I have a much better feeling about this tournament than about GP New Jersey. I think that when someone is asked to HJ a GP, it is not only because of his skills in making rulings. My rulings in NJ were good, but all the rest could have been better, it ran itself and I missed some of the few things in which my actions could have made it better. GP Paris was a much different story: many things went wrong but I made my best to keep it under control...
Like my report about NJ, I’m going to divide this one into three parts: general organization, judges’ management and rulings.
Before the event
At GP Helsinki, I had a meeting with Erwin and Gis about the major issues. The venue was divided into three levels, and we would probably have to divide the tournament into two groups according to the new procedure since Erwin was expecting around 1400 players. He definitely looked like he wanted to make out of GP Paris a perfect event and wanted everything to be perfect. We decided on the venue’s set up (I requested a central stage, which would be easier to manage than the two stages –one at each end of the room- that was already planned). We agreed on the fact that we needed one very fast laser printer for each tournament and that we needed about 7 to 8 computers for registration.
Upon my request, we also agreed that we would split registration into two steps: payment and then registration, so as to make it easier for the computer guys… I also presented my idea to have, on each table during deck registration, a sheet explaining the different steps of deck registration and how to proceed with them. We also agreed on the fact that each group should have a different colour for its decklists, result entry slips and pairings so as to make it simpler for the players.
The deck registration trick
One thing always made me crazy with sealed deck events. Since you have to wait that everyone is finished before making the swap, deck registration has the timing of the slowest player and can take up to 30 minutes if you are unlucky. I am looking for a way to get rid of this for several years and suddenly had The Idea a week before the GP.
What I wanted to do was very simple: have the players register a deck as soon as they register in the tournament. All I needed for this was a way to lead the players to a deck registration area and a few judges to watch them and coordinate all this. I think that this method has no additional risk as compared to the traditional system and could make us save the 30 to 45 minutes we need to sit 1400 players and have them register and swap a deck.
I immediately told Erwin that I had found an idea to save at least 30 minutes and all I needed was about 100 meters of rope or tensabarriers. He looked pretty happy to hear this but didn’t ask any question about the method. In the meanwhile, I told my judges that they were all welcome to give a hand at registration on Friday afternoon, and were welcome to bring a laptop in case they had one available.
Unfortunately, I should have made it more clear to Erwin and explain him the whole method. Since he knew nothing about my intentions, he couldn’t warn the site staff that we wanted to use the venue on Friday evening, so, because of security purposes, there was no way that we could use it. We tried to set up a few tables close to the registration but only had room for 70 to 80 players. That could have been enough, or maybe not, but at this point Felix said we wouldn’t do it. He was afraid that this number of tables wasn’t enough and would delay registrations.
Not doing it on Friday meant we couldn’t do it on Saturday since 743 players registered on Friday evening and would have to register a deck on Saturday morning anyway.
I must still give some credits for the method since I’ve been told that Dorian, from New-York already used this method. Nevertheless, this isn’t an officially approved method and I don’t think it should be used before we receive some instructions about it.
It was divided into three floors, each of them was 1.5 meter above the other, there was stairs on both sides of the room and the stage was, according to what I asked to Erwin when we met in Helsinki, between the first and the second floor, in the centre of the room. The bottom one was the largest, it could welcome about 850 players. The middle one would hold about 500 players, and the top one 200.
As I said, Erwin said we were expecting 1400 players. The French office agreed on this figure so the venue was set up for 1492 players. We had 1500 chairs total. I decided to anticipate the fact that we would have more than a thousand players and so numbered the tables in two separate groups. Because of the way the venue was, one group was occupying 3 quarters of the first floor and the second group was split between the fourth quarter and the two floors above (each floor was about 1.5 meter above the other and there were large stairs on each side of the room).
When we realized that there might be more than 1500 players showing up, Erwin started to look for more chairs and gathered about a hundred extras. We kept them in a staff room and agreed that in case something really bad happened, plan B would be to use the corridor leading to the venue and set up extra tables and chairs, and extra tables clothes on the floor in case something went really wrong.
We started registration at 8 am on Saturday, with 4 computers. Jason Howlet, the scorekeeper, hadn’t set up more computers and I asked him to do so when I realized it at 8:10. Unfortunately, it happened that some of Wizards’ computers didn’t work so I had to give my own laptop. Finally, we found two extra computers and could go on with 7 computers total.
