Pages

Thursday, March 13, 2014

GP Richmond and the Pink Paper Team

Last time out, I discussed some of my social anxieties tied to this mega-sized Grand Prix Richmond. I reached out to Event Manager Jared Sylva and Head Judge Jason Lems independently, explaining my issues. I don't know how much was due to this, but when staff assignments came out, I was in the Ballroom, which later became known as the Pink Split/Room, because our Scorekeeper was Kali Anderson and she procured pink paper to run the event on.

A side note, when a GP is split, it is customary to have different colored paper for each event so things like match result slips don't end up going to the wrong Scorekeeper. This was less relevant to the Pink Room because the Ballroom was in a completely separate part of the Convention Center from the Blue and Green Splits. This made for a much more comfortable event for everyone involved. We didn't have the added distractions of coverage, feature matches, vendors, side events and the associated mass of humanity. We just had a 1600-player tournament. This made things a lot easier on me from an anxiety standpoint, and it was great for the judges and players as well. We even had carpeted floors. All of the feedback I've heard about the Pink Room has been positive.

But a nice room alone isn't enough to deliver a great experience for players. When I spoke to the Paper Team, I made it clear that we were the team that had the most real impact on how fast the tournament finished, something that Head Judges love to emphasize. Sure, time extensions from deck checks could spill over and make for a long round turnover, but whether they did or not was largely a matter of luck. The Paper Team is directly responsible for additional time burned during what Nick Fang describes as the Prep portion of a round here. Paragraph break for emphasis time.

The longer it takes to get the pairings from the printer to the pairings board is time that you, the Paper Team, are exclusively costing everyone in the tournament.

There are a couple of things you can to do get back this time. Even if each of these actions only saves seconds on its own, that's seconds per action over multiple rounds, potential minutes that you can save 1600 players and 30 judges. Isn't it worth that extra effort?
1) start taking pages off of the printer as soon as they start printing. Do not wait for an entire set to finish printing. Most large tournaments will have multiple letter ranges and each range could have multiple pages. It takes time to collate and divide these pages among your team members, so the sooner you start taking pages off of the printer, the sooner you can get to that task.
2) have tape prepped. The easiest way to do this is to have tape prepped at each pairings board. I recommend doing this when you take down the pairings in the previous round. (You should be doing this because it takes time to take down old pairings and put up new ones at the same time. Leaving old pairings up too long can also lead to player confusion.) One innovation that someone suggested at GP Richmond (at least this was the first time I had seen this--and I apologize for not remembering who suggested this) was to have reusable tape loops. This solved the twin problems of tape ripping off when you take down old pairings and having ugly pieces of tape lining the sides of the pairing board in preparation for the next round. The tape loop is simple. You attach the piece to your surface as normal, but at the bottom you loop it so that it adheres to itself, leaving you with a non-sticky loop or tab. You can then lift that tab until some of the adhesive portion is exposed, and slide the piece of paper under it and attach the tape to it.
3) walk fast. Judges are advised to not run on the floor. It looks unprofessional. It is dangerous. However, as I told the Pink Paper Team, when you have pairings in hand is the fastest you should be walking in your judging career. Those seconds matter. And the effort you display to the players matters. You want to be seen hurrying with the pairings because...
4) get players out of your way. Use your judge voice to announce that you are coming through with the pairings. They will move out of the way for you. One surprising thing I found at Richmond. They will even help you put up the pairings. I was having a little trouble with the tape and a player reached out to help me. They just want to find their table number and get to their seat. If you need help to make this happen, they will give it to you, and in the future I'm thinking of being a little more proactive in asking for their help.

Another important factor in saving time is traffic flow. This is all about the room layout and the positioning of the pairings boards. One mistake I see Paper Teams make is automatically placing the boards in alphabetical order clockwise around the room no matter what (because we read from left to right and the letter ranges are displayed this way (A-B, C-F, etc). Why does this matter? It may not necessarily match the way that players are seated for Player Meeting, which is seated alphabetically from Table 1. In the Pink Room, Table 1 was to the right if you were standing on the stage and the tournament expanded to the left. If you place the pairings boards in alphabetical order starting on from the left of the stage going clockwise, you are created an unnatural traffic flow for that alphabetical seating. The A-B name range has to walk to the left to find their seat, then travel all the way to the right to their table. at the opposite end of the alphabet, T-Z will go to the right to check their seating and go back all the way to the left for their table. Hey, at least the middle of the alphabet won't have to go far. :(

This is terrible. Not only are you forcing the 2 ends to walk the maximum distance, they have to cross each other in the middle, leading to a traffic jam in the middle.

There are other factors that can lead to traffic jams. One classic one is space around the pairings boards. For the Pink Ballroom, the entire floor plan inside was taken up by the tables and chairs for matches, plus the main stage and single dealer booth (a small ad hoc one in the corner). If we had chosen to place the pairings boards inside the Ballroom, they would have been up against walls and the distance between them and the tables would have been far too narrow. You need to allocate enough room around pairings boards to allow people to both come and go, keeping in mind that up to several hundred people may need to access each board. Unfortunately, Magic players haven't acquired any kind of sensible etiquette like "enter from the right, exit from the left." Perhaps there's a social engineering opportunity there.

