Perf. Std: Disadvantaged Riders & Regions

[pluvinel]

As some of you may be aware, the 3 data nerds have continued to nerd away. We have analyzed ~45,000 scores from 2008 shows from all regions and all levels.

Some interesting stuff is emerging.

Would people be interested to know that there are some "low scoring regions?"

Would people be interested to know that there are some "low scoring breeds?"

Would people still support a qualifying rule if the lowest scoring region scored, on average 2.5 points below the highest scoring region?

These are not rhetorical quesitons. What should we do with this info?

[molliwog]

Well, there are other data-data nerds on this board.

I think you should post it, along with a brief description of where you found your source data and a description of what analytical methodology was applied to lead you to your conclusions.

Then, the rest of us data nerds can evaluate it for you. Hey...it's not a peer-reviewed journal or anything, but I doubt that the USDF connection or Dressage Today is going to publish your findings no matter how sound your analytical method may be......



(Wouldn't it be nice if our national organizations actually used a data-driven approach before making decisions?????)

[Touchstone Farm]

I agree -- post it here and email to the GMOs for discussion before the upcoming USDF conference. Send to John Long at the USEF too. Decisions need to be made on data...not opinions and anecdotes.

[MelantheLLC]

I'd love to know about it.

It could make a great difference and certainly raise some questions that should be looked at more closely.

Not that I give a hoot about showing but I like patterns and facts. ;)

[yaya]

Would people be interested to know that there are some "low scoring regions?"

Would people be interested to know that there are some "low scoring breeds?"



Interested, but not at all surprised. Would like to know which they are, to see if they match perception.

[fiona]

I'd be fascinated - as a non american, non nerd, non mathemetician - it's interesting but it would need to come with a stats for dummies type explanation!

[STF]

As some of you may be aware, the 3 data nerds have continued to nerd away. We have analyzed ~45,000 scores from 2008 shows from all regions and all levels.

Some interesting stuff is emerging.

Would people be interested to know that there are some "low scoring regions?"

Would people be interested to know that there are some "low scoring breeds?"

Would people still support a qualifying rule if the lowest scoring region scored, on average 2.5 points below the highest scoring region?

These are not rhetorical quesitons. What should we do with this info?


Id love to see it (Im a math nerd too).

[Elegante E]

I vote yes as well. I think it would be helpful. Might help various GMOs figure out where their area stands and could lead to helpful, knowlegeable, improvements.

[Atlantis]

Would people be interested to know that there are some "low scoring regions?"

Would people be interested to know that there are some "low scoring breeds?"


I am curious as well.

I can see the "low scoring regions" going either way. There are the big dressage areas like Florida and S. California, where we'd assume the scores would be higher (regions 3 and 7), but then there are also areas where dressage is not so big and where it is not unheard of to hire generous judges to encourage people to participate at all, and there is also the fact that although regions 3 and 7 contain Florida and S.Cal, they also contain other non dressage heavy states, which could also skew the results in terms of speaking regionally. In other words, I don't know if determining high scoring regions would really mean anything. Nonetheless, I'd still be curious!

It's the same with breeds, we can rather well guess which breeds are higher scoring and lower scoring, or can flip thru any Yearbook issue to see All-Breeds results. I'd still be curious in your specific findings, however.

While I don't know how meaningful the information may be in terms of the big picture, it would still be interesting.

I would also be interested in the average score nationwide, per level, to see how it drops as people move up the levels and the work gets more difficult, if this is information you happen to also have easily available.

[slc2]

Just looking at show scores I've often thought certain regions scored lower, but also that scores from specific shows average very differently even in a higher scoring region. and certain breeds score lower based on the HOY published numbers.

I'd really like to see the results of your study to see if my impressions are really valid overall, or just apply for the small number of scores I've looked at.

Do the scores run higher in california and florida? Are your numbers dividied by professional and amateur?

[pluvinel]

Well, there are other data-data nerds on this board.

I think you should post it, along with a brief description of where you found your source data and a description of what analytical methodology was applied to lead you to your conclusions.

Then, the rest of us data nerds can evaluate it for you. Hey...it's not a peer-reviewed journal or anything, but I doubt that the USDF connection or Dressage Today is going to publish your findings no matter how sound your analytical method may be......

(Wouldn't it be nice if our national organizations actually used a data-driven approach before making decisions?????)

PEER REVIEWED!!!!! This is brilliant!!!!! We welcome as many eyes as possible on the analysis.

We're in the process of finalizing the document. The question was basically what to do with said document. The inital results raised a few more questions on how to slice and dice for further analysis....so we're doing that.

It will take a week or so to finish the report as some of us need day jobs which interfere with this project.

