Some thoughts on online service launch woes.

I haven't played GTA Online.

I guess more generally I was thinking yesterday about the hoopla surrounding the server troubles that accompanied the launch of GTAO and how uninterested I was in their ramifications. I've been through numerous poor launches of online titles, including but not limited to World of Warcraft (all but unplayable for the first week of its release), Half-Life 2 (the first major test for Steam; it hung up on authentication for a large proportion of people trying to unlock it at midnight), and more recent launch day (or week) failures of Diablo III, SimCity, and FFXIV. As I've gotten older, though, these things just seem to be less and less worthy of comment from my end. My inability to play WoW when it came out was a matter of genuine distress; nowadays I just kind of shrug and assume that I'll be able to get into the game a week later.

It's difficult to say whether this sangfroid is a good thing or a bad thing. Generally speaking I don't find myself getting all that upset at anything anymore; if I find something annoying or frustrating I usually just ignore it or remove it from my life rather than give it more time than it's due or write angry forum posts about it, and I don't really buy anything unless I'm pretty sure I'm going to like it. Specifically, though, I generally assume that highly-anticipated games with an online requirement are going to be kinda busted for the first few days of their release. That doesn't reflect well on the competence of major publishers, who really should've learned from any of the dozens of teachable moments that have come down the pipe over the last decade, and maybe I should exercise my right as a consumer to loudly berate companies that fail at seemingly simple tasks, but I just have a hard time getting up in arms about things like launch-day server issues nowadays. It's easier to sit back and unwindulax than fret about this kind of stuff, or play some other game that's currently functioning rather than worry about the one that isn't. These things tend to work themselves out with time.

It is somewhat interesting to think about why this stuff keeps happening, though. I assume that most publishers have their spreadsheets about anticipated server loads over time and simply do some math like that of the narrator in Fight Club, something along the lines of: the server load on launch day will be twice as big as it ever will be in the future. We can EITHER spend twice as much money to ensure a smooth launch, and buy excess infrastructure that we won't wind up using OR we can launch with the infrastructure that we assume we'll need permanently and absorb whatever damage to our reputation that'll come from a less-than-functional launch. I assume that the math usually winds up going in favor of the latter option.

But maybe I'm completely off about all this stuff and it's simply really difficult to judge how many people are going to be using an online service the day it launches. Who knows?

I also kinda just expect them to be broken for a week or so now. I usually just pretend those kinds of online games don't launch until a week later, works great.

I can learn a thing or two from that pup.

Its refreshing to hear someone say something in a calm and reserved manner. Most of the time these launches bring passionate,angry, hateful comments. I do fear though people will become even more outraged when the next gen consoles launch, and their online marketplaces won't work on day one.

It's a bit of a bummer that GTA online isn't working but I know it will sooner or later. Rockstar at least came out and said it's probably going to be busted when it comes out.

I can only imagine what the reactions are going to be when the new consoles launch and things aren't working. These things happen so I don't feel the need to get upset about them anymore. There's plenty of old games to play while I wait for the new ones to start working.

if I find something annoying or frustrating I usually just ignore it or remove it from my life rather than give it more time than it's due or write angry forum posts about it

Sage wisdom that more people should follow.

Good post Rorie I've been saying the same thing about this. Sure, it should work but this is just old hat and I'm not going to waste time freaking out about it.

Been really looking forward to GTA O (love V itself) and I'm just going to wait for a few days or maybe a week and just jump in then when it's hopefully working.

And we say "this should be easy!" but.... is it? Blizzard runs the most successful MMO there is and they couldn't get Diablo III working right when they put it out. It seems like no one has figured it out yet.

Yeah I wasn't surprised that GTAO would be kinda busted for at least the first few days given how many people who own a copy will want to play as soon as it launched. Luckily that game can fall back on a single player component that's separated and I hope that continues to be separated for next gen games.

Which do we think has upset more people: Obamacare's launch day jank, or GTA's?

Ignorance is definitely bliss. I think you're right. The best and least stressful option is to just assume the game doesn't work and play something else until you hear otherwise. The problem I have with that is it gives the publishers and developers no reason to improve. They can assume that the public is fine with the way things are and don't need to push forward.

With next gen around the corner I don't want these shotty releases to be commonplace. If they are they need to figure out a new term for launch day because yesterday surely wasn't a functioning launch.

I don't know what to think of it anymore. I'm a pro-consumer kind of guy (I think we snidely refer to that as entitled now?) so I can totally understand how people consider this unacceptable but after years and years of watching many of the best and most innovative online games and concepts suffer these woes it really does kind of feel like part of the "cost-to pay" at this point.

@rorie"...who really should've learned from any of the dozens of teachable moments that have come down the pipe over the last decade"

That's what I had been thinking too. I actually assumed that these "teachable moments" were the reason Rockstar staggered the single-player and online launches. I'm starting to wonder, considering the complexity of today's games and systems and the competition always pushing to what can barely be achieved, if lessons learned from previous launches will ever be able to solve this problem.

I look at this empirically....

1) The biggest publishers in the industry can't make launches go smoothly.

2) The existence of Amazon cloud services, which would seem to be designed to allieviate these exact problems, doesn't make them disappear.

Therefore I'm willing to believe that it's either incredibly hard, or that literally everyone in a decision making capacity is incredibly bad at making the right decisions. Without taking any potshots, that seems unlikely.

