Discussion: course lengths for M21 (and F21)

William has a comment in the poison oak thread that the Long-O' course last weekend needed to be a good bit longer. When I saw that the winning time was under 90 minutes, I had the same reaction. In fact, there have been a number of A meets this year where the M21 courses have been fine, but too short (and others have gotten it just right).

My guess is that some course setters don't realize how fast the top M21 runners go. (And hopefully they will go even faster in the future, but that's another matter...)

Is anyone -- JJ? -- willing to research the numbers for the last year or two, for the blue courses. Length, winning time, time per kilometer, whether the field was particularly strong or weak, and whether the intention was normal length courses or short and long. Would be useful to know, and might be something we could pass on to the course consulting committee and the course setters. A little education might help get the course lengths we want. (Note that the way to get longer winning times is NOT to bury the controls in the dark green!)

It would be good to check F21 too, thought my sense is that is has not been so much a problem.

Peter

A couple years ago, frustrated by the fact that the rankings were nearly a year behind, I calculated the rankings for Blue myself. In the process, I found that the course norms for blue courses were uniformly in the 60-70 minute range rather than the perscribed 70-75 minute. I would speculate as Peter does that course setters are not adjusting for a general increase in ability among the US elite.

My own method for avoiding this is to first check out all the control locations (so I don't loose any time to error) and then run the course. When I don't make any mistakes, I'm a pretty consistent 80-point runner, so if I run it in much under 90, I know it's too short.

I think Red courses are also a little short on average. I usually complete them in around an hour (should be 68 minutes for my ranking).

Speaking of the Long course, what is the desired winning time? USOF rules specify a maximum winning time of 145 minutes (for blue), but do not specify an optimum winning time! This year's course satisfied the rules just fine, i.e.,it was longer than normal, winning time less than 145.

But, based on past experience, I was certainly expecting a winning time close to 2 hours, not 1.5, so another 4k would have put it in that range.

I spent a few minutes looking at lengths and winning times for M21 courses. I went through a stack of ONA's recording the date, length and winning time for the first U.S. A-meet results listed in each issue. I excluded any short or long events. I ended up with a list of 37 events from 1994 through 2004. Here is what I noticed:

The average winning time was 75 and the median was 71. Courses tend to be on the short side (the current USOF rule calls for 75-80 minute winning time for the the orienteer with 100 ranking points).

I didn't see any pattern over the ten years. I didn't see winning times getting shorter or longer.

I recorded the winning time. If I'd have recorded the best U.S. time, the results would have been a little different. I saw a couple of races where the winner ran several minutes faster than the top American (Holger Hott-Johansen is clearly better than the 100 ranking point orienteer).

23 of the 37 races I looked at were won in less than 75 minutes. 9 of the 37 were won in more than 80 minutes. It looks like the tendency is for course setters to underestimate how fast the top M21s go (as Peter pointed out).

While course setters missed, they didn't really miss by all that much. If you just made the courses about 3 minutes longer, the overall picture looks a lot better.

I was surprised at how many results in ONA don't include course lengths. I excluded results if they didn't include lengths. I wonder if that biased the results.

Just for fun, I looked at the winning min/km. Over the ten years there isn't any obvious improvement. I didn't look at enough results to draw any conclusions about an increase in ability among the U.S. elite (which Eric speculated about).

I don't think either that NA elite is running faster lately.

The formula I use to estimate winning times in NA (that can be obviously used by course setters to calculate course parameters) is as following:
(CL*TDF + CC*0.008)*CS
where
CL - course length in km,
TDF - terrain difficulty factor (1 for open, 1.1 for more green and terrain features, 1.2 for most green and terrain features),
CC - course climb in m,
CS - course speed (100 pts NA or 1000 pts WRE results - 5:00/km for M21 long/classic, 4:30/km for M21 middle, 4:00/km for M21 sprint - different for gender and age groups).

