32-player TAW is not good for Zero-K

no, i don't think you're getting the psychology of it. I have a ton of other things i should be doing. When a game pushes me out when I'm engaged, I'm likely to just leave and go do something more useful.

Also being cycled out may prevent people from enjoying their last game of the day or when they'll have to leave in an hour for some other reason.

Even without this mechanic people already left crowded rooms voluntarily for a variety of reasons, opening up spots for those in queue.

It's bad both in 1-room big TAW and in smaller team games. A player seeds a small teams room, waits until it fills up beyond cap, then next game he's cycled out against their will because a few more showed up to play, something which he has no control over.

This reduces the incentive for people to seed rooms or join as they build momentum and rewards those that got in only when it was already full.

Instead you should be incentivizing people to start filling other rooms. Imo the UI needs a working !promote command that does something visible across the lobby (similar to the matchmaker prompts) so it's easier to draw more people into new rooms.

quote: Also being cycled out may prevent people from enjoying their last game of the day or when they'll have to leave in an hour for some other reason.

If there are N+M players wanting to play in an N player room, M players are going to be left out in the cold.

Why is the experience of the M players so much worse if they played the previous game, compared to if they did not?

quote: A player seeds a small teams room, waits until it fills up beyond cap, then next game he's cycled out against their will because a few more showed up to play, something which he has no control over.

Then next game they get cycled back in.

quote: This reduces the incentive for people to seed rooms or join as they build momentum and rewards those that got in only when it was already full.

Conversely, why should players who live in an "earlier" timezone get to hog player spots at the expense of players who live in "later" timezones?

quote: Instead you should be incentivizing people to start filling other rooms

I don't think that the current system is a lot worse at this, and it might even be better. Compare the followings scenarios:

(1) 22 players are playing a long teams game. Through the course of this game, 12 players arrive in the lobby wanting to play a game. They are arriving one by one through the duration of the game, so attempts to seed a new room are going to be very slow, and they have no guarantee of being involved in the next game in the large room either. Many of the 12 players get discouraged and leave.

(2) 22 players are playing a long teams game. Through the course of this game, 12 players arrive in the lobby wanting to play a game. They know that at worst they will get into the next game, so they stick around. When the next game starts, 12 players of the original game are moved to the waiting list *at the same time*. If they want to continue gaming, they can all start a new lobby together; since they are all ready and present at the same time the chances of a successful seeding go up considerably.

quote: If there are N+M players wanting to play in an N player room, M players are going to be left out in the cold.

It's very likely that either M is relatively small and people voluntarily leaving after each battle will eventually free up slots after one or two battles (my previous experience of years playing ZK) or M is big enough and the extra players should just move to a new room and get at least a small teams game going.

quote: Conversely, why should players who live in an "earlier" timezone get to hog player spots at the expense of players who live in "later" timezones?

Since when is this a relevant problem?

quote: (1)... (2)...

You're assuming players facing queue times are more likely to leave in (1) compared to (2). I doubt it. Guess we'll see.

My point is that the annoyance of the loss of control over whether to keep playing isn't worth the minor problems it tries to solve.

These new "cycley" mechanics could be kept as an option for rooms like the "1v1 pro" room with lots queued and spectating to prevent a pair of players from holding it for too long. Even there, people already kind of manage themselves voluntarily.

Yeey, AI in ZK. For bonus connect an angry LLM to complain about everything and we are done. Same experience as with human players.

Galamesh what I really wanted is something explaining Wednesday. But that feedback is good too.

quote: One thing I'm really not clear on is this: is it intended that a player that just joined the room and played only one match gets benched while other players in the room are on their 1+n match?

Yes, it is the simplest system. Tracking further into the past is going to be harder on its intuitiveness.

raaar

quote: This reduces the incentive for people to seed rooms or join as they build momentum and rewards those that got in only when it was already full.

This is designed to not happen. In fact, if I follow your argument then the new system encourages seeding new rooms. I'd better clarify how the queue is meant to work, before people end up arguing about nonexistent systems.

The player list is sorted by two parameters.

• First it is sorted by whether you played the last game in the host. Players who did so are sorted lower. The last game is any game that ended with an automated map vote, otherwise it is deemed too short to count.
• Next, if two players tie for whether they played the last game in the host, then they are sorted by how long they have been a non-spectator in the host. This is exactly the same as the old waiting list.

If you seed a new room, then you're winning the duration-based tiebreak. So the only way you won't play every game is if 22 people join who did not play the previous game. This seems unlikely, and if it does happen, then there are surely enough people wanting to play to form a new room. This is the most reliable way to play every game. That seems like an incentive for seeding. Previously, to play every game, you just had to be the 32nd person to join the room, so joining a small room early now has an advantage over being a late joiner.

If you jump on the bandwagon as a game is getting large, at around 10-15 players, then perhaps you will be waiting for a game. I'd have to have people track and report this to know how likely it is. In any case, this seems like an incentive to seed a new room. People who join when the room is near-full get to play a game immediately, but then run the risk of making space for other newcomers in the next match. I've yet to hear your explanation for why this is inherently worse than having newcomers wait a whole game or two before they get to play.

quote: Next, if two players tie for whether they played the last game in the host, then they are sorted by how long they have been a non-spectator in the host.

I guess (from the rest) that you mean " then they are sorted by how long they have been a non-spectator in the host, with the player that was a non-spectator the longest given priority to keep playing.

Nice system, but clearly would benefit for some explanation somewhere besides a forum post.

From GoogleFrog:

quote: I'd better clarify how the queue is meant to work, before people end up arguing about nonexistent systems.

Is the source code available so we can look it up ourselves? I'd like that.

I think ability of 32 player should be kept, while not the default game in zk.

Im thinking about to just set taw smaller, while allowing custom lobby to be 32.[Spoiler]

My expectation is that some people will go to taw by default and run small pot, while some people whose demand for large pot will make a lobby themselves. then we have both small pot and big lobpot.

(Well my expectation may not be true.)

The only number to answer the test is whether more people play when the 22 room or 32 room is full. Should look at same days the year before and month before. When the 22 room isn't full, then the test really is not running. Find the times when the 22 room is full and compare to that same time the year or month before. If when the 22 room is full there are more people playing (PvP or PvE) than that same time the year and month before, then the room limit probably helped. The difficulty is isolating for only times when the 22 room is full.

It might also be useful to see if there is a trend up or down in people playing year-to-date and factor that in to the analysis. An upward trend should be subtracted from the 22 room number and a downward trend should be added to 22 room number to compensate for the general trend.



So, which is it?

quote: For me, I am not great at micro, I prefer a more supporting role which is not feasible in small matches as in small matches you have to perform all roles at all times - not my cup of tea, never has been. Those games are for young kids who have something to prove. Now, if we had a higher selection of narrow maps where the front can only support 3-4 players and the map is very long where it's a nice tug-of-war then smaller maps with the roles I enjoy could be more possible, but there currently isn't any maps like this that are in the rotation.

quote: Because in BAR there are multiple roles and is split evenly between front line and back line support - which is not the case in Zero-K as pretty much all maps are extremely wide and gaps between players is very wide and becomes incredibly easy to raid any kind of back line.

Honestly, this sounds like BAR is better suited to what you want to do, regardless of the player limit. Roles like "the techer" and "the eco maker" are specifically and deliberately supported by BAR's game design, and (at least as compared with BA/BAR) are specifically and deliberately de-emphasised in ZK.

(As a corollary, it seems like a dubious strategy for Zero-K to set its player limit to appeal to players whose tastes are fundamentally better served by BAR.)