Ingrid, my wife, was head judge of Grand Prix: D.C. I was also judging. This gave me a fairly unique perspective on the largest North American event ever held.
Ingrid first learned that she was going to be head judge at GP: D.C. sometime late last year. She knew basic details — when, where, what format, etc. — long before they were announced. She let me know early, so I could arrange time off, but of course I could not write about it until the official announcement was out.
Ingrid has a ton of work to do prior to the GP, but very little of it makes for an interesting story. I’ll just cover one area, since it affected me directly: assigning judge responsibilities.
Ingrid had roughly sixty judges working the event, plus several scorekeepers.
Some of those judges had to handle the Grand Prix Trials held the night before. We had about two dozen 32-man GPTs run Friday night. At the start, you need at least a couple judges for each event. As they wind down, some of these can be shifted to other GPTs as they fire. Still, the Friday night GPTs make for a long night of work for several judges, and Ingrid had to consider how to adjust Saturday staffing to reflect the fact that some judges would need a chance to sleep in.
For Saturday, some of the judges have to work the main event, and some handle other public events, like drafts, the iPod tournament, etc. That means splitting off a handful of judges — mainly lower level judges, plus one or two higher level judges to handle questions and appeals, etc. You also need an experienced judge to act and to assign judges to events, and to shift them around as events wind down and others fire. Of course, the exact number of judges needed for sides will vary, depending on attendance. If the event is a bust, you won’t have many people to play in the other events either. If it is huge, like D.C., you won’t have table space for too many other events. In D.C., we managed to squeeze in a couple drafts on the edges of the event, but table space was at a premium.
Fortunately, the tournament organizer had some local judges that could cover some of the duties for events, but we still had a small handful of GP judges working that area on Day 1.
The next question is how to divide the judges into teams. You need a team to handle paper — putting up standings and pairings, cutting and distributing match results slips, etc. With a big event, pairings are listed in name ranges (e.g. A-C, D-G, etc.), and you really need one judge per name range if you want to get all the pairings up at one time. After the pairings go up, those judges distribute slips, then cover the floor and answer questions.
You also need a couple deck check teams. These teams handle deck checks, and they work best if the teams are smaller, so they can keep track of each other. The team leads coordinate between teams. Given the numbers of judges, etc., three deck check teams seemed optimal.
The final team is logistics, which is responsible for everything that isn’t paper of deck checks. GP: D.C. was a Standard event, so logistics was a lot simpler than Limited events. In a Sealed event, for example, logistics needs to prepare and distribute Sealed product and decklists for all players; collect and redistribute the pools once registered; then set up land stations. Constructed is easier, but there is still plenty of work.
At the GP level, nearly every judge chosen to work knows the basics of all of these jobs, and the team leads will provide extra training. However, some judges are particularly gifted at handling certain areas. The head judge may choose to assign those judges to that area — or not, if they are needed elsewhere and enough judges are good at any area. In developing teams, Ingrid had to make sure that all the teams were strong, even if that meant that certain players were not at their best possible position.
Team leads need to be chosen. This can be the strongest, most experienced judge on each team. However, for the long running health of the judge program and the game, choosing a talented level 2 judge as team lead can give them valuable experience. In that case, you need to assign a skilled judge to act as support/mentor for that judge.
In developing teams, Ingrid also had to be cognizant of judge personalities. Generally, judges are a professional bunch, but some people just grate on each other. They can and do work together civilly, but that can be tiring, and an event like a GP is hard enough without adding any unnecessary stress.
An event like GP: DC attracts a lot of different judges of different levels. Ingrid had to make sure each team contained an appropriate mix of levels.
Judges were also arriving from a lot of different countries, and spoke a number of different native languages. Ingrid also had to balance out these factors. Judges should be spread out. One reason judges come to events like this is to meet and interact with judges from other areas and other parts of the world. At an event like this, you really don’t want to work with people you see all the time. I remember Ingrid commenting when, after creating teams with the right mix of strengths, nationalities, levels, languages and other factors, she realized that the judges from Florida, who work every PTQ and Prerelease together, were all on the same team.
Start again.
