In our Side by Side: Hops vs Gitchi article, we compared the two fastest times of all dogs at any height running in the AKC finals and specifically stated that the 20”class was the toughest 20” field seen in several years. During the finals livestream, commentators Sarah Fernandezlopez and John Nys both noted that the 20” field was very tough this year. The finals included the defending champion (Hops), one of the fastest large dog non-border collies in AKC (Gitchi), and a group of handlers that collectively has competed at the FCI Agility World Championship, European Open, and Americas y Caribe, won AKC nationals with different dogs, and appeared in finals at the AKC Invitational and Westminster. The 20” class produced an astounding 16 clean runs in the finals, with just 0.77 seconds separating 2nd through 8th place! An impressive and highly competitive group, however, one reader took exception:
“in the toughest 20″ field seen in several years” – Huh? How so?
Let’s take a look at the yards per second for each dog who ran in the 20” finals over the last three years, taken from the qualifying period for that year’s nationals. The faster dogs are in the upper right corner.
Perhaps this trend occurs in all heights due to improved training methods or better handling. Maybe judges are wheeling courses bigger, leading to inflated yards per seconds over the last three years. Let’s take a look at the other finals height classes and see if dogs are getting faster as well.
The other height classes do not share the 20″ trend, except for 24″ which had a bump up from 2014 to 2015, but not to 2016. I’m comfortable saying that the 20″ finals this year was tougher than previous years and that this wasn’t the case with the other heights, but why did this happen? 26″ entries were down again this year, and many of these 26″ dogs returned to their natural height—20″. 20″ finalists from this year included Gitchi, Karli Renay, and Stella who all jumped 26″ at nationals last year. The 26″ class is largely a self-selecting group of elite competitors generally vying for spots on international teams. Another 20″ finalist from this year, Scoop, run by Nancy Gyes, finished 2nd in the 24″ finals last year and moved to 20″. Many other dogs moved down in height to 20″ but did not make the finals. People aware of this trend realized that the 20″ class would probably be tougher than previous years. With the impending international shift in jump height from 26″ to 24″ comes the next interesting questions: will the AKC eliminate the 26″ class for the 2017 NAC and if they do, will dogs be allowed to jump 24″ even if they measure into 20″?
I’d be interested in the 16″ preferred data as my gut feel is that it had the same bump as 20″ regular this year? (I don’t have the data, I just think that based on watching that class over the years) It feels like there has been a trend of preferred classes getting more competitive from year to year? Not sure how you get the raw data, wish I had it, love looking at this data. Is the raw data out there somewhere for download, like on AKC’s site? Thanks for publishing these charts. I find them very interesting.
You had to be three clean and fast to get into 20″ finals this year. Very competitive with the international folks coming down in height. I think 20 inchers should leave the 24 inch class alone. Have a 24″ international class if need be, but don’t group them all together, IMHO.
I agree. It seems strange to have two 24″ groups but if you’re not going to have everyone jump their natural heights, I think this should be done.
I agree 100% with Maria. It seems very unfair for the 24″ class to be taken over by dogs who should run 20. Just putting my agreement out there — she has articulated the reasons very clearly. 🙂
Interesting that 2014 had so many “slower” YPS, but only in the 20″ class. Was it the footing? Too many good teams in Challengers instead of Finals? 20″ also ran different courses to get to the Finals that year (…advantage to moderate speed dogs over fast dogs)? Gotta love statistics! They tell you what is or what was but not why (correlation but not causation).
The plots certainly seem to support the conclusion derived, but without actual tests of statistical significance there are no scientific “statistics” given at all. No criticism intended except the use of the word statisitics in many applications make things sound more scientific than is supported.
Pointing out that something does “not involve true statistics” is in fact a criticism, which is okay for anyone to share as a comment. My response is that there’s no pretense that we’re writing an article for a scientific journal. There’s no discussion of statistical significance or P values or even the use of the words like “scientific” or “statistical” anywhere in the article (feel free to read it again). In my opinion, it isn’t worth the time to play with the data with different analyses until I reach some arbitrary threshold of a p-value, in order to claim that I’ve stumbled upon some unassailable scientific/statistical truth about the topic. I think the graphs are both helpful for people and compelling. –Esteban
I hope not! Smaller dogs have an advantage running at 24. Maybe they can have a 20+ for dogs jumping 24 and measuring into 20 and leave the 24 inch class for the big dogs. Would be nice to see that at local trials also. It isn’t an accurate representation of competition when the 1st-4th placements in 24 are taken by 20 inch dogs. Yet one more reason my puppy is a 12 inch dog.
I am not afraid of competition, I just want it to be fair. 20 inch dogs jumping 24 isn’t fair. My dog isn’t slow- we won 26 inch GP a few years back at the Southwest Regional, and last year in Reno took 1st in the Premier Run. But when we compete against the 20 inch dogs, their shorter stride and ability to run through tunnels without ducking gives them an advantage over the dogs who must run at 24.
Love your thoughtful statistically based analysis.
The scatter plots are very interesting and are suggestive but they do not involve true statistics which are formulas to estimate if observed data could have occurred by chance alone.
I don’t know exactly which statistics you call “true” but I’m giving you the benefit of the doubt here. Yes, let’s hope no one submits this to Science Magazine. –Esteban