GP Philadelphia 2008 - Head Judge Report

  • Print
Author Image

The letter A!t any event I judge, I try to pay attention at the event and analyze the results afterwards. Grand Prix are especially interesting because they can actually be harder to head judge than a Pro Tour. What I want to share with you, be you a newly minted judge or a long-time certified trooper, aren't rulings or anecdotes from the GP. This report is to share ideas, techniques, and processes that have worked for me and may very well work for you. Feel free to file off the serial numbers and steal what works for you; it's how all judges get better.

From a philosophical standpoint, I believe two things that are especially important for Grand Prix events. First of all: with the exception of a judge's attitude, everything else can be taught. If a judge shows up, is willing to work, and can follow directions, the DCI system can truly sculpt him or her into a better judge. Secondly: good judging is learning good techniques and then scaling those techniques to be bigger and go faster. For a quality judge, good events are not aberrations – they're standard operating procedure because good judging (just like bad judging) becomes habitual over time.

Over Here, Over There

Grand Prix Philadelphia was held concurrently with Grand Prix Vienna on March 15th and 16th. North American GPs are contracted from Wizards of the Coast to independent Premiere Tournament Organizers. In Europe, Grand Prix are handled directly by WotC. That functional difference, the cultural differences, and regional variations translate into very different experiences at GPs on either side of the Atlantic. One purely statistical result is Philadelphia had 30 judges for 969 players – North America's largest GP to date. That's a ratio of 1:32.2. In Vienna, the GP was split, and their ratio was 51 judges to 1154 players, or 1:22.6.

I was asked why GP Philly didn't split like Vienna. The simple answer is that we weren't in Europe. The more complex answer is that because of the different dynamics, it's not easy to split a North American GP, and it's usually not needed. A major factor in this particular case was actually the room itself. While we had plenty of space – about two football fields of floor – the room had no carpet and no dropped ceiling. The result is that the acoustics were quite challenging. Having two demanding events at once with a large audience not accustomed to split GPs would likely lead to a mess where you can't effectively communicate beyond a word or two. I'd be unsurprised to see North America split a GP in the future.

So Many Zebras, All in a Row

When faced with a lot of variables and a lot of activity in a small period of time, preparation is your best option. Several months earlier, I had anticipated a medium turnout for GP San Francisco. Given that, I tried something new for a Carter GP – L2 team leads on Day 1. Each team had an experienced person on board to assist the team lead, so I wasn't too concerned. However, when the tournament got a larger-than-expected attendance, the leads were under more pressure than I would have desired. While that event went well, it seemed wisest to stick with the time-honored plan of L3s leading on Day 1 to set the example and L2s on Day 2 to demonstrate their own skills and practice what they saw. Given that the Northeast has a different flavor of player than the Left Coast, I had my team leads for Philadelphia lined up about three weeks in advance.

I asked the staff what their goals and desires for the event were about two weeks early. With a little under a week to go, I had teams laid out and plenty of additional notes to help me make quality choices in meeting the event's needs and the desires of the staff. There's always some last minute flux as people's travel arrangements shift or things come up, so I kept the final list private until just before the Friday meeting.

A personal habit of mine is to meet with my Day 1 team leads the night before the GP. This allows me to take plenty of time in going over all the details of how I want things to run. Since all the team leads (if available) are there, they'll have an idea of what each other is covering, and they'll all know what my big emphasis points are. This also allows the team leads to ask clarifying questions or sketch out unclear areas well away from the event. The end result is team leads that are confident in how things will be run, and I can shave precious minutes off the staff meeting the next morning because the team leads can disseminate the finer details of things rather than having to cover them as an entire group.

Having my notes on judges with a rough plan laid out in Excel was also helpful as Day 1 ended. Juggling judges between the needs of a main event and side events is often difficult. Figuring this out before the event can lead to poor choices – maybe a judge needs more help, or maybe I notice a judge doing particularly well, or maybe the TO really needs judge X to do Y. Figuring out the Day 2 staff at the event can go poorly because Day 1 is so busy. The hours spent outside of the event sketching out rough ideas of who should do what and getting word from the staff made those choices easier. Ultimately, the hours of work on notes only saved a matter of minutes at the event, but those minutes are precious when you have several hundred players to worry about.

