I would like to see a bit more of the Training level amateur owner–maybe someone who has not competed at Prelim or only up to Prelim. Erin Contino is probably the closest–with adviser Dr. Miller a similar voice. But I would like to see more input from “office manager who competes at Training level” or “retail associate who competes at Preliminary”–even without the association qualifications. But overall I like the list.
2 Likes
@Gardenhorse & @Thames_Pirate exactly my same thoughts. Glad I am not the only one thinking that. But I do think it is a reasonable mix, and includes some of the under represented areas.
regarding the new task MERs task force & EquiRatings…
In my opinion the new task force is better than what led to the previous situation. Hopefully, they will be more thoughtful and we will get a result which is helpful.
I do have a feeling that they already want EquiRatings to be the answer, and they will move in that direction regardless. This conforms with the notion of just having a set of metrics to classify the rider, and one need not think about the quality of the ride. And, one need not have a third party professional sign-off on the move up. It makes it “algorithmic” and takes human judgement out of the equation.
Many of the EquiRatings metrics are simply averages. It’s well known in statistics and performance modeling that averages can be a notoriously poor metric. Please note I am not trying to throw cold water on EquiRatings at all. I think it’s great to have more science and produce more knowledge. We do however, want to understand exactly what the metrics are and are not.
An example, the EquiRatings SJ6 metric… This is described as a “jumping penalty average” aimed at “benchmarking your performance in the show jumping phase”. It is calculated by averaging the rail down penalties over your last 6 rounds. Their example has penalties of 0, 4, 0, 0, 4, 8 in order. This produces a metric (SJ6) of 16/6 = 2.7 jumping penalties. That’s your SJ6 score. Easy enough. Right?
Well, consider the following 2 horse/rider combinations, each of which gets an identical SJ6 score of 2.7:
Horse/rider #1 had rounds with penalties: 0, 0, 0, 0, 8, 8
Horse/rider #2 had rounds with penalties: 12, 4, 0, 0, 0, 0
Both those pairs get SJ6 = 2.7
A mindless, “just look at the numbers approach”, says they are the same. But, clearly they are not.
Horse/Rider #1 in their last 2 outings had rails each time, where they used to be clear all the time. What’s up with them lately? Horse/rider #2 had a lot of rails a while ago, but is clean as a whistle in the last 4 outings.
Which combination is a better candidate to move up a level, assuming the SJ6 score of 2.7 is acceptable? Many would argue that horse/rider #2 is more interesting to move up as they are clean now. Most would also want to know what’s going on with horse/rider #1 lately? What problem has developed?
So… mindless application of averages when the distribution has large variance and/or the data set is small can yield counter intuitive results. Simplistically, put one foot in ice water and the other in boiling water, and “on average” you feel ok.
If I had a vote, I’d rather see a sign-off which also uses professional judgement, rather than only simple metrics
6 Likes
Agreed.
Also, a rider getting into the triple badly might have three rails down for one mistake–so their rating might be 0, 0, 0, 12, 0, 0 for an average of 2, but they had one bad distance that was compounded by it being at the triple. Meanwhile a rider who crashes through a fence vs. one whose horse tends to be a little sloppy occasionally and knocks a rail are two different things.
We’ve all seen or even been the riders who get deep to practically every fence, never get a rhythm, whose saintly ponies bunny hop everything, and who end up with minimal faults. We’ve also seen the riders who get their horses to the right distance every time only for the clever pony to know that he can bump a few (those horses are almost always rock stars XC). Now, I am not saying that horse should move up. I am simply saying that the latter is a better and safer combination, but the ratings might not show that, which is why we need subjective evaluations.
I would be interested, though, to see if there are patterns in the EquiRatings scores. I would also be interested to know how a letter score factors in. If I jump before XC and am eliminated on XC, do I include that jump round? Do I get to choose? If I have six events with clean SJ and then two where I fall in SJ, do those not count since they are letters, even though they indicate a problem has recently appeared in my showjumping that I should address before moving up? What about events that aren’t qualifying rides but have clean SJ?
EquiRatings are a great piece of information to have. Again, detecting patterns is critical to understanding and predicting. But sometimes we as riders know the backstory of a particular event–a thrown shoe, a decision to be cautious and circle in the rain as the footing gets slick, a saintly horse, a dog running through the arena, a stomach virus–that can affect averages but not indicate overall performance. My outlier best dressage score came on a day that was 104 degrees, and I rode at 3pm. I just wanted to get the heck off my horse. I decided that I would just get through it, didn’t care, and suddenly instead of a nervous nellie, I rode the way I am more than capable of riding at home and beat my personal best by at least 4 or 5 points. Yay for my dressage EquiRating!
Anyway, the point is the same one I have made several times–we NEED a system that involves some subjective approval by qualified individuals if we really want to make the sport safer.
5 Likes
Some great points @Thames_Pirate. My gut is that they are looking for a system that has no subjective component to protect themselves from lawsuits.
2 Likes
And the irony is that protecting from lawsuits actually does less to protect riders and horses. Hopefully they will at least consider the subjective.
As the girl with the horrible show nerves in SJ and D, I am the girl who never wins and rarely places (usually only when others screw up). But on XC I feel confident and have a good sense of pace, line, etc. What I need is miles–lots and lots of miles. For me, the MERs would not be the issue. But when I signed up for my first Prelim, I came off in a lesson (first time falling off that horse) and a few days later came off in the SJ at an event, my coach and I decided that my subconscious was too nervous to move up, and we called and bumped down my entry. The next event, what would have otherwise been my last T, was my best ever placing with a rock solid round in all three phases (leaving up SJ fences that coach had pulled, foot perfect and smooth rhythm in spite of crazy weather on XC). Clearly the issue wasn’t anything that would have been objectively discernible, and other than the one freak fall and the mediocre dressage, I had a pretty good record with that horse, including at a T3DE. It was only the subjective–the coach and I talking it over–that caused me NOT to move up and to decide to take the time. In the end life circumstances changed, and I ended up having to sell the horse and never moving up. I still hope to someday go Prelim, but only when I am truly ready.
4 Likes
Wish everyone had a coach like yours and thought about what they were doing and why as thoroughly as you do. Well done!
1 Like
I just wish I had been able to do at least one Prelim! But maybe that just won’t happen. I’d rather live to do 1000 more Trainings, though, so . . . .
1 Like
I don’t know that the committee’s driving force is merely to inoculate themselves against lawsuits. However, remember Ferdinand Porsche famously said, “committees are by nature timid, which is why no committee will ever be allowed to design a Porsche”.
Simplistic averages for metrics (such as EquiRatings), used as gates to control moving up a level, will have edge conditions which are nonsensical. I pointed out a few, as has @Thames_Pirate so nicely. Such metrics are inherently assuming a large sample set, a stationary statistical distribution, with no temporal (ie time-based) factors influencing the data. I would wager that none of those things are true in the problem set being discussed here.
@make_the_time, I love statistics talk.
I don’t think litigation is the committee’s driving force. I truly believe increased safety is. But they will be risk averse, concerned about litigation related to individuals not being allowed to move up*, and so will prefer a system that can be characterized as objective rather than subjective.
- Remember some years back there was a lawsuit by a rider who believed she (I think) should have been selected for the Olympic team. Threw quite a spanner int he works.