But once the report is final, lets get a "gathering of the nerds" and have a data fest.

Give us a couple of weeks, but please continue to give us ideas about where the document should go, if they haven't already been covered.

[JRG]

Nerd away...I would love to know, I am nerdly challanged.

[pintopiaffe]

aahhh... but even regions don't tell the whole story.

My region includes NEDA (MA) and WNYDA (NY) and shows like NEDA fall, Syracuse, etc.

My *state*, and my *area* (5 +/- hrs from nearest big NEDA event) definitely don't fare as well as the rest of the region...

3 recognized shows this year within a 3 hr drive. That's it.

Still, I'd love to see the numbers.

[J-Lu]

Please post them.

I could be wrong, but I'm guessing that the trends follow what has been published in USDF connections in regards to average qualifying and regional championship scores by region and the breeding statistics (performance scoring by breed, sire of get, produce of dam, etc. statistics).

Of course there are lower scoring regions. There will be higher scores in areas of the country where there is a higher disposable income to buy quality horses and take lessons and put the horse in professional training, where there is an abundance of quality professionals, where there is an abundance of breeders-especially quality mare owners, and where english-type riding is the dominant equestrian sport. Likely, the whole of the east coast, especially Florida, and California regions have some of the highest scores.

I'm sure my region (Region 9) is one of, if not the lowest scoring region. As a whole, Western disciplines are very popular (and draw alot of the horse people with money), there are alot of "nontraditional breeds" doing dressage, and we have relatively fewer quality trainers and breeders than, say, New England or the Mid-Atlantic, for the same land mass. There is also less money as a whole in this region. I mean, in between cities here we have alot of *open space* and ranches - not suburbs and small cities. We hear crap all the time about how terrible the region is as a whole.

Of course I would also expect low-scoring breeds. I'm guessing the most popular and populous breeds (hanoverians, dutch warmbloods, oldenburgs) are amongst the highest scoring and everything else would be a whole lot lower? I'm also guessing that those with common import breeds fall into a financial bracket that affords them to pay for professional training and lessons. NO DOUBT that increases one's scores. And that top professionals (who score well) are riding them (year end award and qualification stats support this). I'm guessing less TB and QH owners spend that amount of money on training - or they are doing other disciplines as well as dressage.

I think it is interesting information, but I'm not sure what it really says at the end of the day. We have judges coming to our region *expecting* to see low scoring horses and riders. I also hear people telling others that if they buy XX breed, they'll do better at shows and in the discipline (these conclusions can be supported by the various year-end award stats). This usually ends up with newer and amateur riders being overmounted - exactly the problem that the judges complain about. Who wants to buy an appropriate QH or Welsh cob when they already know they won't score as well as a NA-KWPN? I see this alllllllllll the time. Publishing data about high and low scoring regions, and high and low scoring breeds really needs to come with a real-life interpretation of what the numbers mean. I'm not saying that the stats geeks need to do this :), but that people need to keep it all in mind. There's nothing worse than no stats than misinterpretted stats. :):)

J.

[J-Lu]

PEER REVIEWED!!!!! This is brilliant!!!!! We welcome as many eyes as possible on the analysis.

We're in the process of finalizing the document. The question was basically what to do with said document. The inital results raised a few more questions on how to slice and dice for further analysis....so we're doing that.

It will take a week or so to finish the report as some of us need day jobs which interfere with this project.

But once the report is final, lets get a "gathering of the nerds" and have a data fest.

Give us a couple of weeks, but please continue to give us ideas about where the document should go, if they haven't already been covered.

I'm a scientist nerd, and would be thrilled to be a part of the "gathering of the nerds". i love data fests!

[canyonoak]

I'm not a numbers nerd, but I KNOW this data is relevant and important; that it helps to dispel the myth that U.S. riders need a Standards Rule; that it helps to explain WHY the Standards Rule as presented will not work and has no operating system in place to help make it work.

BUT--

Honestly-- I think the only numbers that will interest John Long and the BOG are the economics.

If the people who vote are made to understand that this Performance Rule will undermine their cash flow; that this Performance Rule presents as elitist and is unattractive to riders and possible future riders...then I think there is a chance they will vote no.

[claire]

FWIW, I vote to publish your data on the BB's. I think you would get some interesting feedback from many of the posters here and on TOB.


