InspectorG, on 02 November 2015 - 01:53 PM, said:
Hypothetical:
You have a player base of 48 players. Enough for 2 matches of 12 vs 12. Problem is 25 drop clan.
BV fixes this how? Who has to wait. Which players own several mech of both tech? How many players only own Clan because ther are better?
How many players want to use IS mechs with SHS?
BV simply wont fix those problems, they really didnt for BT anyhow. How do you weight SHS vs Clan DHS in MWO?
You cant with BV cuz PGI doesnt have a metric and the BT metric wouldnt work either.
The "Just Use BV" crowd will say here "Just assign values to everything, add them up, and Battlevalue!"
Of course, if this were easy, PGI would do it. It's not. It's extremely hard - way harder than in tabletop, where there are far fewer variables and combat is inherently random. Two mechs firing at each other can be modelled in Tabletop very easily, so point values for weapons and whatnot are much easier to arrive at.
[edit: in this context, random doesn't mean "unpredictable", it means via dice rolls which while unpredictable in a given match will play out very predictably over multiple matches. MWO isn't using dice rolls at all.]
In MWO, it's totally different. There are
vastly more variables. Burn duration. Projectile speed. PPFLD vs. Spread vs. DoT.
It's not technically impossible to assign values to things here, but it'd be
extraordinarily difficult. Particularly when you start trying to assign values to, say, engines, and those values need to correlate to the values of weapons and modules.
When you step back, and look at the job PGI does in balancing the weapons (and I'm not saying they do unusually bad work here, just that this too is a large and difficult topic) and how many weapons have terrible stats as it stands now and some iteration and work could really improve them... What hope is there that they could create
another whole line of stats for every piece of equipment in the game, and have them all work when compared against each other? It's practically zero. And a
badly made BV system results in
at best what we have right now, and at worst a terrible, exploitable mess.
But all the above is irrelevant, because you're absolutely right.
It doesn't matter how accurate the BV system is. Lets assume is magically perfect.
If you don't have the right players with the right BV and PSR in the queue when you're searching, it doesn't matter. As things stand, we have battles kicking off using 3 of the 5 PSR tiers because we can't get everyone in a game at just one tier because populations are too low, and that's with the other factor being only a very coarse weight class restriction (solo queue) and nothing at all (group queue).
No matter what matchmaking voodoo you employ, you can't create players out of thin air. This is, and has always been, the fundamental flaw of the matchmaker and it's not one they can fix. So everyone in the queue has a perfect Skill Rating, and a perfect BV rating for his chassis. Without the
right combination of players being available, it'll be another "GGClose MM" anyways and appear to not be working at all.
Edited by Wintersdark, 02 November 2015 - 02:38 PM.