Combined body length in selective breeding thoroughbreds.
Moderators: Roguelet, WaveMaster, madelyn, Diane
Combined body length in selective breeding thoroughbreds.
To increase the chances of breeding an elite racehorse, a genetic dominance tree (GDT) of the prospective dam, and a GDT of the prospective sire, are constructed. A GDT has indicia for determining whether the horse has any dominant and substantially constant representative physical characteristics. Theory by Cecil O. Seaman. Anyone read it? Crap or Probability?
-
xfactor fan
- Breeder's Cup Winner
- Posts: 2212
- Joined: Thu Sep 16, 2004 8:46 pm
-
xfactor fan
- Breeder's Cup Winner
- Posts: 2212
- Joined: Thu Sep 16, 2004 8:46 pm
After taking a look at the website and thinking about the services offered, here's my verdict.
What they are doing.
The measure your horse, in 15 different places, then boil that number down to a point on a graph. Couldn't find anything on the website about what they measure or how the final number is calculate.
This number is then plotted on a bell curve, assigned a number and a color code and you get a pedigree also color coded with all the ancestors of your horse with their very own color coding.
What they are trying to do is relate phenotype to genotype, in other words, what the horse looks like from the outside to what the horse is likely to pass on to the next generation. Not a bad idea in itself. But..there are some problems with the underlying assumptions behind this.
First the nature of bell curves. This is a very common way of looking at the range of any given characteristic in a population. For example a bell curve of human height would have jockeys at one end of the curve and NBA players at the other with the large part of human population (the top of the bell) falling between the two extremes.
Most of the horse population measured is going to fall into the middle section of the bell--red and orange--. So a pedigree is likely to show a large number of red and orange.
The samples up on the website include a successful horse (lots and red and orange) and a unsuccessful horse colors all over the map. This is supposed to correlate to racing performance in young horses, and breeding performance in retired horses.
They don't seem to make a clear connection between body type, and racing success. The website claims add 3,000 horses a year to the database, which sounds like a lot until you thinks about how many TB's are registered each year.
Another concern is that they have skewed the database by adding lots of successful racehorses, and few if any claimers. Going back the human height bell curve example, there is a positive relation between height and success in the NBA, in that there are very few 5' tall basketball players. Just as there are very few 7' tall jockeys.
I'd be interested in seeing the data divided into two bells curves, the bell curve of horse that have earned half a million, and horse that have earned less, and see how they overlap.
You could plot height and get the same type of results, few TB's under 14 hands, few over 17 hands and the greater number of the population falling between 15.3, and 16.3, and average size horses will have average size ancestors, and average size descendants.
Anybody actually used the service?
What they are doing.
The measure your horse, in 15 different places, then boil that number down to a point on a graph. Couldn't find anything on the website about what they measure or how the final number is calculate.
This number is then plotted on a bell curve, assigned a number and a color code and you get a pedigree also color coded with all the ancestors of your horse with their very own color coding.
What they are trying to do is relate phenotype to genotype, in other words, what the horse looks like from the outside to what the horse is likely to pass on to the next generation. Not a bad idea in itself. But..there are some problems with the underlying assumptions behind this.
First the nature of bell curves. This is a very common way of looking at the range of any given characteristic in a population. For example a bell curve of human height would have jockeys at one end of the curve and NBA players at the other with the large part of human population (the top of the bell) falling between the two extremes.
Most of the horse population measured is going to fall into the middle section of the bell--red and orange--. So a pedigree is likely to show a large number of red and orange.
The samples up on the website include a successful horse (lots and red and orange) and a unsuccessful horse colors all over the map. This is supposed to correlate to racing performance in young horses, and breeding performance in retired horses.
They don't seem to make a clear connection between body type, and racing success. The website claims add 3,000 horses a year to the database, which sounds like a lot until you thinks about how many TB's are registered each year.
Another concern is that they have skewed the database by adding lots of successful racehorses, and few if any claimers. Going back the human height bell curve example, there is a positive relation between height and success in the NBA, in that there are very few 5' tall basketball players. Just as there are very few 7' tall jockeys.
I'd be interested in seeing the data divided into two bells curves, the bell curve of horse that have earned half a million, and horse that have earned less, and see how they overlap.
You could plot height and get the same type of results, few TB's under 14 hands, few over 17 hands and the greater number of the population falling between 15.3, and 16.3, and average size horses will have average size ancestors, and average size descendants.
Anybody actually used the service?
xfactor fan wrote:After taking a look at the website and thinking about the services offered, here's my verdict.
What they are doing.
The measure your horse, in 15 different places, then boil that number down to a point on a graph. Couldn't find anything on the website about what they measure or how the final number is calculate.
This number is then plotted on a bell curve, assigned a number and a color code and you get a pedigree also color coded with all the ancestors of your horse with their very own color coding.
What they are trying to do is relate phenotype to genotype, in other words, what the horse looks like from the outside to what the horse is likely to pass on to the next generation. Not a bad idea in itself. But..there are some problems with the underlying assumptions behind this.
First the nature of bell curves. This is a very common way of looking at the range of any given characteristic in a population. For example a bell curve of human height would have jockeys at one end of the curve and NBA players at the other with the large part of human population (the top of the bell) falling between the two extremes.
Most of the horse population measured is going to fall into the middle section of the bell--red and orange--. So a pedigree is likely to show a large number of red and orange.
The samples up on the website include a successful horse (lots and red and orange) and a unsuccessful horse colors all over the map. This is supposed to correlate to racing performance in young horses, and breeding performance in retired horses.
They don't seem to make a clear connection between body type, and racing success. The website claims add 3,000 horses a year to the database, which sounds like a lot until you thinks about how many TB's are registered each year.
Another concern is that they have skewed the database by adding lots of successful racehorses, and few if any claimers. Going back the human height bell curve example, there is a positive relation between height and success in the NBA, in that there are very few 5' tall basketball players. Just as there are very few 7' tall jockeys.
I'd be interested in seeing the data divided into two bells curves, the bell curve of horse that have earned half a million, and horse that have earned less, and see how they overlap.
You could plot height and get the same type of results, few TB's under 14 hands, few over 17 hands and the greater number of the population falling between 15.3, and 16.3, and average size horses will have average size ancestors, and average size descendants.
Anybody actually used the service?
Too expensive for me, I'll stick to the SWAG method. It seems to work just as well.