If designing teams seems difficult, try doing it twice. Until the night before — when preregistrations start pouring in — no one really knows whether the GP will have to be split into two parts or not. That means Ingrid had to create a schedule for the judges based on having one humungous event, or two big events. It isn’t just a case of splitting the big teams in half. For example, the paper team needs roughly one judge per pairings board. The software that runs tournaments can only split the name ranges in up to six ranges — and we would want all six for each side, if possible. This means that paper needs six judges for a big event, and two teams of six if we split. Deck Checks worked best with three teams of judges for a combined event, and two sets of two teams each for the split event. And so on.
Having watched — and worked on — this mess, I much prefer “scheduling” for the small store launch parties I head judge. Since those are staffed by — well, just me — scheduling is infinitely simpler.
How Big?
Friday, one of the local judges had a barbeque for the judges at his house. It was a blast (thanks Ryan!) Over the course of the afternoon, a couple dozen judges showed up. Along with drafts, EDH, Frisbee, and so on, we talked about attendance. Guesses were all over the place.
Personally, I was pessimistic (I always am.) After the disappointing numbers for U.S. Nationals Qualifiers, I was worried that we might be down. My private guess was 600, but I didn’t share that number. (Good thing, too.) The guesses were all over the board.
Late afternoon, we had to head over to the venue. Ingrid had lots of last minute stuff to check with the TO — and we wanted to see how the GPTs were doing. (They were doing just fine. The venue was packed already.)
The venue itself was pretty decent. On the plus side, it was large enough, and had plenty of space between tables. More importantly, the tables were set up for three matches per eight foot table. Speaking as a player, I love having those extra few inches of play space. The venue also had decent lighting and one of the best PA systems I have heard anywhere.
It wasn’t perfect, of course. As a judge, I would have killed for carpeting. I would also have liked acoustic tile on the ceiling. It was LOUD! It was also very humid. We had rain much of the weekend, and the heat and wet strained the AC. It was bearable, but the humidity did cause some problems with foil cards both days.
A quick note on choosing venues. The TO has to juggle the cost of the venue, the cost of everything else, and the cost of entering the tournament. The costs of an event like this are amazing. Here’s just one example: tablecloths. You really need to have tablecloths on the plywood and metal tables at most venues. Those tablecloths are not free — rental, plus cleaning afterwards (which you pay if you rent) can run upwards of $8 per table. With somewhere around 250 tables in the venue, renting tablecloths ALONE would have cost the TO two grand. Instead, the TO bought tablecloths online, and resold them afterwards. Whatever — the point is that running an event like this bleeds money in all directions, so compromises have to be made.
Back to attendance.
By 6pm or so, we had 600 people preregistered, and the flow was constant. After talking with judges and other friends, I wound up keeping the scorekeeper company. He was off on one end of the venue, all alone. I hung around to interrupt and distract him — and to help decipher strange American handwriting. (Henk was European, flown in specially for this event. More on that later.)
Generally, about 60% of the players preregister for a Grand Prix. They don’t have to worry about the line the next morning, so that number is pretty good. For D.C., however, we figured that could be low, since a lot of players in the greater D.C. area could drive to the event the morning of.
By 8pm, we were closing in on a thousand preregs, with preregistration not set to close for another two hours. By the time we closed, we were over 1,200. By next morning, we had 1,932 players in the event.
As you probably know, that makes it the largest Grand Prix in North America, and the third largest in history.
With this number of players, a split was inevitable. It was also something fairly new to North American GPs.
Splitting the Event
By splitting the event, the tournament is divided into two. The two events run separately — two head judges, two seating areas, two sets of pairings, two scorekeepers, two sets of judges, etc. After Day 1 ends, the tournaments are merged back together, but for the first day, they are separate tournaments.
Splitting GPs has been common in Europe for several years, but has only recently been done in the U.S. One reason is that the scorekeepers have to know how to do so. For GP: D.C., Wizards had flown in one of the European scorekeepers with experience doing this. The problem is DCI Reporter, the software used to run these tournaments. DCI Reporter does not support split tournaments in any way. To split a tournament, you have to create a second tournament, then find the data files for the tournament and manually split them into two. This involves not merely cutting them in half, but doing so in a way that ensure roughly the same number of byes, etc, in each half. Then you have to verify that the split was accurate, and that no one is in now playing in both halves of the event.
This is not all that difficult, and many American scorekeepers could puzzle it out, unaided. However, they probably could not puzzle it out as quickly. The European scorekeepers have developed a spreadsheet and techniques to facilitate the split. In D.C., we had the tournament split, verified and functioning within twenty minutes or so of registering the last player.
