Feedback request: Matchmaking services.

[x4000]

Hey guys,

I'm trying to get a good handle on the various matchmaking services that are out there, so that I can evaluate their pros and cons for both AI War and other upcoming Arcen titles.  Google turns up surprisingly few results, so I thought I'd ask the community here what you guys have seen and used in the past.  Here's what (non-defunct) matchmaking services I already know about, and/or have the SDK for:

- Steam
- Gamespy
- Impulse Reactor

What other ones do you guys know of?  I found two or three others, but those seemed to be (or were definitely) abandoned in the last few years.

[PineappleSam]

Battlenet (warcraft III): You create an account and login. I don’t think it even asked for an email address, although it will accept one for password reminders.

There was a *join normal* option that would put you in a random map with (an)other player(s) or you could join a custom game by choosing one from a list.

Gamespy (Battlefield 2142): ARGH. You create an account (which was a pain, and I was never certain what username was what), and when I eventually sorted it out anti-cheating software kept kicking me from the servers until random fiddling seemed to sort it out. THE worst experience I’ve had with online gamine, especially as BF2142 requires logging into an online account before you can play anything

Steam (Half-life Deathmatch, TF2): I’ve had no problems with it. The router used in our house currently blocks the ports required for me to HOST a game, but joining is hassle free.

Those are the ones I can think of offhand. There may be others though.

-Pineapple

EDIT: Why on earth are there strikethroughs?
EDIT: Sorted... "(an)other player(s)" was using square brackets.

[x4000]

Thanks for the feedback!

Battle.net is only for blizzard products, unfortunately, so far as I know.

Gamespy is definitely a possibility.

Steam would be a possibility, but including it as the defacto multiplayer service means that the Steam store is also included, which automatically pisses off every other retailer to the point of them boycotting some games for that.  Which is a shame, because the Steam tools look really nice for this, but I can't afford to be exclusively with any one distributor. 

[vonduus]

Steam would be a possibility, but including it as the defacto multiplayer service means that the Steam store is also included, which automatically pisses off every other retailer.

And a lot of gamers...

[x4000]

Steam would be a possibility, but including it as the defacto multiplayer service means that the Steam store is also included, which automatically pisses off every other retailer.

And a lot of gamers...

Well, there's a lot of gamer rage at every matchmaking service, from what I have observed.  Aside from comments on any specific one, in general it seems that many people don't like having to log into something to play multiplayer, but that's simply what is needed for the sort of matchmaking that a lot of people want.  To make sure I am clear, while we're all on the subject anyway, nothing's going to change with the existing networking support -- nothing is going away from AI War, we're just adding a secondary layer.  For Tidalis, given the audience of that game, we might go with a more matchmaking-centric approach there for its multiplayer, but still having LAN/Direct IP play is the goal also.

Impulse Reactor v2 would probably be the perfect platform that is likely to make the most people very happy, but that's not going to be around for a while with the features we would need.  So that kind of stinks from that end.

[Black]

I already have steam. If I ask every gamer I know whether they have steam, 95% will say yes.

I don't have impulse's matchmaking service. If I ask everyone I know, a very small number will say yes.

You have to choose the solution that works for the most people with the least headache with the best features. Steam is the only choice that fits the criteria. You cannot satisfy everyone. Satisfy most people in the easiest way.

[PineappleSam]

Thanks for the feedback!

Battle.net is only for blizzard products, unfortunately, so far as I know.

Gamespy is definitely a possibility.

Steam would be a possibility, but including it as the defacto multiplayer service means that the Steam store is also included, which automatically pisses off every other retailer to the point of them boycotting some games for that.  Which is a shame, because the Steam tools look really nice for this, but I can't afford to be exclusively with any one distributor. 

Oops, I misunderstood! I thought you were looking to compare the matchmaking services so that you could create your own based on people's likes/dislikes! My bad

[x4000]

I already have steam. If I ask every gamer I know whether they have steam, 95% will say yes.

I don't have impulse's matchmaking service. If I ask everyone I know, a very small number will say yes.

You have to choose the solution that works for the most people with the least headache with the best features. Steam is the only choice that fits the criteria. You cannot satisfy everyone. Satisfy most people in the easiest way.

Steam includes their store, which means some retailers/distributors boycott me.  Impulse does not use their store in their matchmaking service, which keeps me independent of any concerns with retailers/distributors.  Hence, Steam is pretty much not an option for the matchmaking, so it's a matter of finding something that works for my purposes.

