Blog Archives

6/1/07 Accidental generlization

When you decide to map a clump of trees as a small patch of white woods rather than mapping the individual trees, you're making a choice. Even if the decision of how to generalize is so obvious it's nearly automatic, there's a part of your brain that is going through the exercise of forming a pattern to represent what's on the ground. There's another form of generalization that's becoming increasingly prevalent: sloppiness.

You'd think that with all the fancy mapping tools invented (or at least made affordable) over the last 20 years, maps would be getting better. They're not. At the highest level, I'd say the quality is holding, but there's been a noticeable slide even at the A-meet level. Local maps are getting downright bad.

While it's unbecoming of an IT professional to blame technology, the ability to make a map quickly and cheaply is part of the problem. When clubs were forking over big bucks for a set of color plates and then printing several thousand offset copies, there was a real incentive to make sure things were done right. A map cooked up by volunteers and inkjet printed on an as-needed basis is about as low-risk as you can get. The stakes go down; so does the quality. The result is that we have a lot more maps, but fewer good ones.

Even if you don't intend to use it for elite competition, there's no reason your quick and dirty map needs to look bad. Here are some tips to consider when turning your field notes into the final version of the map:

• Get a copy of the IOF specs. You can't conform to a standard if you don't know what it is. The IOF specs are free and can be downloaded from the IOF site. Make sure you're using the right symbol set in OCAD for the spec you're using.
• Learn to make an accurate curve in OCAD. This is the most common area of sloppy cartography. You've already made approximations in the field. If you compound them by not copying your lines accurately, you'll wind up with stuff that's off enough to matter. More on this in a future entry.
• Overlap your areas. Many very good cartographers don't find this necessary. However, I think most amateurs do benefit from having some overlap where different colors come together. This eliminates little white spaces between the areas and also makes your larger areas easier to draw. Know which areas take precedence so you draw the accurate boundary on the top-printing area.
• Related to the above, break large areas up into overlapping smaller areas. Again, this is not necessary, but it makes the task easier. If the task is easier, you'll be less tempted to cut corners.
• Pay attention to junctions. Make sure your roads, trails, streams, etc. all come together cleanly. This typically means making sure that both lines have a "dash" point at the junction. This is particularly important in the case of trails because if the dashes don't meet, that implies that the trail junction is indistinct (if that really is the case, the proper technique is to not intersect the lines at all, since there's no such thing as a "non-dash point" in OCAD.
• Pay attention to line thickness. This is especially crucial on sprint maps where the width of a border can be the difference between a crossable or uncrossable feature. Know which borders to use with which symbols. For example, if you are using a street outline with a pavement area, know if the outline will print on top of the pavement (thus revealing it's full width) or under the pavement (only the half extending beyond the area will print, cutting the line width in half). Don't assume that just because OCAD labels a symbol "street outline" that it's the right one to use on the edge of a street - you need to understand how the symbols work together.
• Take pride in your finished map. Even if it is a quick effort or one that reflects, shall we say, "developing" skills on your part, you'll do a better job if you have it in your head that the finished product will look "nice". Put a title and border on the map. Make sure the scale and contour interval are printed large enough that someone can check them quickly during their run. Add essential information like who did the mapping and when. If possible, put a decent legend on the map (sometimes you don't have room on a sprint map). Bonus points for carving out a space to overprint a clue sheet. Align blocks of text rather than having stuff scatterred around randomly. Frame it all on the page in a way that makes sense.

Although that sounds like a lot, mastering those guidelines will not only improve the quality, but also speed the process (except for the last bit, but if you've done a good job, you'll want to spend an extra half hour on the layout to show it off).

6/2/07 Cliff Cave "Sprints"

SLOC put on a meet today at Cliff Cave. Bill Langton set courses. Because Cliff Cave is a relatively small park, he decided to set a couple sprint courses.

The second course was a fairly typical (albeit slightly long) woods sprint. It had lots of route choice and most of the running was fast. If I hadn't been saving my legs for tomorrow's adventure race, I would have run this course pretty hard the whole way.

The first sprint was more like a middle course. There isn't much route choice and most of the legs demand you run straight on the red line. Furthermore, the woods are really thick at Cliff Cave this time of year so you weren't moving that fast, even if you were pushing.

