Seating Lottery?

[Pendulum^-1]

Originally Posted by Andy A.

I would be largely satisfied if FIRST could ensure that there is, for every match, high quality video feed of the whole field, without any cuts to close ups or drive teams or whatever. Just a single camera with a high angle view of the whole field, uninterrupted for any reason.

Relay that to some projectors and make the video available to download or burn to a disk or whatever. Heck, just provide a AV hookup for teams that want to record that feed as they deem fit. That'd give you a pretty solid base of scouting data to review and work from after matches.

Considering that video equipment is so cheap now that multiple teams (including our team) currently have their robots equipped with GoPRO cameras capable of recording 1080P video for 6 consecutive hours on one 32 GB chip, I fail to see why the above cannot be done at multiple venues for not much cost. Get GoPRO or another video vendor to sponsor it and set it up. The poor video feeds to the internet, and lack of video feed availability in general, really needs to be addressed. And it should help address the scouting/seating issues. Want 20 scouts pouring over match data? No problem. Examine the (multiple-angle?) HD video for hours, using pause, slow-motion, whatever. The access to such match video data should help teams improve their scouting significantly.

[EricH]

Originally Posted by Pendulum^-1

Considering that video equipment is so cheap now that multiple teams (including our team) currently have their robots equipped with GoPRO cameras capable of recording 1080P video for 6 consecutive hours on one 32 GB chip, I fail to see why the above cannot be done at multiple venues for not much cost. Get GoPRO or another video vendor to sponsor it and set it up.

All right, I'll bite.

Let's assume that there are 20 fields, each needing its own camera(s). Let's assume that the powers-that-be want a streaming version (any); each camera is $200 minimum (GoPros in this case). Let's also assume that one is NOT enough--pretty reasonable assumption. Let's randomly say that each field gets 5 (partly because that gives a pretty good view, and partly because that gives some nice round numbers). Let me say again: One is NOT enough! If one doesn't work for instant replay, one probably won't work for scouting. (For one thing, how do you identify the robots from above?)

Anyways, 5 x $200 X 20 fields gives $20,000 to set up the cameras that are now broadcasting on a WiFi signal. This does not include any receivers or recorders. Not too bad--but you have no spares (figure 1/field, for an extra $4000) and potential interference.

Let's briefly assume that instead of sending out on WiFi, the cameras record data--just record it. For it to be useful to a group of scouts, the records need to be pulled out of each camera. 32GB is a lot of data to pull at a time, and now there are 5 of those... per event... going up AND being downloaded (and that doesn't even factor in parsing). Not to mention the multi-view options.

Essentially, it could be done. But it's not going to be easy, or cheap.

The poor video feeds to the internet, and lack of video feed availability in general, really needs to be addressed. And it should help address the scouting/seating issues. Want 20 scouts pouring over match data? No problem. Examine the (multiple-angle?) HD video for hours, using pause, slow-motion, whatever. The access to such match video data should help teams improve their scouting significantly.

I'll also bite here. Yes, the poor feeds need to be addressed. Guess what? FIRST does not run the feeds. Local teams/groups do. But they take the video from the video board at the event--there's typically a drop box near Pit Admin or somewhere like that specifically for that purpose--so you get what's being shown on the screen. (I know there are exceptions--teams place their own cameras for some casts.)

It gets better, though. For webcasting, you need an internet connection. Field traffic gets top priority--FMS sends data to FIRST after each match, FTAs may need to e-contact their know-how to help solve problems, stuff like that--and usually has a dedicated line--which may be the only one in the building. If there isn't another one, webcasting over wireless could be somewhat problematic.


However, I don't see that as helping scouting OR seating in the slightest. I can't say that I know of many teams who use video to scout in the first place--they might record matches and just look at the ones they want on occasion, or use video to help their drivers, but very few actually use the video to scout, regardless of source. Also, now you have the entire scout team sitting down somewhere to watch video (or just sitting down)--now another 10 or so of those 40 seats saved by 4 persons have people in them, so it's 14 saving 40 instead of 10 saving 40. (Or maybe 14 saving 50--something like that.)

[sanddrag]

Originally Posted by PVCpirate

At the Granite State Regional, there is a small section (I think 50-100 seats) right in front of the field that only goes halfway up the lower bowl of the arena, and it is reserved for teams in the current match. There's usually someone making sure people enter and leave in an orderly fashion, but I don't remember any problems with it. The teams that have good seats don't use it as much, and the teams that get stuck along the sides get to watch their own matches up close.

