Discussion: point density

Wow! Looks like a lot of fun!

This one looks so fun! I was curious about point density, and this one had 125 checkpoints in 6.7 square miles, so 18.7 checkpoints/sq mile.

The "average" Street Scramble has about 33 checkpoints and covers about 9 square miles. so that's significantly less: 3.7 checkpoints per mile. The standard Street Scramble is foot or bike, 3 hours maximum.

There are two annual Street Scrambles that are shorter (1.5 to 2 hours) and foot only, the shorter of the two covering about 2.8 square miles, so that's about 11.8 checkpoints per mile. Still not even close to Rochester!

It really was fun.

The high density of controls meant that you often (or usually) didn't have much time between controls. So a good bit tougher mentally to keep being accurate and efficient.

The places my skills could be improved, in addition to just being a stronger biker --

-- Learn (well, relearn) how to ride with no hands, so I could rearrange the 3-page question sheet on the go.

-- Learn to write while biking (might be tough, but would really save time).

-- Get better eyesight so more questions could be answered without stopping. I was pretty good about reading the questions on the go, so most of the time I knew what I was looking for.

-- Learn how to bunny-hop curbs. Would have helped more than you would think.

-- Get better at planning a control ahead, usual orienteering concern, so you know which way you're heading after you find (and write down) each answer. Got lazy on this many times.

I found I could read the map on the go pretty well, including seeing the little red dots that marked (about 95% of the time) the exact location. And do that without running into anything, or really even having any close calls.

And my recollection was that it was a lot easier to handle reading the maps and dealing with the questions and answers last year when I was running. Took a lot more coordination on the bike. But also not nearly as tiring.

Sounds like you could handle a couple of these issues by using Google Glass instead of paper.

As one of the two course-setters, I do realize that we (Rick Lavine was the other) over-did it a little in terms of the number of controls! We both went out on many outings into the city to find possible control points, and we found we had something like 160 between us. Obviously too many. So we starting trimming, but I think we both found that we were somewhat "attached" to many of the potential control points we had found and kind of reluctant to abandon too many.

Plus, our reasoning was that for people less ambitious than, say, PG, who would be covering much less distance, it would be more satisfying to have more controls to find rather than fewer. And in fact we did have some people (generally groups) in the 3-hour category that got fewer than 20 controls (one group amazingly got a mere 6, and they were out for nearly 3 hours! I suspect they MUST have stopped for a leisurely lunch and/or drink somewhere.). So with a significantly lower density of controls, groups like these (3-hour time category, not moving particularly fast) would have had presumably a lot less satisfaction from finding controls.

Nice to see the center dots for the controls. Sometimes used for ski-o and helps identify which side of the trail (or road) to look for the flag. When you're moving fast, it's surprising how you can sail right by if you're looking on the wrong side.

Obviously too many.

I liked having lots of controls, and a bunch of them not too far from the S/F. Good for those not doing too long a route. And made it more of a challenge for those of us trying to get them all. If you had given us 160, well, just more fun.

I agree, more controls=more fun.
It's apparent I need to work on my reading comprehension given 6 of the controls we visited were answered wrong. Or I need to improve my penmanship.
I will say that Dicks recommendation to place the answers on a separate notepad seemed like a good idea for a team but in practice it didn't work very well as it took too long to transpose onto the answer sheet at the finish. As PG stated, writing and riding is definitely a skill that would help.

Or I need to improve my penmanship.

One thing that Street Scramble in Seattle does (that I've also adopted with 'Hood Hunts) is having questions with multiple choice answers. It's quicker for participants to answer, and it usually reinforces that you're in the right spot. Multiple choice is participant-friendly.

It does require more work on the organizers end, though, as now you need to come up with two incorrect answers, which is harder than it seems. The wrong answers can't be obviously wrong, making the question easily guessable, and they shouldn't be ambiguous, and they shouldn't be a correct answer of a parallel feature (eg: another mural a block away contains an incorrect option).

Several years ago, I know that Street Scramble was even thinking about using scantron sheets so that they could tally scores quicker, but that never happened. The Street Scramble franchise in San Francisco is developing SMS-based scoring, where you answer questions via text message, and you score-as-you-go.

Dick, how much time did you and Rick spend on scouting checkpoints? Finding 160 and then whittling down to 125 must have taken a long time!

