Why have betas?

[Kerotan]

In this technical aspect of gaming I have zero knowledge.  

So to the uneducated man like me, I think why do ubidoft feel the need for a beta for rainbow six.  

I use them because I'm just about to start this beta.

A better example would be black ops 3. Chances are the number of people who'll be playing bo3 at launch will be similar to mw2, mw3, black ops and black ops 2. I'm sure ghosts and advanced warfare won't be far off either.  

So surely they know what's required.  Surely they know how many servers they need and how they react.  Are they used as a demo in disguise or a tool to encourage people to pre order.  

Any tech wizards on here who can shed some light?

[Legend]

Well sure!

From a marketing perspective, they work so much better than demos. People get early access to the game, "it's only beta" so they get a free pass for all negatives, and they are only temporary meaning players can't think "the demo is fun enough. I'll just play this instead of upgrading to the full game."


From a dev perspective, all games have betas and alphas. They're just a stage in development. Alpha tests are usually internal only(devs, QA team, friends and family) with the goal of polishing the product into its final shape/form. They last months usually and have tons of revisions. Beta tests take the practically finished game and stress test with thousands of players. Their goal is to iron out any remaining bugs that didn't show up in Alpha, and fix major server issues. With servers there are two primary things that matter: server power and code efficiency. COD devs might have identical server power to the previous games, but it's always possible that some new code is bogging them down.

This is infact what happened with DriveClub. At launch the game could have had twice as many servers, yet it was an oversight in the code that caused issues. A large open beta could have potentially allowed them to catch this "bug" pre launch.

Betas also get user feedback. Even though usually it's too close to launch to change anything minor, it still gives devs a great understanding of how the community feels. They can then patch this stuff in later.

[the-pi-guy]

From a marketing standpoint, it's a great opportunity to get people interested and trying out the game.  
With a new IP, it can be: this is how it plays.  A great opportunity to get people on board for a brand new game, that they wouldn't have bought without trying it.  

With an old IP, this still works.  It can be: CoD still feels great, and we made some new changes, aren't they awesome?  Another great way of getting people on board, even if most of them were going to try it anyway.  

But there's a lot to it.  

A server obviously doesn't run on magic.  You can't just hook up 10 servers and say look it works!  The servers have to run some sort of software on them.  As software this can be optimized in general, and also optimized for different genres.  

First person shooters might send a lot of data, they might send an X, Y and Z for where the player is, and maybe deltaX, deltaY, and deltaZ for where the player is going.  (Or they could be taking a rotation instead.) Z would be how far from the ground you are, if you jump for example.  
Additionally they'll need to send out trajectories for things that you throw or shoot.  

But would a racing game need all that information?  
You're probably not going to be shooting anything in DriveClub.  

For a game like that you would probably just need X, Y, deltaX, deltaY, and a few rotations.  
Although for most games you might end up having the Z anyway.  

Different games have different needs, and even old games might be added to in hopes that it'll run better.  
Whenever a game is added to, there's possibilities of mistakes.  

And if the issue is a memory leak, you might not be able to notice it until thousands of people are playing your game.  
The game might work perfectly well until they reach a certain point.

Like first person shooters, if they try to make the projectiles work better, they might work on the code.  
During testing they might find that the new code is keeping every bullets trajectory.  A player shoots and the bullet keeps going forever and is saved in memory.  
Now the server is keeping 10000x as many bullets as it should be.  That means if 10,000 players could play on one server the last game, now far less can.  
Even though the code works, and the devs were able to play and it worked great, once the larger group gets on, suddenly the server reaches a brick wall.  


Something like that...  

[kitler53]

In this technical aspect of gaming I have zero knowledge.  

So to the uneducated man like me, I think why do ubidoft feel the need for a beta for rainbow six.  

I use them because I'm just about to start this beta.

A better example would be black ops 3. Chances are the number of people who'll be playing bo3 at launch will be similar to mw2, mw3, black ops and black ops 2. I'm sure ghosts and advanced warfare won't be far off either.  

So surely they know what's required.  Surely they know how many servers they need and how they react.  Are they used as a demo in disguise or a tool to encourage people to pre order.  

Any tech wizards on here who can shed some light?

not true.  load testing is hard to set up because if you want to test how your server performs with 100k users online you have to have 100k users and setting up 100k virtuals is time consuming and expensive.   ...and it's not like you can just clone a box because you can't have all 100k clicking "start" at the exact same moment and call that typical performance.

and the tests don't always go great.  driveclub clearly should have had a beta.  SSF V i think recently had issues too.  from a QA perspective a beta is a godsend as perhaps the only way to complete certain tests.


that said,.. the business people have found a way to make beta's a valuable marketing and sales tool too.  many gamers seem to like the early access thing so apparently it is mutually beneficial.  i personally loath the idea of testing someone elses code but maybe that's because i do that as a job.

[7H3]

well if your gonna have alpha testing why not beta testing? (o:

[Kerotan]

Thanks for the replies men.  
You've printed a clear picture in my head.  

It's hard to believe that driveclub didn't have a beta considering how it was the first big and hyped racer on ps4.

And if they did a beta more people would have realised how awesome it is but months in advance.  

The game probably would have worked week 1 or at least after the first week and the first week sales and subsequent legs would have been much bigger.  Game could be over 3m now instead of 2m.

From a technical and PR point of view not having an open beta test a few months in advance was a costly mistake.  

And yeah they are mutually beneficial.  In this age of post demo's they provide us with that much needed early access.  Whether you're an eager fan or a consumer wanting to try it out pre purchase.  

Playing rainbow 6 with my payday 2 crew today was a blast and I'll now be buying that which previously I wouldn't have no matter how good the reviews are.  And it's out in December so my next game purchases are nicely spread out.  

Second half of the year to the end:
Payday 2,
God of war 3,
Rocket league,  
Mgs5,
Uncharted collection,  
Star wars battlefront,  
Rainbow 6.

And then next year,  Uncharted 4 and rachet early on.

Anyway betas are good for the gaming companies and the consumer.  Long may they last.

[Ludicrous Speed]

Marketing, bug testing (sometimes, and sometimes they just ignore the bugs)

For a lot of games the betas are glorified MP demos to drum up hype.