They had this at Los Angeles (Long Beach) also, but the problem is they had one entrance/exit, and it took so long for the whole stands to exit, that there was a pileup of people waiting to enter, and it blocked a lot of area and caused folks to almost miss their match getting stuck in the crowd. So long as there is a defined entrance and exit side, this area works well.

[Pendulum^-1]

Originally Posted by Pendulum^-1

...... I fail to see why the above cannot be done at multiple venues for not much cost.....

Originally Posted by EricH

All right, I'll bite.
.......

Outstanding critique of my half baked ideas about cameras. I now understand more about the technical details and costs. Seriously, I don't know much about either setting up video feeds or running a fully staffed, elite scouting squad.

And that is why I post it here on CD, to get critiqued. So that we can put the less practical suggestions away early.

In order to avoid a multi-post digression on the financial, logistical, and technical details about camera installations on this thread about seating, I suggest that we get back to what can be done about the seating issue(s) in general.

[EricH]

I've seen a lot of seating discussion over the years. I think what we've got right now is the worst possible way, except for all the others.

Open seating--teams save seats (against the rules) and won't let anybody not with their team sit with them. Students also sit in the aisles--not cool. Later-arriving teams get whatever's left; spectators get locked out; potentially, teams get split up. (Current method)

"Cheering section" seating--you have the "fun" of moving 6 teams in and 6 teams out of seats with a good view. You also get more of saving seats, as teams have nowhere to sit when they aren't playing--unless they save a block of seats. The other alternative that I've seen is a standing cheering/dancing section if the bleachers don't go to the floor--usually better traffic flow.

"Scouting section" seating--This one may work, except for a couple of details. First, most teams use somewhere around 6-10 scouts, who may or may not also be the cheering section, if they scout at all. This means that you should figure, say, 240-400 seats at a district event alone (though that could probably be dropped by 100 at any given event). Second, most teams like to sit together--hence the seat saving mentioned earlier. I don't think they'll like to be split up, meaning that the mad rush will be for the seats right behind the scouts.

"Spectator seating"--IMO, a spectators-only zone should be small. My opinion (not that I've been just a spectator at all) is that spectators often get the most fun/inspiration out of the event by sitting WITH some of the teams, asking them questions. Saying "You have to be in this area down front" cuts off that whole aspect--in addition to potentially scooping up team-associated parents.

"Block seat", or "sectional seating", or "lottery"--Ugh. This one's a pretty big mess. If all the teams were the same size, and sat during all the matches, this could work out. The problem is, neither is the truth. There are teams with 60-100 team members--not counting parents, I presume--and there are teams with 10 team members. Then you get the parents added on, and outside spectators. So now you get the problems: Divvying up the seats into blocks, distributing the blocks so everybody gets a fair crack at decent seats, dealing with the complaints from team X that team Y stood during the entirety of their matches so team X, who wasn't in a decent place to begin with, couldn't see a thing, and all that sort of "fun".



All that said, I think there is a solution. It's in the "open seating" model--and yet, it's not in that model at all. It's in the people.

What if...
--Everyone got in line at the doors and walked, not ran, to the seats/pits?
--Seat savers took a smaller amount of seats, leaving some on the aisles and other "boundaries" as "mixing" seats?
--Other teams and spectators were welcomed into the "mixing" seats, not chased away by seat savers?
--Teams that stand through their matches intentionally took the top seats or seats around the sides?


What if FIRSTers actually practiced some Gracious Professionalism in the stands? And yes, there are teams that do that now. How about some of the rest of the teams joining in?

[1306scouting]

From the perspective of a scouting captain who is trying to implement a very large collaborative scouting initiative (CrowdScout):

Having a dedicated scouting area would be nice, but every team has different needs. Many teams (including my own up until this year) would need at least 6 seats to accomodate their scouting team. 7 would be even better. At 7 seats per team, and let's say 50 teams in a regional, that's 350 seats. Many venues couldn't accomodate that many seats without making compromises in seating for other groups (VIPs, cheering sections, and the like).
None of these account for teams that use computer based systems (and the associated power requirements) or that would need to move their scouting sheets from the dedicated area to a processing area.


If the teams of a regional wanted to organize it, I could see it happening. I would never ask a regional planning committee or a Regional Director to try to organize this for the teams, as the logistics are very difficult. Keep in mind this is for a scouting section alone, just the requirements I know from teams at my regional, and ignoring any setups I saw at Championships or here on CD.

