Default player setting in auto-match: make it "3 or 4 players"
As a suggestion, is it worth making the default setting "3 or 4 players" in auto-match?
Getting a 4 player match is the unicorn of all things Catan (short of being able to organise a custom match - but the communication avenues are not effective enough to easily co-ordinate between 4 people). And many people I have spoken to prefer it, and bemoan the difficulties of getting a 4-player match.
The difficulty with 4-player is there is never enough people waiting to form a game. Most people seem to stick with the 3 player, as it seems to be the fastest bet to get a game. But I suspect most people don't mind, so long as they get a game.
Setting the default to the "3 or 4 players" option would increase the pool of players happy to play a 4-player game, while allowing those who want a specific number to change it if they wish.
Would it perhaps
Or give us back the lobby.
Seriously, the problem with matchmaking is that you need hundreds of people wanting to start the game AT ALL TIMES. This game is not that huge. Lobby is the most optimal way to go. @Administrator Can you list the reasons why you want to use matchmaker instead of a lobby?
I do not want to let an algorithm decide which players are the best for me to play with. Especially since so many people quit the game and your broken AI takes over. It will just ruin Catan experience online.
@BlueMythril Thanks for the feedback. We are analysing the data we receive during the early access (games played with 3 or 4 players, scenarios chosen, ruleset chosen etc.) and we will optimize the default settings according to our findings.
@Stroom It is true that the numbers of active players is not that high during the early access. Even though the number of players is not that high we are currently hosting about 3000- 4000 games a day on the platform.
Using a lobby or list to browse through these games would be quite annoying for the user and you would need some kind of search or filter to find the lobby/game you want.
Most multiplayer games these days got rid of the lobby system and gave the user the possibility to set his or her favorited parameters to search for a game (or they got rid of it all together if you look e.g. at Overwatch). Setting your favorited parameters of a game does mostly the same as the search or filter in a lobby browser would do. You should see it that way , i guess.
If you want to have full control of the game settings and who you want to play against, try the Custom-Game mode.
@Administrator Out of curiousity, will that analysis be factoring in the bias towards 3 player games? For instance, I had no idea a "3 or 4 player" option existed until someone told me it did, and so was just leaving it on 3 player all the time, as 4 player took ages and inevitably went to a 3 player game anyway. I suspect many people leave it on 3 player just because it takes less steps to set up, and generally is a faster match making process. But having "3 or 4 player" as the default setting would be just as quick, and would give you (or the developers) greater measurability of gaming choice from the early access experiment, as anyone who strongly wanted a 3 player or a 4 player setting would change it.
I personally don't mind the auto-match, but think it could certainly be improved.
Anything that increased the probability of 4 player options is one improvement, having a better communication option for players who might want to play a custom match (or set auto-match to friends only - the current popular method for playing with friends, given the cumbersome issues with custom match). Friends and I have resorted to using the chat option here in the forum, as alt-tabbing between here and the game allows for better communication and co-ordination of timing when using auto-match (friends only) to play with friends. The lack of a chat room or quicker option than the very clunky message system is a definite frustration and inhibitor of full gaming experience.
The message system is pretty lacklustre:
- you have to leave a waiting list to check it;
- It keeps remarking messages you have read as unread, meaning sooo often I have left a waiting list for a game to check what is flashing as a new message (thinking perhaps someone wants a game with me), only to realise it's actually an old one;
- It doesn't show you the original message someone is responding too. So when you have added a bunch of friends recently, and are still getting to know which one is which, and there has been several days (or weeks) between communication... well, getting a response to something you have said, but cannot remember, makes for awkward conversations.
- It deletes everything so you cannot track down previous messages in response to point # 3. I actually don't mind the deletion - I find it kind of liberating in an age where we accumulate data for the sake of having data. However, without messages retaining earlier thread of conversation, auto-deletion leaves you with no context for recovering the context of whatever conversation and content is in the message. You should have one or the other (either auto-delete but with previous messages included in reply; or else no responses included, and just the reply message content: the first option is the fastest and least infuriating work flow).
(Btw, @administrators, I'm not being negative or aggressive here, simply questioning logic and high chance of biased results)
@BlueMythril Generally there we can say that 3 player games are favored by the community. As you said it is the fastes way to play the game and this is of huge iporntace for our useres. But that finding is also based on the current App and the experience our collaegues at the Catan GmbH made with PlayCatan.
