12/13/07 Finally back
Well, I've probably lost all my readers, but we seem to have a web site back again.
Actually, there may still be a few bumps to come, but for the moment, we're up. I have a
fair number of blog entries written, but rather than post them and just have them roll off
before anybody comes back to see them, I'll re-date them and post them over the next
couple of weeks.
Meanwhile, one very important piece of news is that the Third Thursday Series is done
for the year. Too many people were going to miss the December race to make it worth the
trouble. The series will return in March, 2009.
12/14/07 Intervals on the needle
St. Louis terrain does not encourage you to use your compass much. In fact, Gary
Thompson has turned in some pretty strong performances at our meets without even
carrying one. As a result, I'm not quite as precise as I should be when running on the
needle. This is particularly true in light green terrain (as was evident at US Classic
Champs this year).
Today, I ran a workout that I think I'll start doing regularly. It's basically terrain
intervals, but rather than running along a handrail, I run on the needle. In this example,
the odd-numbered legs are all 500m, which takes me about 2:45 at threshold effort. As
you can see from the course, the actual control location is such that it's reasonably easy
to find, even if off by 50m or so approaching the circle. The idea isn't to learn how to be
spike a control using just a bearing, it's to be reasonably accurate while running full
speed on the needle. By running the "hot" legs at threshold, I'm not letting myself make
adjustments based on reading small details - I have to trust the compass. As the control
approaches, I pick up the feature and make an adjustment as necessary.
This workout has the advantage of honing three things at once. Like any interval
workout, it works VO2Max (it takes some discipline to run at threshold in light green, but
this time of year it's a bit easier to avoid getting hung up). It also improves compass
work. Finally, moving this fast through light green works your running technique in
12/16/07 Year end rankings
The debacle at Classic Champs knocked me out of the top 20, but I can't say I'm
displeased to see my 2007 Orienteering ranking wind up at 23rd nationally among Blue
runners. That's the highest I've ever been ranked on the elite course. It's also nice to see
that only three of those ahead of me are older than me. It seems that, at least for now, I'm
doing a fair job of overcoming age with experience. I didn't do enough events on Red to
get an age group ranking, but the two National Championships make it a pretty satisfying
season on that front as well.
Lessons? Well, the obvious one is you get good at what you train. I focused on
orienteering this year and it showed. A bit more subtle is that it's better to be looking up
than down. In 2003 and 2004 I also focused on orienteering, but I was trying to win age
group races on Red. I had good seasons and won several A-meets, but I wasn't performing
at the level I was this year. Getting beat by the
best is more instructive than winning against lesser competition.
Plans for next year? More of the same. I'd like to make one more run at breaking into the
top echelon (by North American standards; I'm well aware of the fact that the bar is
considerably higher in Europe). To do that, I need to raise my game roughly 10%.
Obviously, at age 45, I won't be doing that with fitness. There is some hope, though.
Three of my A-meet results this year were at that level, so I know that I can turn in that
type of performance with the fitness I've got. The key will be consistency with technique.
It's a stretch goal to be sure, but it should be fun to try.
12/17/07 Possum Trot meet report
... is here.
12/18/07 Not yet, but...
Kate and I saw
Beowulf recently. It's the most serious attempt so far to create an animated movie
that looks like it was shot with live actors. While the movie was pretty cool (and fairly
well acted, in terms of the spoken parts), the Screen Actors Guild has nothing to fear for
the moment. However, take a step back and look at the trend and maybe they should be
Twenty years ago, ray tracing (the fundamental technique for creating realistic computer
generated images) was understood, but the computational demands to pull it off were
such that films from startups like
Pixar were limited to short stories involving inanimate objects in relatively simple
settings. Increases in computer horsepower over the past two decades have eliminated
such limitations and the main obstacle now is that accurately modeling all the little quirks
of human motion (particularly
micro-expressions) is pretty tough without hooking all the motor control to a real
brain. However, with the speed of a desktop PC expected to surpass that of the human
brain sometime in the next 10-20 years, there's no reason to think that hurdle won't be
cleared in the relatively near future. It's entirely possible that by 2020, shooting a live
action movies will be akin to shooting one in black and white today: a novelty technique
that is used to evoke nostalgia, but not one used for mainstream production.
