90th Annual Statuto Race

by Posted on

My friend Rich sug­gest­ed this race, which despite hav­ing tak­en place eighty-nine times pri­or to today had nev­er come to my atten­tion. The tim­ing was good for a new race. The DSE race today was a 5K in Gold­en Gate Park, which would have been fine but this race—starting next to Wash­ing­ton Square in North Beach and going down The Embar­cadero just past the Bay Bridge and back the way we came—was not only new to me, but cov­ered parts of The Embar­cadero I haven’t run in years.

I had trou­ble most of the race keep­ing my weight for­ward. I’d catch myself feel­ing like some­thing was­n’t right, then straight­en my back, lean in, and feel things smooth­ing out, but for what­ev­er rea­son it nev­er stuck. I tried keep­ing pace with Rich and stayed near him for the first mile, but grad­u­al­ly he got far­ther and far­ther ahead. He fin­ished about five min­utes ahead of me.

The race was sup­pos­ed­ly 8K, though my GPS reporte a dis­tance around 250 meters short of that. If I had to guess I’d think it’s more like­ly that the course dis­tance was off, not my GPS. The GPS some­times adds dis­tance (if, for instance, it reports inac­cu­rate vari­ances from what should be a straight path) but I can’t imag­ine it los­ing dis­tance unless it missed some cor­ners. Los­ing a quar­ter of a kilo­me­ter just does­n’t seem rea­son­able. So I’ll clas­si­fy this as an 8K, but I think 7.75K (the Garmin device claims 4.82 miles—I’m not sure where the 7.87K above comes from) is actu­al­ly more accurate.

Published by splicer

8 Replies to “90th Annual Statuto Race”

  1. Garmin or any GPS

    Steve, I think the short­com­ings of any GPS is it’s fail­ure to detect ele­va­tions.  Those three last hills short­er on GPS than actual.

    Some­day tech­nol­o­gy will fig­ure out a way to over­come this bug, if it already has.

    1. Ele­va­tion

      Inter­est­ing. From Bay Street to Green­wich there is 150 feet of ele­va­tion dif­fer­ence over a quar­ter mile. Com­par­ing the path in two dimen­sions to the path in three dimen­sions makes a tri­an­gle (ignor­ing the irreg­u­lar­i­ty of the sloped side. A right tri­an­gle with legs mea­sur­ing 1320 feet and 150 feet would have a hypotenuse of about 1328.5 feet. Since we trav­eled that twice, that means that the dis­tance trav­eled over pave­ment is at least 17 feet longer than the dis­tance trav­eled on the map. That’s a lit­tle over five meters, or about one-fifti­eth of the dis­crep­an­cy report­ed by the Garmin device. Unless my math is wrong, I don’t think the ele­va­tion data (or lack there­of) tells the whole story.

  2. GPS accu­ra­cy

    Most GPS receivers are equipped with WAAS which gives rms errors of less than 3.2 meters ver­ti­cal­ly and 1.8 meters hor­i­zon­tal­ly. If you have LAAS and are with­in 20 – 30 miles of an air­port both errors will be under 1 meter in both axes. There are very expen­sive units that offer far greater resolution.

    Sit­ting off the coast of Maine on a sail­boat, I could tell the state of the tide with WAAS. That’s close enough for me.

    Dad

    1. WAAS

      The Fore­run­ner 305 is WAAS-enabled and on city side­walks it’s almost always accu­rate with­in 10 – 15 meters. It’s rarely accu­rate with­in 3.2. For ele­va­tion, it’s basi­cal­ly use­less. Take a look at http://connect.garmin.com/activity/28857073 and look at the dif­fer­ence between ele­va­tion cor­rec­tions enabled and dis­abled. Ele­va­tion cor­rec­tions use a topo­graph­i­cal map instead of GPS data to deter­mine ele­va­tion. With ele­va­tion cor­rec­tion mile my ele­va­tion from  1.5 to mile 6 is all between 0 and 6 feet above sea lev­el. With­out ele­va­tion cor­rec­tion, I peak at 148 feet at mile 4 and bot­tom out at 71 feet under­wa­ter at 6.38 miles, where the ele­va­tion cor­rect­ed ver­sion has me already head­ing uphill at 26 feet above sea lev­el. That’s a good 30 meters off!

      I don’t know whether being wrist-mount­ed makes it hard­er for the device to keep track of the satel­lites or what. One thing I have not tried is dis­abling «smart record­ing» which drops the sam­pling rate from once per sec­ond to once every four or five sec­onds. Tak­ing more sam­ples would not keep the device from get­ting inac­cu­rate read­ings but the soft­ware I use has an option to fil­ter out noise in the track. Hope­ful­ly hav­ing more gran­u­lar data would help the soft­ware bet­ter find the sig­nal in the noise. Or it could just make it worse.

      1. My Delorme Earth­mate shows 9′

        My Delorme Earth­mate shows 9′ hor­i­zon­tal accu­ra­cy and 39″ ver­ti­cal accu­ra­cy inside the house. When I step out­side it drops to 8′ and 14′. I’m fair­ly wood­ed here so it should get even bet­ter with more sky and more satellites.

        My Garmin 350 shows only one accu­ra­cy lev­el. It gives 31′ in the house and 14′ out­doors. In either place it shows an alti­tude of 588′ and my topo map says 563′. Add the 5′ off the ground I was hold­ing it and that’s 20′ of ver­ti­cal error.

        I think that you are either get­ting bad mul­ti­path sig­nals from the city build­ings or you have a defec­tive GPS. They’re land­ing air­planes with these things! (using LAAS) Are you sure WAAS is turned on?

        Dad

        1. I don’t see any way to turn

          I don’t see any way to turn WAAS on or off. I did change the sam­pling rate on my lat­est run from «smart record­ing» to «every sec­ond.» Look­ing at the lat­est map (in my desk­top appli­ca­tion, not the Google map) it still looks pret­ty bad, but it’s the worst under tree cov­er and next to build­ings. The paths get nice and smooth when I’m out in open space, like at the top of a hill.

  3. I sus­pect the 305 is not

    I sus­pect the 305 is not WAAS-enabled. Garmin lists its accu­ra­cy as +-10 meters 50% of the time and does­n’t men­tion WAAS in the man­u­al or on their web­site specs. The WAAS spec is bet­ter than 25 ft (both hor­i­zon­tal and ver­ti­cal) 95% of the time. Typ­i­cal results in good con­di­tions are clos­er to one meter lat­er­al­ly and 1.5 meters ver­ti­cal­ly. (Accord­ing to wikipedia.) Those are the kinds of num­ber I have seen.

    Of course, not being able to hold the anten­na face-up while swing­ing on your wrist may ham­per recep­tion enough that Garmin sim­ply does­n’t claim WAAS accu­ra­cy. Set­ting it on a table anten­na-up might show a far bet­ter result. What does the GPS accu­ra­cy data field show you?

    Dad

     

    1. 20 feet
      I just turned it on indoors and it says about 20 feet. In the past cou­ple min­utes it has been as high as 22 feet and as low as 16 feet. I’m upstairs but inside. I’ll have to find where it is that I read it is a WAAS device. I can’t find a men­tion of it in the set­tings. I’ll also pay atten­tion to the accu­ra­cy field as I run over the next few days and maybe look at the accu­ra­cy from the top of a hill or the mid­dle of a park­ing lot—somewhere with­out any over­head cover.

Comments are closed.