I started to panic when I saw the line outside at 8:45 am in the morning. We already had more than a thousand players registered and I quickly estimated to more than 400 the number of players waiting outside. Erwin didn’t panic and said we would not reach 1500. At around 9:30, the number of actually registered players was over 1500 and there were still a few players waiting. At that point, we should have triggered plan B.
We had started removing the table numbers of the upper zones for a while. When Erwin told me that we would add extra chairs (in replacement to plan B), I put Lubos Lauer in charge of coordinating with him to renumber the tables according to the chairs that Erwin would add. Nevertheless, they started putting more chairs at around 10, and only on the lower zone. Thus, we had to put back the table numbers on the upper zone and renumber the lower zone. Because they were 1592 players in that room at this point, with people sitting on tables, chairs moved and trades going on, it took us more than an hour to renumber everything. We even had to ask the players to get away from the bottom zone. What happened is that there was a huge misunderstanding between Erwin, Gis, Lubos and me about the way we would renumber the tables, and then had to do it several times. In addition, several people willing to help did their own numbering; all of this creating a bit mess… Thus, even at the end of the process, there were a certain number of gaps in the numbering that we decided to fix once the players would be seated.
Unfortunately, at that point I was more focused on the scorekeeping station than on tables numbering and didn’t notice that things were going that wrong.
Dividing the tournament in two groups
Of course, I was in favour of doing so. Having two scorekeepers to enter the results sounded like a great idea. We had to split the registered players into two groups, each of them having the same number of players with 1, 2 and 3 byes.
What I didn’t knew was that the way to do it was to export the files into excel, use various methods to sort the players and then re-enter them in DCI-R. On Friday evening, Jason told me this whole thing would take 20 minutes. Because of various troubles that included mistakes in DCI PINs or bad variables in the files, it took approximately an hour. We finished this just at the moment when tables numbering was finally done. So I expected not to loose more time.
I had requested twelve pairings boards for each group. Then we would split the pairings in 6 ranges of letters and post two copies of each at two different places in the room. When we posted the seating for deck construction, we had to post both groups’ seating on all pairing boards to make it easier for the players (which also meant we had to coordinate both groups’ letter ranges).
Once the files were ready, we had problems with the ranges of letters for the pairings: they were described as “not active”, and it took us about fifteen minutes to fix both a problem with the printers’ installation and the letter range.
I made it very quick. The instructions for the sealed deck registration, swap and building were on the tables, so I only said “welcome”, mentioned the number of rounds and their length and said good luck. I should probably have said a word on the fact that cheating wasn’t allowed at this tournament… (see rulings’ report).
Deck registration and swap
As I said earlier, I had created a document which included the whole procedure for deck registration, swap and building. Inspired by Adrien Teh’s work, this sheet made my life easier since I didn’t have to repeat all the procedure on the microphone and players had all their questions answered. I received a lot of very positive feedback about this document from the players.
I also asked the scorekeepers to sit the players by first name for deck construction. I had everyone stay seated until the end of the deck construction period so the judges could collect the decklists in order and didn’t have to sort them and check that we had them all. I’m also strongly against having 1600 players queuing to hand out their decklist; last year in Lyon, this kind of mob move caused a 15 minutes delay.
As the sideboard picture shows, a dozen players had to build their decks on the floor. I had my judges register about fifty decks on Friday afternoon to replace defective packs and suspicious abnormalities, so I gave those players a pre-registered deck and so they only had to build on the floor and not register.
I chose to have the players seated by first name to avoid seating members of the same family next to each other during deck construction. I also kept this system for the whole tournament to remain consistent with the first pairing and thus prevent any additional confusion. However, I think I should have switch to last name for the rounds, it would have make players’ life easier since they wouldn’t have had to look for their first name among 10 other similar ones.
I had the judges arrange the swap according to whatever method they wanted to use and it worked pretty well. They usually passed the decks to the following table and redistribute them.
I didn’t leave time between collecting decklists and posting the round 1 pairings. This was one of my worst mistakes of the weekend since, after being kept seated for almost an hour, many players went to the bathroom and many of them weren’t at their seat quickly enough to avoid the tardiness game loss, despite the permissive procedure I set up…
One of the main reproaches I make to head-judges is that they feel like they have to wait for everyone to be seated before starting the rounds so that none gets a penalty. Especially at REL4 where the penalty is a harsh game loss, I often estimate the delay caused that way to 3 to 5 minutes per round.
Therefore, like in GP NJ, I decided to slightly change the way it worked and announced that a game loss would only be applied for tardiness after 3 minutes, which enabled me to start rounds slightly faster.