I don't mention this as a complaint; I found both courses fun and was happy to run them. I'm merely pointing out that sprints are not supposed to be just really short versions of regular courses. Sprints are supposed to focus on the skills of fast decision making and map reading at high speed. The first course didn't do that.

6/3/07 Back in the saddle

David Frei and I teamed of for the Conquer Castlewood Adventure Race today. It's the first adventure race I've done since the Castlewood 8-hour back in January. It was great to finally do one again (even if the format of Conquer Castlewood is pushing the definition of Adventure Race a bit).

Although it was the weakest part of the race for me, I really enjoyed racing the bike again. The trails were really greasy from all the rain we've gotten this week and I went down a few times. Far from pissing me off, I liked the challenge of staying focused when things weren't going right. David and I were in the lead for most of the race, so there was a lot of pressure to get back up quick and not get flustered.

The short story is that is was a very gratifying win. A full report is coming soon.

6/5/07 Drawing a curve

Finishing up the series on making a quick map, I'll write a bit today about something cartographers used to take for granted: drawing an accurate curve.

One of my first jobs was as a cartographer. I was 12 and my dad was doing some environmental impact studies. He'd project an aerial photo up on a big sheet of paper taped to a wall and I'd draw in the vegetation boundaries. He also introduced me to the drafting stereoscope, which I felt was the coolest thing ever invented. I still use that same stereoscope to make base maps today.

While nobody's ever accused me of having good penmanship, it didn't take too long to get reasonably good at drawing the lines accurately. A good set of drafting pens makes a big difference. With OCAD, your line usually looks good, but that doesn't mean that it is.

OCAD removes the relatively minor hurdle of learning to use a drafting pen and replaces it with the seemingly easier task of create curves using a mouse. In truth, I'd say that drawing the map freehand is less work. True, you don't have to worry about messing up hours of work with a single mistake, but any reasonably coordinated person can get the curve exactly right when tracing freehand. Doing that with a computer tool is more complicated and I think a lot of recreational mappers settle for "close enough."

That's not too much of a problem when doing a quick map of an urban park. In complex contour terrain, it's disastrous. When you lose contact with the map in complex terrain, you need to be able to rely on the shapes of the contours to relocate. If the contours have been simplified and/or approximated, this is much more difficult to do.

The rest of this entry is going to get fairly technical, so if you've never used OCAD, you may not get much out of it.

Let's dispense right off the bat with creating curves using the "freehand" tool. It sucks. The documentation admits that it sucks and tells you not to use it. That leaves the curve tool which builds a curve by forming what are known as splines from a set of points and slopes that you enter with the mouse. When the curve is being built, OCAD does its best (and it's pretty good at it) to determine how much influence each point should have over the next spline. Once you get the hang of where to put the points, you can usually leave this alone. However, for the odd case where the line isn't just right, you can adjust the pull from each point which is often easier than redrawing the line or adding more points.

A common mistake is thinking that an accurate curve requires a lot of points. With practice, you should be able to do a single reentrant with 3 or fewer points, sometimes only 1. The key is to place the points properly. Understanding how polynomial splines work is helpful, but the simple rule of thumb is to place a point whenever the rate of change in the radius changes. Put another way, if a curve is getting uniformly tighter, you don't need another point. If it suddenly gets a lot tighter, you need a point. Spotting the place where the rate changes takes practice, but that's all it takes. Once you know what you're looking for you can get pretty good at it without worrying about the underlying mathematics.

For example, if you have a reentrant that is just a little blip on the hillside, you can get away with a single point because the rate only changes at the center of the reentrant. Typically, a reentrant is a little more complicated, with at least one side requiring an extra point to establish the curve between the center of the reentrant and the main hillside.

OCAD doesn't draw the curve until you've committed to the endpoint of the spline. Thus, it's common to place a point and realize that the prior point isn't right. Rather than stop, I keep drawing the curve and come back to it later. When I'm done with a curve, I zoom in to 8 or even 16x and check the entire length of the curve to make sure it doesn't deviate from the line in the notes. If it does, I'll first try to fix it by moving one of the points or adjusting the slope tabs (see below). If that doesn't work, I'll try adding a point. If it's really hosed up, cut the line, redraw the portion that was messed up and then glue the line back together. I strongly recommend gluing everything back together. On the monitor, it may look fine with the two ends on top of each other. When printed, even a tiny misalignment is visible.