And my final words of wisdom: "Dafuq are you playing games at launch, son??"

I definitely used to feel the, "I need this game on the first day" and "I need to be able to play this online right away", but it's all pretty much gone. I'm excited about GTAV Online and really looking forward to when I get to play it, but I'm not surprised in anyway that it's not working right now, nor am I remotely frustrated about it. If it's still as bad as it is now in a couple of weeks, then yeah, I might start to be a little disappointed, but I've got so many other games to play and so many other things to do that I'll survive just fine without it.

I've come to accept that playing online games at launch for a big product is basically that of driving home in rush hour traffic. It also helps that I work in Networking, so I have a decent grasp on what goes into running those servers; one of the simultaneously most amusing and most annoying comments is always, "why don't they just spin more servers up? Use VMs or something!". Really, game launches for me at this point is just feeling bad for the network engineers who are under an incredible amount of stress, doing everything they can on the budget they've got, and having a mostly sleepless week until the activity load lessens drastically and evens out a bit more.

It's really not that none of the developers can have their online servers run correctly at launch, it's that the cost is too high, and it's too much work for too little reward; people certainly don't understand the work of setting up the infrastructure necessary to handle twice+ the player capacity, only to have to downscale it all by half only three days after launch. It's actually an interesting prospect for the next generation, and whether Microsoft/Sony might be able to help with game servers during launch, at a reduced cost to the developers.

The problems that GTA:O face are only going to become more prevalent as we go forward. Right now we have a few games that will be coming out over the next 2-3 years which want to blur the line between single player and multiplayer experiences. Watch Dogs, Need for Speed: Rivals and The Division all seem to be incorporating seamless multiplayer integration into single player.

I can kinda see how this mechanic could grow into a more standardised feature in open world games which could allow for a more immersive experience and bring real advantages to the genre. It's going to be really interesting to see how these publishers try to lessen the impact of that post launch rush to play the game and how they will handle the back lash.

I've been to a few mmo launches and the only one that was positive would've been Tera but that might speak more towards the games popularity than good server management. I think slowly adding more player slots to servers over a weeks time is still the best solution right now or perhaps they could do region launches.

I guess Valve is the only outfit whom has found a way to deal with it. Launching in waves like they did with DOTA 2. I guess people accepted it, because it's a free 2 play game. I wonder if such a thing was applicable for major releases like GTA Online and Battlefield 4.

I don't think people would like it any better than dealing with the usual swamp of issues of a traditional launch. Once one has paid, there's a reasonable expectation that one can play. Waiting for the activation email isn't a thing that I expect to go over well at all with the paying customer.

Just scaling up hardware for the launch traffic is simply prohibitively expensive, or everybody would do it. Nobody does it though. So unless circumstances change somehow, it's never going to happen.

Thats pretty much how I think it goes down. They could spend a whole fortune on servers they would only need for a couple of weeks top or just get enough for a reasonable high demand. I would be really interested to see a breakdown of the money and see if they are being stingy or it is just totally uneconomic/

Been saying the same thing for years now, not just in the context of server issues but by day one patches. Thats the reason i quit buying games day one, there's no reason.

At first it seemed smart to delay the on-line portion of GTA V, that gave RS the opportunity to calculate the server strain base on how many copies they have sold.

And yet, they fucked it up.

The thing is you get complaints either way. Star Wars the Old Republic comes to mind, having server queues that could be lengthy at release, and that was its own point of complaint for a lot of people.

Who else thought this was about the new health care online services?

Don't forget Destiny, probably the upcoming seamless online game with the highest potential sales. I am interested to see which of these games has the most stable launch. I think of the all the games trying this sort of online experience, Destiny has the best shot of being stable right out of the gate.

For one, Bungie has a good track record of having pretty stable online launches for games that support millions of players from the get go. They are also having an open beta on every platform, which should help them get things ready. Finally, if any publisher has shown they can handle the launch of an online game with 15 million plus copies in sales, it's Activision.

Or maybe its a calculated risk. Why spend so much money on server infrastructure that you won't need after the initial 48 hours? I may be going out on a limb, but I feel like RockStar knows their stuff and didn't simply drop the ball. As I read on another thread, why upgrade your bathtub drain for an Olympic sized swimming pool volume of water, when you know that that much water will never need to be drained again?

I like these sporadic, puppy laden blog posts lately Rorie.

I was able to play for a few hours last night without issues so I'm satisfied.

Matt Rorie's blogs are secretly the best content on Giant Bomb.

@brackynews: I agree with you completely, it can't be coincidence that all the largest publishers still have problems with this, and people saying things like "there are loads of prior examples to learn from" and "this shouldn't be acceptable with modern games" have less than 0% knowledge of how any of this works.

Servers are hella expensive too, for games to succeed they need to be profitable and that means they can't blow their entire budget on launch day servers that will go to waste.

What is with this small puppy bias? Where are the Irish Wolfhounds? :(

It's a fact, puppies make everything better.

My theory is that publishers know more or less exactly what their load will be, but they never target their resources toward that anticipated load because that investiment will not pay off in the long term. Instead, they invest in infrastructure for a more reasonable long-term server load. Still, that would explain sub-optimal performance, but not complete non-functionality.