Re: [RC] Completion rate by Wt division - Michael MaulI noticed the difference too and decided that Truman would let me know if I was wrong. He did and I was off a little on how long that would take... :-) Truman is certainly correct. I considered looking at the 12 years of data and whether distances like 100 miles vs LD made a difference but given my present "job", I don't have the time. Doing simple averages as I did is OK when the results are close but real statistical analysis is needed when they are not. What it means - I'm not sure other than it went from lightest - Juniors to heaviest - in completion rates. Mike --- On Wed, 10/22/08, Michael Maul <mmaul@xxxxxxxxx> wrote: From: Michael Maul <mmaul@xxxxxxxxx> Subject: Re: [RC] Completion rate by Wt division To: "ridecamp" <ridecamp@xxxxxxxxxxxxx>, "Truman Prevatt" <tprevatt@xxxxxxxxxxxxxx> Date: Wednesday, October 22, 2008, 10:50 AM --- On Wed, 10/22/08, Truman Prevatt <tprevatt@xxxxxxxxxxxxxx> wrote:From: Truman Prevatt <tprevatt@xxxxxxxxxxxxxx> Subject: Re: [RC] Completion rate by Wt division To: "Michael Maul" <mmaul@xxxxxxxxx>,"ridecamp" <ridecamp@xxxxxxxxxxxxx>Date: Wednesday, October 22, 2008, 9:23 AM While these "numbers look close" there is a significant statistical difference between HW and the other weight divisions.Onemight suspect this since the difference between MW and HW (2.4%) is8times greater than that between the other categories (.3%). The Strong Law of Large Numbers is a wonderfulproperty -it allows one to make mathematically valid decisions from statisticsthatappear "close" when the sample sizes become largewhichis the case here. When you calculate the standard errors for the fourcategoriesand then compare the number of "sigmas" apart the completion probabilities you find for the three categories, FW, LW and MW themaximumsimgage (FW to MW) is 1.2 which is well within expected statistical variation at random. However, when you compare the simage of HW toMW itis on the order of 4.5 sigma - which equates to a probability oflessthan .001 for the sample mean for HW actually coming from the MW category. These data show to a confidence of greater than 99%thatthe pull rate for the population of horses carrying heavyweightriders ishigher than that for any other weight division. These data alsoshowthat the difference between the pull rates between MW, FW andLW arestatistically insignificant based on these samplesizes. Inorder to see if the statistics of the pull are statisticallydifferentfor MW, LW and FW - more data would be needed. To conclude to a 99% confidence that say the pull rate of FW was greater (or less) than asamplesize of over 116000 for each would be required. Truman Michael Maul wrote:There was a recent post - either on themember'sforum or RC - thatcommented on the completion rate by wt. division at the NC. It showed some differences by wt division so Idecidedto see if thatwas true for a whole season's rides. In 2006 Div starts completions % Comp FW 4720 4104 86.9% LW 6558 5677 86.6% MW 4747 4096 86.3% HW 3087 2589 83.9% Jr 815 715 87.7% There's a slight drop in completion rate forHWand a slight increasefor Juniors but overall - it looks pretty close. The averagethatyear for all riderswas Mike=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=Ridecamp is a service of Endurance Net,http://www.endurance.net.Information, Policy, Disclaimer:http://www.endurance.net/RidecampSubscribe/Unsubscribehttp://www.endurance.net/ridecamp/logon.asp RideLong and Ride Safe!!=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-- "The trouble with fighting for human freedom isthatone spends most of one's time defending scoundrels. For it is against scoundrels that oppressive laws are first aimed, and oppression mustbestopped at the beginning if it is to be stopped at all." H. L. Mencken =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Ridecamp is a service of Endurance Net, http://www.endurance.net. Information, Policy, Disclaimer: http://www.endurance.net/Ridecamp Subscribe/Unsubscribe http://www.endurance.net/ridecamp/logon.asp Ride Long and Ride Safe!! =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
|