For example,
2004 USA Short Champs M21:
(4.4*1+250*0.008)*4:30~29:00
2004 USA Long Champs M21:
(11.9*1+680*0.008)*5:00~1:26
Both estimates are close to winning times. That shows that USA Short champs were right on and USA Long Champs were ~4 km and 200m climb shorter (for 2 hour target).

As an example, here are estimates for M21 USA Champs in Telemark:
Day 1: (9.01*1.1+360*0.008)*5:00~1:04
Day 2: (12.19*1.1+460*0.008)*5:00~1:25

Although some subjective factors are used (TDF and CS) I think this formula can be safely applied to most of the NA orienteering events. Try it.

Course planners can't control the weather. A good example was WOC this year. Times in the long terrain were longer than expected. That area received a lot of rain. The middle and short terrain did not get as much rain and times were bang on (or faster). I remember when we hosted the 1986 World Cup that we were certain we had the distance right. "As long as it doesn't rain a lot or is very hot", stated Ron Lowry (event director). Well, it rained all week leading up to the race and then was +33C (hot for mid May) on race day. Times were a little slower than expected...

As Brian noted, USOF rules only specify a maximum for the long course. The organizers of the Long Champs indicated in the meet info that they were shooting for 150% of classic winning times. At the shortest this should mean around 112 minutes for M21.

The Green was long enough for me. Bruce Wolfe won it in 70 minutes though, which is a little bit fast.

I just wanted to bring up a counterpoint. If the Long-O champs had been 4km longer then my time would probably have gone up to 185 minutes or so. So some of the slower orienteerers (like myself) might be discouraged from competing and you'd lose some breadth. Then again, I could of course have signed up in M-Red or whatever shorter open course were available.

I'm very glad to see this discussion. I was waiting for this since before the event, when I saw the course data.

This might be the time to point out another course that I think is chronically miscalculated in the other direction, and that is the short green course. This tends to be the green course for women's classes, and I think this length is regularly off by a significantly greater percentage than the red and blue courses.

More times than not, when there are two green courses, the short green is virtually the same distance as the long green, which not surprisingly, makes the times consistently long. To get appropriate winning times the distance of the short green should, not surprisingly, be about midway between between the brown, and long green. Yet this concept seems to be lost on most course setters. In simple terms the distances should be approx 3.5km, 4.5km, and 5.5km respectively for these courses in average NA terrain.

EricW, Good point about the Green Y course. When I set courses for Telemark, it didn't occur to me to set one Green course shorter than the other (like I did for Red). But, in looking at the winning times afterward, it was quite obvious that I should have.

A simple statement in the USOF rules (which I was attempting to follow) that Green Y and Red Y should be slightly shorter than their X counterparts would help guide course-setters.

Note please that Eric did not refer to the shorter Green course as "Green Y". Particularly when the classes are split into Men and Women, some people find it irksome if Women are assigned to Y and Men to X, when their chromosomes are the other way around.

Agreed that Eric did not refer specifically to Green Y. But, at Telemark, it was the Green Y course that had the longer winning times and should have been shorter. And, given that the USOF rules suggest an X/Y split that generally (but not exclusively) puts the women on Y and the men on X, I assume this is typical.

If some find the X/Y split irksome, then they should change the suggested split in the USOF rules. [And while they are at it, they should stop suggesting F55 run Green Y, when they actually run Brown.] As a physicist, it didn't bother me in the slightest to assign women to the variable Y and men to the variable X.

Now, the fact that OCad orders the courses alphabetically (i.e., Blue, Brown, Green, Orange, Red, White, Yellow), that is irksome.

Since the USOF rules are written in terms of the time of a 100 point runner, doesn't that mean that to follow them means green X and Y being the same length? The winning time does not matter, because all green classes are lumped together in calculating the rankings, so the 100 point runner is the same speed on X and Y. The rules talk of splitting when participation warrants it, but not into long and short.

