32-player TAW is not good for Zero-K

quote: Use and support !split. I doubt games are going to naturally split on their own.

I feel this conflicts with acknowledging some players have preferences for larger games. A split will annoy people that prefer a larger game. I would prefer something like lxa4000 solution, because if a host would start "while waiting" that means most people agree to start (maybe we need higher voting threshold for the alternate), and even people that prefer larger games will not feel that you just "stole their large game".

As someone who likes both small teams and big pot (for fairly obvious reasons.... One cant be a "try hard" or "slacker" all the time, some variety is both fun and necessary to enjoy the game), I do not think lob pot is "bad" for zk. The very ironic situation here is that many people claim they want small teams games.... but if you cant gather up 5-12 people to seed your own lobby do all of you REALLY care about small teams only?

Anyways, a solution should respect both sides as it is ultimately a preference thing. How about setting team sizes based on time? Like say large lob pots On Monday, Wed, Friday, Sat and Sunday and Small team lobby on Tuesday and Thursday. Weekends are an exception where we often get 50+ players, so you should be able to start multiple rooms.

I see the screenshot of lxa4000 as a sketch of what could be, not necessarily how it should be done in detail. But based on the positive feedback, I see that a vast majority welcomes UI changes to make multiple matches easier.

maybe might need a sanity limit so that it doesnt look like 40 rooms are full of players =P

I have three basic problems with the current incarnation of TAW

The first is the one that's gotten the most talk in this thread: A full TAW lobby makes it very difficult to find smaller team games to play.

The second is a worse problem with TAW specifically: An overfull TAW lobby makes it difficult to find 16v16 games to play, too, because of how long it takes to get off the waitlist.

The third merely compounds the second: The lobby often selects maps I'm not willing to play on, so even if I do get off the waitlist the reward is often passing on the game and going back onto it.

ill play fields of isis with you

quote: Like say large lob pots On Monday, Wed, Friday, Sat and Sunday and Small team lobby on Tuesday and Thursday.

Please no, it's already bad enough when you're not in the peak hours timezone to get into a game, I don't need it to be a scheduling mess as well.

quote: As an idea, double view room, you can be in a waiting list in 2 room simultaneously.

I had a think about this, and I don't see it working. At least, I'm not looking to put the work into generalising the current concept of being in a battle room, only to discover it doesn't work. If someone else wants to implement it, then we could test it, but just don't be surprised when it fails.

My thinking is that corralling a room to pass a start vote can already be tricky. Trying to do it when everyone has their attention split between two rooms sounds like a nightmare. Many players will have a preference about which potential battle gets played, so will try to stall the start of (or just leave) the battle they don't want. And when one battle starts, a bunch of players are removed from the other one. Big game enjoyers are going to want to be in both rooms at once, because whichever starts first is most likely to be the largest battle, especially if they are there stalling out the smaller one. So the incentives flow towards being in both rooms, but when everyone is in both rooms, the duplication does barely anything!

Think about it, if there are 40 players looking to play, and a 32-player game starts, what are the chances that the remaining 4v4 of players will be a viable battle?

So much for the case where both rooms are between games. What if one room is ingame, while the other is looking for players. In this case, where is the big difference between the current system? It is currently possible to watch a game while in a different room. The lobby chat is sent ingame, in case any activity occurs. Double-battling would be a UI improvement, in the sense that you could sit in a spectator slot of the large game while being in the player slot for the smaller one. But if this is the problem with the current system, then that seems solvable with a much simpler bit of UI. Some sort of "spectate other teams room" button on the UI, or something.

Double-room does not seem worth the cost. Even ignoring the upfront technical cost, which someone else can pay if they want, it seems like a costly piece of UI. It is going to confuse new players. Does each room have to be left deliberately, or does it operate on a FIFO basis? I would not fault any player for losing track of a room they are unaware that they were in, and accidentally stalling the game, or causing a restart when they realise that they have to leave.

I think I see the source of the support in this thread. It sounds like a neat poweruser tool for seeding games. Maybe it works if only 3% of the playerbase know about it, and use it for good. But if you put it in the hands of the average player, I fear that the incentives it generates leads to confusion, then essentially nullification of the system. As for poweruser tools, I think people use them less than they expect, because we already have plenty of tools. Right click on games to spectate them without leaving your current room. Try and make !proposebattle a thing, I know the UI is lackluster, but if people want to actually use it, then they will bite. Use !notify to see when the larger battle ends.

I see one clear benefit of double-rooming: it would let people seed games without losing their place in the main room. Perhaps the waiting list has created an incentive against seeding, which double-rooming addresses. But if so, then there would seem to be simpler solutions. So I did a bit of server development and (should have) made the waiting list deprioritise players who just played the previous game. So it should be possible to seed a game, then give up and find a spot in the large teams room when the game ends.

All that said, I've decided it is worth giving smaller rooms a try.

• Rooms cap out at 22 players.
• Split is available at 32, creating two 8v8s
• The queue de-prioritises people who played the last game, so seeding is not so penalised. Maybe people who played the last game and end up in the queue will be motivated to make split work.

Unfortunately, it is not a painless change. When 23-31 people pack into the main room, looking to play, some of them will be on the waiting list. But this is going to happen at some point as people pile into the room. The fundamental problem is how to get 40 people to split nicely into two 10v10s. Making the rooms smaller makes this more likely, provided people have the patience to reach the threshold. The logic behind 22 players is as follows.

• A limit of 16 or below replaces the problem of creating a second room, with that of creating a third room. Walk before you run.
• Splitting on multiples of 4 seems best, and 32 creates 8v8s, which seem to count as the smallest "huge game".
• If we get two 22 player games, then that is 44 players playing, which is a significant improvement over 32, so it might be worth the pain.
• Only needing 10 waiting players to split feels fairly safe. If the requirement is too high, people won't hang around to reach it. The current requirement for 32 players is 48, which I don't recall ever being reached.

I don't know how long the experiment lasts. If it works it works. If people complain and try to ruin it, I'll probably just have to dig my heels in until it is sufficiently tested, in spite of complaints. But if it seem to have perfect conditions, yet fails dramatically, something might need to be done early. So give it a fair go.

Bring back the 32 player limit! Personally I only like to play big battles and not small. Waiting list just means I wait until someone quits, which usually doesn't take that long when there are 32 people who can take a tea break! I prefer waiting over playing small battles. PLEASE GIVE THE CHOICE BACK TO THE PLAYER BASE!

Plus the usual annoyance is the lobby dying instead it overflowing... and that's an issue of the map picker and not the room size. Please return the room size back to 32.