I think you are also right when you say we should try to make a default setting to "3 or 4 players".... Just to see how this turns out. We discucced this option here before but i will bring it back onthe table.
Have you and your friends considered to form a guild? That way you have a seperate chat in the game which can be used from the main menu.
We will have a look at the messaging system. It was implemented at the beginning of our developement process and we will have a close look if it still fits the demands of the useres. Thank you very much for your feedback on that issue.
4000 games per day, over 24 hours... About 5 minutes to get a game started (Worst case number, in PlayCatan a game is usually created in 2 minutes in my experience).
It means that an average of 14 games are pending creation in the lobby at a single time. Even if you get 10000 games per day, it would be really easy to see them all on one screen. Lobby like this is NOT annoying. Maybe find someone who can create a usable and compact lobby UI.
And you can set the filters yourself to filter them down even MORE! For instance: use the matchmaker filters but show all pending games for the players so the players can manually select a game if your matchmaker takes too long time to find a game. Maybe even order the games based on the matchmaker filter, but SHOW THE PENDING GAMES LIST. Mostly players can find a game much faster than the matchmaker. And they are happier that they got to do it themselves rather than waiting for 5 minutes.
Also since you are using matchmaker, you are removing the possibility for a human to substitute another player. Substitution was a really great feature. Playing vs AI basically means that other people will quit the game too.
As I said, using the matchmaker removes the possibility to make truly custom games with strangers. The game will never start. With a lobby, you will at least get curious people TRY to play with a custom setting. This is really bad because people will all resort to using 3 player friendly robber games and that totally takes the competitive level out of playing online Catan. With the matchmaking there is no incentive to experiment with customization. Hell, even getting a 4 player game going will be killed. The game is best with 4 if you want to play competitively. You are taking it away with no lobby. Might as well stop with real life Catan tournaments because people can't practice playing online anymore. For the very least also include a lobby and let the 3-player guys make their own games. It is not feasible to create a custom game with a matchmaker. Matchmaker ONLY works for one set of rules. For anything else, you'd rather go buy the actual game, get a bunch of guys from the street and play in real life.
I don't want to be mean but that 3000-4000 number is probably caused because your games are breaking so much that players have to restart the games over and over again. Can't finish a game at 20 minutes but the games usually take 40 minutes to complete means that the players create 2 games in the same time so the number of started games is a bad metric.
Big games like Overwatch and League of Legends use matchmaker because they have 10000+ times more users than Catan. You can not possibly compare yourself to them. Even for these games people want a lobby because they get tired of the same set of rules dominating the game. Matchmaker is too slow.
What good is the guild for? If you have a global chat lobby you can communicate better. Especially with this kind of low player count. Once you get bigger, think of guilds. Having a guild with 20 people is pointless if you can't get enough online at the same time to play the game.
Considering all this... And you still have a LOT to develop, this game will probably be in "early access" even one year after this. Just open registration to PlayCatan again.
@Stroom What you are saying is that a matchmaking algorithm is slower than a person looking at a list containing various entries each entry with at least 4 variables listed in them (scenario; expansion; dice mode; turn timer; number of players). This excludes info like the ELO rank, the Karma level, the region the player is coming from (that we considering btw. in our matchmaking process.
If playing on a competitive level my main goal is to compete against players with about the same ELO-Rank or better. Should i browse manually through a list of games to see if there a suitable competitors for me? Wouldn´t a list need info about all the players in there, for me to make an educated decision. Or wouldn´t it be nice if this was done by an algorithm for me?
Think about the new player starting the game and getting matched up with a veteran. I would guess you will not have a good time playing the game if you look at it from a competitive perspective.
But there are two ways of thinking about the issue, i guess. We decided to go with a matchmaking system which takes multiple things into account, as I mentioned. Other games including ours have proven that this is a solid way to go.
@Administrator Your matchmaker can NOT find a suitable game for all players. Online Catan does NOT have enough simultaneous players to make everyone happy at the same time. What would make people "happy enough" would be to start a game within 5 minutes not 1 hour. Hell, if you have a matchmaker, the threshold is 2 minutes! A lobby I can handle but the reason to use a matchmaker is to get the games started FAST. So far, I have stopped waiting for matchmaker and played vs AI.