[x4000]

Oops, I misunderstood! I thought you were looking to compare the matchmaking services so that you could create your own based on people's likes/dislikes! My bad

NP!  We have done some work toward creating our own, but server load is going to be a big pain, and there are a lot of things that have to be worked out with a service like that.  And if Arcen ever does go out of business, I don't want the matchmaking servers to immediately disappear. Sometimes it is better to just reuse someone else's premade thing, but there are drawbacks to all of them. In the case of Gamespy, we have an indie license to use that, which doesn't cost us anything, but our retail publishers would have to pay, which is a problem.

[Collic]

Unfortunately I can't offer you any better solutions than what has been mentioned so far, except to say.. Gamespy is horrible ! As third party match making services go it's the worst I've ever used. I convinced a friend to buy Men Of War through steam with the aim to play co-op, and his understandable frustration and anger at gamespy (the specifics aren't really relevant) has meant that has never happened.

Though perhaps if you're using it in conjunction with the existing functionality it wouldn't be so bad.

Of all the options (except impulse, I haven't used that so I can't comment) I'd say steam is the best, but I have no idea what kind of support they offer third parties other than the basic steamworks stuff. It's definitely likely to be the most popular option people are already using if it is viable.

EDIT: Ahh okay, I wasn't aware of the steam store problems. It's a shame they wont seperate that out. I love steam, but I can see it only causing more problems for them in the future.

[I-KP]

Gamespy is a pretty dreadful experience IMO.

[Black]

Mm. Well you are the only one qualified enough to make the decision regarding steam. If Arcen had a larger portfolio of games and was more well known, depending on how much of your income comes from distributors that would refuse to do business with you if you incorporated steam services into the game, I think it would be a reasonable possibility to go with steam.

But you know more about business and specifically Arcen Games' needs than anyone else here, so I understand your decision.

And if Arcen ever does go out of business, I don't want the matchmaking servers to immediately disappear.

If this ever happens, you can release the source code to the match making server so that users can set up their own. Just make it easy for the user to modify which address your games will look at to find the central server. The community is strong enough that someone will throw up a server if Arcen Games ever goes under.

[x4000]

Mm. Well you are the only one qualified enough to make the decision regarding steam. If Arcen had a larger portfolio of games and was more well known, depending on how much of your income comes from distributors that would refuse to do business with you if you incorporated steam services into the game, I think it would be a reasonable possibility to go with steam.

My best guess is that it would be a 10%-20% loss of income, if not more.  And it could be as much as a 50% drop in income if it annoys retail publishers that we might secure, as I've seen word that it can.  Which sucks, because I want to stress that the Steam solution is absolutely great looking from what I have seen of it, and I have full access to the entire deal through Steamworks.  But from a business side of things, I am faced with either a) making matchmaking only for Steam customers (which will piss of the rest of my customers, and make the other vendors grumble), or b) not using their service because the store is included in it.  Frankly, that's not a political battle I want to get in the middle of. *sigh*

And if Arcen ever does go out of business, I don't want the matchmaking servers to immediately disappear.

If this ever happens, you can release the source code to the match making server so that users can set up their own. Just make it easy for the user to modify which address your games will look at to find the central server. The community is strong enough that someone will throw up a server if Arcen Games ever goes under.

Well, that is a good point, and that's what I'd do.  But you still have all the load issues (look at Demigod), and then bandwidth costs if I go with some sort of cloud solution.  Either this is costing Arcen a bunch of money per month, or our servers are unprepared for if a future title (such as Tidalis) becomes really popular.  This is the challenge I've been wrestling with for a while, which is why I haven't chosen a vendor yet.

My background is in web-based system design, which would seem to be ideal for just making our own matchmaking service, but there's just not time in the development cycle for me to do that for Tidalis, which is what needs matchmaking more imminently.  Fortunately, on the other hand, Keith's background is also in web development stuff, so I could theoretically make that his big project to go alongside AI War bugfixes, with a service being geared towards both Tidalis and AI War.  Then I'd just have to figure out the hosting and costs issue, which would also impact the sort of technology we'd use to make it (PHP/MySQL is good for some stuff, but not if this is more cloud-based). 

I've used Amazon's cloud services before, and I like them pretty well, but they are not free, not a RDMS (not a killer, but less than ideal), and anything I develop

[keith.lamothe]

I'm happy to write something, the tricky part is paying for the infrastructure, and for that point alone I'd rather deal with the vagaries of Impulse v1.  If it came down to write-our-own vs. gamespy, though, I'd have to say write-our-own.

[Dathknight]

I would do a own System, because both Gamespy and Steam are hated by many people (I hate Gamespy, which is unstable for me).

But I don't care what you do, I am fully satisfied with the singleplayer at the moment