So, why am I rambling about movies without the usual disclaimer that today's entry has
no racing content? Because orienteering and adventure racing face a similar dilemma,
although very few competitors want to acknowledge it. Currently, GPS is banned, but
nobody really pays much attention because it wouldn't help the elites go any faster
anyway. The main reason GPS doesn't help is that all it does is keep you in contact with
the map. The top competitors rarely lose contact and when they do, they can relocate off
features about as quickly as they can read GPS coordinates and plot them on a map (this
is less true in adventure racing where night navigation and rough maps make errors more
frequent and relocation more difficult).
But, that's not all GPS can do; it's just what's currently built into most GPS watches.
Bring in a tiny digital camera to snap a picture of the map at the start of the event.
Add a processor that can make good routing decisions based on the map information.
Hook it to a heads-up display that would project the best path and distance to the control
onto a pair of glasses. Now, the competitor merely needs to follow the arrow into the
control. You could even add voice activiated feedback so the unit could adjust the route if,
for example, the light green turns out to be slower than assumed.
Science fiction? Hardly. All of this can be done today. It would be expensive to
implement in a small package, but all the technology currently exists. In another 5-10
years, there's no reason to think it would be all that expensive. Look another 10 years
down the road and there's no reason to think that a good portion of the population won't
have GPS-enabled communication devices implanted in their bodies with direct
connections to the brain.
Is our sport doomed? I honestly don't know, but the fact that nobody's even seriously
considering the consequences of these developments leads me to believe that we'll be
caught flatfooted by these changes.
12/19/07 Dante's Interval
In a bit of a departure from my usual reading list, I was looking at Dante's Inferno
the other day. I spotted one thing that escaped my notice when I struggled through the
text in college: Dante knew a thing or two about hill training. Check out this verse which
is right there on the first page:
After my weary body I had rested,
Take a minute to form the image in your mind. On a level trail, the foot on the ground is
always lower than the foot in the air. Going uphill, that's not the case. If the foot on the
ground is always lower, that means you've got enough spring in your stride that the back
foot has come up a fair bit before your front foot hits the ground. Given that he's already
implied that it's a pretty substantial hill he's ascending, this description is basically
telling us he's hammering the climb pretty hard. Throw in the line about resting
between efforts and it seems that Dante was an advocate of hill repeats several centuries
before they came into vogue.
The way resumed I on the desert slope,
So that the firm foot ever was the lower.
E. Molitor (AKA
Yukon King asks, "Yesterday in our local little paper they had a reference to a
'slingshot' move in bike racing - what is that?" The short answer: tons 'o fun.
There are several forms of the slingshot maneuver. The legality depends on the type of
racing. The most effective form (and the form that is always forbidden, except in
Madison events on the track) is properly called a handsling. In a Madison race (named
after Madison Square Garden in New York, where the event was hugely popular in the
late 1800's), riders pair up and alternate turns on the track. The exchange is done by the
current rider tagging the hand of their teammate. The exchange can take place anytime,
so a big part of the strategy is deciding when to tag off.
It didn't take long for riders to realize that instead of just tagging the hand of the rider
coming in, the contact could be used to transfer some momentum as well. The most
effective way to do this is a handsling. The rider coming in gets himself going and then
rolls down the banking (gaining more speed in the process) straightening out just above
the sprinter's line (1m from the inside of the track). The inside (left) hand is held out,
even with the hip. The tagging rider passes on the inside and grabs the hand. Both riders
then pull hard and the incoming rider is literally thrown forward while the exiting rider
slows by a roughly equal amount. Having been accelerated up to a pretty good speed
without expending much effort, the incoming rider proceeds to kick as hard as possible,
achieving a speed that would be difficult to attain without the sling.
This is typically done on sprint laps with the incoming rider finishing off the sprint. It can
also be done to create a breakaway. For obvious reasons, this sort of thing is taboo in all
other races. The only reason it's not disasterous in the Madison is there typically aren't that
many teams in the race and everybody is expecting it. Despite the ban, it happens quite
frequently in road races. If done in the middle of a big pack, it's not an easy thing to
catch (unless it's botched and causes of one of those spectacular pileups).
Another form of the slingshot is the "butt push" which is basically the same thing except
the push is done from behind rather pulling as you pass. This is also illegal, but it's so
common on the road (both in sprints and on long mountain descents) that it generally gets
ignored unless it's extremely flagrant.
