At the European Go Congress 2023 in Markkleeberg near Leipzip / Germany, I was a co-tournament director of the side tournaments 9x9 and blitz. I am experienced with directing these or 13x13 side tournaments at EGCs (and was the major tournament organiser in 2000). This year, however, organisation faced some difficulties. 9x9 occurred on earlier days than blitz.

City bureaucracy demanded closure of the venue at 22:00 so that schedules had to be tight and prolongued schedules had to be avoided at all costs. Early starts at 17:00 or 16:30 helped a bit but were insufficient for the blitz tournament. 9x9 was played as group qualifications and KO finals on two evenings so the schedules fit (with less than desired thinking time for 9x9 though: we wished 30' but must set 13').

At first, the congress organisation planned to skip blitz until I volunteered and was admitted a schedule slot of only one evening. This made group and KO stages impossible but blitz had to be played as Swiss (or McMahon in the case of many participants). I chose the pairing program MacMahon. Swiss pairing for a blitz tournament had occurred once before in 1999 and posed problems for fast pairings between the rounds. I knew that this would be the central problem of organising this tournament so ensured enough helpers collecting and reporting the results to me. During the tournament, I and the other co-tournament organiser were permanently busy entering the results and publishing the pairings. Just for a few seconds, I took the freedom and looked into playing rooms to see some of the tournament games in progress. 9x9 organisation was much more relaxed and I could also participate in that tournament.

Most of my earlier directions of side tournaments at congresses I had done as a single tournament director, although sometimes assisted by referees. It had gone smoothly and usually punctually. Registration deadlines could be two hours before a tournament start with expected up to ca. 150 players but since 2012 three hours with 250+ players thanks to the congress director's wise recommendation to me. This year's major tournament organiser insisted on co-directing my two side tournaments. This presumes mutual willingness to cooperate - no problem. Two directors can do more work at the same time so a co-director helps progress with organisation in principle. We used his computer for blitz. However, the existence of a co-director also created organisation problems explained further below.

The congress organisation used additional online pairings for major tournaments and registration for most tournaments. For the side tournaments, players could register in paper registration lists or online. At congresses, I travel without electronic devices so, at first, was unaware of the alternative online registration for side tournaments. Roughly at the registration deadline of 9x9 three hours before the tournament, I learned about online registration from the co-tournament director. Apparently, exactly one player had registered online. I prepared the group pairing sheets and the tournament was about to start punctually. Then, however, a few dozen additional participants appeared who had also registered online. The programmer of online registrations had made a bug so that access rights prevented too many relevent organisers including the co-tournament director and major tournaments organiser from accessing almost all registrations. Accordingly, the start was delayed and several additional groups had to be formed ad hoc. We learn the obvious: if new methods of communication shall be invented to tournaments, they also require thorough procedural testing.

At the end of the 9x9 group stage, I collected the groups sheets and was faced with two problems: the ad hoc groups were not numbered so I had to ask the co-director for the tentative number of groups so that I would collect all sheets. It turned out that one group was missing. We reconstructed that two under-occupied groups must have dissolved and merged but I needed a few nightly hours to verify this from context information of names on groups sheets and in lists of registration and prepared pairing.

During a round of the 9x9 KO stage, I had to interrupt three games early. In each of these games, one player had allegedly won his previous round game due to a missing opponent. It turned out, however, that language problems let one of the referees declare that those opponents were supposed to wait in the next room to ensure reasonable silence in a particular playing room. As the chief referee for this tournament, I interrupted my own game to decide that the three previous rounds games still had to be played and, if the same player would win, he could then continue his previously started next round game. One player first wanted to call the appeals committee but, on seeing that the other two players were playing their previous round games, eventually agreed on doing alike.

Later during the congress, the appeals committee decided not to arbitrate for side tournaments but they did not know that the EGF Rules Commission had clarified this after the 1999 congress that the appeals committee is also in charge for side tournaments of its congress and must judge even if only one of its members is present. In 2003, such an incident had occurred when I as the only present member of the appeals committee had to arbitrate as second instance in a side tournament. On another occasion that year, the appeals committee could judge normally with the spectacular ruling "win by 1 or 3 points" by almost reconstructing an area scored game.

Blitz faced its own problems of organisation. Some players registered twice on paper and online so we had to figure this out by deciphering names. The real problem, however, were reckless players registering for both blitz and torus go, which was another side tournament at the same time. I would have solved this by emphasising an early start of round 1. The co-tournament director, however, insisted on enabling as many played games as possible. There were also players leaving shortly before round 1 or after some later round. During rounds 1 and 2, this contributed to delay while we could handle such swiftly during later rounds. The MacMahon program does not allow changes of players and pairings of a round quickly so a delay of an additional circa 20 minutes occurred.