An average 75 minutes separated a “you may begin” from the following, with a high of 95 minutes for a round (3 or 4 if I remember well).
I spotted some way to improve my pairings’ teams procedures, since they were usually waiting for all their team members to be ready before they went to post the pairings. This was taking about 5 minutes so I asked them not to wait, and thereafter everything went perfectly well.
I had opted for an original system on this side of the Atlantic Ocean: judges didn’t have to sign the result entry slips. They were still in charge of collecting them to prevent players from loosing them and I had specifically asked that the judges still ask for confirmation of the result. I think it was quite a good thing because it saved some time and the judges could focus more on some other issues or even collect the slips more efficiently during “crisis” periods.
We used the “boxes” system at scorekeeping stations; judges had to sort the result slips according to the result: 2-0 / 2-1 / 1-2 / 0-2 / “drops” and “others”. That system is easier for the scorekeeper, who doesn’t need to focus on the result at the time he enters it, but tends to create judges’ clusters before the boxes since it takes a while to sort the slips…
I announced on round 7 that the subway was closing at half past midnight and that we were expecting to finish round 9 at quarter to two. We actually finished the tournament at half past one. We lost more than two hours (time to re-number?) as compared to an optimistic “everything goes fine” schedule that would have been:
- 09:30 end registration
- 10:00 post seatings
- 11:30 post 1st round pairings
- 23:30 end round 9 (about 1h15 per round)
Having two head-judges
I wasn’t in favour of this system, for two reasons.
First, having two judges wearing the red and black shirt didn’t seem like a good signal to send to the players. There could be risks that players do not understand that the ruling given is final, or feel that there could be a difference in the approach that those two head-judges have. So, they could be frustrated by a ruling and think that it would have been different with the other head-judge.
Second, there is a lot a personal pride involved. I was really happy and proud to have been chosen to Head-Judge GP Paris, and I felt somehow “stolen” to have someone having the same role. Also, all judges knew that Gis is a judge with more experience than me, and I was afraid that they would have a tendency to refer to him, since he appeared to have the same role as me in addition to his two extra levels. Since Gis couldn’t attend the meeting on Friday night, I wasn’t sure of what to announce to the judges and felt like it would create even more risks on that matter.
Since I seemed not to have the option to decline the offer, I played the game and even told the players about this system during the players’ meeting…
After having experienced this system, I must say that I mostly changed my mind. Gis remained in his role: handling the appeals when I couldn’t do it, advising me when I needed it and redirecting the judges to me when I was available. On Sunday, Gis even took the initiative to let his red and black shirt at the hotel room since it wasn’t necessary anymore, which I greatly appreciated. Of course, under those circumstances, I can only say that it was a very positive experience.
Merge the two groups for top 128
After round 9 was over, I made a short debriefing with my judges and we went to bed at about 2:30 pm. Fortunately, Erwin and the staff were very efficient and had set up the venue for top 128 very quickly, even before round 9 was completely over.
Upon some player’s request, I postponed the beginning of the tournament to 9am instead of 8, to allow them to get more than 4 hours of sleep. That wasn’t the case of Jason, who spent the night merging the files for DCI reporter and who finally succeed at around 8am, right in time for the seating.
Blame it on the Boogie
Day two went fine. 4 players didn’t show up for the first draft, but one of them came in thereafter, so I had him re-entered for the second. Rounds went fast, drafts went fast until…
Round 14. After we post the pairings for this round, 4 players came to me to say that they were paired against someone who wasn’t in their pod. I came to the judge station, where I met a payer who told me he wasn’t on the pairing anymore. I came at the computer, close to Bart, who showed me another weird thing: a player was paired alone at a table…
Due to this important number of strange things, I decided not to start the round and announce that we would have a 15 minutes delay, time to fix this, and that we would repair. Bart Moulinier, the official scorekeeper, tried to fix the pairing problem, but we only had access to the 32 first tables in the ordered pairing. I started to worry and asked if we could come back to the latest stable point: round 13 backup. We then re-entered the results for this round and started to re-enter the drops. From the morning, the computer was really slow: it took more than a minute to make a pairing each round, more than this to make a ranking after each round and at this time, it took about a minute for each drop to enter. I asked to stop doing it because it was going to take too long and wanted to come back to round 14 pairing, delete it, re-enter the players who had dropped (we suspected that the problem came from a 7 players pod in which 4 players had dropped at the end of round 13), make a new pairing and fix the drops issue thereafter. Unfortunately, we couldn’t delete the pairing since the software only allows a re-pair for draft algorithms and has no option to delete a pairing that has been made. We could have switch to swiss so as to delete it and re-enter the dropped players and then switch back to draft to make a new pairing, but unfortunately, we didn’t think about that solution.