Due to the lack of space in the room, we placed the boards outside the Ballroom in the lobby area. At first, we had all 6 boards in the main lobby. The Ballroom was rectangular in shape with 4 sets of double doors as entrances from the lobby on the long side of the rectangle. The main stage was on the opposite long side. Again, this meant a long walk from printer to boards, making speed of the essence.

But wait. After placing the 6 boards in the main lobby, I began to worry about the traffic through those 4 double doors. As with space around boards, doors can be a serious impediment to traffic, especially because players do not follow any orderly traffic pattern as a norm and will just try to walk through each other going both ways. 4 double doors for 6 boards seemed low. I spoke with a few of my team members about my concern and they agreed with me. We noticed that there was a set of double doors on each of the short sides of the room. They were closed, but we confirmed that they could be opened and did so, placing the A-B and T-Z boards at the appropriate sides to match the numbering layout (counter-clockwise).

It's important when doing any job not to get married to the plan just because it is the plan. Flexibility, especially in light of actual logistical issues is important. It would have been a simple matter to stick to the plan of all 6 boards beyond the 4 main doors. Looking back at the way that traffic ended up flowing just with the 4 remaining boards in the center, this clearly would have been awful, and I'm glad that we made the adjustment.

Speaking of adjustments, it's important to watch the actual flow once players start moving and see if you can identify any problems with it. Theory is all good in... theory, but what matters most is what actually happens. In our case, we immediately noticed that the P-S board was taking a considerably longer time to clear and get seated. We went out in the lobby to examine the issue and discovered that the problem was that the P-S board was standing right next to an ill-placed pillar. This caused awkward in-out traffic flow issues since players couldn't use the pillar side of the board, essentially cutting the traffic capacity of the board in half. We moved the board away from the pillar. Even though it meant a longer walk from the Ballroom, getting the traffic flowing better meant that it kept up with the other boards.

Speaking of traffic flow at the pairings boards, we were lucky to have the new double-sided zip banners. This meant that we could put pairings on each side of the banner, doubling our capacity per board. Frankly, even without the double-sided banners, I would have probably put a second set of pairings on the backside anyway. When you're dealing with 1600 players, you have to do everything you can to divide people up into smaller chunks for the process of seating. Twitter pairings are a great innovation for this, but the buy-in has been slow and being dependent on Convention Center WiFi or having their concrete bunkers block signals has caused issues at more than one location. I briefly considered putting up a third set of pairings somewhere, but couldn't find a sufficient location that would reduce the stress on the existing traffic points or not cause other logistical issues to crop up. My understanding is that the Blue and Green Events (running together in the main hall) only had 1 set of pairings each at the start of the day for about 1300 players apiece. That's certainly one among many reasons why we finished so much faster.

I tracked the amount of time it took from when the printer started printing pairings to when the round started for our event. As I said, this is time that we the Paper Team cost everyone between just getting pairings collated and posted, and traffic issues. Some of it is on the players actually getting to their seats quickly, and yet another factor is the Head Judge and his or her willingness to start the round when enough players are seated, absorbing some number of tardies or "you're okay because you were on your way."

In Round 1, the printer to round start turnover was 12 minutes. Part of this is probably a natural player education period. It takes a few rounds for players to get used to the room layout, find good traffic alleys, and learn the general layout of table numbers. (btw, putting up signs or maps with table numbers is another way to potentially cut down on this turnover time.) Part of it was also the Head Judge making a few more announcements. For the rest of the day, the printer to round start turnover averaged under 8 minutes. It wouldn't surprise me if an equal-sized tournament with only one set of pairings took 10-15 minutes on average. Given that, it's likely that the decision to only post one set of pairings initially (my understanding is that this was corrected in either Round 2 or 3) cost those events around 5-10 minutes.

Wow. I've already written quite a bit and I haven't even talked about cutting result slips or how we took care of our Scorekeeper. I guess there's going to be a Part 2 to this episode. Stay tuned because you don't want to miss the greatest innovation in Scorekeeping technology of all time. Before we go, I want to give a huge shout out to the Paper Team, Ward Warren, Rick Salamin, Ben Klein, Jarrod Williams, Xander Forral, and Michael Mills. They bought into my messaging and their experience and willingness to engage me in conversations about how to improve our processes led to a lot of the things I've discussed here. The day was so hectic that I didn't note who said or did what in most cases, but each of you added something important to the team that made it better, and my hope is that future Paper Teams can learn from the example and bar that you set in the Pink Room.

2 comments:

  1. Riki, I was in the "pink" flight and you guys did an amazing job.

    ReplyDelete
  2. Too many comments here: http://www.whysogrumpy.com/2014/03/14/papers-please/

    ReplyDelete