Also, these 2007 vs. 2008 stats you posted on TOB were interesting. The 2008 scores are actually lower at T-4th levels (statistically significant :confused: )


In 2007, the average scores across all regions was (truncated at 1 decimal place):
Training-62.3
1st-61.7
2nd-60.5
3rd-60.1
4th-61.3
PSG-60.3
Int-60.1
GP-59.1

In 2008, the average scores across all regions was (truncated at 1 decimal place):
Training-61.9
1st-61.7
2nd-60.3
3rd-60.160.0
4th-60.9
PSG-661.0
Int-62.0
GP-60.7

And of all the scores, in all the regions, in both 2007 and 2008, only 0.03% of the scores were 40% or below.....

Remember, the score of 4 means "insufficient"....ergo, the judges are NOT sending the message to the riders that their riding is BAD....

The descriptors of "bad" starts at 3-"Fairly bad"...then 2-"Bad".....then 1-"Very bad"....so if the judges want to have the riders get "the message" then the score need to reflect bad riding when it so merits."

[quietann]

One data nerd here, would love to see this!!!

[Tamara in TN]

[QUOTE=pluvinel;3531195]As some of you may be aware, the 3 data nerds have continued to nerd away. We have analyzed ~45,000 scores from 2008 shows from all regions and all levels.

Some interesting stuff is emerging.

are there also any juges who consistantly give low scores all round ??? also I know in some other breeds/sports there are judges who are hired just because they are cheaper than all other folks...and sadly their opinions matter:( as they shape the year end scoring...

best

[molliwog]

Let your fellow nerds know when the data are ready.......would be glad to have a look.

[slc2]

"are there also any juges who consistantly give low scores all round ??? "

yes.

"cheap judges"

the low scoring judges aren't necessarily bad judges.

[MEP]

This should be the text that helps interpret the raw data. More can always be said, but I think this about sums it up. :yes:

Please post them.

I could be wrong, but I'm guessing that the trends follow what has been published in USDF connections in regards to average qualifying and regional championship scores by region and the breeding statistics (performance scoring by breed, sire of get, produce of dam, etc. statistics).

Of course there are lower scoring regions. There will be higher scores in areas of the country where there is a higher disposable income to buy quality horses and take lessons and put the horse in professional training, where there is an abundance of quality professionals, where there is an abundance of breeders-especially quality mare owners, and where english-type riding is the dominant equestrian sport. Likely, the whole of the east coast, especially Florida, and California regions have some of the highest scores.

I'm sure my region (Region 9) is one of, if not the lowest scoring region. As a whole, Western disciplines are very popular (and draw alot of the horse people with money), there are alot of "nontraditional breeds" doing dressage, and we have relatively fewer quality trainers and breeders than, say, New England or the Mid-Atlantic, for the same land mass. There is also less money as a whole in this region. I mean, in between cities here we have alot of *open space* and ranches - not suburbs and small cities. We hear crap all the time about how terrible the region is as a whole.

Of course I would also expect low-scoring breeds. I'm guessing the most popular and populous breeds (hanoverians, dutch warmbloods, oldenburgs) are amongst the highest scoring and everything else would be a whole lot lower? I'm also guessing that those with common import breeds fall into a financial bracket that affords them to pay for professional training and lessons. NO DOUBT that increases one's scores. And that top professionals (who score well) are riding them (year end award and qualification stats support this). I'm guessing less TB and QH owners spend that amount of money on training - or they are doing other disciplines as well as dressage.

I think it is interesting information, but I'm not sure what it really says at the end of the day. We have judges coming to our region *expecting* to see low scoring horses and riders. I also hear people telling others that if they buy XX breed, they'll do better at shows and in the discipline (these conclusions can be supported by the various year-end award stats). This usually ends up with newer and amateur riders being overmounted - exactly the problem that the judges complain about. Who wants to buy an appropriate QH or Welsh cob when they already know they won't score as well as a NA-KWPN? I see this alllllllllll the time. Publishing data about high and low scoring regions, and high and low scoring breeds really needs to come with a real-life interpretation of what the numbers mean. I'm not saying that the stats geeks need to do this :), but that people need to keep it all in mind. There's nothing worse than no stats than misinterpretted stats. :):)

J.

[unbridledoaks]

I am curious as well. I think that it's always interesting when something like this is done.

[CatOnLap]

as a trained statistician, I would love to see the raw data, including all the "n" numbers to achieve significance, and any conclusions. We have enough conclusions on this thread already without any data. It will be fun to see how many of those guesses match the data.

So spill already Pluvinel!

[SGray]

is there a list on USDF site of all riders in a region that are qualified for the GAIG Champs.?

[DressageGeek "Ribbon Ho"]

I do peer review all the time...and even do some stats. This would be very interesting, because it would help uncover possible bias or demonstrate little bias (eg against specific breeds).

I'd love to see it too.

[yaya]

is there a list on USDF site of all riders in a region that are qualified for the GAIG Champs.?

