Discussion: USOF A-meet blue courses: too short?

I've noticed a trend in the last few months that blue courses have been consistently a little too short. This is a polite request to future course-setters to be careful in considering the course-length guidelines.

For USOF 2-day classic A-meets, the format that we have right now at most meets, the course length is supposed to be such that a USOF 100-point ranked runner takes 75-80 minutes.* The easiest way to check whether this is being met is to look at the course gnarliness value which is used in calculating the USOF ranking. If the 100 point runner takes 75 minutes, the GV should be 75*100 = 7500. Looking over the last 12 months and discarding sprints, middles, and 'extended', ie ultra distance events or WREs advertised with longer-than-USOF classic length, we get a set of 24 GVs.

Tabulating these 24 courses:
too long
3 courses around 8200
correct length
3 courses in range 7500-8000
too short
5 courses in range 7000-7500
3 courses in range 6500-7000
9 courses in range 6000-6500
1 course less than 6000

Median GV: 6804. Mean GV: 6921.

Therefore, the median course should be between 10% and 18% longer to hit the 7500-8000 target, or roughly 1.3km.

On the other hand, looking at course lengths, all courses but three fall within the 8-14km guideline that is published as a suggestion. As do the correct lengths, usually. I suspect, therefore, that a lot of course-setters are using the '8km-14km' guidance too closely and not recognizing that running speeds differ a lot across different terrain, so that course lengths nearer 14km are sometimes required to get the correct winning time.

The most extreme examples: one meet at 8.9km should have been 13.5km, one at 11.3km should have been 14.2km, one at 10.1km should have been 12.6km. I realize in a few cases special circumstances make it difficult to set longer courses, or undesirable to do so, so I don't want to name specific examples: the pattern is pretty general.

This is intended as constructive criticism, not a complaint. Future course setters should consider: blue courses should maybe be longer than you think at A meets! And please don't take hitting the 8km-14km guidance as your primary aim if you are setting a traditional 2-day A meet: take 75-80 minutes for the 100-point runner as the goal.

I'd like to thank course-setters for their hard work: the problem is that I like what you do so much, I want to do more of it... ;)

* = There is a slight inconsistency in that this is also referred to as the winning time, but because of the way in which the rankings are normalized, the top runner usually has a ranking around 103 and not 100. I'll ignore this. I'll also ignore the fact that some meets are not in USOF 2-day format any more. In my opinion, more middle/IOF long meets would be appreciated. Be that as it may, the point of this comment is that we should stick to the rules we have for the meets we are putting on right now.

** List:
2007
Flying Pig classic: 7798
DVOA day 1: 7143
DVOA day 2: 7370
HVO day 1: 8232
HVO day 2: 6994
USMAOC day 1: 6261
USMAOC day 2: 6265
Colorado day 1: 6157
Colorado day 2: 7115
UNO day 1: 6591
UNO day 2: 6468
BAOC day 1: 8290
BAOC day 2: 6221
CAOC day 2: 7362
LAOC day 2: 7801
US champs day 2: 8201
2008
GAOC day 1: 7555
GAOC day 2: 6092
SDO day 1: 7467
SDO day 2: 6229
NTOA day 1: 5095
NTOA day 2: 6384
TSN day 1: 6402
TSN day 2: 6614

I set the courses at the HVO Surebridge Challenge last year. Day 2 was about 5% shorter than day 1, but the winning time of day 2 was about 15% shorter. I had tested both blue courses (and almost every other course, too), but with some spring snow. Assuming I had my red speed, I deducted 1.5 minutes per k (ignoring the fact that I knew the controls). This is how I decided on the course length.
I think the faster times on day 2 are caused by the slightly shorter course but also by the fact, that the runners were tuned in to the technical terrain.

The courses were longer and harder in the earlier stages, but I decided that they were too hard for many runners here (sorry to say this, but I have to adjust to local customs), so I allowed for easier elements and even some trails (the original version had about 90% off trail navigation).

In essence, I believe harder & more technical courses would make sense (and add a lot of fun), but we would lose people doing this.

Thanks for the analysis, feet, as I wanted to bring this up within the Tucson club because of last weekend. Now I can just plagiarize. ;-)

Interesting analysis, feet.