At that point, since I was figuring that many manipulations would be necessary to fix the problem I tried to open the tournament backup on another computer. That worked, and much, much faster. We had probably lost half an hour at this point.
This backup was at round 14, after the pairing had been made. I decided to order pair it and once again came with the issue of having access to 32 tables only. At that moment, Bart and I realized that I probably came from the fact that Jason had merged two tournaments where he cut the top 64, that is to say only 32 tables. Within a few minutes, Jason found in which file that variable was, we corrected it, and had access to all tables!
I then decided to remake a whole ordered pairing by hand, switching to the swiss algorithm because we had to delete the pods to reset the pairing control. I was announcing the pairings that we posted on the pairing board to Jason who was entering them, whereas Bart and Riccardo were controlling that the players were on the same pod. We ended up with a few illegal pairings that we manually fixed at the end. It took us more than an hour (probably an hour and a half) to fix this problem, which could probably have been fixed much faster, had we known its reasons. It appeared to me that there were too many errors in the pairing to let the players play and fix the problem in the meanwhile. However, two of the four players who claimed not to be in the same pod were actually in the same pod according to other players from this pod. I probably wouldn’t have repaired if I had realized that only 3 or 4 players were involved in the pairing bug instead of the 5 or 6 I first heard of.
At that point, I was somehow reinsured about the tournament. For a while I thought that we would loose the tiebreakers and have to end the tournament manually, this is why I was insisting so much in using the software for that round 14 pairing. But we weren’t fully safe: round 15 was coming and we had no more pods since the tournament was running under the swiss algorithm. We couldn’t remake the pods since we figured that the whole problem came from the fact that we had forgotten to reset the potentiality that two players that already played together earlier would play again together in the second draft. We would then have ended with the exact same problem in round 15: either weird pairings, either players playing each other for the second time in the second draft, and none of those options seemed acceptable.
At this point, Jesper Nielsen suggested that two groups would work on two different solutions and that whoever would end first would print a pairing for round 15. It could have worked, but it seemed really dangerous to me, especially because everyone was really tired and I was afraid to loose again 2 hours on that pairing.
And then came the most concerted idea ever. I had the idea to gather the players and make a pairing by hand: I would use blank papers with players’ names on it and, pod after pod, would pair the players manually by putting those papers on the tables; Bart would then have time to enter the pairing manually during the round. Jason said we could use blank slips instead of those papers. Bart suggested that we used several judges so that we could make all those pairings at the same time. Gis added that we had to pair them by points and not by number of victories in that pod. Jesper finally had the idea to gather them at their draft tables to make a pairing with the help of a standing and a random method such as lands of different kinds. That whole solution took about 15 minutes to implement, which was much faster than any other idea we could have had.
It went fast and easy, nothing worth mentioning, except that a French lost, once again, in the final… We ended close to 1am.
I had written a briefing a long time in advance, so I send it to my judges a week before the event (which I think is a deadline if you want everybody to read it…) I also included the paper about deck registration that I planned to have on the tables.
I gathered the judges on Friday night and a short meeting where I went threw the key points of the written briefing I had sent. I had everyone introduce him or herself, which wasn’t very original, and highly recommended them to get some sleep on Friday night since they probably wouldn’t have the opportunity to sleep a lot on Saturday night.
I wanted to have only two teams per group. One would be in charge of deck checks and the other in charge of logistics (pairings and slips). Because we were 32 judges total, that would make teams of 8 judges. The team leaders had to be good at bearing pressure, good at efficiently manage a large team and have a good experience of such tournaments so as to coordinate as much as possible without my instructions.
Several level 2 asked me to have a chance to team lead on Saturday. However, I wanted to be certain that my team leaders could do without me at all times in case I wouldn’t be available. Therefore, I chose to take the four more experienced judges of my staff as team leaders: Adam Cetnerowski, Jesper Nielsen, Lubos Lauer and Riccardo Tessitori. I know all of them for long and know I can trust them. They didn’t deceive me and did an impressive job, just like all the judges. Also, it is now a settled procedure to have the level 3 be “only” floor judges and I found it nice to have them show a surge of leadership there…
Sunday would be easier and shorter. At 11pm on Saturday, I had made up my mind on who would be team leader on Sunday, and I then decided to go to those people and offer them to go to bed at midnight; I wanted them to be fresh and efficient after this hard day and I already knew that Saturday would end at around 1:30 am. I also made the same offer to some of the judges who looked really tired, in exchange for their promise to go straight to bed.
