O Log - Older Archives
[28-Feb-02] 

    There is alot of chatter on the O Net lately about where results should be on sheets of paper or strung out on individual punchcards. Its amazing how passionate people can be and how upset people can get about something, that to me, seems so trivial. I think all most people want is prompt results. Choose the form that gets the results out promptly, and spend all that energy considering more interesting issues.

    One thing that came up that is important is prompt results on the web after an A-meet. Choose the form that makes this more likely. This should be the primary goal of a results system. I've been to A-meets where the results took 6 weeks to get on the web. Prompt results maintain interest levels.

    Along those lines, it seems important that the rankings are done promptly. I'm not sure why, in principle, the rankings cannot be done after each race. WRE rankings are done after each race, and the algorithms appear to be along the same lines. This is possible in the WRE system because each runner has a unique identifier, and results are required in a standard format. I don't think these things are beyond the ken of USOF over the long term. Its easy to write software to convert any sort of results page into a "standard" format. Unique id's a a little trickier, although we already have USOF membership numbers. I think the snag is that with family memberships, every runner in the family has the same number, but this problem doesn't seem intractable.

    As for the work of associating ids with runners, this is a job of the registrar, and would not be much extra work. I've been a registrar -- the registrar already has the member id on the registration form, and must already compare it against a spreadsheet provided by USOF to make sure the registrant is eligible for a championship and the USOF discount. It would not be any extra work to put this number in the database eventually used for meet results.

    Then, you should be able to chug the rankings algorithm after the results are published. I'm sure I'm missing something, but I think it would be cool to have this in the long run.

[21-Feb-02] 

    The results from Avalon made me happy. The average min/K was very high, so I wasn't as slow as I thought. The biggest thrill, however, was learning that I finished ahead of VB. To me that is the unthinkable. I don't care if he had a huge boom on one control, and that perhaps it is not legit, you take what you can get. It still is a big thrill. Interesting that our times would have been within about a minute without that boom, so if I throw away my largest boom, I still come out ahead. If it was only a one minute boom in question, it would seem more legit. Weird. (I don't care about this lame self-indulgent blather, you have to get excited about this stuff. It was enough feeling good to take almost a minute off my Mt. Misery trail run time.)

    I've been watching a bit of the olympics. I don't care for the "judged" sports such as figure skating. I like to know who's winning, making a comeback, etc., as I watch. Some the the skaters I think have beautiful routines get lame marks, while routines I think are dissonant and choppy win medals. The knowledge needed to view figure skating and understand who is doing well seems so esoteric. I wonder if the same could be said of O if it were televised to a mass audience.

    What amazes me are these sports where medals are won by 10ths of a second. How do you train for that? Does the same athlete always win by a 10th of a second, were the race run 10 times? It would seem, on the surface, that luck would have to play a factor. The ice in racer A's lane may be just a bit wetter in one patch due to variations in sunlight. Racer A missed breakfast by 20 minutes because his waiter was a ditz. Couldn't timing of when you eat affect your time by a 10th of a second? I dunno. I know my running is affected, in some way, anyway, by when I eat.

    It seems like luck would be a factor in O. No matter how clever you are, there is often the unmapped patch of green, deadfall, or rock that you hit that slows you down, that a runner who went 3 meters west didn't hit. Did early runners have to deal with a strong headwind. Etc. This sort of luck would seem to matter in a race decided by seconds, but I imagine it does not. I imagine it evens out.

    One bit of luck that doesn't even out, or so I'm told, is the start time draw. Its been estimated that late starters have a 3 minute advantage over a course of WC classic distance length due to elephant tracks. One way to mitigate this perhaps would be with forking, but then the argument would be that forking is not fair, because everyone runs a different course. But it seems it could be made "more fair" than a three minute advantage (in other words, I think it could be designed so the "luck" of which fork you got was more of the deadfall variety, and evened out over a long course). You'd have to do alot of forks, I guess, to make the elephant tracks barely visible. I wonder if that three minute estimate has any basis in science, or if it is just a guess.

[18-Feb-02]  Avalon, MD [Blue, 9010/410, 84:54, 9.42]

    Francis Hogle set a good, physical course today. I was tired. The course seemed alot more physical than the numbers indicate. I'm not sure why that is, as the woods were mostly very white. I think it has to do with the fact that the climb was in clusters. The fact that it was not more spread out over the 9K course seemed to take more a toll. Minutes were wasted climbing a stack of contours while not eating up the distance. I could not make up the time lost on the flatter or downhill sections. This is also tiring. Also, since the park was hilly, it may have led to less than direct route choices, which lengthened the course more than normal.

    Or, perhaps my route choices were lame. I don't know. I ran strong, thought I made good choices, and only had a couple of small booms, perhaps a total of 3-4 minutes lost at most. So, I should have been in around 8 min/k. I cannot really figure out why I was slow. I hope to look at the splits and times of the field when they become available to see if it was me, or unexplained difficulty of the course.

    The course had alot of long legs. I was happy with my ability to execute long, complex route choices on the legs without error and hesitation. A couple of times, though, I made poor micro route choices. I guess I'm still at the stage where I'm happy to remain in contact for the whole race, and I don't worry about these little things. But they are important.

    For example, the plan is to cut thru the green as fast as possible to get to a trail run. I know if I head in a westerly direction, I'll hit the trail. But I get lazy, if I'm off by 20 degrees, I'll still hit the trail eventually. Can't miss. Problem is that that extra 20 degrees in the green is lost time. But it doesn't show up as a boom. I think there was too much of that going on today. Perhaps that is why my time was lame. I can think of three legs where that was a problem -- relying on the collecting feature, but not caring about absolutely optimizing the route there. It is something to work on. Looking at the splits will reveal how much of a problem it was.

