Atlanta, Georgia
Time: Friday December 14th – Sunday December 16th 2023
Main Event: 1300 players


Friday – Scheduled Sides Lead

Empirically Confusing
Lately TOs have been realizing that running side events is not only important, but also hard. In general, being sides lead requires a relatively high logistics score. Being sides lead at Dreamhack even more so. You see along with the regular challenges of leading side events, Dreamhack presents some unique challenges as well. One of these is that events have a somewhat ephemeral quality, they may appear at random and their qualities are constantly shifting, prize support is often different depending where on the website you find it, and whether it's single elimination or swiss is dependant on the whims of the area lead at the time.

I knew I would need to do a lot of work before the event. Dreamhack is not the kind of event where you can roll in the day-of and make things work. If you go in without a strong plan, everything is going to fall apart. Alternatively, if you go in with a strong plan, everything is probably going to fall apart anyways.

I was given the sides lead position about a month before Dreamhack, but didn’t know the exact events that would be on the schedule until about a week and a half beforehand. I organized a detailed schedule which accounted for breaks, which can be seen here. There were several changes before the day of as staff members phased in and out of existence.

Be Prepared
I brought my team in half an hour before our first event on Friday, and myself in a half hour before that. When I walked in at 9am on Friday I immediately noticed that the room wasn't numbered. This was pretty concerning as LCQs would be starting in full force at 10am. The next most concerning thing I noticed was that it appeared we didn't actually have prepared table numbers, and that some judges were working on sticking numbers to triangle tents. I couldn't help with that activity until the numbers were actually assembled so instead I familiarized myself with the meanings of the various gathering points:
White - Main Hall, directs players to mtg
Red - Registration
Black - On Demand Events
Green - Scheduled Sides
Blue - Competitive REL Sides
Planeswalker Logo - Prize Wall

After I put the gathering points in the right places, I put basic lands by the land stations, and by then enough numbers were made that I could help number the area. First I was told to place number 1 near the coverage area. I began doing that, then I was told this was wrong, and thought I was told to put table 1 by the gold gathering point. After about a row was placed someone else came up to me and asked wtf I was doing and to instead put table 1 by the coverage area again. I sighed and took them all down and began renumbering the area by coverage again. While I was renumbering the same area multiple times, I remembered back to RC Dallas, we'd had a similar issue at the start of that show, where the numbers that had been ordered didn't arrive on time. Luckily after RC San Diego, some forward-thinking individual had packed up the numbers just in case this very thing happened. I recalled that as teardown for RC Dallas was happening the numbers were being thrown out. I put forth that maybe the next show might have the same problem, and it might be prudent to save some table numbers. I was told this would be unnecessary and the numbers would for sure arrive for Atlanta.
....
My next self-delegated task was moving some pairings boards and having some judges put tape on them. Pairings boards were a struggle. We started with four and I sent a judge to get more, since that was clearly not enough for today, let alone Saturday. The judge came back and told me we couldn't use the other three they had because they had the old Dreamhack logo on them. I asked the judge to go back and see if anything could be done to make the boards useable since we wouldn't really be able to function with just what we had. After some negotiation we managed to gain control of the other three boards leaving us with a healthy 7 for the event. It wasn't optimal but would do for Friday.

A Trial of Trials
LCQs were a mess from the minute they started. Players were able to self-register in events, which is great, but MTGMelee would allow players to register past the cap (ie. 42/32) for some reason. I believe this was due to a combination of many players attempting to register simultaneously, and registration staff manually enrolling players that were paying in person but not noticing that the LCQ they were putting them in had filled up. This meant that people needed to be removed from the current LCQ and registered in the next one, unfortunately as soon as that one was made, it was public (since you can only have events either be public or in 'test mode') and people began registering in that one instead. This resulted in huge delays and a lot of players waiting.

The next issue was the fact that there wasn't a really a consolidation plan for LCQs. Whenever I've lead LCQs in the past I would just kind of poke people when their LCQs were small and get them to move. Upon reflection, I knew this event wouldn't be the kind of event where the lead had enough time to do manual consolidation and I think two rows should've been reserved at the front of the room, and as LCQs entered T8 or T4 they would be handed over to a single judge in that area.
The reason this mattered is that the judges ended up being so spread out they couldn't realistically help each other with floor coverage. Also the LCQ lead was constantly bothering me for side event space (which would further compound the "being spread out" issue, as well as trying to steal more pairings boards, to which I had to say no to a few times.