The legal form of the slingshot doesn't involve any actual contact, but is simply a way to
maximize the draft. The lead rider begins the acceleration and the following rider jumps
just slightly less. Through the lead rider's push, the gap continues to open so that when
the lead rider tops out, there is a 2-3m gap between the two. The following rider
(benefiting from both the draft and a slightly less violent acceleration) continues to
accelerate through the draft, pulling to one side at the very last instant to get around the
Jon Torrance responded to
my 12/18 post on technology with the following points:
I'm almost certainly vastly ignorant of the state of the art in software tasks analogous to
interpreting an orienteering map and making route choices but my sense is that right now
developing such a navigation system, even without trying to make it small and light
enough not to greatly affect running performance, would be the sort of task DARPA
might be up to but neither they nor anyone else with the resources has much incentive to
try. Technological progress may change that faster than I imagine but I can't see that the
sport is doomed until it's plausible to imagine that the resulting system would be invisible
to the naked eye when in use.
I'm curious though - what do you imagine we might be doing differently if we were
seriously considering the possible consequences of such developments?
BTW, I have a hard time imagining there isn't going to be a significant time penalty
involved in obtaining a sufficiently good digital picture of the map and converting it into
a computer-usable and adequately precise representation of the terrain while on the clock.
Though I suppose having a reasonably good digital elevation model of the surrounding
area already in the system before the start might allow the system to accept a distorted
image of the map and still produce a good model of the terrain.
OTOH, I can't see why the system would need voice feedback to be told an area of light
green is slower than anticipated. If it's got a clock, real-time GPS data and the smarts to
make good route choices, it'll presumably know how much the light green is slowing you
down better than you will.
While Jon may not be an expert on software, he certainly knows a lot more than me about
orienteering. Thus, careful consideration of his views seems appropriate. At the heart of
the matter is the question: what should we do differently? I'll get to that, but I'd like to
clarify the smaller issues first.
On the first point of developing a navigation algorithm, he's absolutely correct. While it
can be done today, nobody with resources adequate for this task is going to put them on
such a niche application. However, who said that an algorithm needs to be written?
Currently, we develop algorithms to compensate for the fact that computers are so much
slower than a human brain. That will only be true for another decade or two. Once
computers are matching us in speed, there will be no (or at least greatly lessened) need to
write algorithms. Computers will be able to solve problems the way we do: trial and error
with feedback. Exposing a neural net to several thousand maps with routes taken by
WOC-level orienteers would give an excellent set of base rules that could then be honed
through actual competition. This is not particularly hard to do. Assuming that one will be
able to purchase "off the shelf" neural nets that can be used to make decisions based on
visual patterns, any amateur programmer would be able to do this with a few months of
The packaging of the system is a non-issue in my mind. I expect that peripheral device
and knowledge implants will be available by 2020 and fairly common by 2030.
Excluding such things from competition will mean excluding a significant chunk of the
Scanning the map at the start is probably the biggest current limitation, but the
easiest one to solve in 5-10 years time. Consider what's happened to the digital camera
market in the past 5 years. We just bought Yaya a digital camera for Christmas. For $50
we got a camera in a weather and shock resistance package that a four-year-old can easily
operate. The images rival the very best digital rendering from earlier this decade. Just as
we correct for distortions by taking repeated glances at the map, the interpretation
algorithm could use multiple images to reduce problems with the capture.
The bit about voice feedback is valid; that wasn't a good example. My point was that the
information doesn't need to flow in just one direction. The navigation program should be
viewed more as a partner with whom you converse. Depending on the implementation,
the mode of communication may or may not be verbal and could range over any number
of topics, just as teammates discuss many aspects of the race in team events. Whether the
program would also be willing to talk about all the irrelevant stuff that comes up at 3AM
when you're just trying to keep the team awake will be a deferred design decision.
So back to the big question: what to do? In general, there are three approaches. The first
is to pretend that this stuff isn't coming. That's what we're doing now and I think it will
result in the collapse of the sport if followed for too much longer. The other two are to
fight technology or to embrace it. I'll look at each in the next few days.
12/23/07 Fight it
Let's examine the preservationist (or, if you prefer, Luddite) road first. I'm going to take
the following points as axiomatic in 10 years time, though some would debate them:
Computers will navigate better than people. By "navigate," I mean the ability to know
where you are at any given time and which direction you should proceed. This point
could be debated today (though I'd certainly be willing to defend it), but it will be
indisputable in 10 years.
Computers dedicated to solving problems in a reasonably narrow domain (such as
navigation), will be able to make good decisions faster than people.
Computer decisions will not degrade significantly under the stress of competition.
