TrueNicks : Products, Validity

To Region or Not to Region

admin — Wed, 16 Apr 2008 18:04:00 GMT

We were recently forwarded an e-mail containing an advertisement from another pedigree consultancy that mentioned "... other pedigree information services that attempt to dazzle you with vast database resources without discriminating the information that matters from the information that doesn't ...." It also asserts that it chooses not to include the results of restricted stakes - even though that can include such valuable and prestigious races as the Canadian Queen's Plate or the Japanese Derby.

Without conjecturing whether or not this refers to TrueNicks, it is certainly the case that TrueNicks has "vast database resources." However, TrueNicks - developed by industry professionals with extensive international experience in all aspects of racing and breeding - certainly knows how to sort and provide the information that matters.

In any case, the comments do raise some interesting points. One is that the fundamental difference between the TrueNicks database - that of The Jockey Club Information Services - and others is not so much that it is vast, but that it is thorough. It is this comprehensive database that allows TrueNicks to take into account all starters and foals bred on a cross, thus reflecting true - as opposed to hypothetical - opportunity. (As a bonus, it also means that TrueNicks ratings are always calculated using the most current data possible.)

Other nicking services work from a selective database that includes some stakes winners, from some countries, over some time period - and have devised convoluted arguments to explain away shortcomings in their data.

In fact, they strain to make a virtue out of a vice, and suggest that examining hypothetical opportunity among a select class of runners is going to yield more accurate results than comparing real achievement with real opportunity across all known foals and starters. Of course, this is patently false. The email cited earlier also alleges that

"... patterns of successful breeding are evident only in runners of genuine quality"
"... certain methods of inbreeding, sire-line crosses, or combinations of ancestors that never yield a listed stakes winner may often show up among the winners of restricted stakes"

The first assertion is patently wrong: patterns of successful breeding turn up in good and bad runners alike (as do less successful ones), and it is only when looking at all foals that we can determine how successful a cross or pattern is relative to opportunity. TrueNicks factors in both frequency of attempts and quality of material involved.

The second contention needs to be looked at in two parts. Firstly, let's consider the assumption that the unrestricted races included in our competitors' compilations of stakes winners are always going to be superior to restricted stakes winners. Upon even a surface examination, their criteria produce some strange anomalies.

For example, a horse would not qualify for their calculations by winning the Japanese Derby (a restricted race) - but a Japanese gr. III winner, who was soundly defeated in the same race, would qualify. It is also strains credulity to believe that all stakes winners in countries such as New Zealand, Italy (including Sicily), and Germany, or all gr. III winners in Chile or Peru, would be superior to the winners of, say, the Canadian classics. Even looking at Australia, where racing is of a generally higher standard, the winner of a race such as the Tasmanian Derby at Hobart (a group event), would in most cases struggle to get anywhere near the winner of the Canadian restricted turf classic, the Breeders' Stakes.

Similar anomalies apply when we consider racing in the U.S. There are restricted stakes at major tracks where the winning performances are clearly superior to open company stakes events at some smaller tracks. With these inconsistencies in mind, it should be clear why TrueNicks chooses to follow the industry standard and considers all winners of black type events as determined by the International Cataloguing Standards.

It's true that restricted and regional stakes winners are drawn from a wider gene pool, and as result sometimes have inbreeding via ancestors different from those found in more commercial pedigrees. However, in our experience of studying Thoroughbred pedigrees - which now extends back more than 35 years - nicks, crosses, or pedigree patterns that consistently work at the highest level tend to be beneficial further down the tree. For example, Giant's Causeway has done very well with Mr. Prospector-line mares, so we might expect that his brothers Freud and Roar of the Tiger (TrueNicks, SRO) (who stand in New York and Florida, respectively), would benefit from being bred to Mr. Prospector-line mares in those regions. The reverse can also be true: a nick that establishes itself in a regional program, and with modest stock, may well be worth recreating with higher-quality individuals.

Moreover, the sheer quality of individuals found at the higher levels can often compensate for lower degrees of pedigree affinity, a fact that highlights a major flaw in selectively-compiled databases.

For those who argue that it's problematic to compare highly-commercial sons of a particular stallion with less-renowned sons of the same sire, and to help breeders determine the level at which a nick has success, the TrueNicks page has a unique feature - a list of the top five horses bred on the given nick.

Last but not least, omitting restricted stakes winners severely limits competing products' suitability to be relevant to regional breeders. Unlike TrueNicks, they are not able to accurately identify and report specific regionally-successful nicks .

TrueNicks, with its comprehensive database, is the new industry standard in nick ratings.

As Easy as A, B, C ?

admin — Tue, 29 Jan 2008 02:01:00 GMT

Stallion managers and broodmare owners have offered some interesting insights since the release of TrueNicks.

One stallion manager told me that he would come across clients who would reject matings that rated A on another nicking product because they wanted A++. In an e-mail conversation regarding a prospective mating, a broodmare owner told me that a planned mating that had a very clever pedigree had a "disappointing TrueNick rating." When, I looked, however, that rating was a B, with a variant of 1.60 - that is, 60% better than might have been expected by opportunity.