Kickstart it and Forget About it
The plan was basically, kickstart team starts an LCQ, then hands it off to a nearby judge to babysit. This alleviates the need to have a dedicated turnover team (which would be unwieldy with the amount of LCQs that were being launched), and makes any given judge responsible for 2-4 events, which should be manageable, as they shrink quite quickly. However, KS had a pretty rough lot dealt to them. When they first came in it was a struggle to locate prize tix, promos and decklist registration sheets (for sealed). Prize tix in particular must've been lost in same void that the table numbers were lost in, because the ones we ended up using all weekend were just printed out on business card paper with the printer on site, this made them kinda stick together and need to be manually punched out initially, and after that they wouldn't separate easily in your hands due to their texture. This meant that when the LCQs were finally ready to launch everything still wasn't entirely ready and there were also five that needed launching all at once. I wanted to help more with LCQs but by this time we were nearing the launch of my first scheduled side and I needed to assist with that to ensure it ran smoothly.

Master of the Small
We had a free entry Minimasters event, I got some details about how it would be run before the event, but I think that information was incomplete. What I thought it was going to be was players all receiving 2 packs of LCI and building a 30 card deck, then playing three rounds of swiss. It actually turned out to be single elimination, in addition to this most players weren't in the event on MTGMelee. As the scorekeepers were still dealing with LCQs, I decided it would be better to run this event without the use of software entirely. Each player was given a small sheet of paper to put their name and email on so they could be later entered into the event (if people claimed to be in the event, we just believed them). The judge assigned to the event had run Minimasters at a recent convention and offered a structure that circumvented the need for software. Players build and play round one, then once they're done, they call a judge, who gives the winner a pack directs them over to another area and they're paired league style for round 2. If they win, they're given their second pack and directed back to the area they played in to begin with for their third and final pairing.

Saturday – Scheduled Sides Lead

We Only Have Space for 9ks
The first thing I noticed when walking in was that there was not a lot of space for side events. The second thing I noticed was that the little space that did exist was supposed to also facilitate a 180 player 10k in addition to my side events. Luckily, it appeared there was already an effort to fix this, a bunch of chairs were being stolen from the tabletop gaming area. I wanted to ask why we weren't just numbering the tabletop area, but we didn't really have time for that. We were also renumbering from the back of the room not the front, which meant the numbers wouldn't be contiguous with each other, I also wanted to ask why we were doing that, but again, no time. So me and my team (as they trickled in) helped move chairs and table numbers, and eventually the 10k started about 45 minutes after the scheduled start time. After all was said and done I found out that the tabletop gaming area tables weren't great for Magic as they were significantly thinner than the tables in the Magic area, and the reason we hadn't numbered contiguously was because we didn't have enough table numbers for that (a few had been ruined by someone writing the wrong numbers on them). I recovered the ruined table numbers, and a few pieces of obfuscating tape and some sharpie later we had shiney new numbers. Once we solved that problem I had the area renumbered to mesh with the rest of the room.

Sides Lead: the Soviet Mind Game
The 12pm standard event happened uneventfully with like 6 players. Then there was a 2HG event, that was 34 teams, slightly larger than anticipated. We renumbered a section of the room. However, after 2HG, we only had 30 tables of space left, and the modern already had 60 players in it. I knew the current round on the 10k was turning before modern was slated to start, which would open up a little more space as some players had dropped. Unfortunately right in the middle of that space was the standard event and a 20 player flesh and blood event. Luckily, we had enough lead time to get both standard and FAB to relocate to the tabletop area, and they were both run like FNMs without table numbers. This let us fit a cool 75 players in the modern event. That was the crunchiest part of the day, as after that the main event shrank each round, giving us an ever-growing amount of space in the back of the room.

Wonky Formats
One of the events was Commander Draft with Commander Legends, Battle for Baldur's Gate and Commander Masters. I haven't interacted with commander draft in a while so before the event I reviewed all the commander-specific rules (2 cards per pack, 60 card deck minimum, 1-of rule doesn't apply, but the color restriction rule does. You also have access to Prismatic Piper and Faceless One). I brought some printouts of Prismatic Pipers and Faceless Ones just in case. However something that did come up was the fact that apparently in Commander Masters all single-colored and colorless legends were considered to have the partner ability, making it easier to have two-color pairs in the set. When one of the players brought this up I was a little thrown off balance because I hadn't been aware of it, but felt like it would be appropriate to allow this.

