The Chronicle of the Horse
MagazineNewsHorse SportsHorse CareCOTH StoreVoicesThe Chronicle UntackedDirectoriesMarketplaceDates & Results
 
Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 51
  1. #21
    Join Date
    Apr. 30, 2009
    Location
    Canada
    Posts
    2,949

    Default

    If you look at the averages this year, there was some talented horses so the averages were high. The horse in question had scores lower than the average by a bit. The bell is suppose to adjust for testing and scoring differences between test. So if the average is high then the curve assumes that the scores are scored high as opposed to having a group of overachievers. The further your scores are from the average, the bigger hit you take. That same horse might get significantly better final marks if the averages would have been lower. There would also be play between the disciplines. So if you are weak in dressage but above average in jumping but were in a class with outstanding jumpers and they are fairly good movers too, you are going to take a bit hit in final marks. The good thing is that it rewards overall athleticism more than a specialist. That is why you want a large class, the horses will be more typical of a range of abilities. This year many of the averages were 8's. On the other side, it would be hard for horses to get the big marks. A test with lower averages would result in a larger final score for some of these horses. In other tests some of these horses would have got at least 130 for jumping scores. Again, the raw scores are more informative.



  2. #22
    Join Date
    Mar. 28, 2006
    Location
    Oregon
    Posts
    3,373

    Default

    It makes no sense to me to score horses against a bell curve and not against a set standard. Why do you penalize SOME stallion owner every year? How does that make fair approvals? I'm not saying any stallion does or doesn't deserve approval, but to the outside observer, it doesn't make a lick of sense...
    Family Partners Welsh Ponies - Home of Section B Welsh stallion *Wedderlie Mardi Gras LOM/AOE http://www.welshponies.com
    Click here to buy: A Guide To In Hand Showing of Your Welsh Pony


    4 members found this post helpful.

  3. #23
    Join Date
    Nov. 5, 2000
    Posts
    9,466

    Default

    If you look at the raw scores of the horse at the bottom of the pack, he fell way below the other stallions in willingness to work, ability to work, walk, trot, stadium jumping, cross-country gallop, cross-country jumping, etc. And even his canter was below the average. Plus, it looks as though he was scored fairly consistently by the various "judges" - training judge, test judges (2), and test riders (4). So - without having been there and seeing the horse go - it would appear that the combination of very sub-par jumping ability and sub-par gait scores were what put him out of contention for a passing score. IOW, if the horse can't or won't jump nicely, and he doesn't have good gaits, then does he have the qualities needed to make a good sire of sport horses?



  4. #24
    Join Date
    Nov. 5, 2000
    Posts
    9,466

    Default

    Also - I haven't been able to find the names on the website of the test judges or the guest dressage riders. Does anyone know who they were?



  5. #25
    Join Date
    Dec. 14, 2007
    Location
    Wilsonville, Ontario, CANADA
    Posts
    4,374

    Default

    It makes no sense to me to score horses against a bell curve and not against a set standard. Why do you penalize SOME stallion owner every year? How does that make fair approvals? I'm not saying any stallion does or doesn't deserve approval, but to the outside observer, it doesn't make a lick of sense...
    This ^ exactly. Thats the part I am having trouble "getting" and accepting


    1 members found this post helpful.

  6. #26
    Join Date
    May. 20, 2010
    Posts
    100

    Default

    Regarding the bell curve thing, the reason I would see is that, even if you try as hard as possible to compare all horses to a standard, you end up comparing each horse with the ones you judged just before.

    Having been involved in some testings, you will always note that, depending on the panel of judges and even sometimes their mood, the scores for a same horse would vary from one day to another. For exemple, on a good day, the first horse could gets 7 all across the board, and these marks would remain the benchmark for the rest of the day. If the judges have been to generous for that first horse, the rest of horses will be marked higher as they cannot score them lower then that first one. In testing process where their is no bell curve theory applied, I saw days where 11 out of 13 mares judged where scored as premiums where , in fact, the quality overall was average. By applying the bell curve marks, you would have the opportunity to put these high marks in perspective and make the corrections to more realistics results.

    Of course the impact of those higher scores are less important in 70 days tests as the judges may have several days to correct the scores but I think it can explain the whole question of judging by the bell curve.



  7. #27
    Join Date
    Feb. 20, 2006
    Posts
    177

    Default

    Have no despair! The upcoming blog posts that we are preparing (and translating) go into detail explaining all of the scores giving during the Rideability/Gaits judging days by judge riders and the scores given by the judges on the ground. Here is a small excerpt explaining what is judged by the character score:

    "Character
    Amiability and interaction with humans, obedience, acclimation to testing facility, Posture (such as defensive reaction, ear positioning, and behavior while brushing, saddling, bridling, mounting, or dismounting)

    Unacceptable: Turning away and pinning the ears, resistance when touched, threatening to hurt you, lifting hind legs to threaten kicking, biting, kicking or rearing"

    The judges were Georg Van den Boom (Germany) and Dr. Antonius Bornemann (Germany). Guest judge was Paul Gummelt. Rideability/Gaits judge riders were Jessica Wisdom (USA) and Kim Pfieffer (Germany). We left the names off the website until after the final days to prevent allegations that riders/judges could be bribed in advance.