I think the main reasons for limiting the total number of controls to just about what the fastest few participants can reach are: 1) To allow for repetition of the general scheme for quite a few years without (m)any actual repeats of individual control locations. Enough imagination has been used in the three events so far so that a significant fraction of the control locations actually give a really interesting glimpse of some aspect of the city's history or culture that one leaves with a feeling of having a better grasp of why Rochester is a bit of a special place. That can likely go on for quite a few more editions of the race, but gets harder as more of the really "prime" spots get onto the used-up list. 2) To make it a significant challenge to get to all or most of the controls just inside the time limit for the top competitors. With practice, people get a little better at the game, and especially with more people taking to the bicycles this year, it was nice to have things right at the threshold of the possible for the attendees. This year there was one finisher visiting all controls, and still falling just shy of getting all answers correct. Last year with 2/3 the number of controls on basically the same map, probably 2/3 of the bike field in the 5-hour event got to almost all the controls. If there are 50% more than anyone can get, then there is a little less incentive to push really hard for the entire event. I thought Rick and Dick did an impressively good job of getting things just about perfect this year, while last year on the other hand, was just about perfectly optimized for those going out on foot.

My guess is that the little centering dots probably made a difference of several controls in the number that could be properly identified in the time limit for most competitors. Certainly for me it meant that finding the relevant bit of building was almost immediate, whereas in the previous two versions often half a minute or more was often devoted to trying to figure out where the specific target was actually located. I thought they were a great addition (far more important than I have found them on ski-O maps, partly because of the smaller scale map, partly because the object to be identified is not always the same bright orange-and-white flag, and hence immediately identifiable from some distance.

It's not clear to me how the extra notebook would have helped anything in the writing in of the answers. It was still necessary to find the line on the 3-page clue sheet to determine what the question was that needed an answer. That seemed to be the time-limiting step, and with the thumb already on that line, it is then as efficient writing the answer in there as using a separate sheet of paper.

More later, but regarding multiple choice questions -- yes, we have thought of that, but also, yes, it would be a lot more work for the organizers! As Pink Socks points out, the wrong answers can't be obviously wrong, which would definitely create a bit of a challenge especially in some of the questions. (No real problem with questions that have numerical answers, but relatively few do.) And we'd have to do what schools typically do on multiple choice tests, which is to deduct a fraction of the wrong answers, so that there would be no advantage to random guessing.

I'll check out from my AP logs how many hours and miles I spent finding controls. Rick L. isn't on AP and my guess is that he doesn't have a record of what he did.

deduct a fraction of the wrong answers, so that there would be no advantage to random guessing

That's what Street Scramble does. 50% point value penalty for wrong answers.

For many years, I was the scoresheet triple-checker for Street Scramble, so I was able to spot trends with certain types of checkpoint questions and answer choices. For one thing, people are really bad at estimating heights of things. So when I started organizing my own 'Hood Hunts, I had a lot of experience with Q&A.

Before the very first 'Hood Hunt, one of my friends downloaded the map and Q&A, and he wagered that he could make educated guesses and get more than 1/3 of them correct, since he figured that I wouldn't be very good at creating incorrect answer choices. He only got 4 of 20!

'Hood Hunts are small, 0.3 to 1.0 square miles, and when I host one, I tend to be a completist, and I go up and down every block, looking at everything! It usually takes several hours. I take a photo of every checkpoint location, and I think about the question and incorrect choices on the spot. At first, I would limit myself to just 20 checkpoints each time, but I keep increasing. The 1st anniversary event had 26, the 2nd had 29, Kansas City and Wichita had 40 and 38 respectively, and the 3rd anniversary also had 40.

With regards to the center dots, the Cuyahoga Falls FallsQuest event also used them and I found it immensely helpful to know what side of the street or which corner something would be on. It was especially useful there because most of the clues involved yard or house decorations - nothing that would easily stand out if you were just going down the street.

If it's allowed, I bet an tandem would be an awesome weapon in this sort of event. Captain in front and stoker/secretary in back.

No reason a tandem wouldn't be allowed!

A downside of the center dots, from a course-setter perspective, is that it makes it quite clear when the circle is misplaced even just a little. :) (As I've found, in retrospect, on a number of our control circles for the map adventure.)

The only one I noticed being off was the fireplace on U of R campus. And I can verify the 6 point group may have stopped at a pub or two on the way...

Thinking more about multiple choice questions -- it wouldn't have to be all or nothing. Probably a lot of questions would lend themselves well to multiple choice questions, while some may not (too difficult or time consuming to come up with good "wrong" answers, or maybe some other reasons I'm not thinking of at the moment). So for next year, we'll strongly consider having multiple choice questions but possibly not for 100% of the questions.

In terms of time needed to set the course, it turned out to be less than I might have expected. It was a total of only 20 hours (involving 60 miles of walking), and that was spread over a number of months (although most of it was in April). That included vetting controls that Rick had found -- the vetting was extremely important by the way, because probably for 25% of his controls (and vice versa for mine) it was decided that the question needed to be worded more clearly, or a hint was needed, or in rare cases, the control just wasn't a good one for some reason. I don't know how much time Rick spent, but I'm guessing it was comparable to a little less than what I spent.