Another simple tool I use for events is a staff roster. On the roster, I put the name, level, and DCI number of each judge. I ask each judge to sign and initial the roster. This is handy because it then goes to the scorekeeper. Since it's pre-printed, the names will be in an order that makes sense, and the list gives your scorekeeper all the info he or she needs to enter the staff. Your staff can verify that the information you have is correct, and the signature and initials give the scorekeeper a handwriting sample in case a judge forgets to put their name under a penalty. I also include a few blank lines on this list for late additions.

And They're Off!

The biggest challenge with any tournament is getting it launched. Anything can go wrong, but it takes a lot of things going right to get the event moving. I knew getting the decklists counted quickly was vital. We always want to get them done during round one so everyone gets hit with their penalties at the same time – as round two starts. A slow count just drags the process out, and you have to get counts done at some point, so I decided to bite the bullet and push for them to be done in round one regardless of the player count. This meant having a very strong emphasis on the deck checks teams. Philadelphia was divided into five teams: Paper, Logistics, and Checks 1-3. I sat down with Adam Shaw at GP Vancouver three weeks prior, and we talked about how Vancouver was going and how I wanted Philadelphia to go. Because of that talk, I knew Adam had to be a Checks lead. With a quick count required, the Paper team was assigned to cover a very large floor, and Logistics would assist whichever area needed them the most. Adam, Shawn Doherty, and Eric Smith united their teams into one mighty counting monstrosity, and they completed 969 counts with 20 minutes remaining in round one – with only 11 bad decklists.

The low number of decklist errors was due in part to stolen technology. Seamus Campbell at GP Vancouver paused his player meeting (seated alphabetically for easy sorting, of course!) for one minute. During the minute, players were instructed to count their lists and make sure their names were on their lists. While I don't like slowing down for anything, the benefit to judges and players for such a simple exercise seems to prove itself pretty highly. I'd recommend you give it a shot at your local PTQs and see if you like it as well.


Early in the day, I worked with Logistics to find optimal positions for our pairings boards. Sadly, we had four to work with, so we staked out the tournament area with boards along the edges, but we pulled them in about a quarter of the way along the long sides of the room. The large turnout meant we also needed to shove tables out of the way so people could easily approach and leave the boards. Still, that's a lot of people, and two things could have been done better.

Most importantly, we had large columns (six feet per side) in the room, and the site personnel had not forbidden attaching things to the walls or columns. (Always check with the TO before you whip out the wall tape or thumb tacks.) What we wound up doing after gauging the player flow was to abandon using actual boards stage-right and put pairing on three sides of two columns. This meant we could take those boards over to the stage-left front board area where there was no column and triple that posting, too. At the stage-left back area, we combined a board and two column sides to have all areas posted in triplicate. In considering whether to divide the letters into twelfths instead of triplicate quarters, I felt is was best to keep the areas consistent with what the players had seen already despite the added pairings copies. This dramatically improved the flow of traffic and thus the all-important round-turn-around time.

The other logistical issue came up as we shrank the number of active players. Once we were down to half the players, we lost the corresponding number of tables. Due to the massive size of the room and the buffer area to let people move in and away, the pairings area furthest from the stage was significantly far from the active tables. A timely start for most players meant the K-P players had to scramble. Several tardiness penalties were overturned because the "boards" were, in my opinion, exceedingly far away for that round. The Logistics and Paper leads coordinated as needed to get the pairings relocated along the new edge of the event.

Further rounds were logistically smooth. The triple-pairings arrangement switched over to two copies of pairings, one on either side of one set of standings, for the last two rounds. Only one standings copy per post is needed in this arrangement because many people are only concerned with the pairing. Judges were stationed to move players along after they've had time to check their pairing and relative standings. Standings are not there for players to contemplate in detail, and they are never to be taken to a table.

One Down, One to Go

Day 1 wrapped up with little fanfare. In case this report isn't obvious enough, I can go on and on about judging for extended periods (no pun actually intended), so one of my goals is to keep meetings at events short. In keeping with that, I went through a very quick (by any standards) wrap-up meeting and gave everyone their assignments for Sunday. I decided to leave the room untouched for the next day's setup. This let judges go eat and sleep sooner. It also meant the Day 2 Logistics team could meet the new day with a task ready for them.