It's clear from interactions such as these that there is often some misunderstanding about what a nick rating is and isn't. TrueNicks is an expression of one specific aspect of the pedigree, that of the sire line/broodmare sire line cross. From our studies in developing TrueNicks, looking at over 100,000 horses, it is clear that there is a strong correlation between the sire line affinities and success, as expressed by wins in stakes events.

That said, there are a number of reasons not to blindly follow a rating without due consideration of class, conformation, and aptitude. While the rating is a wonderful tool to apply when trying to determine primus inter pares - first among equals - common sense must be used when considering the class of the pedigree.

For example, all other factors being similar, a yearling by Storm Cat, Distorted Humor (TrueNicks, SRO), or A.P. Indy, out of a grade I winning mare, with a TrueNicks rating of B has more chance of becoming a high-class runner than a yearling by a $5,000 regional stallion out of an inexpensive mare. When it comes to choosing between horses of similar class backgrounds, the system does reflect rather efficiently.

For example, around 42% of Distorted Humor's foals are TrueNicks rated A or better, and 62% are rated B or better. Looking at his stakes winners, however, we find that 74% are A or better (a 76% improvement on opportunity) and 85% are B or better (a 37% improvement on opportunity).

Mention of the "B or better" group for Distorted Humor underlines something else. That is, while very high variants are appealing, they are often the result of some good results generated from a relative small number of attempts. So, while A++ ratings are highly positive, A+, A, B+ and even B ratings represent a significant improvement over opportunity. What also has to be taken into account, from the pedigree perspective, is the presence of positive inbreeding and linebreeding in the pedigree (the interpretation of which is admittedly a somewhat subjective exercise). For example, the mating that our breeder friend described as giving a "disappointing" rating (actually a B), gave an intriguing inbreeding to the excellent mare Wild Applause, 3 x 3 through three-parts brother and sister.

In our book, a B with a powerful inbreeding pattern is a very positive mating. TrueNicks was calibrated to show more C and D - or even F - results than A's and B's, so there's never a question of a B being anything other than an above-average rating.

The TrueNicks program provides additional information to give even greater flexibility in interpreting the rating, as the TrueNicks page for a named horse or hypothetical mating is the only one to give the top five horses bred on a cross. This sometimes reveals a good individual bred on a sire line/broodmare sire line cross that has generally provided modest results. Exploration of that individual might determine that it is worth taking the opportunity to emulate a specific successful version of a cross that in general has provided disappointing results.

After all, we must never forget that at the end of the day, we are attempting to breed racehorses and not ratings.

Curiouser and Curiouser

admin — Fri, 18 Jan 2008 04:01:00 GMT

"Curiouser and curiouser" cried Alan. The anomaly of widely differing nick ratings on an identical cross is a very strange one.

When TrueNicks rates Unbridled's Song (TrueNicks, SRO) with a Miswaki mare, the result is a "D," based on the direct cross of Unbridled's Song with Mr. Prospector-line mares, as there is insufficient evidence (opportunity) to rate matings between Unbridled's Song and mares by Miswaki himself. The top five horses bred on the cross include Silver Lord, who became a first time stakes winner at the weekend.

When we consider Value Plus (TrueNicks, SRO)-another son of Unbridled's Song-with a Miswaki mare, the TrueNicks rating rises to a "C" as there are some sons of Unbridled's Song who have stakes winners on the cross, improving the stakes production rate relative to opportunity.

I looked at the same crosses as rated by a competing product, and ran a hypothetical pedigree for Value Plus with a Miswaki mare: the program returned the mating as an "A" rated nick (based on the Unbridled/Miswaki cross), albeit on the basis of just one stakes winner. It might be expected that the one stakes winner would be Silver Lord, who is by Unbridled's Song out of a Miswaki mare. This, however, is where things get very "Alice in Wonderland."

Looking at that system's list of stakes winners from this cross, the only one we see is Awfully Smart, who is by Anees (a son of Unbridled) out of a mare by Black Tie Affair (IRE) (by Miswaki). So this means, were one to enter Unbridled's Song with the dam of Silver Lord, one would presumably get the same "A" rating; would be told that the rating is based on one stakes winner (something which I would think is a pretty shaky premise for a rating)-but the one stakes winner wouldn't be Silver Lord, who in fact is the only stakes winner bred on the Unbridled's Song/Miswaki cross, but another horse, who is neither by Unbridled's Song nor out of a Miswaki mare!

I think the problem must be that the system is not updated anywhere near as frequently as is the TrueNicks program (which includes new stakes winners within a matter of hours), and so is "unaware" that there is actually a stakes winner bred on the Unbridled's Song/Miswaki cross. This might be regarded as a serious failing, as, if you are going to rate horses on the basis of one stakes winner, it would probably be better if that stakes winner were the horse actually bred on the cross!