There's a whole lot more to using OCAD than I care to cover, but I think this particular point is often overlooked. Back in the days when maps were made by hand, it was obvious to everyone that the quality of the map was highly dependent on the skill of the cartographer in properly recreating the field notes. Somehow, that message has been lost in the switch to OCAD. Just because the computer renders the line nearly perfectly doesn't change the fact that the line itself is still created by a person. If that person can't "draw" the line accurately, it's wrong.

6/10/07 Conquer Castlewood race report

is here.

6/11/07 Steady State

A run that used to be a distance training staple is noticeably absent from many training plans. It's called the steady state run. This run is like a tempo run in that it usually gets put in the middle of a longer run. It's longer than tempo and run at a slightly slower pace (but still significantly faster than normal training pace). For example, a tempo run for me is 15-30 minutes at around 6:25/mile. A steady state run is 40-70 minutes at around 6:35/mile.

I only do steady state runs during base periods or the very early portion of buildup. The reason is that these runs are tough and take a few days to recover from. There are several ways to make a steady state run more interesting and more valuable. The first, as mentioned above is to slide it into a longer run. Run for 40-60 minutes at normal training speed, do the steady state run, then finish off with another 30-40 minutes at normal speed.

Another form is the "fast finish" run. This is a run that starts with a reasonably long run (defined by your current training context, for most adventure racers this intro would be 12-15 miles) at normal training pace. Then bump up to steady state pace for 40-50 minutes. Finish with a real race effort for another 15-20 minutes. When I do this form, I usually actually enter a race in the 10-mile range. I'll run to the event, then run the first 7- 8 miles at steady state and try to finish as best I can. It's fun to do this in a real race because you end up smoking all the people who took it out to hard in the last bit. Passing folks helps keep you going and you need all the help you can get for this form - it's tough.

My favorite form is a variant of the first except that I replace the steady state on roads with park orienteering. My steady state pace is pretty close to my competition speed in a longer park event, so it makes for a good orienteering workout as well. Today, I ran over to Forest Park (40 minutes each way) and reran the course from SLOC's 2005 summer meet. I finished the course in pretty much the same time that I did it in the race (another 40 minutes for 2 hours total). I think I'm just a touch faster now than I was in 2005, so I'd probably do this course a couple minutes quicker in a race, but it was a reasonable simulation of competition.

6/12/07 Fall schedule

I suppose I should start thinking about what races I'm going to do this fall. Looks like there's something pretty appealing every weekend:

If I go to all of these, I'll be coming home to an empty house (or at least one where the locks have been changed). Decisions, decisions.

6/13/07 Nervous

Peter Gagarin wrote an interesting little bit about being nervous before an event on his training log. I'm not sure what conclusions (if any) to draw, but I always find it interesting that world champions go through all the same emotions as everybody else. And, as in Peter's case, it doesn't have to be a big event to set that off.

I remember telling Greg Lemond once how bad I felt on a climb where he and a few others had got clear of the group I was in (and, by clear, I mean a horizon job). His response, "It will always feel like that, the only difference is that when you get better, you go faster." As that was in 1990 (the high-water mark of my cycling career), I never did get any better, but it was some comfort to know that my experiences were not that different from those at the top.

6/14/07 Getting' large

Last week I was mapping Carondelet Park and ran into a section that really threw me. It's a really tight little section of depressions and complicated vegetation patterns. I wasn't able to see the vegetation boundaries from the aerials, so the base map was useless. The old map was even less informative.

I started from scratch, drawing everything by hand, but was frustrated with my inability to accurately render the detail at 1:5000. Finally, I gave up. I could say I got saved by the bell, as it was time to get back for dinner and Kate was making one of my favorites.

I went back today with a 1:2500 base map and it was much easier to get things right. Of course, the final map will still be 1:5000, but sometimes you need to zoom in a bit when doing the drawing. This is one of the reasons I'd like to do my fieldwork on a tablet PC of some sort. Then, you could zoom in when you needed. (Other reasons would be the ability to switch between base map and aerial photo, ability to view just one color at a time, and never losing a pen). Unfortunately, I haven't come across a tablet PC that has all the features I'd need (lightweight, high-res screen, excellent outdoors viewing, long battery life, and rugged construction) that doesn't cost a month's salary.