We also had an American scorekeeper learning the process, for next time.
The other half of splitting a tournament is gluing it back together for Day 2. Once again, the process of reintegrating the player databases is manual. The combined database needs to include not only player data, but the player won loss data and each player’s tiebreakers. Once again, this is not something DCI Reporter does. It is more something DCI Reporter treats as database corruption, and you have to perform a number of work-arounds to get through it.
American and European Ways of Running a GP
You may wonder why the Europeans first developed the methods of splitting a GP. Part of it is because the European mass transit system, coupled with their relatively higher population densities, made it easier for European GPs to get gigantic. That is part. The other part is that European GPs are managed and run directly by an arm of Hasbro. In Europe, a crew of Wizards employees delivers the GP kit — the banners, stage setting, scorekeeping system, etc. They also set it up and manage the event. In the U.S., the TO, TO’s local staff, and the judges do all of this. The U.S. GP kit is a pair of wheeled crates that Wizards ships to the venue, and that the TO and judges then unpack and assemble.
The U.S. GP stage setting is considerably less extensive than the European counterpart. It fits in two wheeled crates. The main kit is about 1.5 meters tall, 2 meters wide and almost 2 meters deep. The crate for the banners is a half meter or so tall and wide and 4-5 meters long. In America, judges tear down the stage setting and banners, and pack up the kit.
Seeing the venue losing the players, the tablecloths, the banners, and all the rest of the paraphernalia is sad. It is really clear when the event is over. The venue stops looking like Magic and starts looking like an empty convention center.
For the judges, though, the event doesn’t end when the venue closes. Saturday night, once we finished game play, results entry and prep for Sunday, Ingrid, John Carter, and I ended up leaving the hall slightly after midnight. We had no car, so the only options for dinner were walking up to the Taco Bell drive-through, and Hooters. Hooters it was. The Hooters waitresses have some obvious assets, but intelligence clearly was not one of them. It took over 40 minutes and three visits to the hostess stand to get someone to take our order. When she finally appeared — well, she did not sparkle with wit and brilliance.
Sunday night, the judges left the hall about midnight. This time we had more judges with us, including several with cars. However, the options for after midnight on Sunday night were limited, and we ended up at an IHOP. The food was not terrible, and the big advantage was that IHOP does not kick you out. We hung around discussing stuff judgely and non-judgely until after 4am. We finally got back to our room close to five, then slept for an hour before heading out to the airport. We did get some sleep on the flight home.
We got back to Madison late Monday. We unpacked, washed our judge shirts, and started repacking. I worked Tuesday, then did judge prep until 2am, slept for an hour, then got up for the drive to O’Hare airport and flew to San Juan. I’m there now, writing this on the hotel balcony, looking out at a huge thunderstorm. I’m also noticing that I am barely coherent, so this will almost certainly appear a week late.
I should discuss some rulings before I quit.
First, if you are in extra turns, and the whole tournament is waiting for you to report your match, when it is done, sign the match result slip immediately. you can desideboard, discuss your play, etc., afterwards. This event, I had one pair of players that had had a time extension, then dawdled their way through the extra turns. I had already given both players slow play warnings, but at the end to the match one player wanted to argue about why it wasn’t their fault that they were delaying the tournament. As he is sitting there, holding the last match result slip, he wanted to argue that he wasn’t holding things up.
After six years as a judge and over a decade as a player, I have realized an essential truth: we players are morons.
The most common ruling I had Day 1 — amazingly enough — was something I have never seen before. A player plays Bloodbraid Elf and cascades into something that fetched lands — e.g. Rampant Growth or Borderland Ranger — then shuffle away his hand while resolving the fetch. I had this happen three times on Day 1.
The penalty for this is pretty simple. Cards have gone into a zone that they are not supposed to be in. The infraction is a game play error — game rule violation. The penalty is a warning. Worse yet, for the player, we do not try to reconstruct their hand. The player continues the game with no cards. (That’s the only consistent option. In some special cases, you might be able to reconstruct part of some hands, but the goal of judge policy is to be consistent. Since we cannot recreate every lost hand, we don’t recreate any.)
On the plus side, now the player is pretty much immune to Blighting.
More next week!
PRJ
“one million words ” on MTGO