    1 members found this post helpful.

  8. #28
    Join Date
    Oct. 4, 2003
    Location
    Oklahoma
    Posts
    7,412

    Default

    It defeats the purpose of having a test, if all stallions "pass" the test. It IS a test. The sole purpose of the bell curve is to help standardize the testing from one test to the next. The bell curve does not change how the stallions place in the test and it does not make them fail. The lowest scoring stallion will always be the lowest scoring stallion. If there was no bell curve, instead of having just one or two stallions fail to meet the minimum licensing requirements, four or five could "fail". The standards set by the registries for acceptable scores are based on the test scores calculated with this formula. If they were accepting raw scores, I can assure you that their minimum accepted score would change to eliminate the bottom scoring stallion(s).

    From the testing stand point, stallions do not "fail". The test gives the stallions an index score and it is up to the registries to determine their minimum accepted score for licensing.
    Silver Creek Farms - home of Apiro & Validation
    Visit us on facebook!



  9. #29
    Join Date
    Aug. 15, 2008
    Location
    Vermont
    Posts
    126

    Default

    IMO the objective for the stallion testing is to approve only exceptional individuals for breeding. Stallions with a sub par performance should fail.
    If mediocrity is what you want then you might as well do away with the testing all together.


    5 members found this post helpful.

  10. #30
    Join Date
    Aug. 15, 2010
    Posts
    1,659

    Default

    Quote Originally Posted by rideagoldenpony View Post
    It makes no sense to me to score horses against a bell curve and not against a set standard. Why do you penalize SOME stallion owner every year? How does that make fair approvals? I'm not saying any stallion does or doesn't deserve approval, but to the outside observer, it doesn't make a lick of sense...
    I agree! I think a "fixed" standard makes more sense. In all reality, most years some stallions won't pass, there are only a few extraordinary horses that should earn this honor. But what if you have a year with 10 extraordinary horses - shouldn't they all deserve to "pass"? ANd what if you have a bad year, and there are 10 not-so-extraordinary horses - the bell curve says some of those still have to pass, even if the "top score" is still less quality then the lowest scoring horse from the year before? Yes, I know this is an unlikely scenario, but there will be years when you have a really high quality group of horses - and bell curve can "damage" that kind of group.


    5 members found this post helpful.

  11. #31
    Join Date
    Mar. 5, 2007
    Posts
    1,704

    Default

    Quote Originally Posted by showjumpers66 View Post
    It defeats the purpose of having a test, if all stallions "pass" the test. It IS a test. The sole purpose of the bell curve is to help standardize the testing from one test to the next. The bell curve does not change how the stallions place in the test and it does not make them fail. The lowest scoring stallion will always be the lowest scoring stallion. If there was no bell curve, instead of having just one or two stallions fail to meet the minimum licensing requirements, four or five could "fail". The standards set by the registries for acceptable scores are based on the test scores calculated with this formula. If they were accepting raw scores, I can assure you that their minimum accepted score would change to eliminate the bottom scoring stallion(s).

    From the testing stand point, stallions do not "fail". The test gives the stallions an index score and it is up to the registries to determine their minimum accepted score for licensing.
    I don't see how it standardizes anything.
    It depends more on the quality of horses in the test than it does on a judging on an individual horse against standard requirements.
    And there is nothing wrong if more stallions failed the test. So what if all the horses in a particular test 'pass' or 'fail' based on the individual registry requirements.
    Too much luck of the draw. An exceptional horse who had the bad luck to compete in a year that all the horses were exceptional could fail and the reverse is true. A poor quality horse in a below average group could pass.

    Makes no sense.


    2 members found this post helpful.

  12. #32
    Join Date
    Nov. 28, 2003
    Location
    MO
    Posts
    4,574

    Default

    Quote Originally Posted by showjumpers66 View Post
    From the testing stand point, stallions do not "fail". The test gives the stallions an index score and it is up to the registries to determine their minimum accepted score for licensing.
    This is the key. It is up to the registries to decide what score they will accept. In the past, certain registries have used their discretion to accept the lowest scoring stallion, so it can happen.
    I know I keep saying it, but this is why you should go watch in person, if at all possible! This was my first year attending, and it was an eye opening experience. I think we would have less of these "it's not fair discussions" if more people could go and see how some of these stallions perform in person.