6/15/07 Hammerin'

I don't do too many structured sessions on the bike anymore. I save that for running on the track. However, just as running only in the forest can lead you to kid yourself about your running speed, it's a good idea to objectively measure your bike fitness from time to time.

Objectively measuring bike times is tough because there are so many environmental factors that can complicate things. Today was a bit warm, but the air was completely still. That made it a good time to head over to the flood plain and crunch some big gears. There was nothing particularly special about the workout, I just did repeat 500m full-on with about three minutes of recovery between each. Still, the information was valuable.

Most of the news is good. Despite being lighter this year, I can still turn a big gear (I rode these in the 53x13 with a cadence in the high 90's). My speed is off a bit from my prime, but not by much. It's certainly no worse than last year - maybe a touch faster.

There are some things to work on, though. It seemed like I was using more quad and not as much glut as I have in the past. From the way my pants are fitting, I'm guessing I've lost some muscle mass back there since giving up on squats. Working those back into my routine might pay some dividends.

6/18/07 What was the thinking?

OK, mapping is subjective, but check the difference between the old Carondelet Map (left) and what I just got done making. Do ya think the vegetation might influence route choice?

6/19/07 Back on blacktop

I've been averaging about an hour a week running on roads over the last few months. This is generally a good thing - I've been getting in a lot more running in the terrain. However, with the summer road racing season upon us, I might be down a little in the speed department. It will be interesting to see how things shake out this weekend at the Connection 5K. It will be my first race on roads since the marathon last February.

6/20/07 One more time...

For various reasons, UMSL isn't going to work out in July as a sprint venue. That means I'll need to make another map for the July event. I like mapping, but I'll be ready to put the pens down for a few months. I could use Willmore, which would cut down the work a lot since I already have a decent map; I'd just need to convert it to sprint standards. Unfortunately, I don't think there's enough terrain there for a double header.

I guess I'll go with Kirkwood Park, although that will be no small amount of work. At least I'll have another prime area mapped when I'm done.

6/21/07 Photo finish again

Rob Wagnon put the hurt on the field to win the first sprint tonight. The second sprint was a mass start with forking (known as a Farsta - I'll explain more fully tomorrow). Rob trailed Mark Geldemeier for most of the second lap, but put on a big surge up the final hill to finish just one second behind. But wait, a quick inspection of Mark's card showed that he took the wrong fork on the second lap. So that big sprint was for nothing and Rob gets the sweep.

I'll have a full report and results tomorrow.

6/24/07 Carondelet results

... and meet report are posted on the Third Thursday page.

6/25/07 Training with Spike

Michael Eglinski (AKA Spike) and Mary Jones were in town this weekend for a combination training camp/weekend getaway. On Saturday, I joined them for some sprint training in Forest Park. On Sunday, Spike and I did some more training at Carondelet.

At first glance, it might seem that Spike and I are pretty evenly matched. That's true, in the sense that if we're entered in the same race, it's pretty much a toss up (we're right next to each other in the national rankings). A little closer inspection shows that he's the better navigator while I have a bit more speed. But, even those differences are not so big that you can predict the outcome just from the type of race. At Team Trials, he beat me in the sprint and the long, while I was faster in the middle distance - just the opposite of what you'd expect.

Despite that, training with Spike never seems like training with a peer. The reason is that, unlike my other training partners, Spike has been at the top of the sport (at least by North American standards). He's been on the US Team several times and won the overall US Championship in 1987 and 1994. He's competed in many of the big European races and been to World Orienteering Championships. That sort of background leads to a much different perspective on goals, training, and what it means to really push your limits.

More than the actual training, what I really like about training with Spike is sitting down afterwards and talking about what we did and training in general. While he fully understands the commitment required to be one of the best, he's also a realist and is very good at spotting ways to improve within the constraints of holding a full time job, having a family, etc. I'll share a few of the things we discussed over the next few days.

6/26/07 Relevance

Orienteers talk a lot about the "relevance" of terrain. By this, they mean the degree that skills required to perform well in one area are the same as the skills required for another. For example, if you are preparing for a sprint race in an urban park, you'll probably want to get a hold of some urban park maps and practice short legs with lots of route choice and direction change. Training classic legs in the woods will have value, too, but it will be less relevant.

