AKC's normal rankings have two major flaws:

1. The use of bonus points strongly favors dogs that mostly trial in parts of the country where trials are small. Bonus points also distort the rankings by encouraging some 24in Labs to jump 26in, where bonus points are often easier to acquire. The data also shows that bonus points are a little easier to get for 24in Labs than for 20in Labs. Eliminating bonus points, which is what our second ranking does, eliminates both of these problems.

2. But both AKC's normal ranking system and the "no bonus points" ranking still have another major flaw: Both hugely reward dogs that trial a lot vs those that trial maybe once a month or less. Showing up for lots of trials does nothing to prove that a dog is better at agility than a dog that trials less often. Both our "Percentage of Dogs Defeated" ranking and our "Average YPS" ranking eliminate the problem with bonus points and the problem of rewarding entering lots of trials. The first of these two alternative rankings is obtained by dividing the total number of dogs defeated in all of the dog's qualifying runs in 2009 by the total number of (other) dogs entered in these same runs. Similarly, the "Average YPS" ranking is obtained by first averaging the YPSs for all standard qualifying runs in 2008, averaging the YPSs for all of the dog's JWW qualifying runs in 2009, and then taking the average of these two numbers.

By taking these percentages and averages for the entire year, we have eliminated both using bonus points and rewarding the pressure to trial "every other" weekend.

3. A given yps tends to produce a higher percentage of dogs defeated for 24in Labs than for 20in Labs; this suggests that the ranking by average yps is the "purer", and hence the better ranking.

4. Finally, none of these rankings (except very slightly AKC's bonus points) directly reward Labs that have a high qualifying rate.