Piutepro - Why do you think that making the blue courses harder and more difficult would result in losing some people? Couldn't they just run red instead?

I also think that we are not in danger to loose people to the sport that already run Blue courses. Correct me if I am wrong, but if you have made the effort and took the time to learn the sport and are running blue courses, you are already soooo hooked. Retaining people and gaining new should be the task for the yellow/orange/green courses.

Yeah, take it easy on the poor blue runners. However, feel free to amp up the technical difficulty on the red course please!

I believe you could also lose people who feel the blue course is not challenging enough (I know I find myself questioning things when I spend a bit of travel money and am disappointed by the challenge). That argument may cut both ways. Perhaps the board has considered the effect on participation when setting the guidelines. I dunno.

I agree that the blue courses are too short. I feel they have been too short ever since I've been running blue. I don't think it has changed that much. I started running blue simply so that I would not feel over my head when running my age class in Europe, which told me something was a bit out of whack. I'm not sure how to fix it. USOF does not have event controllers or a course consulting chair (that I'm aware of). In TSN's defense, they made a request in terms of a course consultant on the sanctioning app, but there simply isn't a course consulting process (that I'm aware of). Don't know if the data support the notion that courses are closer to spec when there is an assigned USOF course consultant. I believe the VP of Comp is aware of the lack of course consultant process (not sure if it is being addressed or not).

One suggestion may be to add a 21E class, and make the present de facto blue the M21A course (and perhaps W21E). That may at least alert course setters that something longer is warranted for the top runners. Dunno. Would also address the problem of athletic adult beginners who are intimidated by finishing at the bottom of the pack -- a 21A course allows one to work in to it with that being less of a factor.

I realise suggesting adding a course or two may be as popular as the suggestion to drop a championship or two, but WTH? Perhaps USOF has already tried this before my time. It sticks vaguely in my memory, but I dunno.

This is an interesting discussion topic since I was just talking to someone else about the lack of complexity of US ski-o courses versus international events. I hear there's a document that recommends a winning time for each division, but what about number of controls? Have you seen recommendations for foot-o? I found the IOF Ski-O Rules 2007 and various other documents, but don't see any winning time or number of control recommendations.

It's not uncommon for international events to have 40-50 controls for a long course. The Empire State Games were just held in New York, and the longest course had 11 controls (~20K skiable distance). The 2008 Sierra ski-o championships had one event with 25 controls on the long course, but all the others were in the low to mid teens.

For both foot and ski-o, if our domestic course setters aren't setting courses that help us train for international events, what are the other options?

Yes, Whatever happened to the USOF Course Consultant Program? Isn't it time to get that program up to speed again? As an occasional A Meet Course Setter, it is a secure feeling to have that experienced course consultant as a partner throughout the course setting process.

(Happy to learn I set the gnarliest Blue course of the year. BAOC Day 1)

Yes, BAOC is fortunate to have some very gnarly terrain. I remember one A-meet near Santa Cruz where the Blue course was under 4km as I recall! On this terrain, the GV could be drastically altered by adding just a few meters to course lengths.

But should clubs seek out that type of terrain for their O-meets? Do we want to move toward steep, thick and gnarly maps? I'd like to see some consideration for the other less-quantifiable aspects of our courses. Namely the exhilaration a runner enjoys when competing in the woods with his peers.

I'd guess that those high-gnarliness days are short on exhilaration...for just about everyone.

a high GV does not mean the terrain was unpleasant, just that the course had a relatively longer winning time.

I predict that the Blue courses at the US Champs this summer in Wyoming will have GVs that are not too low. I also expect comparatively little climb, and no vegetation to speak of (maybe a little sagebrush?). And I expect very high speeds, and probably a modicum of exhilaration. How can you have all of these things? With the longest courses ever offered at a regular US Champs. Whaddya think, 16 km? More?

chitownclark is wrong in his assumption that the high GV of last year's BAOC Meet was because of steep and thick terrain. Just ask anyone who was actually there. Boggs Mountain is nice and challenging continental style terrain. Some say maybe the best of it's type in the US.

