NSTRA Results: Dog Rating

Post Reply
User avatar
mrgordonscott
Rank: Just A Pup
Posts: 11
Joined: Fri Dec 25, 2015 8:54 pm
Location: Colorado Springs, CO

NSTRA Results: Dog Rating

Post by mrgordonscott » Wed Feb 10, 2016 7:06 pm

I am working on project that rates dogs and predicts when a dog would champion and how often. I'd love to rate your dog for you. If you have placed (1st-4th) I can look your dog up on the NSTRA site, if you have not yet placed but want me to rate your dog as well please let me know the scores from all the dogs trials.

Friendly disclaimer: I can't guarantee you will like what numbers get spit out but it will give you an honest look at your dog.

User avatar
ezzy333
GDF Junkie
Posts: 16625
Joined: Fri Aug 20, 2004 3:14 pm
Location: Dixon IL

Re: NSTRA Results: Dog Rating

Post by ezzy333 » Wed Feb 10, 2016 7:36 pm

mrgordonscott wrote:I am working on project that rates dogs and predicts when a dog would champion and how often. I'd love to rate your dog for you. If you have placed (1st-4th) I can look your dog up on the NSTRA site, if you have not yet placed but want me to rate your dog as well please let me know the scores from all the dogs trials.

Friendly disclaimer: I can't guarantee you will like what numbers get spit out but it will give you an honest look at your dog.
I am confused on how you rate a living animal that you can't rate since they are all different and then say you can rate then mathematically. That jus6t doesn't compute in my book. I need more info. Can you help us?

Ezzy

User avatar
Sharon
GDF Junkie
Posts: 9113
Joined: Fri May 09, 2008 4:46 pm
Location: Ontario,Canada

Re: NSTRA Results: Dog Rating

Post by Sharon » Wed Feb 10, 2016 9:42 pm

mrgordonscott wrote:I am working on project that rates dogs and predicts when a dog would champion and how often. I'd love to rate your dog for you. If you have placed (1st-4th) I can look your dog up on the NSTRA site, if you have not yet placed but want me to rate your dog as well please let me know the scores from all the dogs trials.

Friendly disclaimer: I can't guarantee you will like what numbers get spit out but it will give you an honest look at your dog.
Your hypothesis would appear to be that a dog's future success depends on its past success. Is that right? With so many variables in a trial/test I wonder if that hypothesis is valid?

User avatar
mrgordonscott
Rank: Just A Pup
Posts: 11
Joined: Fri Dec 25, 2015 8:54 pm
Location: Colorado Springs, CO

Re: NSTRA Results: Dog Rating

Post by mrgordonscott » Thu Feb 11, 2016 10:11 am

@ Ezzy The rating is actually on the dogs training and that points straight to us as trainers. So far what I have found is there are some great dogs out there that are just a few points away from being a champion but I don't think their trainers understand what is going on.

I will give you a quick example: Calret is a dog that no longer runs trials but when she did she averaged 792.14 points each time she trialed! In her region she often place but didn't get her Championship. My project shows that she was a mere 8 points on average shy of getting that Championship. If her trainer could see this, the trainer could have easily added a few points by working just a little harder on one area.

@Sharon My hypothesis is that we can become better trainers if we see our dog how others (in this case NSTRA judges) see our dogs. I think a dog will only perform as well as it has been trained.

User avatar
ezzy333
GDF Junkie
Posts: 16625
Joined: Fri Aug 20, 2004 3:14 pm
Location: Dixon IL

Re: NSTRA Results: Dog Rating

Post by ezzy333 » Thu Feb 11, 2016 12:34 pm

So you are taking away all natural ability and relying strictly on what it has been taught. I don't think you can do that.

cjhills
GDF Junkie
Posts: 2529
Joined: Sat Jan 28, 2012 7:37 am
Location: aitkin,mn

Re: NSTRA Results: Dog Rating

Post by cjhills » Thu Feb 11, 2016 6:48 pm

ezzy333 wrote:So you are taking away all natural ability and relying strictly on what it has been taught. I don't think you can do that.
Might be training to get the most out of natural ability.
I doubt that there has been a dog born that can win at NSTRA on natural ability without training
He might be on to something given a chance......Cj

ncpointers
Rank: Senior Hunter
Posts: 124
Joined: Sun Jun 06, 2004 6:57 pm
Location: NC

Re: NSTRA Results: Dog Rating

Post by ncpointers » Thu Feb 11, 2016 9:14 pm

Try my dog Nixon. He just completed his 1st championship. Does he have another one or two in him?

User avatar
displaced_texan
Rank: 5X Champion
Posts: 1003
Joined: Fri Aug 17, 2007 10:57 pm
Location: Mobilehoma

Re: NSTRA Results: Dog Rating

Post by displaced_texan » Thu Feb 11, 2016 10:00 pm

If only it was that simple...

Ninevehn
Rank: Just A Pup
Posts: 18
Joined: Sat Feb 01, 2014 6:36 pm
Location: NE Kansas

Re: NSTRA Results: Dog Rating

Post by Ninevehn » Thu Feb 11, 2016 11:27 pm

What's the methodology of your system? Are you testing predictive power to confirm the validity of the ratings you're using? I'm all for adding a little more mathematical rigor to the hunting world (I'd settle for people saying how close 'close' is and how cold 'cold' is), but I'd like to understand what you're trying to accomplish.