Being a Flotation Device

Day 2 of a GP is a chance for me as head judge to work more hands-on with people and hopefully have a minute to chat. At GP Vancouver, Seamus assigned me as a "floater." What is a floater? That's a judge whose job is basically to head off problems before they need a head judge intervention and to operate on behalf of the DCI to give personalized care and direction to judges on staff. Floaters tend to be L4s and L5s because of that personal attention to developing judges (either at that event or in talking about how that judge's own area is developing back home).

The key for any judge is to stay focused on the event. In Vancouver, the staffing level and event size were such that I could give judges hands-on time as part of solving problems early, so the event and mentoring needs were met. In Philadelphia, I knew the demands would be different and that we'd have to emphasize taking care of the event. I was glad to see that at little moments on both days, but especially in the more pressure-free Sunday, many judges had opportunities to share knowledge and ideas. The head-judge-me felt a little pang that there wasn't a flotation-me running around with the time to show everyone that works these GPs how much the DCI appreciates their efforts.

A Few Odds at the End

One of my staff asked an interesting question. He rented cards to players for a percentage of their winnings and wanted to know if that was okay at a GP. This is where your judgey-sense should be clanging "conflict of interests!" Two things I'd strongly recommend in any case: 1) Don't rent cards at an event you are judging. Renting in days prior at least gives you some distance from the event, but it's still not pretty. 2) Definitely don't rent cards for a percentage – that just asks to be abused. I'm not saying that this judge or any certified judge would favor someone, but there's bad and then there's really worse. For the Grand Prix itself, I let him know that any form of rental was verboten. Frankly, it just looks bad; and as judges, our reputations for impartiality are sacred – the GP is a place to build your reputation, and you want to do it right.

At GP Vancouver, Jason Ness – a premiere event TO as well as a certified judge – was trying something I thought was useful for his GP Trials. Two factors had to come together to get this technology to work. One, the event has to be constructed; and two, the event has to have a decent turn-around time. Since Last Chance GP Trials are now single-elimination ("grinder" style), turn-around time is not an issue for half the field every round.

The particular piece of technology is handing players back their decklists after they drop. It seems simple – we don't need the list if they're not in the event – but the effect is important: that player can then have his or her decklist immediately ready for the next grinder to launch. The simple way to handle this is to have one repository of lists separated by Trial. When a player drops, he tells the list keeper his name and his Trial, and that's all there is to it. TOs should be happy because people jumping into the next event means more revenue for them, and players appreciate not having the write their list several times a day.

Another piece of tournament organization tech was originally described to me as "Tokyo-style" draft. As I understand it, PT Kobe would be the actual origin, but Kobe draft doesn't have the same ring to it that Tokyo draft does. We've tried this organization method in Seattle with much success. Here's the way Tokyo/Kobe draft works:

Instead of players coming up to a registration desk and signing a piece of paper to draft, a table is set off to the side of registration. If a player wants to draft, he or she sits at one of the eight chairs by the table and signs up on the form on that table. (A clipboard with many copies of the form works best.) Once the table is full of eight players, a judge swings by and carries them to a spot for the actual draft while the sign-up table refills.

The utility here is that it requires less supervision. Since you don't launch without eight players right there, you don't have to use effort or microphone time tracking down that last stray player. Once a few players sit, they often will occupy themselves with a casual game or trading as the table fills. Once the table nears the limit, they'll even begin recruiting for the last player or two so they can start – good for them and for the TO.

The key with any system of registration is that it fits with your staffing levels and the behavior of your players. If you can, test things out and go with what's best for you, your TO, and your players. When your draft sign-up is very slow or your crowd is small, you might stick with a more traditional method, but at a large event (perhaps Regionals or a Prerelease), consider the option of borrowing technology from our friends in the Land of the Rising Sun.

I hope this report has piqued a few thoughts or illuminated new ideas. If you have any questions or would like more information, I can be reached at

Be seeing you.


  • Planeswalker Points
  • Facebook Twitter
  • Gatherer: The Magic Card Database
  • Forums: Connect with the Magic Community
  • Magic Locator