Strangely, as much as it's a subject of discussion, a look at the training logs of most North American orienteers would indicate that not many of them believe in the concept. The bulk of their training isn't even orienteering, much less relevant orienteering. In one of our conversations last weekend, Michael likened the typical training philosophy to that of a basketball player who trained by swimming and playing baseball. Training those two sports will improve fitness and coordination, and those are skills that will transfer to the basketball court, but nobody would suggest that such a plan was an optimal use of time.

There are some simple things that can be done to make training more relevant. First and foremost, move as much training off-road as possible. This is easier than it sounds. I work in the middle of a mid-sized city. If I didn't think much about it, all my midday training would be road running. However, by planning my routes carefully, I can get most of it off-road. I run in the center median of boulevards and parkways. I run through city parks and vacant lots. I run along railroad tracks and the banks of rivers. This trains the stabilizing muscles to be more efficient (and less likely to get injured) than banging out miles on pavement.

Second, and maybe just as important depending on where your nav skills are, run with a map. It doesn't have to be an orienteering map; a USGS map or aerial photo will work just fine. Even a regular street map is better than nothing. Get used to keeping track of exactly where you are at all times. Practice reading the map in short glances and keeping track of your place with your thumb.

Finally, practice concentration. Yes, sometimes it's nice to just shut off your brain and run. Sometimes a training run is a good time to think about other things. That's fine, but don't forget that somewhere in your training you need to be practicing the skill of keeping your mind on what you're doing. Learn to keep irrelevant thoughts out of your head. Think about how the terrain would be mapped (even if it's just city streets there are lots of mapping decisions to be made). Think about what direction you're going and how far you've traveled since passing a landmark. When running up or down a hill, estimate how many contour lines you're crossing.

Most athletes understand that to see improvements, you have to include "quality" workouts to your schedule. This is almost always interpreted as meaning harder physical efforts that require recovery. By increasing the relevance of your recovery workouts, you can make them quality training as well.

6/27/07 Think about it

Normally, one is taught that you want to practice routines until they are subconscious habits. Whether it's regularly checking your compass or riding a mountain bike over a log, if you have to think about it, you won't do it as well as if it's an automatic action.

While that's true, there may be some actions that we are too quick to put on autopilot. Because we just do them by rote, we fail to get the value from them that we otherwise might. Spike got an interesting article from Swiss orienteer Martin Lerjen looking at the reasons for taking a look at the map. You can check out Spike's blog entry on it if you want an overview. The full article is here. The translation is a bit rough - English is not Lerjen's first language.

Spike was telling me that he was actively thinking about why he was looking at the map so he could make sure that his purpose was accomplished. Of course, this is just a training exercise. In competition, he'll want to be reading the map automatically. The goal is to further develop the action and then return it to a habit in it's enhanced form.

That got me to thinking about other things that become habits and then get forgotten when they might be taken to another level.

• Navigation: Purpose of reading the map; map simplification; terrain visualization; changing speeds; aiming off; etc.
• Running: Form in general; jumping obstacles; ducking under obstacles; rocky terrain; etc.
• Cycling: Pedal technique; riding position for climbing, descending, roads; maintaining speed on rough sections; etc.
• Paddling: Paddle entry and exit; proper rotation; cadence; turns; etc.

6/29/07 Cut

I've never had any kind of surgery. I've had broken bones set and wounds stitched, but nobody's ever needed to go inside to do anything. Until now. Seems I've got a hernia.

Actually, I've had one for a few years. I knew something was funny, but I thought hernias were always in your lower abdomen. This one is really high - almost at the stomach. Anyway, it needs to be fixed, so I'm going to get cut.

The getting cut part doesn't really bother me any. It's the general anesthetic that I'm not thrilled about. Seems like if they can do a C-section with a local, they should be able to fix a hernia with a local as well. The doctor wants to do a general. I know that the chance of something going wrong is very small, but I still don't like being put under by drugs. Maybe it's a reflection of my bias against drug use in general.

The real bummer is that I'll be out of training for a few weeks. That pretty much wrecks the summer schedule. I'll still go out to Long and Short course orienteering champs in early August (non-refundable tickets - might as well), but I obviously won't be at the top of my form.

I suppose the break might do me some good. I never really took any time off over the winter because I was prepping for the marathon. Now I'll have no choice but to get some rest. Their going to cut me open on July 10.