It would be very helpful if guidelines accompanied the rules, defining the courses in terms of course length ratios. So, for instance, blue/M21 could be used as a base with a length ratio of 1.00, the distance that a '100 point runner' would run in 75-80 minutes. Red would then be defined as, say, 0.68 of this length, green 0.51, etc (numbers from the top of my head). Any course planner could get base distances by test running, estimating the speed of a 100 point runner on their own course, then scaling accordingly. They would, of course, make tweaks to account for differing use of the terrain, climb and aesthetics. (Using length ratios like this is not my idea - see eg British OF guidelines on their website). At present, any course designer has to effectively calculate or guess these ratios, and I'd suggest that accounts for the discrepancies in course length from meet to meet. I don't believe anyone has an accurate feel for the relative speed on each course, but it would be easy to calculate this once and for all using a few years of ranking data.

I say all this as course planner for day 2 of US champs 2005 for which we've split brown, green and red into two more or less equal courses X and Y (courses are already planned and vetted as the terrain will be under snow until next July). We have also made red/blue 10% short on day 1 and 10% long on day 2 (WRE), as at Telemark, so are stretching the rules somewhat there already ;).

Brian, the way I found around this irksome feature is to name my courses starting with a number: 1 Blue, 2 Red X, 3 Red Y, etc. or whatever order you want the courses sorted in.

As a woman running Green Y, I have no problem with being on Y. I really hate it though when we are Green Y, Y, Y and then go to a meet and get assigned to Green X (like West Point tends to do...).

andyd, I think this is a case where thenUSOF rules don't quite provide enough guidance. Let me try to explain ...

The spirit of the rules is that for older people, the winning times should be shorter than for younger people (i.e., winning times should be shorter for M/F 65 than for M/F 21). Hence, courses should be shorter for older people than for younger people.

When we lump people into courses, we make some compromises. On a given course, some people are going to have somewhat shorter winning times than ideal and some people are going to have somewhat longer winning time than ideal. When we split a course into X/Y, I think it makes sense to eliminate some of the compromise we would have with unsplit courses. The recommended splits of X/Y are essentially into stronger/weaker pools of runners, so it makes sense to make the courses X/Y courses longer/shorter, respectively.

Consider that (for the men) ...
Blue has M21
Red X has M35, M40
Red Y has M45
Green X has M50, M55
Green Y has M60
Brown has M65, M70
It makes sense to have a continual decrease in course length going down this list (i.e., so Red X > Red Y; Green X > Green Y). One could include women in the same chart and it would suggest the same.

As for the 100-pt businees, I see this as basically being a mechanism for taking out the outliers. Consider it a way to deal with the "Peter Gagarin" and "Sharon Crawford" effect.

Brian

Yes, I absolutely agree, Brian. Splitting the X/Y lengths is a better way to do it. I was interpreting the rules to the letter as I see them. Pity they're so fuzzy, because that means everyone has their own idea of what is right. That's why I'd like to see length ratios as a guideline. Optional splits to X/Y would be easy to incorporate and standardised so everyone would know what the deal is. And, coming back to the original subject of this thread, it would be easier for a green-running course designer to estimate an approriate F21/M21 distance.

Andy

I started this thread on the assumption that the blue course has often been too short, and that, as I have often heard, we need longer courses to prepare for international competition.

JJ hasn't yet produced another of his statistical masterpieces (I assume it's in the works), so I did a quick version -- just the three best times for each day.

The goal is 75-80 minutes. Excluding some courses that were purposely short (30-35 minutes), many of the course setters did a pretty good job of hitting the target. There were a few courses, which I've marked in bold, that could have been another km or two longer, but a healthy majority were just right.

As far as preparing for international competition, I would suggest:
(1) We just need to be faster, no matter what the distance.
(2) The times of our male runners in the Long Qualifier at WOC-2004 were 75, 78, and 85 (vs. best times of 60-62). To be strong at the end of 75-80 minutes, it helps if the distance feels short, both physically and psychologically. That's reason enough for some events with winning times in the 90-100 minute range (plus a lot of training runs longer than 90 minutes!).
(3) When we start making the long final (winning time this year 105), then experience at racing 2 hours plus will be essential.
(4) We've talked about having more variety in the length our courses -- shorter Saturday plus maybe a sprint, a little longer on Sunday. That's something we should keep pushing (and it can be don for the blue and red courses without upsetting the status quo on the other courses).