[BJC]

Originally Posted by EricH

All right, I'll bite.

Let's assume that there are 20 fields, each needing its own camera(s). Let's assume that the powers-that-be want a streaming version (any); each camera is $200 minimum (GoPros in this case). Let's also assume that one is NOT enough--pretty reasonable assumption. Let's randomly say that each field gets 5 (partly because that gives a pretty good view, and partly because that gives some nice round numbers). Let me say again: One is NOT enough! If one doesn't work for instant replay, one probably won't work for scouting. (For one thing, how do you identify the robots from above?)

...

Essentially, it could be done. But it's not going to be easy, or cheap.

Eric,

Have you ever seen the FiM Archive run by 2337? This requires one GoPro, one fisheye lens, and one painter’s pole. In 2337's case matches are generally uploaded within several minutes of being played. This is the best recording of matches that I've seen. You can see what any of the six robots are doing at any time. This year I used these videos to keep exact tallies on how many disks were shot, how many were made, auto mode, and the endgame for multiple teams that I was keeping track of/benchmarking to compare to 33's numbers and identify improvement points -- that sort of analysis is impossible for most match videos and is really a testament to how great this single view is. Basically, I would like to respectfully disagree that you need more than one camera to have watchable match video.

Regards, Bryan

[D.Allred]

Kim,
Thanks for posting your proposal. Ultimately prearranging block seats will create its own issues. I think a limited scouting section would be good. But that would still require me to save contiguous seats for my scouts.

The root cause in my opinion is teams want to sit together, the rule says you can't save seats. These are opposed forces if you take a legalistic approach.

Originally Posted by EricH

...
All that said, I think there is a solution. It's in the "open seating" model--and yet, it's not in that model at all. It's in the people.

What if...
--Everyone got in line at the doors and walked, not ran, to the seats/pits?
--Seat savers took a smaller amount of seats, leaving some on the aisles and other "boundaries" as "mixing" seats?
--Other teams and spectators were welcomed into the "mixing" seats, not chased away by seat savers?
--Teams that stand through their matches intentionally took the top seats or seats around the sides?


What if FIRSTers actually practiced some Gracious Professionalism in the stands? And yes, there are teams that do that now. How about some of the rest of the teams joining in?

I agree with EricH's approach. Some seats need to be saved for a group. You have to train your team AND your visitors to follow these practical rules.

[jvriezen]

Answers to some questions and challenges to my proposal:

Are teams required to contribute crowd-scouts?

No. Teams would check a box at registration time to join crowd scouting and all teams would know who is crowd-scouting before the event. They can collaborate on logistics/technology prior to the event.

Do you share data with teams that are unable to contribute?

They can optionally share data with non-participating teams-- It would gracious to share with a small team that has barely members enough to drive and staff the pit, but reasonable not to share with a large team that should have scouting eyes to share but chooses not to help the crowd.

If yes, will teams abuse this? If no, who makes that judgement call?

Again, its up to those participating in crowd scouting to determine if/how to share as a group decision.


How is the data managed, averaged, and shared?

The crowd scouting teams collaborate before (or even during) the event and teams that care enough will provide/promote a system that serves their wants. Note that crowd-scouting need only gather and share the raw data. Each team participating can take the raw data and have private data crunching that they do not share. Or the crowd teams may decide to have crowd-crunch and share. This would be a major 'coopertition' opportunity! (or would you rather have coopertition game elements in the game rules? )



Do I get to take the data home Friday night?

I see no reason why not.


Could the data be poisoned before and/or after averaging? Will it be?

I'm going to assume this will be an extremely isolated case and not that difficult to detect, as most scouts should have very close counts of the raw data.

How do you keep scouting sheets circulating between 36 people? (already difficult to do with 6)

This is an engineering problem that an enterprising team will need to solve. There is no requirement that you have six sets of eyes on each bot. Your crowd scouting strategy may be two or three sets of eyes on each bot, others doing some data aggregation from the prior match-- again, this is invention that is not dictated in any way. FIRST says if you agree to crowd scout, you work with other teams to decide who occupies seats (probably a rotating schedule) and you share at least raw data gather by scouting observers-- that's it.

As to the complaint that teams won't be able to 'scout' the way they want (i.e. 18 scouts in 2009 to watch bots, trailers, humans)... Consider this:

The ideal scouting setup for your team is to have as many scouts as you care to, and have them all stationed right beside the field with Wi-Fi connected handheld cameras and input devices, and also able to record the conversations behind the alliance wall. Obviously the extreme ideal is logistically impossible to provide for every team. Well, the current model of each team having its own full scouting team with decent near-field seating is also logistically not feasible, even if you want it to be. That's why we have this thread.