Computer input devices will be at least as good as our natural senses.
Computing devices (including sensory input) will be small enough that they can be
carried (or concealed) with no performance penalty.
If any of those offend you, just change the time horizon. Computers will do everything
better than we do by the middle of the century. It's not a question of if, but when.
If the goal is to preserve the sport in more or less its current form, all this technology has
to be banned outright. There are three problems with this. The first is that we already
allow two very important pieces of technology to be used: compass and watch. As
anybody who's applied for a patent in a high-tech area knows, it is becoming increasingly
difficult to define devices using simple nouns. Watches do a lot more than just tell the
time and there is no shortage of compasses that have additional features built in. The
acceptance of these devices in current competition will make it more difficult to write
rules that exclude the extras. We'll either have to ban compasses and watches altogether
(the more promising course in my view) or put up with increasingly complex exclusion
rules (and the accompanying protests).
More serious is the fact that, banned or not, these items will be used in
competition. Once made sufficiently small (or implanted), it will simply not be feasible to
catch cheaters. One could argue that in an amateur sport cheaters are only cheating
themselves and the rest of us can simply abide by the rules, but just because there's no
money on the line doesn't mean it's fun to be beat by someone who cheats.
But, the biggest problem by far is the fact (presumed, but a reasonable presumption) that
the bulk of the population will have these devices implanted in a way that will not be
easy to turn off. Nearly all cell phones carry GPS now. Navigation systems in cars are
becoming increasingly popular despite a pretty hefty price premium. Those two facts lead
me to believe that, once available, people will flock to have nav-enabled communication
devices implanted. I don't see why that would be much more than 10 years away. I'll
concede that there may be some hesitance because people are naturally queasy about
hardwiring something to their brains, but it will be technically feasible. At any rate, when
it does happen, the preservationist road starts to look pretty bleak.
So, does this path make any sense at all? I think it does, if only to buy us an extra decade.
Here's what I'd suggest to get us pointed that way:
This buys some time, but sooner or later maintaining the sport as a "human only"
endeavor will collapse under the weight of greater societal and technological changes (we
haven't even talked about the fact that folks with artificial legs will no longer be
"handicapped" but "enhanced".) But, the time may be useful, because implementing the
alternative strategy will take time. I'll take a look at the "embrace" road next.
Ban all electronic devices except for punch units from major competitions. Yes,
that includes your watch. You don't need it if you're epunching since you'll get your
splits later. If we must keep compasses (and I suppose we must, just because it would be
so difficult to separate the sport from the manufacturers who have spent the last century
developing it), insist they be based on a magnetic needle with no electronic enhancement.
Continue the emphasis on sprint and middle distance for individual competition. These
courses typically feature short and/or complex legs with lots of micro-choices. These are
the types of legs where computer nav will take the longest to catch up with human
Ditch the individual long events and replace them with mass start "goat" type events
where pack tactics and fitness count for more than rough navigation.
Discontinue the practice of using GPS/mobile phone technology to track the top
competitors at international competitions. You can't have it both ways. If you don't want
the technology to be used, don't use it.
12/24/07 Resistance is futile
Sports, like just about every other human activity, evolve. This is not necessarily a bad
thing or a good thing. Sometimes it makes sense to push back on the evolution. Other
times pushing back just results in silliness. I think the foremost example of the latter is
"traditional" cross-country skiing. There's nothing traditional about using poles taller
than you are to push your way around a course on skis with no kick wax. If the powers
that be wanted to roll back the clock on the sport, they should have stopped grooming the
trails. Of course that's ridiculous, but that's the point: you can't roll back the clock. Once
everybody realized that the diagonal stride was inferior, there was no bringing it back.
