Supersmacky, on 12 March 2014 - 09:07 AM, said:
To translate: let's just do it without any testing and see what happens.
Thus, my reply to that comment which let to your comment. So, it is really you that is not keeping up and creating straw men. My replies are not in regards to a reasonably tested product change, but to the "let's just see what happens" mindset of the original poster. I actually agree with you about testing and it was never my position otherwise. So, I would appreciate it if you would stop creating straw man arguments in order to try and win this one. Unless, of course, you believe untested functionality should be released to a public production environment in which case I thoroughly disagree with you.
The problem with using the Test Server is twofold. Firstly, it takes just as much development time to put up a functioning Test Build as it does to put it on the Live Server.
Secondly, There are simply not enough players who make use of or are even aware the Test Server exists.
The reason why i advocate shipping the 3/3/3/3 MM into the Live Build is due to that limited and easily manipulated player pool.
With the new MM system in the Live Build, you access 100% of your active players.
With the new MM system in the Test Servers, you access a couple hundred players if you are lucky.
On the Test Server, you are not only accessing a tiny part of your active players, but also at a very specific time of day for a very limited time.
It is simply far too limited in scope to truly assess the viability of the new MM system.
In the Live Environment, you can very clearly see (from PGI's end) the kind of drop class trends that would make or break the new system.
There are many times i would have preferred PGI use the Test Server. The new UI and DX11 could have had far more testing than they did. Weapon Balance changes should go through there like crazy. But those are the sort of things that small scale testing works for.
The MM System does NOT work in the extremely limited environment of the Test Server, and should be pushed onto the Live Server with a plan for quickly rolling back if the data shows that it wont work properly.
Finally, it is most certainly NOT exciting in random matches to discover that 8 of your opponents 'Mechs are Assaults. Yes, we will know there WILL be 3 Assaults on the other side.
But what are they? Awesomes? VIctors and Highlander meta poptarts? 3 DDCs? LRM boat Stalkers? Banshees? What about the Heavies? Will we be fighting BoomJagers or Gausscats? maybe a Dragon Brawler? Quickdraws? What kind of Lights are needingt to look out for? Locusts or Firestarters?
Scouting will be just as important for your team as it ever was. Not that many full PuG drops even know what Scouting truly means anyway.