User avatar
oldbeek
Rank: 5X Champion
Posts: 766
Joined: Mon Jun 09, 2014 8:47 pm
Location: Lancaster CA

Re: NSTRA Results: Dog Rating

Post by oldbeek » Fri Feb 12, 2016 12:10 am

Where does luck of the draw come in?? You draw braces after real good dogs that clean the field. Only 4 birds to find. You draw a brace when breeze is dead. Or you get a draw after some poor dogs and the breeze has just started up. Field is full of scent and birds. Naturally you are going to get lots of finds. Just lots of variables.

User avatar
nhachman
Rank: Master Hunter
Posts: 243
Joined: Fri Feb 18, 2011 9:39 am
Location: Kearney, MO

Re: NSTRA Results: Dog Rating

Post by nhachman » Fri Feb 12, 2016 6:58 pm

I agree with what has mostly been said in the other posts. Increasing average score through training is only one of possibly a hundred different variables that would go into the formula for winning a given trial. Plus, training can only increase the score on finds and retrieves if the dog finds birds. What sort of information goes into your figuring that the dog in your example was 8 points from a championship? Do you mean that the dog just needed a win, and the average first place score is 800 points, so her average was about 8 points below that? That's all fine and good but averages are only good when you have normalized probability, which in my opinion would be after hundreds of trials. I think a lot of dogs on the cusp or "just a win or two short" are usually ones that have had a decent number of breaks that happen to build up some championship points. These are not the dogs that show up to a trial every weekend and expect to finish in the top 5-6 of the field.

User avatar
nhachman
Rank: Master Hunter
Posts: 243
Joined: Fri Feb 18, 2011 9:39 am
Location: Kearney, MO

Re: NSTRA Results: Dog Rating

Post by nhachman » Fri Feb 12, 2016 7:05 pm

Also, how will you rate a dog just based on the scores on the NSTRA site? Those scores are just for 1-4th place, as you stated, so what about all the other scores the dog achieved when it didn't place? Not including those in your average discredits the math right away.

User avatar
Luminary Setters
Rank: Senior Hunter
Posts: 168
Joined: Sat Jan 07, 2006 7:41 am
Location: Spring City, Tennessee

NSTRA Results: Dog Rating

Post by Luminary Setters » Sat Feb 13, 2016 8:32 am

It would mathematical model that is data driven. No doubt it will have flaws that will have to be worked out to account for the variables.

Rather than bust his chops, why don't some of you offer some advice to help him with the project. It wasn't that many years ago it was said mathematical models to predict the weather were an impossibility, but now, while not perfect, they do a pretty good job.

User avatar
ezzy333
GDF Junkie
Posts: 16625
Joined: Fri Aug 20, 2004 3:14 pm
Location: Dixon IL

Re: NSTRA Results: Dog Rating

Post by ezzy333 » Sat Feb 13, 2016 10:26 am

Luminary Setters wrote:It would mathematical model that is data driven. No doubt it will have flaws that will have to be worked out to account for the variables.

Rather than bust his chops, why don't some of you offer some advice to help him with the project. It wasn't that many years ago it was said mathematical models to predict the weather were an impossibility, but now, while not perfect, they do a pretty good job.
If you think asking questions and questioning things is busting someone's chops then nothing of much importance has ever been accomplished with out busted chops. If you want o see something that was done without anyone being allowed to question it look at our healthcare and foreign policy programs. Or closer to my home at least is the 53 million our city controller borrowed and closer yet the plaque on my wall from the company I worked with thanking me for questioning things we were doing that insured everyone could account for things they were trying to do.

Good ideas that work have to be able to stand questioning and criticism though I don't believe we have seen much of that. My question was how do you mathematically account for natural ability and all of the variables that happen day to day. If it can be done, wonderful, but no one has ever been able to yet, so lets hope it works if that is the goal. I'm just not even sure I want to be able to sit at the table and figure how good my dog is without participating and see who wins. However, if it works, we won't need to compete but we can just enter and then see which dog has the highest score. I'm being factious to a point but it just doesn't seem right. Why doesn't the same dog win two days in a row? I'll wait and see and help where I can but there are so many variables that will have to be answerewd.

Ezzy

User avatar
displaced_texan
Rank: 5X Champion
Posts: 1003
Joined: Fri Aug 17, 2007 10:57 pm
Location: Mobilehoma

Re: NSTRA Results: Dog Rating

Post by displaced_texan » Sun Feb 14, 2016 9:03 pm

Things such as where the birds are in the field, and which way the dogs head on breakaway cannot be mathematically accounted for or predicted.

User avatar
kninebirddog
GDF Premier Member!
Posts: 7846
Joined: Sat Dec 25, 2004 12:45 am
Location: Coolidge AZ

Re: NSTRA Results: Dog Rating

Post by kninebirddog » Wed Feb 17, 2016 9:08 am

Thay is what the scores in NSTRA are for of people would look at them they could see for themselves where they may need to Polish up thier dogs instead of whining about the score the judge gave them. Look at the over all scores maybe ask What the judge saw and why. Some times that can be tricky when a judge has sat in the saddle all day judging 16 braces

So I could see how looking at the averages could give a rough guess at approx how much longer it could take if all things are approx the same. Law of averages could apply

Post Reply