At the moment, statistically about 14 games are pending creation at any given time. That is not very much. Also your matchmaker will be SLOW. People want games to start FAST. Your matchmaker will be much SLOWer than people would expect a matchmaker to be. This will drive people away from your game. If you had a lobby it would make people see that there actually are people playing this game and people are trying to organize games. (Again, let's admit it that this game is not as big as Overwatch!) With a lobby, you would just wait at least 3 minutes and then a game suddenly starts. If you see a lobby, people creating games etc, your sense of time will be reduced much more than looking at a loading bar, waiting.
You mentioned in a German thread that your matchmaker also includes language barriers - if you don't want to play with English speaking people, they are not included in matchmaker. If you include these kind of filters, your matchmaker will be too inefficient. How don't you get it? The games will take an hour and you can NEVER start a game with custom settings. The only solution is to MANUALLY join each guild, search the forums and find people who would be willing to play with the settings you'd like to play. This takes days in some cases. Instead, if you had a lobby it would probably take 20 minutes max. Why should people form guilds to bypass your shitty implementation of a matchmaker? Make a form of organizing games that works! I am talking about lobby here, if you didn't get it. A lobby solves SO MANY PROBLEMS that you currently have. Want to start a game with obscure set of rules? Lobby. Want to see how many people are online? Lobby. Player leaving a game - search the lobby to join as substitution. Matchmaker is meant for BIG GAMES. Catan is small.
I even wonder if it is profitable for the owners to sell the license. Actually, selling license is profitable, using the license probably isn't.
If you have only 14 games pending creation, a human will be faster at finding a suitable game. Even if it is not most optimal, he will be more happy that he found a game rather than waiting an hour constantly looking at the screen to get the game going. Your algorithm will suck. It includes a "waiting time" which is most likely larger than average person's tolerance.
Lobby case: The game creator can set an ELO range so only people with the acceptable ELO can see the game. This way people who join the game can be sure that the game they join is the right one for them. Instead of only thinking in matchmaker vision, think in the vision of players setting the filters and... Just look what PlayCatan did and improve the filters!!! Seriously, there was NOTHING wrong with PlayCatan lobby and no matter what you do with your matchmaker implementation, you will not increase your player count by more than 50%. Even if you get 3x more players, the improvement of filters will be enough to enable players to find the good games for themselves.
Your examples about inexperienced players and pros being put together are bullshit. Game creators should be able to set an ELO rating so new players can't join these games in the lobby. No need to remove lobby for that reason.
Look at boardgamearena.com lobby. You have 20-40 DIFFERENT games with different settings pending start at the same time. And it is really easy to find a suitable game for yourself if you use the right filters.
You might argue that using matchmaker is easier than using a filter. Your matchmaker filter is much more complicated to use than the one other games use. Just let people filter the pending games themselves. Players are not idiots.
But why even argue, you are never going to make major changes to your implementation of the game based on player criticism. You only implement stuff that is easy to do or that improves current things.
Your matchmaker will forever be slow, over-analyzing things and causing problems. When you mention "other games" then these games have thousand times more users than the game you are developing. You have a developer mentality, you do not care about the players, you just want to write the code that the other games are using to "fit in". This is an old-school game. Old-school solutions work better. Ticket to Ride is using a lobby and it is easy to use. Catan is not much larger than that community.
I agree that there should be defoult 3 or 4 player game in quick match. I believe most players would prefer 4 player games but as it was said before people just want to get in matches quicker. By givin 3 or 4 players defoult option, there would still be 3 players matches in case of small server population, but also there would be 4 players games in case there are enough people.
I also believe (although I don't have data to prove it) that 3 player game is a lot more biased. I noticed that because there is no 4th player, the third player still gets 2 pretty good spots and can easily wins the game. Also there are more better spots left on the board at the end of the settlement phase, therefore road strategy (wood-brick strategy) has massive advantage (racing for the great spots left on the board). For the end I would like to say that in 3 player game it is harder for the 2 players to stop the leader, while in 4 player game it is easier.
So to sum it up: give 3 or 4 player games as a defoult option, most people will be OK with this and it will also restore the balance in games.