Yes, it's here: http://www.usdf.org/competitions/competitions-championships/regionals/qualifying-rides.asp

[ironbessflint]

Another twist on who might appreciate the info would be breed organizations. I for one work hard to promote the Haflinger as a dressage horse. I know from a quick look online that Haflingers qualified for All-Breed awards at 2nd level (just to pick one), average higher scores than the top five qualified Hanoverians! Now, from there useful data goes down....many Haflingers could and do go on, but there's only one qualified at PSG. But what I'm getting at is if a smaller or more obscure breed seems to do well, actual DATA might be helpful in promotion!

[SGray]

thanks yaya -- will copy data for a comparison and report back

[DownYonder]

Yes, it's here: http://www.usdf.org/competitions/competitions-championships/regionals/qualifying-rides.asp

Bear in mind that the USDF site runs 3-6 weeks behind in posting show results, and therefore also in updating qualified rider lists. There are a bunch of riders who have qualified for regionals within the past several weeks but are not yet listed.

[SerenaGinger]

Bear in mind that the USDF site runs 3-6 weeks behind in posting show results, and therefore also in updating qualified rider lists. There are a bunch of riders who have qualified for regionals within the past several weeks but are not yet listed.

They also don't include horses that are not individually owned, like ones owned by an LLC.

[ClaraLuisa]

Just adding another vote for more statistics fun.
Nerds need feeding here.

[NoDQhere]

We would love to see the statistics too :). As a "lowly" Region 4 person I would also love to know where I can find the number of entries in each Regions Championship. Region 4 is the lowest scoring Region, yet the USDF Region 4 Championships are, I believe, one of the larger Regionals as far as number of entries. Any ideas on where I can find these numbers?
TIA

[slpeders]

We would love to see the statistics too :). As a "lowly" Region 4 person I would also love to know where I can find the number of entries in each Regions Championship. Region 4 is the lowest scoring Region, yet the USDF Region 4 Championships are, I believe, one of the larger Regionals as far as number of entries. Any ideas on where I can find these numbers?
TIA

Contact Lloyd Landkamer -- he organized the Championships I believe or go to dressageshowinfo.com and count the entries maybe?

[millerra]

I would argue that it's not "just in the numbers" because judges are human.

My dressage coach has shown in the midwest, east coast and florida. She has said, more than once, that a similar ride would score better in a big show (or area) because the judge expects you to be better than if you are showing in po-dunk area.

Case in point: a friend/dressage trainer/coach, who also rides FEI in Florida, was showing a 3rd level horse at a backwoods (literally) local show and scored a 5 for rider (and a 50 something %) for the test on the first day of the show. By the end of the show, he was again scoring, from the same judge, his typical scores in the mid-60's and getting 7's (+) for rider.

So, IMHO, the real test is to video rides from different parts of the country, being ridden by "unknown" riders, and have judges score the rides based on whether they think the rider is showing at Devon or in the back woods.

It would be a "simple" experiment - and we would see if a national scoring system would be fair (or not)...

[RedMare01]

I would also be really interested in seeing the data...

Caitlin

[NoDQhere]

Contact Lloyd Landkamer -- he organized the Championships I believe or go to dressageshowinfo.com and count the entries maybe?

Yes, I figured that. Its just that I am in the middle of 3rd cutting and I had hoped there was a "place" I could go that would list the number of entries in each Regional Championships.

[mjhco]

Don't count on the list of qualifying riders for the GAIG championships to be an accurate assessment of how many actually qualified. In a region as large as Region 5 (Canadian border to Arizona), when the regionals are in Arizona for example some of the riders don't bother to pay the $10 fee to have their scores counted for the regionals. I suspect our region is not unique in that respect.

[sm]

Would people still support a qualifying rule if the lowest scoring region scored, on average 2.5 points below the highest scoring region?

These are not rhetorical quesitons. What should we do with this info?

I think all charts/data should be supplied as part of the main study, in different sections as needed. For two reasons:

- As far as I can tell, correct me if I'm wrong: the DC still isn't properly analyzing their current project, or they would have figured some of this out for themselves. So the more hard data that can be provided to force them actually think through the impact of what they are proposing --- the better.

- We're also guessing WHY a rule change is needed and the initial study is responding to that perceived WHY, however WHY was never clearly stated that I can see. So, once again the more data the better, the broader scope of the study submitted the better. Does that make sense, am I rambling...

Major kudos to everyone working on this stuff.

[sm]

Would people be interested to know that there are some "low scoring breeds?"


Yes. It would probably highlight the infinite wisdom of the international rule that 50% is sufficient. There is SUFFICIENT and then there is ABUNDANT.