Discussion: Relay best practices

Assume this question is hypothetical...

Are there any online resources for relay best practices? In particular, forking? Forking while using Condes?

If one had, say, three legs, of roughly equal length, how do you ensure that every team does all the control legs, while producing a reasonable number of overall permutations, such that it is very unlikely that a small relay will have runners on the same leg run the same forking?

I guess one could produce A, B, and C, and then you have six possibilities. But, what if you wanted the last leg to be identical across all teams? Probably you are not left with much flexibility.

I think I may be over analyzing this, but am curious to hear what people think.

Thanks.

OCAD does relays adequately, including essentially arbitrary complexity.

Which one of the above questions was the hypothetical one?

>I think I may be over analyzing this,

Yes, Canada will still win.

I just planned a (one man) relay with 3 legs for the Vancouver Sprint Camp. It has forking and a common section for the second half of the last leg. If you are not planning to come to this race, and promise to keep the files to yourself, I can send them to you. I'm not sure I'm using "best practices", but it might give you some ideas. Send me an email if you are interested.

The sample map with Condes has relay course examples. If you select the "all variations" option for the relay course, you will see the overall course structure, and the drop down arrow shows individual variations. It's a clerical procedure to ensure that all teams run the same overall course, and to assign course combinations so that there is a reasonable mix of variations.

Thanks everyone...

I think pi's experience is most relevant to what I'm doing. Having a common last segment will probably do the trick.

But, I am curious if anyone has come across some interesting (or even just basic) forking schemes online. I have several relay masters from WOC and one from Jukola, and those are great starts. I am curious to talk to someone who has set a major relay about guiding principles for forkings.

pi--I am reserving the right to come to the camp, so I waon't take you up on your kind offer yet!

Whatever you do, don't look at this year's JWOC courses. The basic principle that they missed was that each team has to run the same course, just with the leg order shuffled amongst teams!

Yes--that JWOC situation is exactly what I am trying to avoid.

To avoid the JWOC situation, you only need to follow one rule: ensure that you have a common control for all 3 leg runners at the beginning and end of every forking.

The only software that I've used for relay course design is Ocad, and it works very efficiently. It spreads the total number of teams / competitors evenly across all the potential leg combinations to avoid grouping.

For guiding principals, the norm for most international-level relays seems to be that legs fork for only 1, or a maximum 2 controls. There are thus a lot of common controls, which suits the IOF's goals for maximizing the spectator element, with lots of head-to-head racing where team positions can be compared. Generally 1 of the 3 relay legs is either longer than the other 2, with any extra distance normally added to the longer (often third) leg after the final spectator control, which tends to be 70-80% of the way through the course. This last segment of the longer third leg is thus not forked.

Personally, I like to plan my relay courses so that the common controls are less predictable.

Forks don't have to be simply an option of which control in a cluster each runner has, either. Some other possibilities used at WOC93 are:
a) a long leg between common controls, with a control either near the start, near the end, or in the middle
b) a control-picking section, where each runner omits one of the controls
c) two pairs of controls (a vs. b and c vs. d), where the runners have ac, ad, bc, or bd. (This obviously doesn't work for a three-leg relay.)

c) two pairs of controls (a vs. b and c vs. d), where the runners have ac, ad, bc, or bd. (This obviously doesn't work for a three-leg relay.)

And not for a two-leg relay either. They did that mistake at 10mila 2008. (But it's excellent for four-leg relays.)

Forks don't have to be simply an option of which control in a cluster each runner has, either.

I set a forking at the 2000 Flying Pig that to my knowledge doesn't have a Swedish name for it. It's perhaps a better fit for a one-man relay than for an actual relay, since less repetition is achieved at the cost of more controls. Imagine two parallel
courses with no common controls (it wasn't what actually happened), and a third course switching back and forth between the two (looks like KI).

OK--time to revisit this...

I think I've worked out the forking scheme, but I am still using Condes wrong, somehow.

There are three choices: farsta forks, farsta bound forks, and relay forks. I think I want the latter.

If I have three forks on a course, with 2 branches, 2 branches, and 3 branches, don't I have a total of 12 possible courses?

Also, what does the number in parentheses mean in the Variation drop down in Condes? For instance, I have aaa (0), abb (12), and bac (19).

I think the (0), (12), etc, is the number of the variation starting with zero rather than one.

The problem is that Condes can't handle (to my knowledge) a 2:2:3 arrangement. If you need 3 forks in the third grouping, then you have to carry along 3 forks in the first and second as well.

That means you've got 27 variations, and the parentheses numbering will go from (0) to (26). I think.

That makes a certain amount of sense. But, I think I got it now.

Thanks!

I like to fill the unused options with codes and clues completely outside of the normal range---say if we own stands from 31 to 255, I'll not only number the unused forks with say 900s, I'll use gibberish clues that any orienteer would immediately notice. That way if you accidentally print a wrong map, you've got a good chance to notice it.

I haven't seen the new fork-styles used yet. I wonder if there are example files available?

In Condes, the number of variations is the number of persons in the team raised to the power of the number of forks in the course, thus a three person relay with three forks has 27 variations (you can see this if you define each fork as a simple Farsta fork).

Defining a fork as a bind fork or a relay fork limits the variations displayed in the drop down box, but the variation number in brackets is derived from the theoretical maximum number of variations.

Long time since i used condes - I think the answer is things like 'bind' and 'relay' forks that Hawkeye mentions.

But would this help as a way of visualising it:
change the 2-2-3 to 3-3-3 then make two of the variations in the first two forkings the same

eg first forking

a: S- 31-40
b: S- 32-40
c: S- 32-40

second forking
a: 40- 33-50
b: 40- 34-50
c: 40- 34-50
but then you might have problems making sure each team runs the same combinations I think.

Maybe what you are intending is to have two forkings of two and you want each runner in a team to run a different combination through those two forks. So make it into a single forking within condes.eg:

a: S- 31-40-33-50
b: S- 32-40-34-50
c: S- 32-40-33-50
you have one combination left over that you don't use (31-40-34)
the fact that all three forking go to 40 doesn't matter - you don't have to treat it as a common control and the runners don't know you have done that.

Hawkeye, how many arrows can you let loose in one minute?

Rob's solution is good--that should do it.

An old IOF guide to relay course setting can be found here

That IOF guideline is great! Thanks.