On the other hand, drag racing has done an outstanding job of limiting technology in a
way that both preserves the sport and allows people to look for improvements. Jet-powered
dragsters have been around for quite some time. If the
NHRA was to embrace jet dragsters as the premier
class of drag racing, I expect the sport would die a fairly quick death. Jet dragsters are
exceedingly costly and dangerous, which means that few pros and no amateurs would get
to drive them. They're not that interesting to watch, either. As a novelty act, they do a
nice job of filling the night time intermission with their pyrotechnics, but then the
attention is turned back to the classes that really get the fans going: Top Fuel and Pro
Stock. These classes capture the imagination because they resemble what you do every
time you leave a stop light. Sure, the cars are purpose built and go through several gallons
of gas in the four or five seconds that elapse, but they are doing the same thing a normal
car does: converting heat to speed via an engine, transmission, and some really
sticky tires. A jet dragster is faster, but we have no frame of reference to assess the
If the problem was just one of limiting technology on the course, I wouldn't have any
worries about orienteering. The steps outlined yesterday would be adequate, perhaps even
overkill. Unfortunately, there's a bigger issue looming: the ascension of silicon-based
life. You can plug your ears and sing "la-la-la-la-la" if you don't want to think about it,
but purely "biological" humans are only going to run this planet for another 30 years or
so. After that, the dominant "species" will either be a heavily augmented form of Homo
sapiens or some sort of completely new artificial life (assuming the development of the
former, I don't expect the latter to be
superior until 2060-2070, but some who know more about it than I have predicted
One might rightly wonder if the future of orienteering really rates as a serious concern in
light of such a change, but I would argue that we ignore these "small" issues at the peril
of not being able to deal with the big ones. If we can't figure out how to incorporate an
artificial life form into a simple pastime, how are we going to respond to its request to buy
a house, vote, get married, adopt children? These changes are going to be such a shock to
our society that I think it makes sense to start small when thinking about them.
And, just like everybody else, I think I'll put that off until tomorrow, because I've got a
biological life form of my own making that needs my attention right now.
12/25/07 Embrace it
If we're looking for a model of a sport that has successfully embraced technology, one
could do a lot worse than sailboat racing; particularly the offshore variant. To the casual
observer, ocean racing still pits crews against the elements in an effort to get across the
sea using only what nature gives you. All the drama (and a good bit of the danger) is still
there. On closer examination, these boats are closer to the space shuttle than their 19th
century progenitors. And, you literally could launch a satellite for less than the typical
budget of an America's Cup team.
Ocean racing boats make extensive use of electronic instrumentation. Aside from the
obvious use of GPS and weather instruments, sensors monitor the attitude of the boat, the
shape of the sails and relative position of other competitors. While the ultimate decision
still rests with whoever has their hand on the wheel, the computer indicates the best
course to windward, position of the lay line (the point at which you can sail straight to the
next mark without tacking), and assist with sail selection.
You might think that all this takes some of the skill out of the sport, but it doesn't. I've
crewed on ocean racers and can attest that there are still plenty of ways to get it wrong
(and all the gadgetry in the world doesn't take any of sheer terror out of sailing in a gale).
If anything, having being relieved of the guesswork, the premium has shifted to strategy and
preparation. Having better data is akin to having a better map in an orienteering race: it
removes some of the luck, but you still have to know how to apply the information. A
good crew will wax a bad crew every time unless one boat is significantly faster. The
rules are tight enough that such speed mismatches are rare.
So the question is, can orienteering incorporate these types of devices while still keeping
the fundamental challenge of the sport? My take: sort of. The sport will be much
different, but it will still be a challenge. The specific direction will be determined by what
parts of the sport are valued most by the participants. We'll need to take a step back and
come to terms with that before looking to the future.
12/26/07 What are we trying to preserve?
Lisa Carr solicited the Attackpoint crowd for reasons they orienteer. You can read the
thread for yourself if you want the details; here are the main positives cited:
Combination of physical and mental challenge. Correlated to the mental aspect were the
opinions that orienteering was less boring than most endurance sports and that one
could continue to improve later in life.
Feeling of freedom. This one was taken many ways - freedom to choose your own route,
freedom of running your own race, and the more general "free" feeling one gets when out
in the woods.
Sense of purpose in finding controls.
Fascination with maps and puzzles.
Camaraderie of the orienteering community.
While Attackpoint is an excellent sampling of competitive orienteers, one also needs to
consider the desires of the recreational participants. Without them, the sport would
simply not have the numbers to exist. I wouldn't say I'm particularly well plugged into
that group, but my time on the SLOC board gave me some exposure to their preferences.
Their list is not that different, but one would put a lot more emphasis on the
second and third points. These folks are mainly looking to get out and enjoy the woods.
Finding the controls gives structure to their activity.
Any future version of the sport that I can contemplate satisfies the last three desires. It's
the first two that are tricky. The reason for this is that, when combined with the coming
technological shifts, they are at odds with each other. Not to the point of being mutually
exclusive, but certainly competing priorities.
