pjb45 Posted June 1, 2010 Share Posted June 1, 2010 (edited) Ok, Jake, thanks for the kind thoughts on lamblasting me. I must be feeling rather masochistic today. This is not about grading on a curve, the world population, randomness, etc. USPSA is the population. It is not a sample. As a population the first hypothesis would be does it have a normal distribution? The null hypothesis would be USPSA does not have a normal distribution. There is a difference between inferential statistics and descriptive statistics. In a normal distribution, wouldn't the top 5% of the population would be in the 95% percentile? Without doing the math, the numbers that Chuck gave us would lead me to believe we do not have a normal distribution curve. Our distribution is skewed. It would appear to be a positive skewness with the GMs further along the x-axis then the Ds. This means there a fewer GMs than would be expected i.e., <5%. Often times a model is set up, this case being the methodology for determining Class, without significant empirical data. 18K members is pretty strong. Even if 50% are classified, 9K is very strong data. 9K in a sample does not infer randomness either. Does the model need to be re-examined for validity now. Not really to add drift to this thread but a lot a GMs talk about there should be an additional criteria for GM, like placing in the top 5 of a major match--oops probably a different thread on GB. Edited June 1, 2010 by pjb45 Link to comment Share on other sites More sharing options...
Jake Di Vita Posted June 1, 2010 Share Posted June 1, 2010 It has nothing to do with percentiles of the population. It is percentage against a number that remains unadjusted over time. A GM isn't someone who is in the top 5% of all USPSA shooters. A GM is someone who shoots at 95% or above on a pre-determined number for classifiers. Link to comment Share on other sites More sharing options...
Merlin Orr Posted June 1, 2010 Share Posted June 1, 2010 (edited) How did the predetermined number become "the" predetermined number if not by performances of/by a group of (elite?) shooters? Edited June 1, 2010 by Merlin Orr Link to comment Share on other sites More sharing options...
Jake Di Vita Posted June 1, 2010 Share Posted June 1, 2010 That's a good question. One I'd like to know the answer to myself. I've shot 140% on classifiers before. That shouldn't be possible if that number was based off of elite performances. Link to comment Share on other sites More sharing options...
bbbean Posted June 1, 2010 Share Posted June 1, 2010 (edited) I've shot 140% on classifiers before. That shouldn't be possible if that number was based off of elite performances. Jake - Next time do the required reload, shoot it weak hand like everyone else, and don't have your girlfriend run the timer! PB - try thinking of USPSA as the shot put. The current world record is roughly 76 feet. If you can throw the shot put 73 feet, your score is (73/76)*100=96% of the high score. It doesn't matter if every other shot putter in the arena also shot 73', or if no one else made it past 60'. Distribution is irrelevant. The only two points that matter are your distance and the world record distance. If you can throw the shot put 73', you're a GM. Comprende, mi amigo? Edited June 2, 2010 by bbbean Link to comment Share on other sites More sharing options...
Corey Posted June 1, 2010 Share Posted June 1, 2010 i would also like to know how these determined HHFs are figured and if people are topping them why doesn't it change. I'm just about to join B class so I've still got some work to do to get to the elite top but its a question id be interesting in knowing. Link to comment Share on other sites More sharing options...
pjb45 Posted June 2, 2010 Share Posted June 2, 2010 Still a cool discussion! Lets assume: On classifier XYZ the top hit factor is 9. Then you shoot a 9.9 hit factor. You have now shot it at 110%. Ten percent above the previous top score. At this point, 9.9 is the top number. All those who previously shot it with a hit factor of 8.6 would be classified as a GM level score. But now with 9.9 being the best/top, those who shot it at 8.6 would be classified as M class. This is the part that gets mathematically ugly, all the data for the hit factors would have to be re-normalized. People who previously were GM might no longer be GMs based on the new and improved training and practice techniques of our newer shooters. I suspect someone at USPSA looked at the consequences of Hit Factor improvements in the Classifier system and made some administrative decisions so people previously classified as GM (or any other class) would not be moved down to a lower class based on the new and improved shooting ability of our membership. From an observational stand point, I still see the top match GMs still at the top of their game, regardless of how they may perform shooting classifiers. BTW: the distribution is relevant because it is the basis of our classifier system. Our Area and National Championships are the second method of getting to GM which is more to your analogy of the shot put contest, given certain constraints. No se amigo. Yo quiero un cervaza, mas, mas, mas Link to comment Share on other sites More sharing options...
Merlin Orr Posted June 2, 2010 Share Posted June 2, 2010 Generally speaking...Once a GM (or D,C,B,A or Master) always a GM..etc. With a few exceptions for various things that can change your classification.... Link to comment Share on other sites More sharing options...
bbbean Posted June 2, 2010 Share Posted June 2, 2010 Still a cool discussion! Lets assume: On classifier XYZ the top hit factor is 9. Then you shoot a 9.9 hit factor. You have now shot it at 110%. Ten percent above the previous top score. At this point, 9.9 is the top number. All those who previously shot it with a hit factor of 8.6 would be classified as a GM level score. But now with 9.9 being the best/top, those who shot it at 8.6 would be classified as M class. This is the part that gets mathematically ugly, all the data for the hit factors would have to be re-normalized. People who previously were GM might no longer be GMs based on the new and improved training and practice techniques of our newer shooters. It's not ugly at all. You move up based on the HHF at the time your classifiers are run. It doesn't matter what someone shoots after you've moved up. Our system is designed to move people up, not down BTW: the distribution is relevant because it is the basis of our classifier system. The classifier system is not based on distribution. It is based solely on HHF. Our entire scoring system is based on HHF. Distribution is never a factor. Link to comment Share on other sites More sharing options...