I was there. Its good terrain - only moderately thick, but it is indeed steep. The quoted climb on Blue for day 1 was 50% lower than actual based on the measured routes of the 4-5 blue runners that we looked at that evening. In my opinion, the high GV for that day was due to the climb. Mis-counting the climb doesn't make the terrain less steep. Also the rocks there are the same color as faded control flags. The vegetation mapping at Boggs is dodgy in places.

Actually I attended the Boggs meet also...but not on Blue. And I agree: Boggs is one of the nicest maps to which BAOC currently has access. The Tahoe maps are good too. But they're all quite a distance from the Bay Area.

Boggs Mountain is nice and challenging continental style terrain and well worth a trip. I look foward to going back (whereupon I hope to navigate to many controls that are hung fairly.)

danf wrote: On another note, maybe I'm wrong here, but the distance targets seem to be not fully compatible with the target winning times for blue as well - what would an 8km course with a 75 minute 100-point runner winning time look like exactly?

It would look, according to the GVs, like the day 1 HVO meet at Surebridge (for which 8.00km is what I calculate as the ex post target), or the day 2 LAOC meet (7.94km). Others under 9km on the list: UNO day 1, BAOC day 1, and HVO day 2.

Links from randy's site for two of these:
HVO day 1
HVO day 2
Correction: these are red courses. Blue was similar... ;)

This is something that should be brought up to the USOF VP of Competition,

Good luck :)

Now that I look at the website, there is actually a name listed as the course consulting committee chair.

I was told that Elis resigned quite some time ago, and was not replaced at the time. I have not received a response on this matter from the VP of Comp.

I actually think there are course consultants, I'm just not sure who they are. I think the way things are working now is -- if you want a course consultant, it is catch as catch can.

If someone gives me a list of USOF course consultants, I would be happy to randomly assign one when a meet is sanctioned, but it is my understanding that the process was (or should be) a bit more sophisticated than that. Unfortunately, there are several black holes in the process presently, and there is only so much you can do before you give up and just invest your energy elsewhere.

Off topic - is there any place where the duties of the USOF board members are actually enumerated?

Robin Shannonhouse's "Virtual Binder" (accessible from the USOF site's Club Resources page) has a link to "position descriptions."

how a course setter should go about determining how long the courses should be to hit the target

If the course setter himself is ranked, he/she should run a course he has designed (perhaps a couple of times to get an average time), and then do the math to relate his time to that of a 100 pt ranked runner (or thereabouts) in his class, to get the closest approximation (I understand the concept, but will let the mathematicians here tell you how to do the math). Then you're dealing with whatever the gnarliness factor might be (steepness, thicker woods, etc.).

If the course setter isn't ranked, what is he/she doing setting courses for an A-meet?

What used to happen when Elis was in charge (and Piaras before him, if memory serves), was that he'd send out an email once or twice a year to all of us on the list of consultants, saying what the upcoming A-meets were that needed consultants, and asking us to volunteer. If he got more than one volunteer for a meet, I guess maybe he'd just pick one, and if he got none, he'd send out another request saying that he still needed one. And I think he probably exercised some judgement in terms of which events he thought probably needed a lot of expert help, and which ones would do fine with minimal supervision.

If the course setter isn't ranked, what is he/she doing setting courses for an A-meet?

Is it a requirement that one be a USOF member to be an A meet course setter? Interesting question, as a present bid proposes a course setter that doesn't even live in North America (USOF membership unknown, but unlikely).

Being ranked, per se, may be a proxy for experience, but it isn't the same thing. Maybe it is the exception that proves the rule, but there are several people I know who are great course setters who are not ranked. Of course, they are experienced.

Again, while I don't like arguing by piling up anecdotes, when I have set courses, I did not do this sort of math. In fact, I have set lots of courses that have hit the winning times I was aiming for without running them at all.

That being said, I think hitting winning times can be somewhat misguided as a first order criterion of good course setting.

If hitting the winning time is a goal (and controlling variance), the course setter should think more qualitatively and subjectively rather than running the course and doing math. (To me this is obvious, but perhaps it requires some explanation... maybe later.)

Like portfolio management, good course setting requires art as well as science (or "optimization.")