"No Rating" Doesn't Equal "No Meaning"

admin — Mon, 14 Jan 2008 19:01:00 GMT

A reader wrote in to ask for clarification on the No Rating return after running a TrueNicks hypo-mating report for Niigon with a Richter Scale mare. Niigon is a son of Unbridled, by Fappiano, by Mr. Prospector. Richter Scale is a son of Habitony, by Habitat, by Sir Gaylord.

Question: I just purchased a TrueNicks hypothetical mating and received a report with No Rating. Hardly worth the $20. I presume this is because of insufficient data. Not much value here so why charge for it? By the way, another nicking product had this mating at A+. Why the difference? -Breeder in Canada

Answer:

Scot: Your question is an important one, and the answer involves the way nicks are calculated in different rating systems.

The TrueNicks No Rating score is actually a valid and meaningful return. It indicates that this cross has not yet had 15 runners or stakes winners from at least two different dams, even taking into account the third generation on the sire side and the fourth generation on the broodmare sire side. In effect, a No Rating return tells you that you're in uncharted territory. While the cross has not been proven or disproven, it is one that has had insufficient opportunity to determine a statistically well-founded evaluation-which is, in itself, useful to know.

TrueNicks avoided the practice of digging further back in the pedigree or "relaxing" the grading standards for less-common crosses, because such procedures devalue the calculation. While it might feel reassuring to see a letter grade, it really isn't much help to score such distant relatives as Mr. Prospector over Sir Gaylord, for example. (For what it's worth, the TrueNicks score of Fappiano over Doubledoor-effectively, the grandsire of Niigon over a full sister to Habitat, Richter Scale's grandsire-yields a "B," with a 1.99 variant score that is right at the cusp of "B+.")

In the particular cross of Niigon over a Richter Scale mare, we see some promise for future attempts at the broader cross. Your TrueNicks report shows the five best foals from this cross, including a grade III-placed mare and another stakes-placed mare-fairly impressive given limited opportunity.
Alan: Firstly, I would argue that the TrueNicks page actually gives more information than other nicking products, whilst essentially coming to a very similar conclusion.
Both reports give a five generation pedigree, with a dosage profile, the TrueNicks page also listing inbreeding.

We have the TrueNicks score as NO RATING. This means that-even when one goes back to the Fappiano/Habitat cross-that there is insufficient evidence on which to base a rating (specifically: there have been fewer than 15 starters and/or two stakes winners out of unique mares). The other nick product shows A+, but note that this is accompanied, in parentheses, by the notation (1 SW). This means that there has only been one stakes winner on the cross. Thus the other product rates this mating an A+ on the evidence of only one stakes winner. At this point, it is important to note that the other program does not know how many times this mating has been tried, so the A+ rating could be based on one stakes winner from one foal, 10 foals, or 100 foals-so, I would suggest it has little validity. It also doesn't tell us anything about the standard of the stakes winner or other foals bred on the cross.

When we come to the TrueNicks report page, however, we have the additional information which tells us that the one stakes winner bred on the cross is a graded stakes winner, and that there is also a stakes-placed horse bred on the cross. A little research also tells us that the third horse on the list has won almost $150,000 with a best BRIS Speed Figure of 94. We now know that the cross has produced three useful or better horses from a maximum of 14 attempts.

With this knowledge, if I liked other aspects of the mating, I may well consider it worth trying.

So, I would suggest that the TrueNicks page actually provides all of the information that other nicking products provide and more-other than a rating which is based on only one stakes winner.

Marking a Mark

admin — Sun, 06 Jan 2008 22:01:00 GMT

Over time TrueNicks ratings do change, but it is how they change that makes it an interesting study.

Affinities can establish themselves quickly but the TrueNicks system requires certain thresholds to be met for the cross to be considered valid. This means that while it can take some time for a nick to establish itself, the actual success of the nick can in turn weigh down or depress a score. A good example is that of Unbridled’s Song (TrueNicks, SRO) and Storm Cat, a combination that has combined very successfully with the likes of Magnificent Song, Buddha and Half Ours. While this nick has done very well and established itself as an “A” rating nick, it has fuelled an extraordinary number of Storm Cat mares visiting Unbridled’s Song.

As the TrueNicks calculation considers all the live foals bred on the nick–failures and successes–it would not be hard to see that, if a stallion has initial success with a certain broodmare sire and in turn meets a remarkable number of mares by that broodmare sire, the score could become depressed when these resultant foals become considered foals of racing age and form part of the calculation.

View Report

We have, however, implemented some logic within the program to account for a successful nick being attempted an exceptional number of times (there are for example some 30 mares about to foal down Unbridled’s Song foals this coming spring alone), giving appropriate weight to the fact that there are the number of foals bred on the nick, but also giving them the appropriate opportunity to perform on the track to prove or disprove the efficacy of the nick.

A really top class nick should continue to outperform, however, and the Unbridled’s Song/Storm Cat nick got a new stakes winner on the weekend in the form of the Mr Prospector Stakes (gr. III) winner Noonmark so it has again proven its worth.
View the TrueNicks report: Noonmark.pdf