COMATZD Posted June 2, 2010 Share Posted June 2, 2010 It's not ugly at all. You move up based on the HHF at the time your classifiers are run. It doesn't matter what someone shoots after you've moved up. Our system is designed to move people up, not down I've wondered about this, why is it that the system doesn't really account for moving shooters down? What if a GM works hard, makes the grade, then after a couple years relaxes, only shoots a few matches a month, goes fishing more..blah blah. Why shouldn't he (or she) get moved down if their performance doesn't make the grade anymore? The current system does seem to bias the curve upward (although I'm not sure I'd expect a normal distribution). Link to comment Share on other sites More sharing options...
Corey Posted June 2, 2010 Share Posted June 2, 2010 maybe we can get someone with the inside know-how on how often, if at all, classifier HHF are changed. Theoretically they could change on a monthly basis... Link to comment Share on other sites More sharing options...
rvb Posted June 2, 2010 Share Posted June 2, 2010 sure, each month each hhf could be adjusted based on the best score posted... That would make all the classes harder to get in to. But with only 1-2 percetile of shooters making GM, would that be a good thing? How small do we want to make the upper classes? What's the goal of the system, and is it not being met? The only problem currently, imo, with the system (and it's pretty minor) is the disparity between classifiers. Some are relatively "easy" and some you have to wonder if anyone has ever hit 100%. Sometimes moving up in class depends on getting a lucky string of classifier... As for not automatically moving people down, the reason should be obvious... "gee, there's a big prize-table match coming up... maybe I'll toss a mike on the next couple of classifiers..." -rvb Link to comment Share on other sites More sharing options...
Corey Posted June 2, 2010 Share Posted June 2, 2010 I dont want the HHF to be increaed, haha, just curious as to how often they do. As for the classifiers being different in difficulty, i agree. While as a whole, they will cover all the aspects of the game, a person in my shoes who has only experienced a small hadnful of them either gets lucky or they dont (depends how you look at it as to whether you are lucky or not). I consider myself a C shooter in SS, but with the string of classifiers i have shot lately I will most likely be up'd to a B in a couple weeks. "Can you Count" and "Paper poppers" were two that I did fairly well on and will help bump me up. I know for a fact that if I had encountered an WHO classifier I would most likely not be moving up. Im much better at reloads than I am weak hand shooting. Just the way the game goes i guess. Link to comment Share on other sites More sharing options...
Steven Cline Posted June 6, 2010 Author Share Posted June 6, 2010 maybe we can get someone with the inside know-how on how often, if at all, classifier HHF are changed. Theoretically they could change on a monthly basis... They just may change monthly. But USPSA doesn't re-run everyone's classifiers by the newest HHF. If you scored a 75% based on the HHF two months ago and next month the HHF increases your 75% remains a 75%. If you were to shoot the same classifier in two months and hit the exact same HF it would rate lower (say 73%) so your old HF would remain. It is even possible that you could shoot a slightly higher HF on a classifier which would rate lower against an updated HHF. I recall reading all this someplace official, but I can't recall where at the moment. Link to comment Share on other sites More sharing options...
Jake Di Vita Posted June 6, 2010 Share Posted June 6, 2010 THey don't change monthly. Classifier calculators would have to be re-calc'd every month in that case. I'm pretty darn sure they aren't. Link to comment Share on other sites More sharing options...
Nik Habicht Posted June 6, 2010 Share Posted June 6, 2010 maybe we can get someone with the inside know-how on how often, if at all, classifier HHF are changed. Theoretically they could change on a monthly basis... Realistically they can't. HHF for recently introduced classifiers have been set at Nationals or Area matches. Why? Two reasons: 1. To ensure that there's a representative number of GMs present and shooting at GM level. 2. To ensure that the classifier was set up properly. That second bit is why HHF can't (really shouldn't) be updated every month --- USPSA has no way of knowing whether or not a classifier was set correctly, whether or not people shot it over and over until someone burned the run they wanted, etc. Now clearly there's adjustments made occasionally --- we're no longer accepting a ten second El Presidente as the hallmark standard.... Link to comment Share on other sites More sharing options...
Coach Posted June 16, 2010 Share Posted June 16, 2010 PJB, What you are saying is called reification. You're treating an abstract as if it were reality. While the 5% of shooters being GMs may have been the original idea, ideas and theory rarely conform to reality in every way. Wow! Most people don't know what reification is and I certainly was caught off guard seeing it pulled like a knife here. Jake I am guessing you have been blasted in class sometime for reification??? Link to comment Share on other sites More sharing options...
Jake Di Vita Posted June 16, 2010 Share Posted June 16, 2010 I've sure been guilty of it but never called out on it. I do a lot of reading haha. Link to comment Share on other sites More sharing options...
norbs007 Posted June 16, 2010 Share Posted June 16, 2010 theres only 9 SS GMs? That's what I said... Link to comment Share on other sites More sharing options...
COMATZD Posted May 10, 2011 Share Posted May 10, 2011 Hey Steve I looked up the classifier ratios again. Looks like your in the top 4.4% now! I'm still working on the 30% Danno Link to comment Share on other sites More sharing options...
COMATZD Posted May 11, 2011 Share Posted May 11, 2011 BTW, I just realized that the data came from the old USPSA site (but dated May 4, 2011). The numbers may or may not have been mined from current data sets, I don't know which for sure. I do know that the data was relevant at some past date Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now