    The testing should let "the cream rise to the top" and I would think all breeders here would want a very stringent testing, so that we can breed with the best, vs. "just good enough".
    Courage is going from failure to failure without losing enthusiasm."
    --Winston Churchill
    https://www.facebook.com/pages/Hills...h/112931293227
    www.HillsideHRanch.com


    1 members found this post helpful.

  13. #33
    Join Date
    Jan. 29, 2000
    Location
    Brownsburg, VA
    Posts
    2,966

    Default

    I believe in the bell curve. We are supposed to be improving with each generation, not to maintain the status quo. Grading on a bell curve accommodates that generational improvement.
    "No matter how cynical I get its just not enough to keep up." Lily Tomlin


    1 members found this post helpful.

  14. #34
    Join Date
    Apr. 30, 2009
    Location
    Canada
    Posts
    2,949

    Default

    Quote Originally Posted by showjumpers66 View Post
    It defeats the purpose of having a test, if all stallions "pass" the test. It IS a test. The sole purpose of the bell curve is to help standardize the testing from one test to the next. The bell curve does not change how the stallions place in the test and it does not make them fail. The lowest scoring stallion will always be the lowest scoring stallion. If there was no bell curve, instead of having just one or two stallions fail to meet the minimum licensing requirements, four or five could "fail". The standards set by the registries for acceptable scores are based on the test scores calculated with this formula. If they were accepting raw scores, I can assure you that their minimum accepted score would change to eliminate the bottom scoring stallion(s).

    From the testing stand point, stallions do not "fail". The test gives the stallions an index score and it is up to the registries to determine their minimum accepted score for licensing.
    I agree that it is suppose to standardize the test BUT that is a function of variation and test size. It won't work on smaller populations. And I am not sure that when you have such a small population that the test has the purpose of having some not pass. A large group, yes. Not saying the lowest horse in this test should have passed but in other years where the test was smaller, a horse could have failed but if they would have taken the test in Germany, they would have passed with average marks and been given a chance in the breeding shed. Picking the best 3yr old doesn't have a high correlation to the best sire. It is one step along the way. I guess that is the heart of the issue, that you could possible take a horse out of the breeding population that may have been a valid contributor. History shows that many decent sires had average results in the test. The order may remain the same but it is the elimination that is the issue. And the order is not that important once the horse is breeding. Plus the idea that same stallion, in a larger test, would have been given that chance in Europe, where there are more sires than we have here.
    I also agree with Cumano that judges are human and prone to bias. But with a 70 day test and smaller groups, that possible bias should not be weighted as heavily. A mare inspection is one day. The interesting thing is that most people can see the top horses and the bottom, it is the middle placing that becomes hard to judge.
    This year, one horse got some 10's for jumping. Now either that score is too high or the horse genuinely earned that score. If he did, it should not be diminished by the bell curve. Which is why I think Silver Creek did a really good thing putting out the raw scores. And none of this is directed towards their efforts, which were excellent. And even them taking the brunt of the criticism of the math which they have no control over. It is a critic of the methodology. And I don't think the baby should be thrown out with the bathwater, as testing is a valuable tool and needs to be available in NA. Someone should rethink how the numbers are used, (well apparently they have but the changes are not out yet).



  15. #35
    Join Date
    Nov. 28, 2003
    Location
    MO
    Posts
    4,574

    Default

    Testing does not, in and of itself, take any stallion out of the gene pool. You can still pursue approvals through performance, or you can forgo being "approved" at all. There are plenty of people breeding to stallions that aren't approved in any registry, and their offspring only get COPs.

    I just don't know why we have to re-hash this every year, but I suppose it is human nature. We need this testing as it is recognized by European stud-books and for those of us who desire those papers for our horses this is the system that we have.
    Courage is going from failure to failure without losing enthusiasm."
    --Winston Churchill
    https://www.facebook.com/pages/Hills...h/112931293227
    www.HillsideHRanch.com


    3 members found this post helpful.

  16. #36
    Join Date
    Aug. 19, 2010
    Location
    Ottawa, Ontario
    Posts
    369

    Default

    Quote Originally Posted by S A McKee View Post
    I don't see how it standardizes anything.
    It depends more on the quality of horses in the test than it does on a judging on an individual horse against standard requirements.
    And there is nothing wrong if more stallions failed the test. So what if all the horses in a particular test 'pass' or 'fail' based on the individual registry requirements.
    Too much luck of the draw. An exceptional horse who had the bad luck to compete in a year that all the horses were exceptional could fail and the reverse is true. A poor quality horse in a below average group could pass.

    Makes no sense.
    Stats nerd interjecting here...In order for the standardization of the testing to be fair (not arguing whether or not it is, just stating facts), it has to be operating on the assumption that the sample (the stallions at this particular testing) is representative of the population (all warmblood sport-type stallions). If this is true, the standardized scoring should be applicable.