And there may be many reasons why a course setter isn't ranked... Injuries or personal circumstances that make A-meet travel difficult come to mind. I've had a bunch of those in the last 24 months that have caused me to be unranked for the last two years. I doubt that they have affected my potential as a course setter in any way.

Side note: I believe that the bid in question has a course designer from abroad, but a local (and experienced) course setter who is separately listed; not being on the board, I may be thinking of the wrong bid, however.

Would someone do a similar analysis of F21 times? I've been luckier getting the M21 times correct. There is some pushback when the Red course gets up near 9km.

I believe that there is a disconnect between USOF "Classic Course" rules and the IOF "Long Course" rules of many minutes longer for the IOF-Long, but I don't have the time right now to go check the rules.

Okay, I admit I wasn't thinking about overseas o'ers, or those who have been injured for more than a year. I take back that comment. :-)

In my last few months as VP Comp I tried to fill the course consulting chair position. I got two inquiries and no takers. My most recent attempt was with Jan Urban last weekend in Tucson. He almost accepted - everyone pressure him now! I also tried to contact Elis to get the list of consultants and got no reply. One thing that could help in a small way is for anyone who is a course consultant to let Randy and Bob Paddock know so they can help clubs find appropriate consultants.

I personally set one of last year's blue courses that was both on the shorter side and had an appropriate gnarliness value. It was indeed "gnarly." I'm sure it was more than 4% climb with notable areas of difficult vegetation. Other areas of nicer running. The reviews of Camp Scherman were mixed. Some folks liked the variety of slower, thicker stuff and some faster terrain. Some didn't and only wanted the faster stuff. The Day 2 (classic) Blue and the Day 1 (middle) Brown were the only two advanced courses that had correct GVs. Everything else was too long - with the worst offender being the middle Blue, GV 5537, almost twice as long as it should be.

I will confess to not running any of the courses for time. I have done pretty well at estimating winning times in the past without running the courses myself. This time around I was a little wary because the vegetation was different and I wasn't really sure how fast the more elite folks might move through it. I intended to run the Green myself for time, but with too little time and too much to do, this is what fell by the wayside.

I believe that hitting the winning time is an important goal - especially in the way the rules are written. Many of the courses I've run where people complained the most about it being too short or too long were cases where the course setter aimed for the distances given in the rules with no regard for winning time. I do think we want some sort of standard for folks to try to set to and, given terrain variety, winning time is a better ruler than distance.

The trick then is for someone to explain HOW one determines if a course they set will fit the winning time. I don't think anyone has tried to put that into words (besides my weak attempt above).

About loosing people on a technical & longer course: If I would set a European style classic course with a (European) winning time of 90-105 minutes, this would result in a US winning time of 110-130 minutes. And the second half of the field, well, say they would take much longer.

I don't test run courses for local meets. I take the percentage off/on trail traveling on the ideal route choice, I assign an estimated time, e.g. 6 min/k on trail and 12 min./k off trail (for a red course) plus additional time for highly complex technical controls. This is usually close to the intended time.

As an outsider my comment is that M21 "winning" times of 75-80 minutes for both days of a 2-day seem on the long side by international standards (not that there are many international multi-days left these days where M/W21 run similar distances on all days - mixes of longs, middles and even sprints are more common these days). The combination of course lengths at the Georgia Navigator Cup seemed to me to be pretty much in line with what I'd expect in Australia for a 2-day event where one of the days wasn't explicitly a middle distance. (I didn't run day 2 there, but feet's times on the 2 days provide a pretty good reference point for comparison).

Judging winning times, though, can be challenging, even on a well-known area. At WMOC 2002 we test-ran a fair number of courses (mainly using national-level M21 runners who we expected to be on a similar level to world-class M35s and M40s). We got the longer courses pretty well right (within 6 seconds for both M35 and W35), but were very badly wrong on M60 and M65 (both were supposed to be 45 but ended up in the 60s). In M65 all the top placegetters made at least 15 minutes of mistakes which is not exactly what you'd expect in a world championship final, but I'm still puzzled by M60 - I test-ran the course in 37 and based on that thought that it might be a bit over 45, but not above 50.