Reader Q&A--Why Not "Sire and his Sons"?

re: Reader Q&A – Why Not "Sire and his Sons"?

russf — Thu, 26 Aug 2010 12:14:55 GMT

many thanks, Byron, for this very thorough and thoughtful response, it's greatly appreciated!

re: Reader Q&A – Why Not "Sire and his Sons"?

brogers — Tue, 24 Aug 2010 20:52:10 GMT

Russf,

Interesting comment. We may lose a couple of readers here as we go along discussing this, but I think that this is a very interesting line of thought, and certainly one worth further comment. Apologies in advance if you feel at any time that you have been misquoted and selectively referenced and feel free to do the same to me if I have missed a point that you are trying to make. Please also accept my apologies if the answer below is a little long.

The overriding question that you seek an answer to is if I “think that any rating bias exists when there is one prominent son of a sire who appears to behave very differently from the other sons.”

In our experience, borne out by the data we collected while creating TrueNicks, sirelines tend to operate in a similar way over several generations. To give an example, take a look at Fappiano/Unbridled/Unbridled’s Song with In Reality and his sons and grandsons, and you can appreciate that over time these two sirelines have generally worked well together. Right now if you mate a mare by Diktat (great-grandson of In Reality) to Zensational (great-grandson of Fappiano) it is an “A+.” Similarly, there are very few branches of In Reality that don’t rate well with sons of Unbridled's Song (grandson of Fappiano).

But just as Fappiano was a different type of Mr Prospector, and operated in some ways differently than most of that horses sons, so there are other sons of Fappiano who have created their own distinct branches. The Cryptoclearance branch, for instance doesn’t seem to work as well with In Reality, but that doesn’t mean he doesn’t share some common affinity with his general sireline. The few sons and grandsons of Cryptoclearance that are at stud have moved in a slightly different direction genetically, more towards another sireline that Unbridled’s Song has also shown affinity towards, for example Storm Bird/Storm Cat.

An example in reverse, where a son has shown a totally different affinity to his sireline, is Point Given. He doesn’t act at all like his sire or his grandsire. Neither Gulch nor Thunder Gulch have shown any particular affinity to Seattle Slew line mares, indeed they have underperformed with the Seattle Slew line mares to which they have been bred, but Point Given has been outstanding with the Seattle Slew line. He is an exception though, as other sons of Thunder Gulch have pretty much behaved like the sireline in general when crossed with Seattle Slew line mares. Our feeling on this is that there are some other phenotypic interactions taking place here with Point Given that make him a unique son of the sireline.

Subsets based on number of times tried can be difficult, because at times it is neccessary to go back beyond of the most relevant subset to gain sufficient numbers to generate a rating. Alan and I discussed and researched the possibility of stopping at selective distinct ancestors but we found it is horrendously difficult to consistently apply, especially given that some crosses seem to be affected by the phenotypes involved and some less so. In other instances, phenotype seems to be a key to genotype behavior (for example Point Given, who is more physically reminiscent of his broodmare sire, Turkoman, emulating that stallion be crossing well with Seattle Slew) in the same way a mixed breed dog tends to have behavior traits most similar to the parent it most physically resembles.

As far as a bias existing, we have also seen examples where there is compensating data. If you take an unproven son of Storm Cat and mate him to a generic son of Mr Prospector it returns a C+. This rating includes the data of a son of Storm Cat that has done really well with Mr Prospector in Giant's Causeway (A+) and one that has not done so well in Forestry (D). In the abscence of any data to the contrary and give a relevant rating, TrueNicks, and for that matter any pedigree rating system that is based on sireline/broodmare sireline affinity, works on the presumption that a stallion will generally appreciate the same bloodlines that his own sire has been successful with. In the main, this is the correct assumption to make but there are exceptions and when there are exceptions, because TrueNicks uses live data that is updated daily from foals, starters and stakes winners from around the world we are quickly able to identify those stallions that are not behaving like a “normal son” and create their ratings based on their own results. We believe that this is the best way to handle the stallions, and for that matter broodmare sires, that don’t behave like the rest of the horses in particular circumstances with the latter word being the key. Getting back to Forestry as an example, while he is deplorable with Mr Prospector line mares in general, he is good with Fappiano line mares and like a lot of sons of Storm Cat, great with Damascus, so in some ways he is only "different" in certain circumstances, a nuance that helps breeders and stallion managers alike make informed decisions.

Byron

re: Reader Q&A – Why Not "Sire and his Sons"?

russf — Tue, 24 Aug 2010 14:54:39 GMT

Thanks for that response, Byron.

That leads me to ask if you think that any rating bias exists when there is one prominent son of a sire who appears to behave very differently from the other sons. Consider the cases of Diesis/Sharpen Up, Distorted Humor/Forty Niner, and A P Indy/Seattle Slew.

If Diesis/Sadlers Wells gets a C, while Sharpen Up/Sadlers Wells earns an A, that seems to imply that Diesis is dragging down the otherwise favorable rating of his sire when matched with Sadlers Wells line mares. Do these different ratings imply that Sharpen Up line sires - who don’t come through Diesis – have a very different affect – a much better one - over Sadlers Wells line mares than do Diesis-line sons? As a result, it may not be appropriate to infer an affinity for Diesis and his sons from the broader base of Sharpen Up and his sons. The apparently superior results of Kris, Selkirk, et al, are masking the less than stellar results for Diesis, himself, and his sons and grandsons.

Similarly, can the reverse case be made for Distorted Humor/Forty Niner? Forty Niner-line sires - other than through Distorted Humor – don’t seem to work with Seattle Slew line mares. Consider that Funny Cide’s pedigree produces an A++ score (Distorted Humor/Seattle Slew), while Ide over Belle’s Good Cide gets a C+ (Forty Niner/Seattle Slew). Thus, despite the favorable results contributed by Distorted Humor with Slew-line mares, Forty Niner and all his sons and grandsons can only produce a C+ rating with Seattle Slew and his sons and grandsons. Those non-DH sons of Forty Niner must not work at all with the Slew line. Put another way, Forty Niner and his sons may consist of two substantially different subsets; one for Distorted Humor, and another consisting of all other sons and grandsons. – and the composite rating may not be accurate for any son of Forty Niner. Seattle Slew may behave similarly with A P Indy and his sons and grandsons in one camp, versus all other sons and grandsons of Seattle Slew in another.

My point is that it may not be appropriate, in certain cases, to use the broader ratings of a sire as a proxy for each of his sons. The published rating for Debussy is an example where the rating may be biased high because of this effect.

Does it seem logical to interpolate between True Nicks scores like this? I realize that the data can get pretty thin when we start to look at subsets of sire line/sire line crosses.