Peter

When USOF adopted the basis of their current course structure, Bill Cusworth prepared course guidelines that can still be found on the web at http://us-o-team.us.orienteering.org/USOF/rankings...

This table includes a distance ratio as Andy suggests. I've used it when setting A-meets and it's very helpful. I can set what I know is an appropriate Green course based on my own ranking and speed and then scale up to the right lengh Red and Blue based on that.

Somehow when the proposal was adopted by USOF, this column of the table didn't make it into the rules.

Clare

Words of wisdom from Peter. Incidentally, a couple of days ago i was having a gook at some of the training logs, and there weren't (m)any who had done as much _running_ as he.

Nobody (or almost?) is training enough to run a 100 minute race well, me included. I'd view six or seven hours a week of running as an absolute minimum. And consistency is IMPORTANT. If six hours is currently too much, do four _every_ week until you can do more...

Focussing on one point of Peter's, it is necessary to be strong physically and mentally for the whole race. For this, one has to be able to run _fast_ without the head being involved. The goal is legs moving fast, body from ribs up relaxed and head clear.

With the terrain you have in the US, there is no excuse for not attaining qualifying standard regularly at WOC classic distance. There are enough of ye who know how to orienteer... and there are quite a few with the athletic ability... if ye can find the time to go terrain running regularly...

sorry, must stop there, i have a plank in my eye that's bothering me.

If the US should adopt the French 35 hour work week, and we could all train much more!

Thanks for the link, Clare. That's just what I was hoping to see!

mindsweeper, do you want to prepare the ground, we'll send some guys over to liberate yiz....

I'd settle for cutting back to 40 hours a week.

Yes, 40 hours would be nice also.

Well, if someone is seriously looking at becoming an elite orienteerer, (s)he needs to start looking seriously at 500-600 hours/year (10-12 hours/week) training process and how to get there from 3-4 hours/week currently being en mass. And close to 50% of that should be in woods.

Just remember that it will take 2-3 years of sustained efforts only to come to this level. Good luck and motivation for many years to all hopeful! We have coaches in this country who can help you.

From my own experience, I can only train that much if I do a lot of cross-training, so 50% in the woods wouldn't work for me.

During the summer of 1999 I managed to get up to 60 hours per month, which was quite doable with a 9-5 job. I wasn't orienteering at that point, so my training concisted of mostly:

* Running
* Swimming
* Rollerskiing / rollerblading
* Weigt lifting

I definitely think the US work-ethic of long days and short vacations are an obstacle to becoming a non-professional elite athlete.

If I could comment on Brian's and Andy's point about the X/Y split on Green, the general trend has been to put men on one (don't care what letter is used) and women on the other. The exception of recent times was the US Champs where the M60 class was lumped with the women. Personally, I prefer the gender split and be able to run with peers, as in M55 or M50, as opposed to the F classes.

If course planners would post the intended split in advance, we could then decide on the course to sign up for. Once we pick up packages it's a bit late altho I did that at SVO last year.

Well, I don't really understand the US course structure, but I would coment that women need to train less than men (not that we are weaker in any way, but they seem to have shorter winning times for us) 8-9 hours in base training season should do... (not that i am at that, but will try harder in jan...)


See http://www.orienteeringbc.ca/HighPerformance/Train...

for a training template by ted.

George, Good point about M60. At Telemark, I just followed USOF rules, which suggest lumping M60 with the women on Green Y (see http://home.comcast.net/~rshannonhouse/Rules2004.h... section 4). Maybe that suggestion should be changed to include M60 on Green X with the other Green men.

Brian, I am aware of that suggestion and think it might be a remnant of a different era. There seem to be a strong contingent of M60s that surface at A meets periodically that can be substantially faster than the F classes. I would suggest that the course setters should use the target winning times for the men and the women separately, although that is a women's issue and not one I care to get into. The men are already taken care of; the women must live with what falls out in the current scheme of things.