We cooperate in the pits and share tools/parts/skills, why can't we cooperate with collecting raw scouting data?

[who716]

I like it and see it happening up u l till the Craigslist part, I always thought an alliance selection for seats, when you register for the regional you are randomly assigned a section.

[EricH]

Originally Posted by BJC

This requires one GoPro, one fisheye lens, and one painter’s pole. In 2337's case matches are generally uploaded within several minutes of being played. This is the best recording of matches that I've seen.[...] Basically, I would like to respectfully disagree that you need more than one camera to have watchable match video.

I have not. Incidentally, I believe the GoPro stock lens is a fisheye, so no extra lens needed.

I would actually argue that you need more than one camera--or at least more than one memory card. If nothing else, it gives you a spare. More likely, it can speed up "cycle time"--the time to get a match up. If you start a camera recording, then grab the other one, dump the memory card to your computer/upload device, and get the camera set up, the match could theoretically be up before the next match starts. Dependent on internet at the venue, of course!

I would also point out that I was also responding to comments about the webcasts as well as recording for the scouts--while 1 camera could work, I do like the occasional close-up shot (particularly of high-intensity action), which in my experience no GoPro can do. For that sort of application, 1 camera is definitely going to be suboptimal.

[connor.worley]

Originally Posted by jvriezen

...

Okay. Just asking the hard questions. I'm interested in seeing how crowd-scouting is implemented in the future and how groups decide to approach it.

[Andrew Schreiber]

Originally Posted by EricH

I have not. Incidentally, I believe the GoPro stock lens is a fisheye, so no extra lens needed.

Confirmed on my GoPro Hero3 and the Hero2 I borrowed at CMP they were both fisheye lenses.

I didn't check the Hero2 but you COULD use the micro HDMI out to stream the footage back and stream it. I haven't tried this though.

[s_forbes]

Maybe sporting an unpopular opinion here:

Is there actually anything wrong with the current free-for-all system? I've never had any trouble getting a good seat for a few matches during qualifications. If you are courteous and pleasant about it, pretty much anyone is willing to give up a seat they are not using for a short period.

Teams trying to save seats do not impact much of the seating space. Do you really want to create more rules that dictate where you can/cannot sit?

[Pault]

Originally Posted by EricH

What if...
--Everyone got in line at the doors and walked, not ran, to the seats/pits?
--Seat savers took a smaller amount of seats, leaving some on the aisles and other "boundaries" as "mixing" seats?
--Other teams and spectators were welcomed into the "mixing" seats, not chased away by seat savers?
--Teams that stand through their matches intentionally took the top seats or seats around the sides?


What if FIRSTers actually practiced some Gracious Professionalism in the stands? And yes, there are teams that do that now. How about some of the rest of the teams joining in?

This. I agree 110%. It sucks, but there really is no option to try and fix the seating problems through rules and regulation. It just won't happen. Every team has their own needs, and regional planners already have enough on their hands. Just some food for thought, though: The Boston Regional had a magnet board with a picture of the stadium on it and a magnet block for every team with their number on it at the entrance. Teams could put their magnet over the blocks of seats that they reserved. As far as I know, it was a very lax system, nobody tried to enforce it. But I do remember that my team lost the seats we were sitting in on Thursday during Friday, so atleast in that one situation the magnet board was not really obeyed. I wasn't in the stands that often, so I can't offer more information about if that my team's case was an isolated case or fairly common.

Also, about crowd scouting: I don't think there should be any bias towards teams who choose not to participate. My team is pretty weak at scouting, so we would probably benefit from this. However, I can put myself in the shoes of teams with good scouting, and I imagine that for them it is a competitive advantage. Scouting can often make a huge difference, so teams that put in the effort to be good at it will use that to gain an advantage in the same way that they would use a great robot to gain an advantage. Now, I am sure that they are willing to help other teams get better at scouting, and maybe even let teams use their system (think any scouting app available on the android or apple App Store), just like a team with a good robot would help out other teams to make better robots. But they wouldn't just hand over all of the scouting data, like the team with a good robot wouldn't just hand over all of their CAD and code so that other teams could easily build the exact same robot. So if teams want to collaborate on scouting, that is great and it will really bring up the bottom scouting tier. But there is nothing wrong with a team that wants to keep their data to themselves.