Since I played in the congress's main tournament and could not organise during this time, I had set the registration deadline for blitz to 20:00 on the previous day to enable a punctual start despite Swiss pairing. My co-tournament director promised to manage the online registrations (the bug was fixed by then) on that evening or else during the next morning. Two hours before the scheduled start of blitz, he admitted to have overloaded himself with too many tasks of organising other tournaments. Since he had access to his computer and online pairings, I could only do the paper work but not help him with initial entering of names in the program and the online pairings. This resulted in his readiness only 20 minutes after the scheduled tournament start and we had to reconfigure the parameters in the pairing program again. What can we learn? If a tournament organiser suspects his time trouble, he must be willing to accept more help from other organisers in time. I could have helped him more if he had informed me about his own time trouble earlier. I co-directed this tournament with its 41 minutes delay of round 1, it was not my fault but, nevertheless, it feels bad to have been in charge when my own standard is always punctual start.

According to one of the referees / helpers, there have not been any disputes during blitz. My clear tournament announcements including references to the EGF General Tournament Rules §5.5 do wonders as they discourage attempts by players to create trouble. They know they would not succeed with provocing random arbitration.

Occasionally, the torus go tournament director would simply plug our USB cable to the printer while we were busy entering results hopefully flawlessly...

The delays and the very limited schedule of the venue, however, resulted in only 6 instead of 8 rounds Swiss of the blitz. Except for possible ties (jigo and shared final result places without tiebreakers were possible due to my tournament system settings) and pairing peculiarities, 8 rounds would have enabled a unique winner. After 6 rounds, we had three players with 6:0 wins and quite a few players with 5 wins. We wanted to solve this problem by letting the top players decide whether they wanted a shared first place or play a mini-KO on the next morning. They preferred a KO. Instead of four players in the KO, the Koreans suggested a cute three-player KO used in Korea and relying on the pairing tiebreakers (here: SOS-SOSOS) after round 6 as follows: the top three players were sorted as players A, B, C. In the first KO round, the pairing was A - B. In the second KO round, its loser played against C. In the third KO round, the two winners decided the first two places. This way, a KO with one player too few avoids a BYE and still pairs fairly by giving C the least chances to be in the final game. If he is the strongest player, he can win the tournament nevertheless.

Announcing pairings on paper and online is all fine and well in principle. However, I abhor delays of paper announcements several to many minutes after online announcments. This discriminates all offline players like me, who can reach their tables only later unless they beg for smartphone access from other players. Unfortunately, such delays were common especially for the main tournament. Universally accessible pairings (those on paper) ought to be published first. If organisers want to avoid huge crowds in front of walllists, publish pairings early, such as on the evening before. Such is possible and was sometimes done in the past! Planning pairing publication only at, or after, the supposed starts of rounds is a bad attitude towards organisation.

As usual, it turned out to be useful that my heavy rucksack always contains copies of the rules and tournament rules. On the second Wednesday during the poker tournament, the under-demanded participant Pascal Müller, who was the main tournament director (the most important but least appeciated job of congress tournament organisation presumably every day as busy as I was during blitz) and chief referee, asked me about handling disputes when stones are moved along the board surface to its supposedly intented intersection. My related advice for EGF tournaments is: on the first occurrence, explain the rules; on the second occurrence by the same player, issue a warning; on the third occurrence, issue the second warning meaning a default loss.

There can (read: should) never be enough pin walls for announcements! Publishing the 9x9 KO qualified players on the hidden back of a wall was suboptimal. Before the congress, I had declared demand for large cardboard, on which to draw the 9x9 KO tree. Nevertheless during the congress, at first none was there but there some organisers would swiftly buy everything other organisers needed. Accidentally, I learned that organisers would get food coins. This did not work on my first day of organisation but worked on later days. Accordingly, I got hold of a few dubious meals (one per day) and hopelessly overpriced, small but tasty coffees.

In my opinion, the most remarkable aspect of tournament organisation was the fairly large number of side tournaments. There were more than I could witness. Germany was assigned the congress only in November 2022 but, IMO, overall we have produced a reasonably organised event nevertheless. Perfection, however, can hardly be achieved within such a short time of preparation.

The tournament organization was a bit of a mess across the board. I feel bad for saying this since I know the organizers had very little time and worked very hard to make this congress happen but the constant delays, confusion and conflicting information were unpleasant for the players. Perhaps a smaller number of side tournaments would've been a better option. I feel like they wanted to offer the players as many options as possible but stretched themselves too thin during the process. I don't think anyone would've minded much if they'd skipped such exotic variants as miai go and keima go.

The online pairing announcements didn't work that well most of the time. I assume it's because the website couldn't handle 400 people refreshing it every few seconds to get the pairings. So, having a smartphone didn't help much. I like the way this was handled in Brussels: we received an email with our pairing a bit before the start of the round. I'm still hopeful that the future organizers will adopt this method and I definitely agree that the paper pairings should be published at least half an hour before the start of the round, if at all possible. I see no reason for delaying this until the last possible minute.