Preserving the feeling of freedom is easy. Just let everybody use whatever they want and
send them out on the course. However, without some limits on what constitutes legal
means of progress, this would hardly even qualify as a sport, much less one people would
One could preserve the mental challenge easily enough by switching to a virtual
environment. There, GPS, or any other real-world navigation aid, wouldn't do
much good. We'll have to see where virtual reality goes, but it's possible that a virtual
environment could adequately test the physical aspects as well. However, I don't think
many of us would find this nearly as satisfying as being out in the actual terrain.
Finally, we need to come to terms with just what is meant by physical and mental
challenges because these very concepts are going to be at the heart of the larger
technological changes. We must assume that, as with technology today, there will be
early and late adopters. This includes everything from communication and navigation
implants, to physical and sensory enhancements, to brain enhancements, to outright
re-hosting of individuals into virtual or cybernetic bodies. The difference in performance
between the early and late adoption of technology is typically 2-3 orders of magnitude.
Pitting a jet against a bicycle doesn't make for a very interesting sport.
My take is that the virtual reality route is the only one with long-term promise, simply
because I expect that everything else about our lives will head that way as well. The need
for the physical world will likely diminish as we become more accustomed to being free
of it. In fact, it's quite possible that future generations will find life in a 3-dimensional
space where things happen so slowly quite dull. Nature may be fascinating in slow
motion, but when our brains are a thousand times quicker than they are now, it will
appear to stop altogether. To challenge such processing capacity will require some sort of
So, back to the original question: where to go from here? I'll look at that in a final post.
OK, I lied. I implied yesterday that I had just one last post on this subject. However,
when I started to write it, I realized that one more piece of background was needed. To
consider where we should go in embracing the ascendance of technology, it's helpful to
look at a sport that has already been overrun by it: Chess.
Maybe you don't like calling Chess a sport, but it's certainly a competition and one
where computers are better than people. All people. That's an important point. It's one
thing for us to dismiss navigational systems claiming that the elite are better than that.
That's what the chess community did for many years. As late as 1993, World Champion
Gary Kasparov was clinging to the notion that computers couldn't play at the top level
because they lacked the "fortitude". When he got beat in shorter games, he (rightly)
claimed that the restricted time limit favored the machine. But, in 1997, he was
beat outright using the time limit for international games and there was no denying
that human supremacy in the area of Chess was over for good. Well, actually Kasparov
did try to deny that for a while, but it was pretty obvious to everybody else.
While that may have been a blow to our collective ego, it didn't immediately threaten the
sport. Computer chess was allowed to advance on it's own with separate tournaments.
Nearly all players, even at the Grandmaster level, used computers to train, but over the
board it was still one person against another. However, less than 10 years later, the
strength of Kasparov's rival, Deep Blue, was mainstream technology. Suddenly, anybody
with a laptop, an off-the-shelf chess program, and an earpiece could cheat their way to a
major tournament win and
I played my last Chess tournament in 1999. I was never that good. In 1994 I was actually
the median tournament player for the entire country and since most tournament players are
young, that meant I got beat by high-school kids pretty frequently. Computers had been
busting my chops for over a decade when Kasparov went down to Deep Blue, but the
knowledge that nobody could beat a machine certainly lessened my interest in
continuing (although, I was pretty much losing interest by then, anyway).
It didn't take much foresight to see the cheating thing coming. For the moment,
tournament organizers are relying on detection. That's not going to work for much
longer. Whereas there are some real physical constraints that currently make cheating at
orienteering problematic, cheating at Chess is currently easy and about to become trivial.
It's hard to imagine that the greatest game in the Western world will vanish overnight,
but it won't surprise me at all if tournament Chess as currently practiced is completely
undone by cheating in the next 10 years. The only form of competition that won't be
completely bogus will be computer tournaments. The World Champion will not be the
best player, but the best programming team (or engineering team, if hardware is not
That may sound bleak, but consider this: the World Champion has always been
the best programmer - all that's changed is that we're programming silicon rather than
neurons. The way you program a Chess computer is to expose it to lots of patterns and let
it play out the position. That's exactly the way masters have trained since tournament
This is the point that needs to be understood if we are to carve out a future for
orienteering. What we do when we train is the same thing that will be done to train
navigational systems based on neural nets. Most experts in artificial intelligence believe
that neural nets are the only way computers will surpass human intelligence anytime
soon. Thus, turning a sport into a programming contest doesn't change the mental aspect
of things nearly as much as one might imagine. Preparation still caries the day and the
more training, the better.