Additional Discovery
NAP controls a Thalia, Guardian of Thraben and AP uses Quintorius Kand's -3 ability to Discover 4. They reveal a Lightning Bolt. How much mana (if any) will they need to pay for Lightning Bolt? {1}, Discover allows AP to cast Lightning Bolt for free as an alternative cost, however additional costs will still need to be paid. (601.2f)

Decklist Dysphoria
NAP accidentally presents the wrong decklist (one from their previous round opponent) AP notices a Sheoldred, the Apocalypse on the list and aggressively mulligans for an answer. AP plays an Island on turn one. NAP plays a tapped land. AP plays a mountain and a Ledger Shredder, at which point AP realizes that the list they saw is incorrect. There aren't a lot of great fixes for this, and the IPG doesn't really cover failing to reveal the correct decklist (or anything at all about open decklist events). The commonly accepted deviation is no warning and restart the game. I'm not a huge fan of this (restarting the game is fine here, but could suck a lot theoretically). I feel like some kind of warning would feel appropriate here, but this doesn't really fit under anything in particular.

Pop-Up Round!
Friday night, it was announced that there would be 10 rounds on Saturday instead of 9. The reason for this is that the hall was officially closed to attendees at 7pm on Sunday evening. Meaning if Main event started on Sunday at 9:30 and was 5 rounds cut to top 8, that's easily 8 hours, causing the event to end at 17:30. While an hour and half of bleed time seems like a lot, most rounds, especially in such a large event have about a 70 minute turnover rate. As it stands, even with only four rounds and top 8 on Day 2 the event ended at 19:15. The idea of cutting a round entirely was tossed around but with 1300 players WotC wasn't okay with that.

Sunday – Scheduled Sides Lead

Pop-up 5k!
I ended up losing two people from my team to a popup 5k so the schedule you see is the reorganized schedule. On the bright side I actually got to do a team meeting on Sunday. I felt like we would be getting out early-ish due to the fact that the 10k had specific judges assigned to it and I figured those judges would just sunset the event on their own, instead the plan was to have sides judges replace 10k judges for the final rounds of the event, since they came in earlier (this makes sense, I just wasn't really expecting it and should've asked) I ended up sending about half the team on 1/2 round breaks and the other half took sitting breaks during their scheduled side events as a last minute remedy to the fact that I hadn't planned for half round breaks, as I assumed we'd be getting out earlier on Sunday.

Total Breakdown
The other big issue was that the renumbering for team trios sealed ended up taking longer than expected, since the event ended up being 45 teams, also because the number of chairs per row wasn't cleanly divisible by 3 (leaving it as is would cause us to waste four chairs per row, which was bound to cause some issues later on). This caused a cascading issue, since some breaks were pushed back a little bit due to the activity. This caused other breaks to be pushed back and so on and so on, which in the case of minimasters, left the HJ only 15 minutes to prepare instead of the normal 30 minutes. This meant we didn't have really enough time to confirm the rules and structure of the event. Because of that, I simply assumed we'd run it the same way as on Friday. But after we announced the structure to the players they protested, saying that MTGMelee said something different. When I spoke to the area manager for Side events, they told me that the structure was different than Friday and also different from MTGMelee. Players were given two packs and 30 minutes for build. They were then given 30 minutes for each round. A pack was dropped on the table and the winner would take it and begin modifying their deck as soon as their match was done. This one was run through MTGMelee and was three rounds of swiss instead of single elimination.

...In Conclusion
Overall I felt like Dreamhack went well, better than expected even. I should've brought my team in earlier on... basically every day. We constantly needed more bodies in the mornings. Another thing I think I'd change in the future is how LCQs are run, a finals area I think is necessary for this scale of LCQs. I think on Friday I did the best I could but things were kind of just messy all around, Saturday I felt like I did my job well and there were very few issues. Sunday I felt like I dropped the ball a bit, particularly with Minimasters and 1/2 round breaks. Lately I've been having a great time rooming with judges and staying up crazy late to talk about policy, but I think it's severely impacting my Sunday performance, and I need to get a handle on that if I want to be a reasonable employee on Sunday.

All in all even with all it's flaws, I had a great time working Dreamhack. Events like this are great for testing where your limits are and finding issues with processes, I can see things to iterate and improve on, which is a good feeling.
I'm sad I won't be able to hit the next Dreamhack since it's on top of RC Ottawa, but I'm looking forward to hopefully getting to work the one after!