as common as the first pattern though (great first crop, not so great afterward).

That first pattern partially accounts for the slide in prices from 4-5 to 6-7 to 8-9. Prices then turn up with the 10-11 group. By then the original group of first-crop sires have enough progeny on the racetrack to determine their true worth as sires. Some will  make the grade. Most will not. The former will have a lot fewer sales foals by age 10-11 or will have completely dropped out of the pool of sales sires, having been banished to some other (cheaper) locality.

It is the process of attrition. The truly good sires survive. Some even thrive and start receiving even better mares. Prices for their progeny perk up from the doldrums. But most sires do not survive this process of attrition, and their removal from the pool of sales sires causes prices to perk up for the remaining (surviving) sires around age 10-11.

So as much as I would like to criticize the adulation of first-crop sires, the numbers do not back me up. I cannot blame anyone for dealing with the world as it really is rather than as it ought to be. I cannot criticize sellers for favoring first-crop sires. Their progeny sell really well. I cannot criticize buyers for favoring first-crop sires. Their progeny perform really well (even better than their prices indicate they should).

But what about the prejudice AGAINST sires in the 6-7 and 8-9 groups???? Is it really justified???? In order to answer that and some other questions, the relationship between prices and results is summarized below.

Age          Foals          Price Index          PPI (Result)          Difference

4-5           1,963              1.26                        1.65                      +0.39

6-7           11,775             0.91                        0.95                     +0.04

8-9           10.131            0.79                         0.90                    +0.11

10-11         7,493            0.95                         1.15                    +0.20

12-14         7,113            1.09                          1.03                   –0.06

15-19         5,313            1.16                          0.96                  –0.20

20+            1,774            1.21                           0.76                  –0.45

As you can see, the 4-5 group has a difference of +0.39 (price of 1.26, result of 1.65). That is by far the best of any group. The 6-7 and 8-9 groups are actually not too bad relative to their prices. The former has a price of 0.91, a result of 0.95, and a difference of +0.04. The latter has a price of 0.79, a result of 0.90, and a difference of +0.11.

The 10-11 group is even better than 6-7 or 8-9. It has a price of 0.95, a result of 1.15, and a difference of +0.20. From there it is all downhill though. The 12-14 group has a price of 1.09, a result of 1.03, and a difference of –0.06. The 15-19 group has a price of 1.16, a result of 0.96, and a difference of –0.20. The 20+ group has particularly dismal results, partially due to Storm Cat and A.P. Indy, as I noted in my previous post. It has a price of 1.21, a result of 0.76, and a difference of –0.45.

So when it comes to age of sires, it is not quite as simple as younger is better. If you were to graph the differences above, it would be a wavy line, not nearly a straight one. Nevertheless, it does appear that older is not better, especially for the bottom three groups (12-14, 15-19, and 20+) in the chart above.

On the other hand, perhaps this is NOT a universal truth. It may have more to do with this particular marketplace and all its peculiarities (and pinhookers).

I started out this research trying to determine if a chronological bias exists and how strong it really is. Evidently it does exist and is pretty strong, strong enough to affect the results of examining a distant name in pedigrees and how often it occurs, a name such as Phalaris, La Troienne, or Nearco. The last of those three will be the subject of my next post.

Posted in La Troienne, phalaris | Tagged , , , , | Leave a comment