With that established, we can finally look at where we should be heading when the
stopgap measures of the "fight it" path are no longer feasible.
12/28/07 Where from here?
In 2004, I directed an A-meet at Hawn. I was approached by the
St. Louis Area Geocachers Association about putting on a combined meet with the
idea that our memberships might have a lot in common. For those who don't know,
Geocaching is basically recreational orienteering with GPS. Well, it turns out the two groups couldn't
be much more different. The geocachers were a fun lot and both camps were friendly, but we
just didn't see things the same way. This is why it's so important that we start
directing the future of our sport in light of technology rather than letting technology
shape it. If we don't, we might wind up with something we don't like very much.
Obviously, this direction will not be one person's vision. It will be the result of many
ideas being tried and the best ones identified and pursued. So, without any further
explanation or apology, here are some ideas. All of them are flawed, but if everybody
started just getting thoughts out, we might find a few that lead us to a place where we're happy.
To do otherwise pretty much guarantees the opposite.
We already have a pretty good orienteering simulator in
Catching Features. The program's author,
Greg Walker, doesn't bill it as anything more than a game, but it is widely used as a
training tool and has hosted some moderately serious competitions. Nobody is claiming
that proficiency at Catching Features is equivalent to skill in the actual terrain, but the
correlation is fairly high once one gets past the initial learning curve.
Aside from being a well-written simulator, I think there are three things that have helped
Catching Features become the de facto standard for computer orienteering. First, the
program operates online allowing for the creation of a community of users. Rankings are
given not only to individuals, but clubs and countries as well. Second, the game speeds
up time. This may seem like a small thing, but it's not. Orienteering is hard because you
are doing it under duress. When the physical aspect is removed, you need something else
to keep you under pressure. This will become increasingly important as our brains are
augmented with additional processing power. Finally, and most importantly in the long
run, the game is truly virtual. It comes with a map editor (OK, it's clunky, but it does
work) so you can create just about any terrain imaginable. "Fantasy" maps frequently get
high ratings by online competitors who enjoy the challenge of something completely
different. I expect that, as we become more accustomed to virtual spaces, orienteering in
higher dimensions may provide the needed additional challenge.
Performance is the outcome of preparation. All great athletes know this. However, sports
are not deterministic. Upsets occur. Computer programs can be made to appear random,
but they are not (at least not with today's technology). Given the same data, a computer
will reach the same conclusion. "Randomizing" the program doesn't change this assertion
- the random seed is part of the data. That's generally considered a good thing, but
whereas a person may produce a surprisingly good or bad result, programs tend not to.
Whether this will change as increased processing speed allows for tackling increasingly
difficult problems remains to be seen. What is clear is that the quality of the program is
hugely important in producing a good outcome. This is also true with humans, but the fact that
we are constantly programming our brains to handle all sorts of problems (and
generally doing a good job of it) often masks this. We tend to talk a lot about "skill" and
"talent" when we really should be noting perseverance. While it may well be possible in the
near future to purchase a neural net designed to navigate, without some real effort in
training it, the results will be quite dismal.
My hope is that the development of these systems parallels what is being done with
technique training today. The basic nets may be developed by clubs or national
federations (just as clubs and federations currently hire coaches) and then the individual
athletes work on training them to their own preferences. Just as elite competitors are
generally open with sharing training tips with the rest of us, I'd like to see some open
source development with various elite competitors uploading their own experiences. This
may seem like giving the game away, but Chess Grandmasters have been doing exactly
this for years and the result has been a tremendous improvement in the quality of play at
all levels. It's true that an opening novelty is usually kept under wraps until used in
competition, but once played, the inventor of a successful move typically publishes their
analysis and others are quick to add their own comments. If this can work in a
professional sport where knowledge is everything, it should be quite feasible among us
Much of the drama of sports comes from the perception that much is at stake. In some
sports, such as auto racing, mistakes can actually be fatal. Usually, the penalty for error is
significantly less, but the important thing is that it feels like mortality is on the
line. Look at the language we use to describe races gone bad: "I died at the end", "that
mistake killed my chances", "there was no saving it", etc.
Hyperbole aside, the emotions of competition are often more intense than the emotions of
real life and death struggle. I've been in three situations that one could reasonably call
life-threatening: a head-on car crash, caught above tree line during a thunderstorm, and
having a heart attack (OK, that last one was a false alarm, but I had no way of knowing
that at the time). All three were stressful and, particularly in the case of the thunderstorm,
there was some very genuine fear. However, I wouldn't rate any of them in my top 5
emotionally-charged moments. Those slots are all occumpied by situations when the
race, not my life, was on the line.