    Typically, there is a minimum sample size necessary in order for the distribution of the sample to be considered normal (and mirror that of the population, with the sample mean equaling the mean of the population). This principle is known as the central limit theorum. I'm not sure what this number is for this particular testing; I believe there was some debate about that last year (for psychological statistical testing, an n of 30 is the minimum to assume a normal distribution). Without the minimum number, the standardized scores cannot be applied, as the distribution will not necessarily be normal and the mean of the sample will not necessarily equal that of the population. This would mean that the comparison between the sample and population would not be valid (you would essentially be comparing apples to oranges).

    Therefore, if the minimum sample size criterion is met, the distribution of the sample can be considered normal and standardized scoring can be applied, AND an increase in sample size will not significantly affect the distribution. Obviously, the more the better, as the larger your sample, the more likely it will represent the population at large. However, the argument that the much greater numbers of stallions at stallion tests in Europe will produce highly discrepant results when the standardized scores are applied compared with tests with smaller numbers *should* be meaningless. According to this principle of statistics, if the minimum sample size criterion is met (let's pretend it's 30), it should make no difference whether there are 31 or 301 stallions in the testing: they are both being compared using the same normal distribution with the same mean (average) that equals that of the population with which the samples are being compared.



  17. #37
    Join Date
    Apr. 30, 2009
    Location
    Canada
    Posts
    2,949

    Default

    Quote Originally Posted by HickoryHill View Post
    IMO the objective for the stallion testing is to approve only exceptional individuals for breeding. Stallions with a sub par performance should fail.
    If mediocrity is what you want then you might as well do away with the testing all together.
    That is assuming that the test is a definitive evaluation of the stallions potential for a sire. This has been shown to not be true, in fact most of the best stallions did not finish in the top places. If the test was only to determine the best 3 year old than yes.

    I believe in the bell curve.
    I also belive in using normal distribution....correctly. As Tradewind pointed out, you need to know the minimal sample size that represents the population of tested stallions.
    And it gets talked about every year because it is important to the test. It is not a law written in stone, the people that govern it can change the process. It is a simple calculation. But if no one speaks up, things do not get better. Or do you think we should not improve the system? Or do you just trust that people in authority aways have the right answer. To people that use stats, this is a glaring issue.



  18. #38
    Join Date
    Oct. 4, 2003
    Location
    Oklahoma
    Posts
    7,412

    Default

    According to the person in Germany who created this formula, 5 stallions is the minimum number to calculate results and 12 is the minimum number for the results to have value and be accurate.
    Silver Creek Farms - home of Apiro & Validation
    Visit us on facebook!



  19. #39
    Join Date
    Nov. 5, 2000
    Posts
    9,466

    Default

    Quote Originally Posted by SilverCreek View Post
    The judges were Georg Van den Boom (Germany) and Dr. Antonius Bornemann (Germany). Guest judge was Paul Gummelt. Rideability/Gaits judge riders were Jessica Wisdom (USA) and Kim Pfieffer (Germany). We left the names off the website until after the final days to prevent allegations that riders/judges could be bribed in advance.
    Thanks, Summer! And totally understandable for them to not be on the website beforehand!



  20. #40
    Join Date
    Nov. 5, 2000
    Posts
    9,466

    Default

    Quote Originally Posted by stoicfish View Post
    That is assuming that the test is a definitive evaluation of the stallions potential for a sire. This has been shown to not be true, in fact most of the best stallions did not finish in the top places.
    No one is saying it is a definitive evaluation, but it is an indication whether a stallion possesses traits desired in a sire of sporthorses, or whether he does not possess those traits.

    I would argue that a stallion whose raw scores in most categories are WELL BELOW those of his peers, is probably NOT a good candidate for being a sporthorse sire. IOW, how can he reasonably be expected to pass on desireable sporthorse traits if he himself doesn't possess those traits?


    2 members found this post helpful.

Similar Threads

  1. 2012 North American Stallion Test Blog
    By SilverCreek in forum Sport Horse Breeding
    Replies: 86
    Last Post: Nov. 15, 2012, 05:23 PM
  2. Replies: 148
    Last Post: Nov. 15, 2012, 08:24 AM
  3. Rolex 2012--Official thread
    By vineyridge in forum Eventing
    Replies: 137
    Last Post: Apr. 26, 2012, 11:43 PM
  4. Stallion Test Results?
    By PippiLongStockings in forum Sport Horse Breeding
    Replies: 11
    Last Post: Feb. 12, 2011, 03:10 PM
  5. Stallion Test Results
    By Kyzteke in forum Sport Horse Breeding
    Replies: 126
    Last Post: Dec. 31, 2009, 10:13 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •