Home / Opinions / Game industry 'ship it and fix it' attitude needs to change

Game industry 'ship it and fix it' attitude needs to change

Andy Vandervell

by

GTA Online

TrustedReviews deputy editor, Andy Vandervell, shares his views on this week's biggest stories, including GTA Online's implosion, Samsung's big flop and HTC's ongoing troubles.

Wouldn't it be nice if games worked from the start?

What are we to do about buggy online games? To Diablo III and SimCity, add GTA Online (and GTA 5 to a lesser extent). Rockstar's ambitious online expansion of the GTA universe hasn't so much stuttered as imploded. Our GTA Online Problems article is the most read article on TrustedReviews this week. Even the mainstream press have begun paying attention. When your woes are on the BBC homepage you know things are going badly.

But the worst thing about this episode is its predictability. At this stage it's hard to think of any game with an ambitious online component that hasn't imploded spectacularly at launch. Previously the debate mainly focused on how such elements were really DRM in disguise, but GTA Online's problems are simpler. It just doesn't work.

And a quick glance at future releases suggests we're destined to relive this nightmare over and over again. Destiny, the current and next-gen shooter from Bungie, has a huge online component woven right into the fabric of the game. The Crew, Ubisoft's next-gen racing title, is billed as a large-scale multiplayer and single-player game where co-op and team play are core parts of the game. I hope they'll fare better than GTA Online has, but history doesn't fill me with confidence.

Does this mean I'm against co-op and online integration? No, absolutely not - I prefer not to knock ambition and innovation where it appears. But it's not enough to aim high and then deliver chaos, to ship it and fix it later.

It's unacceptable that any game, online or not, go on sale in such a state. I'm sure Rockstar will sort out the issues, and we hope to deliver our verdict on GTA Online when it does, but its size and popularity doesn't make it immune from criticism.

The Xbox One and PS4 are bound to encourage more online complexity and innovation, so developers and publishers (mainly the latter) need to start learning the lessons from these debacles or they'll quickly lose our faith and trust.

SEE ALSO: Xbox One vs PS4

Samsung panicked and blinked first

We published our Samsung Galaxy Gear review this week and suffice it to say it doesn't make pleasant reading for Samsung execs. It's clear Samsung was convinced it had to beat Apple to the smartwatch market, but that's where its strategy ceased.

It makes no compelling case for why anyone should buy a smartwatch. It fails to even surpass startups like the Pebble smartwatch and MetaWatch Strata for coherence and vision. It's very rare we review product this bad anymore, it's even rarer coming from an established name like Samsung.

It will survive this embarrassment, of course, but shouting 'FIRST!' is rarely a recipe for success.

HTC feels the pressure of Samsung's rampant success

If Samsung feels sore about its smartwatch flop, its domination of the Android smartphone market, supported the the recent Samsung Galaxy Note 3, is a decent tonic. And nothing demonstrates this better than the worrying news coming out of HTC. It just announced its first quarterly loss ever, a loss of $101 million after tax.

This isn't a disaster yet, but HTC has all the hallmarks of a company with serious problems. Since 2011 its stock price has fallen by 90% and its share of the global smartphone market is now less than 3%, below the likes of ZTE, Lenovo and Huawei. It peaked at over 10% in 2011. That's a big drop.

The sale of its stake in Beats will guarantee a pretax profit in Q4, but that won't plug the massive chasm between its marketing muscle and that of Samsung. A good follow-up to the HTC One is vital, because HTC will only get so many chances to turn its fortunes around. It only needs to look at BlackBerry for evidence of that.

A note about comments on TrustedReviews

Finally, a short comment about comments. We think it's important the quality of discourse on TrustedReviews is high. For that reason, we've decided to take a less permissive approach to comment moderation on the site. We're not going to publish a big set of rules, we just ask that people use some common sense. We're not interested in gloating, rants, bad mouthing and fanboyism; we want quality, insight and a genuine interest in debate. And if you can't make your argument without foul language, CAPS and point scoring then there's a delete button with your name on it.

Next, read our Samsung Galaxy Note 3 vs Note 2 comparison

andyvan

October 4, 2013, 1:42 pm

PS: I promise to write about some 'good news' next week!

kroguk

October 4, 2013, 2:12 pm

It does seem to be the norm nowadays to release less-than-fully-tested software. GTA online is probably the most highly visible culprit, certainly in gaming circles. But I've noticed disquiet in the Apple camp as well with a selection of bugs being noted in its recent release of iOS7. Do you think that some bean counter somewhere within these companies has published a spreadsheet to show that its actually cheaper for the public to test the software rather than employ a full team of dedicated de-buggers to check the software prior to its release into the wild ? Certainly from my perspective, when I read of the issues that people are having, it does tend to lead me away from buying that particular companies products until a full solution has been issued to resolve the problems !

andyvan

October 4, 2013, 3:47 pm

I don't think anyone expects perfection, but there's a big difference between 'a few bugs' and 'doesn't work' and GTA Online so far falls into the latter category. But, as you say, the more often new products/software updates result in a load of problems, the less I'm likely to trust them in future. It doesn't take long for loyalty to be eroded.

Gordon Kelly

October 4, 2013, 4:26 pm

It's a strong argument to make. There have been good examples... Gmail and Spotify's years of beta status, for example.

But as you say, increasingly it is simply a way to rush out products and deflect all criticism by screaming "BETA!"

Even the most famous of polishers, Apple, has started to do it with the likes of Siri and Apple Maps. It's a concerning trend.

dpanch_89

October 4, 2013, 9:15 pm

I tried to log into online, since i was bored of the game after completing it. It never worked. I then traded it in, and made £5 profit as i sold it for more than i bought it for!

Monzy

October 5, 2013, 10:19 am

Most games are like this now. Take FIFA 14, lurking under the mess is a decent game but it is so flawed I wonder what QA they have in place or are they just under pressure to release into the market.

Matthew Bunton

October 5, 2013, 12:56 pm

This all started with the last generation of consoles with game patching and updates. Originally we never encountered this on home consoles it was reserved for PC gamers. However, once the consoles were fully on-line it was inevitable that they would follow suit.

Yes it is annoying but as games have become more complex there has to be a margin of allowance made to developers. After all we would rather have games without annoying glitches or crashes which somehow got past testing. What I really object too though is that a lot of developers are releasing Beta or Alpha software and having the gamers do their testing and quality control for them.

As for GTA V if this was any other game there would be a bigger upset about it and a lot more rage. For now Rockstar seem to be able to get away with anything they want. The franchise and it's games are so popular people are prepared to put up with far more than they would normally.

Admittedly GTA V is technically an incredible achievement and adding the on-line modes is even more so. People will be patient as they want all the features that Rockstar are promising. A lot of consumers are happy to just have the next GTA single player game and wait for the on-line component to be finalised.

I do agree that we will likely see more of this in the future, However, only the best and most popular games will be forgiven and thus able to get away with it.

ACR

October 5, 2013, 2:02 pm

Imagine buying a brand new car that wouldn't Start.

Matt

October 6, 2013, 5:17 pm

Whilst I completely agree that there is no excuse for selling software that doesn't work, it must be a significant technological challenge to launch and run something like GTA V online. It is simply not possibly for the Rockstar QA team to replicate the number of concurrent users (15 million+ copies sold so far?) during testing to guarantee a stable product from the get go.

There are more unforeseen challenges in launching an 'organic' product such as GTA V online and as a result I don't think its fair to apply black and white thinking - i.e it must work flawlessly from launch.

mothergoose85

October 7, 2013, 9:12 am

So this is an interesting debate. I'm a QA Lead at one of the national newspapers and we also come across this sort of issue.

The thing is, it needs to be explained why it happens.

Most development teams work in an Agile method, whereby we develop and release small bits of functionality very frequently (if you use some sort of Continuous Integration setup, other teams use Versioning [specifically the iOS team due to the fact that it takes two weeks for Apple to process stuff]). This means we get small bits of stuff to test and make sure it works properly.

This is good news.

The problem is, a lot of teams don't actually do it properly and still work in mini waterfall iterations where it gets dumped at the end of the 2 week sprint and the QA's are completely swamped so start having to take a risk based approach (agreed with the business) in order to ship it the feature. If we get behind then it knocks us off for the following sprint and the cycle continues. This is poor and happens in a lot of companies the world over. Of course, it happens where I work too!

So what's the solution, exhaustive testing? Well that's how we used to work. It never found all the bugs, took ages and cost a fortune (if you were trying to find %100 of bugs, the last 10% often costs as much to find as the first %70 of issues). So now we tend to make sure the critical path for a users experience works and then if anything goes wrong in the wild we fix it ASAP. It's called Fail Fast and it's cheaper to go around fixing issues that genuinely exist rather than trying to find them first.

So financially the imediate cost is less but the question I often raise is "does this damage the brand and therefore actually cost us long run?" - it generally seems proven that it doesn't which is why we work this way.

And as @8de6f37f3482173f930613e794ce259d has said, Rockstar can't simulate 15m users with any great ease. They know there would be a peak at the beginning and they will have spun up extra servers to cope with it (it could even be autoscalling) but there's a finite amount of resource they'll pour in to something which will, over a week or two, level out.

Karl Buckland

October 7, 2013, 12:18 pm

Rockstar should have launched it as a closed system and slowly invited/allowed more and more customers to join. Start at 100,000 people, then 200,000, 500,000 and so on. That way, they could have dealt with the issues a lot better and probably within a few weeks, opened it up publicly. This is what Facebook and Google+ both did - they grew organically (in their case, via invites).

It is a supreme technical challenge to launch a system this complex and suddenly expect hundreds of thousands (or even millions) of simultaneous users.

andyvan

October 7, 2013, 1:00 pm

Yes, it is an interesting debate and your experience of Agile chimes pretty much with mine. Everyone likes the ideal until you don't budget enough time to do things!

I think with Rockstar, and games like this in general, some slightly different thinking is needed. As you say, it's hard/impossible to simulate 15m users and what impact that will have - does anyone even know how many have been attempting to log in and play GTA 5?

I won't pretend to have THE answer, but there's a scale of acceptable bugginess that GTA Online crossed. May extinct says the developers would have wanted to prolong and expand the beta, while the publisher wanted to launch it asap. I can't believe someone on that team didn't expect these problems.

andyvan

October 7, 2013, 1:00 pm

You're right, of course. I don't expect flawless, but a lot better than broken would be good idea!

mothergoose85

October 7, 2013, 2:29 pm

I'm absolutely sure they would expect there to be issues - they'll have hardened it and then thought "right, this *should* be enough, lets see what happens.

So the similar thing happened with EA recently didn't it?

comments powered by Disqus