Why are indie competitive multiplayer games so rare?

@Kiwasi : This is exactly why my game is going the Dedicated Server route (Player Hosted Servers).
I'll probably host 1 server my self 24/7 (that way there's something going on all the time).

However, this is the problem with today's world. So many people get enticed by these nice fancy features with cloud network approaches, that they fail to see the long term, and that is, having a long sustainable game. That even if it doesn't sell 1 mill copies the first week, that you can at least keep working on it, without stressing that you're gonna go out of business, because of the fees associated with cloud based solutions that warrant the price of paying for them. Nor getting any anti-cheating measures will kill your game quicker than you going to the publish page and publishing it.

So in my not-so-important opinion, going the dedicated server route, not only does it not cost your company money for people to host servers, it allows you to focus on core aspects, and thinking of the future in terms of marketing where money can be better spent on getting more players....

I really did try to give photon a chance, really I did. But have you seen me post a release of a game with it yet? Nope.
I've made at least 20 games with it, and never released them. All because of one concerning fact. Do I really want my customers to pay E.G. 20-40 dollars for a game. And in a few months, have to shut the game down because I didn't get enough sales? At least with a dedicated server approach, my entire company could shut down, and people could still be hosting and playing my game 20 years from now. Too me, having a long sustainable game, whether it sells copies or not, is more important than an easy quick fix, that later down the road, if I used nothing but cloud based solutions, in a 30 year period I'd be looking at potentially millions of dollars that could have been spent elsewhere. Such as employee's, marketing, etc.

 

One aspect of multiplayer development that is far more time consuming is debugging. Instead of simply running the build quick in the editor, a multiplayer dev often needs to coordinate multiple instances of a build while testing. For example, a typical client-server style game will often need a server instance and a couple client instances to test. That makes causes iteration to take longer.

 

We have a stubbed version of our network connection that bounces back messages. So it can be tested localhost only and without overworking yourself. We can also unit test the network code. Some manual testing is always needed before a new release to stable offcourse

 

I understand what you are saying. But you missed completely my point by miles. I haven't stated that single player is, or multiplayer is harder / easier. I wrote, that is not always true.

To your argument, I will go a bit extreme, but valid point by picturing a situation. Imagine you want to make a creative single sand box, where AI build something fancy and unique. This feature probably would take much longer, if feasible at all, than implementing multiplayer, where players do things instead. Look for example minecraft.

 

This is where many older games shine, with no dedicated master servers / clouds. And still working. And in fact, you probably own a transferable copy / installation. Where for example in case of steam games, if steam goes off, or close down, or you loose access to internet, probably halve of games start fixating, without connection. Not to mention, you don't have way to install on different OS / PC, if need change for any reason.

 

You have a valid point. There are some examples where a single player game can be more difficult to implement.

 

comparing clouds to bare metal doesn't really make sense. You can get dedicated bare metal in the cloud.

Cloud vs not cloud isn't a cut and dried thing. The scales can tip in favor of either depending on the situation. You need to consider all costs, including things like how much time does managing your own infrastructure cost in terms of labor and take away from game development.

Cost differences are not that much if you are taking everything into account and also compare the best pricing points that make sense for a real game. Ie not some hobby project that isn't making any interesting money anyways.

Those best pricing points are generally for higher end hardware. And cloud providers have better hardware then places that just rent/colo in almost all cases. If you want to actually rent premium current gen hardware you have to lease. Because colo providers have no other way to get full utilization and make their money back, not enough to compete with clouds.

 

@snacktime Why does everyone keep thinking I'm talking about master servers. Almost everyone confuses the terms between dedicated server and master server. Dedicated servers aren't global relay servers. Master servers are. Dedicated servers cost absolutely nothing for the developer, and for most people, it cost absolutely nothing to host them either.

I can right now, this second, launch a dedicated server from a game over 20 years old that the company no longer supports, that literally has no master servers, host a server, and people can join in. That is a dedicated server. Which is the best parts of dedicated servers. People just type your IP/Hostname and connect, done. Despite them types of issues of IP/Hostnames, there are a lot of websites dedicated to finding them, just for them reasons.

I've hosted resource intensive games on a cheap crappy laptop, packed full, and never really amounted to much of a hiccup. So for most people, so long as you're realistic, there's really no reason to 'rent' these massive servers, yeah they might give you that slight edge over the cheap crappy laptops, but the point is, so long as you're realistic with what you want to do, most people really don't need to 'rent' powerful servers. So in the long run, it cost your company nothing, nor does it cost a lot of the people who like to host servers money.

I mean certainly MMO's are entirely different. But most games are match based, E.G. - complete a task, go to next level. So I mean obviously depending on the game, it makes sense to have server-to-server hosting.

But for while a game is still a live, it's not very hard to match people together at a cheap cost effective route through HTTPS. So you can create lobbies, I'd rather spend 20 dollars a month on a decent SQL Server, rather than 5 grand a month having multiple master servers with low settings.

Start Server > Send info to SQL Database HTTPS.
Client starts game > Request info from SQL Database HTTPS.
Client shows info, and when they click join, they join.
Joins.

And if you ever get tired of running the SQL operations.
Offer a direct IP/Hostname connection. Done.

And people wonder why their indie games never profit. They spend money on all this unnecessary stuff.
I don't claim the SQL stuff is the best route, but it's the most cost effective and affordable route that can be used
until you can afford all this expensive hardware. And of course, if you go the SQL route, you need to know PDO, etc. To avoid SQL Injection. But that's simple.

 

We always wanna make multiplayer competition games since eSports is very hot in ASEAN here.
We have a dedicated server. In additional, we find no tutorial about Ad-Hoc multiplayer for Unity.
If we could make a competitive game like Fun Run series, we would be very happy.

 

I wasn't assuming any specific setup or architecture. I also wasn't addressing games so small they don't make enough money to be interesting, because well who cares costs aren't going to be much anyways in that case.

 

Indies have a hard time selling large numbers of games. You need large numbers of players to have a successful (only)multiplayer game. That is the largest problem. No one wants to play matches against the same 10 players all the time.

 

Because multiplayer games are hard, and keeping things responsive enough to be "competitive" is even harder. That's why I stick with co-op If your teammates movement stutters once and a while, no biggie. In a competitive game, if that was the case, get ready to get bashed in the review department

Also there's just a lot of maintenance that goes along with it. Most indie devs just want to release S***, make sure it works, and put out the occasional patch. Not wake up at 3am to deal with issues. Kind of defeats the purpose haha