A lot of this has to do with the fact that large chunks of your brain shut down when your
pulse approaches maximum. In physically stressful situations, we revert to our more
primal impulses. I'm sure this was useful when not getting eaten was a daily concern for
our species. Now, it's just a rush and it's a lot of what we like about participating in
sports. Whether we find some way to make orienteering work in the terrain or whether it
becomes completely virtual, it will be necessary to force the competitors into stressful
situations or the exercise will seem less like sport and more like doing math homework.
So, where do we go?
Obviously, I don't have a short answer. If I did, I would have stated it a week ago. But I
do believe that the general direction should be to implement the "fight it" suggestions as
soon as possible and start seriously developing a parallel competition track based on
virtual reality. A middle ground, for those who insist on having such a thing, would be to
start embracing the use of technology in the terrain, if only to get a read on how much it
is threatening the fight it path.
In closing, I'll go waaaaay out on a limb and predict that around 2017 an Orienteering
World Champion will lose some sort of exhibition match to a very good runner using
navigation aids. If that is our wakeup call, the sport will die almost immediately after. If
we've seen it coming and are already heading down a different road, we may wind up
with something even better.
12/30/07 2007 Wrap
To say 2007 was a good year for race results would be a colossal understatement. The 21
wins tied my career high and two of those wins were National Championships. I also got
my first "Silver" level national ranking among elite orienteering competitors.
Perhaps the most gratifying
finish was finally breaking the hour at Pere Marquette. All three of those were on my list
of "reasonable" career goals, but when they haven't occurred by your 44th birthday, you begin
to wonder if they will.
Unfinished business? Of course! There's always something more to shoot for. The two
big ones from this year were the very near miss at a sub-3-hour marathon and not making
the US Orienteering Team. The 3-hour marathon still seams like a reasonable goal.
Making the US Team is looking more and more like a stretch. However, it's good to have
a few stretch goals, so both of these get carried over to 2008.
Under the heading of year-end silliness, I went to the trouble of designing a couple control
picking courses today to push my annual control total to 3000. I didn't actually get a chance
to run them, so I'll have to settle for 2928. While there's nothing magical about hitting a round
number like 3000 (particularly when the total is padded with city park workouts like
what I drew up today), I think keeping a control count does serve some value. I've definitely noticed
that I'm sharper when I'm hitting 50-100 controls a week than when I'm getting in fewer
than 25. Running to a specific point is better training than just running on a map. Running
on a map is better training than just running. Running is better than doing nothing. The
bottom line: totals do matter.
Pursuing a high control total just for the sake of meeting a training goal will bias you
towards putting a lot of short legs in your training. This is not really a bad thing. Just
about any long leg can be broken into a series of short legs. If you can run short legs well,
you should be able to run long legs well. The opposite is not true. You can get away with
a lot of sloppiness on long legs simply by running the "green light" sections fast, but
short legs have to be precise because even small errors are significant.
Of course, there are skills that are best trained with longer legs. Rough compass, route
choice, and terrain speed are not particularly important on most short legs. However,
compared to the time losses incurred when map contact is lost, the difference between
being proficient versus merely competent on such skills is fairly minimal. Running short
legs forces you to stay in contact all the time.
Aside from weekly and yearly totals, I believe the lifetime total is pretty significant as
well. Every time you hit a control, your brain matches what it sees in the terrain with
what it sees on the map. After a few thousand of these, the pattern matching becomes
pretty automatic and you are rarely surprised by what you see. However, there's a big
difference between being able to read and true fluency. It takes many thousands of
map/terrain matchings for your brain to automatically construct a 3-dimensional image of
the terrain every time you take a glance at the map. You should be doing this all the time;
not just at control locations, but most people benefit from the discipline of doing it at a
I'm often asked how I train navigation without hanging markers. I believe that training
without markers is actually more beneficial because it forces you to look for the terrain
feature. The downside is that sometimes you get taken out of your rhythm because the
map isn't quite right (or you just didn't read it right) and you have to spend some time
verifying where you are. To me, that's a small price to pay for having the ability to verify
your location any time on a course and not just when you're at a control. If you
have an active local club, you can get in plenty of training with actual markers by
attending all the local meets.