    Splits and comments on AttackPoint

[13-Feb-02] 

    I haven't had alot to write about lately.

    I'm starting to get more comfortable with field checking at Mt. Misery. I've found drawing the contours seems to be the hardest part. In 11 hours I've done a 4x4 cm area, except for some vegatation. The area is complicated, and the vegatation is complicated, but I hope to finish today at lunch. I really think I enjoy field checking. The vast majority of the map is not complicated, so I hope it goes faster.

    I've found a 3 day paddling trip that can fit nicely between the Yukon meets and APOC. It needs at least 3 people. I don't know who reads my olog, but if you are planning to go to both of those meets, and want to look into the paddling trip, e-mail me.

[10-Feb-02]  Riverbend, VA [Red, 7100/210, 53:14, 7.5]

    I still think the woods of Northern Virginia are the most beautiful in the world. Not that I've been everywhere, but I cannot even conceive of more pretty, fast, and open woods, and without rock or greenbriar. So I always drive 2.5 hours to the Virginia meets. Plus, it seems like good training for the US champs.

    Since it was a short course, and the woods are extremely fast, run hard with no hesitations was the game plan. I was happy that I ran fairly clean -- I guess I can't complain about being close to 7.5, though it is discouraging to be smoked by people breaking 7. If I ever break 7 in wooded terrain I think I'll throw a party.

    I had some bizarre problems. I headed for the wrong control, throwing 2 minutes down the toilet. Error of aggression. I think I made this mistake because I backdoored the control before it, and my mind marked the exit from that control as the same direction as the line coming in, but did not account for me backdooring it. After backdooring, I needed to make a dogleg, but there was no dogleg drawn, so the natural flow was to run thru the control to the wrong one, which turned out to have the same geometric relation to the current control as the correct next one. An optimization the mind makes to speed getting thru the control, that backfired in this case.

    I lost my control card, and boomed a control on bearing. I rarely take bearings anymore, but this control was in a bland, low visibility area. It was only a 200m bearing hit. I just can't take bearings anymore, it seems. Not to worry, I'm getting better at reading the map, tho some leg down the road may require a precise bearing again. I used to be good at that, but that was when I was much slower.

    Splits and comments on AttackPoint

[09-Feb-02] 

    I've been asked what a "lollipop" feature is. I guess its a contour feature that is big, round, and sweet. Specifically, biggish hills and spurs that are easily recognized and used to greatly simplify navigation. They sort of look like lollipops too. It actually comes from an interview I heard as a child of a baseball player, who while in a streak of hitting well, commented that the ball "looks like a big lollipop up there".

    This is in contrast to bland areas, or overly technical or detailed areas, or areas where you use non-contour features to navigate by, most of which often merit caution. The hallmark of lollipop terrain is that you can run as fast as physically possible among the contour features and not worry about losing contact, as long as you can keep track of which one you are at, and which one you can see next. They can't be too big so as to be useless, though (e.g., the big hill mass at French Creek West).

    As a corollary -- in detailed terrain, it helps, I think, if you can strip off all the detail and see the underlying lollipops. Take the wrapper off. I'm not particularity good at this, as I tend to be distracted (and slowed down) by detail, but I guess its a useful way to look at simplification.

    Most terrain around here is not really lollipop terrain, I think I see more of it in the southeast and west.

[07-Feb-02] 

    Next time I check "did not like map" on AttackPoint, I'll have to keep in mind the sorts of maps they used in the early days of O in the US.

    I've scanned a copy of the map from the first (I think) meet in the US, at Valley Forge Mt Joy/Mt Misery. The "quality" is from the original, not due to copying or scanning. The long course was 5K, control descriptions are on the left. The start triangle is at Washington's HQ.

    VF O Map, Nov 5th, 1967 (30K)

[05-Feb-02] 

    Good news from Canada. It looks like they moved the North American Rogaine Champs to Tue/Wed from Wed/Thu of the week between APOC and the Rocky Mountain 1000 day, at least according to the APOC web site. This is a much more comfortable fit for someone wanting to do both the rogaine and the first couple of days of the RM 1000 days.

    Lots of A meets this spring. Booked a hotel for Raven Rock today. Not my favorite park, with all the hurricane deadfall, in fact, the last time I was there two years ago, I hated it, only one of two O races I've ever hated. But I figure to give it another go -- I think small doses of unpleasant O terrain are healthy for training purposes, if nothing else. It will also emphasize areas I'm weak at, route planning around problem terrain, poor visibility O, and minimizing drift in problem terrain.

[03-Feb-02]  Little Bennett, MD [Blue, 9100/300, 68:43, 7.55]

    Little Bennett is my kind of map, lollipop features and excellent visibility. I was able to get close to 7.5m/k. Outside of Mt. Joy, its the best I've ever done on a course greater than 5K. I was tired at the end of the race, which is good. I have to look into "goo" like things. The last time I half-heartedly looked into goo, they all had caffeine, unfortunately.

    It will be interesting to see how fast the winning time was. I had the best time when I left, but there were some fast people still out. Times across the board were fast. Eric said they had a US champs here a long time ago. It got me thinking that it would be cool to have a web site listing the past results of all US Champs races.

    I had one boom of note. The map was a color copy, and not a particularly great one, and it was hard to read in places, especially in this flat area of changing types of yellow. You had to read open vs rough open vs scattered trees, and the map was one big blob of yellow. It would have been a fun area for a control had the map been perfect, but as it stood, it seemed a bit bingoish to me. I'll blame it on the map since I had a good race.

Home