I’m making a game that’s multiplayer only, so at least two people need to connect over the internet to play, but the issue I think i’ll face is that no one will play because no one plays? How do people overcome this issue?
Get the word out. Make YouTube videos sowing your progress to get people interested, announce features you implement and note the ones that are unorthodox, or adding something new. And if you really want to sell it, make a cinematic or gameplay trailer for it. This is a problem you will face with ANY game, not just online ones and so the ‘getting it out there’ part is no different.
An absolute crap-ton of marketing work.
I think I read of a small-budget multiplayer game that recently ran free for a few months just to build a community. I’m not sure how that panned out for them. The “no-one plays because no-one plays issue” is a really tough one to deal with. One of the reasons why we haven’t attempted any multiplayer-only games.
And network code is hard! 
Your game’s going to need to be really good, i dont mean in a high production value way, more a get people hooked way, like some famous examples im sure you can think of, and youll need to do a lot of promotion and talking and you’ll need to be ridiculously lucky
I personally would avoid it if i was financially dependant on its success
add single player support with bots… like other games have done
also good for coop and multiplayer when there is not enough other players to make it look like there is someone playing
ai is not that hard lots of nice assets and so is navigation …
I don’t like the idea of bots, I think that within the game they would kind of make it boring. As I’m not that good, I doubt I could make AI that reacts the same way as people in the game.
I think I’ll leave the games release until I’m super famous then everyone will play it 
Even if that’s rather vain.
I played a soccer game, that if i went in at strange hours, you had to wait for minimum 11 players to start.
I think that at the start you just need to keep them busy until someone else comes online. They had a training version and a world to explore, but that took alot of time to make probably. I also stopped playing because of the waiting 
Something interesting in the lobby, but i don’t know what that should be.
The best way to get your work noticed it by firstly getting your friends to test it out. If they are liking what they see then you have a good chance that people will play it.
From then, you can get further feedback by posting threads here: http://forum.unity3d.com/forums/34-Works-In-Progress or here: http://forum.unity3d.com/forums/11-Showcase (remember not to spam)
have a single player or bot mode they can play while waiting for people to connect.
It’s a classic chicken and egg problem. Typically you tackle it by marketing a specific launch date and having new users discover your game while its flush with people from your “grand opening”. You announce specific times that the server will be up and work hard to insure that enough people will be there for the game to not only function but bring them back for more.
Use metrics to determine when you can extend your server time. It’s a careful balancing act. Wait times beat empty servers for sure.
I believe extra credits did a video on this topic but can’t find it for the life of me.
Waiting in the lobby for someone to play is not that weird. If the game is good, the community will create itself.
Your game has a lobby, right?
Its not hard just very time consuming and demanding. I have seen experts squeeze and squeeze every little bit out of the code to get it as optimized as possible with excellent resources and a big server pipe and yet they still have problems with a guy down the street getting booted and lagged out every 15-30 minutes, while their friend 4 states away is fine with very little to any lag and doesn’t get booted for 10 hours of straight game play.
I think in order to make a game thats popular you need to chase after the lowest common denominator, for example look at all capcom games where they turned there IPs into generic shooters trying to go after the call of duty of crowd - everything from resident evil to dead rising.
Unfortunately multiplayer games the basis of being “good” often will rely on players.The issue I’m asking assumes my game is good but still will struggle to gain enough players.
Clearly based on my post count, join date and the fact that I have numerous posts in both those sections I can say that whilst these are a good way to show off the game these forums don’t have enough traffic to sustain a multiplayer game player base.
The strategy I intend to use is to design the game to be played with friends rather than with Internet randoms. You can sell the game in a 2-4 pack and encourage players to buy more copies.
Hire a few hundred people to play the game for six weeks after launch.
Bots, write AI bots that take the place of players, so regardless of the number of players there is always a playable game.
My crazy idea is instead of selling the game, I would sell it as community service for low annually price of $10. Service would provide:
-Ongoing developer support, mostly to better service/game, maybe more content too.
-Ladder/match making, help players form teams locally and worldwide, so they can compete in ladders.
-Mod support, something like steam workshop, easy to find and install mods.
-Cycle money back into community by prize competitions and paying modders for certain amount of downloads they get.
That’s the basic outline of the idea. Besides making the game as fun if not funner than it’s AAA counterpart, I think this would keep players coming back daily.
Of course no one would play if no one was playing - because they couldn’t.
Multiplayer only games need to be heavily advertised, promoted, patched and incentivized on a regular basis. There are a bunch of multiplayer-only games on steam that are flops simply because they are multiplayer only. You go to read the comments on the community hub and the first two pages are people whining that they can never get in a game to play and there is no ai support - if there is, it usually sucks.
There are successful Multiplayer only games obviously, but they’re heavily supported and advertised. Exposure and replay value make those games work.