Despite the day almost ended at 2am, I wanted to make a debriefing with all my judges. It’s always hard to make a debriefing on Sunday and, especially because I know how frustrating it is to learn on Monday, while reading the sideboard online that there have been 4 disqualifications in a tournament you were at.
We discussed a few issues and some of the rulings of the day. I made this quick because of how late it was and sent everyone to bed at 2:30. Main event judges had to come at 8:30 and side event judges at 9:30.
I was very embarrassed with making teams for day 2. Out of the 32 judges I had, only two were level 1 judges, 5 were level 3 and all the others level 2, most of them being level 3 candidates. That made more than 20 judges who were looking for this kind of experience. I also consider that having to lead a level 3 is a good experience for a level 2, especially because the level 3 has usually more experience in giving useful feedback (and recommendation?).
However, there was only 128 players. At most 16 judges (1 per draft table), and ideally 12 so that they have something to do during the day. I asked advices to Gis since I wanted to find a solution that wouldn’t frustrate anyone. He had quite a good idea: switch team members before the second draft. We would keep the same team leaders all day long, keep 2 of the most experienced judges (Jesper and Riccardo) and just rotate 7 judges between sides and main. It worked quite well, except that the team leaders had a hard time gathering their second flight of judges.
I like to give the opportunity to some quite experienced level 2 judges to call drafts. It’s basically a very easy task, but has a lot of pressure and it’s therefore a good exercise.
I used top 8 tablejudging as a reward for the judges who did particularely well. I also offered to some judges that I didn’t select for day 2 to do it.
Unfortunately, because the tournament ended so late, I didn’t have a chance to gather my judges to make a debriefing about day 2. Nevertheless, I sent them all an email some days after the even to thank them all and explain why we lost two hours on Sunday.
I could write pages on this part since I had about 4 appeals per round, most of which being unusual and interesting situations. However, this report is already very long, and I don’t think this part is the one that interests the most those who are supposed to read the report.
2 disqualifications for adding cards to sealed deck
That was two blatant cases. One of them had a weak excuse (“I mixed some cards with a friend of mine”) and the other left right after we took his deck for the deckcheck so I didn’t even had to disqualify him.
I wasn’t really good in handling the first DQ. I didn’t ask for his friends’ name and didn’t look for him, which might have led to another DQ or a proof that the guy was telling lies. I was focusing too much on his attitude, which surprised me because he was very calm: he told me he perfectly understood that I had to disqualify him…
2 disqualifications for randomly determining the outcome of a match
I was a bit embarrassed with this case because those two guys flipped a coin while there was two judges sat at the table. One of the players was definitely sincere and didn’t know that it was forbidden; on the other hand, I was certain that the other knew about it. I made up my mind and opted for the double DQ. I don’t believe that we can consider ignorance as an excuse for breaking the law in such a case.
Too nice with the decklists errors
On Saturday, I chose to give game losses instead of match losses to the players who had forgot to write down their basic lands, other illegal decklists would be worth a match loss.
On Sunday, we had time to count all the decklists before we began the first round of each draft. I then told my judges to get the players with illegal decklists before they start playing, correct the list according to the deck and give only a warning. When Gis heard of this, he made me understand that it wasn’t a good idea, of course because it wasn’t consistent with the Guidelines, but also because it wasn’t consistent with what we did on Saturday.
There was a slight misunderstanding between Gis and me and I thought that he suggested that I could go for game losses if I wanted to be nice. No need to say that he was quite disappointed by the fact that I went once more for leniency and inconsistency on the second draft…
An interesting miscommunication case happened. According to a judge who was watching a game, player A was mocking him by not understanding something. Player B had tapped his creatures to attack, and said a word in his own language; A, who looked very tired, reacted by untaping his permanents and was about to draw a card. Upon my judge’s request, we discussed the opportunity to disqualify him for something we could have called “deliberate misunderstanding of the game state in order to create confusion”. After some investigation, I wasn’t convinced at all that the player did anything wrong, other than being involved in a common case of miscommunication so we let him go.
All the rest…
I also had quite a lot of situations of misunderstanding between the players, some shortcuts abuses, a few issues with sleeves or insufficient randomization, but I wrote this report mostly in order to talk about the general organization of it, so I’m going to stop there…
Hope you enjoyed reading!
David VOGIN firstname.lastname@example.org