Narrowing Tier Spreads Discussion - How to Fix Matchmaking
Дата: 26.07.2011 16:52:56
captking, on Jul 25 2011 - 05:55, said: Thanks for the time to talk to us and give the perspective of the
Company.ChrisK: My pleasure! 
Quote I do have a hard time believing that creating a product and not wanting to make profit off that product does seem a bit dubious.
ChrisK: Of course we want to make money! But we won't make money in the
long run if everyone comes in, spends $300, then quits. It's better
for us if everyone feels they are not spending too much, enjoys the
game over a long period, and invites their friends. And we don't
"rig" the game to make people spend money. One of our core values
is that you can play almost everything in the game for free.
Spending money speeds things up and makes it easier to advance, but
otherwise you can play 90% of the tanks and equipment for nothing.
It's as if World of Warcraft suddenly became free for almost
everything, but they charged to reduce the experience cost of
levels and you had to buy purple items with money. There has to be
a way to make money, but we don't have to force everyone to pay
(typical WoW-type subscription model) or somehow "trick" them (if
you don't want to pay, then don't - it's that simple).
Quote You talk about cultural differences, the NA player-base don't want to dominate, they just want a fighting chance. Driving into a battle that you have NO chance to even damage the enemy, is not fun.
ChrisK: I've been in plenty of battles where I was the bottom tank, but
I've either used what opportunities I had to scout or try to get
behind opponents and threaten cap or arty, or I have found ways to
engage heavier vehicles while they were also engaging our heavies
(greater threats). I don't always survive, but somehow I make
credits and XP. Fun? Sometimes. Challenging? Always. When I lose, I
don't feel bad (I started outclassed), but I do feel a lot better
when I get a kill or we win. For some reason, I think most players
equate balance with being able to kill any opponent, which it
isn't. As someone pointed out in an earlier analogy to Team
Fortress 2, there are some enemies (e.g., a tank with a medic) that
require teamwork to bring down. As a demo (SPG) or scout (light
tank), I can't do that myself, but I can back off, communicate, and
try to help out when our own heavies arrive.
Quote How does WG justify putting a M3 LEE into a Match with a Pz VI Tiger?
ChrisK: The M3 would be a fool to try to take the Tiger in a 1:1 battle,
but since this is a team game, there's certainly an opportunity for
the Lee to take on the Tiger's peers of Tiers IV-V, cap, scout, or
hunt artillery. Again, all valid uses of your tank. If you only
expect to beat the enemy's best tank in every match, you are
certainly not looking for balance!
Quote How does WG justify balancing a TD gun advantage then taking it away by putting into higher tier matchs? IE: T40 Vs KV(107)
ChrisK: The KV's 107mm gun is a Tier VII and the T40's best gun is a Tier
V. Once again, you're comparing apples to oranges. You should be
comparing the KV to the T1 with a Tier VI gun or the T40 to the
Hetzer with a Tier V gun, that would make more sense. If what
you're saying is that TDs are overmatched by same-Tier heavies,
then yes, in general the armor of the heavy is better, the main gun
is about the same, and the heavy is slower. However, the TDs are
there to overmatch the mediums on the same Tier, not the heavies.
It's a rock-paper-scissors effect (Heavies > TDs > Mediums
> Lights > SPGs > Heavies) at the same Tier, with variety
and challenge thrown in by mixing the Tiers.
Quote How does WG justify making changes to some tanks for "Historical" reasons and others are not changed?
ChrisK: In many cases we're discussing prototype or planned tanks whose
actual performance can only be estimated. In these cases, play
balance will always take precedence over mathematical guesswork
about penetration, speed, damage, etc. In the case of real
vehicles, we do our best to model the tank accurately to the extent
that we can (no tank crew can fix their tracks in 15 seconds from
inside the tank!). We do a lot of research, and even the research
doesn't always match up between respected sources. In these cases,
we use the (valid, historical) information that provides the most
interesting gameplay mechanics or balances with other vehicles of
the same type and/or Tier. In the end, this is a game, not a
simulation, but we do also try to use the most accurate data we can
find.
Quote Is the population of the SuperTest server that high? I was under the impression that the super test would be a small group, so any testing of matchmaker balance would not yield any fruit due to the limited number and the inability of the matchmaker to deal with small population. You yourself have commented that if we could get 130,000 of our friends to play then the matchmaker would work.
ChrisK: Supertest is a small group. I believe it is limited to 100
experienced players. The main idea in this group is to test new
vehicles and maps for unfair advantages against other vehicles and
find bugs. However, because we more actively and directly solicit
feedback from this group (on a much less crowded forum), it's much
easier to get your points noticed, and to get feedback from other
highly experienced testers (and devs) with a lot less trolling and
n00bish advice. 
One common thing I have noticed with all of the concern about matchmaking is that it only relates to the current game type (Random Battle). We're adding more game types as we move forward, and correlating the matchmaking system to each game style, so perhaps once we have games that aren't as random (and possibly more control over which games you join or start), there will be less of an issue with MM overall and we can move on to something else. Like spotting.
One common thing I have noticed with all of the concern about matchmaking is that it only relates to the current game type (Random Battle). We're adding more game types as we move forward, and correlating the matchmaking system to each game style, so perhaps once we have games that aren't as random (and possibly more control over which games you join or start), there will be less of an issue with MM overall and we can move on to something else. Like spotting.
Narrowing Tier Spreads Discussion - How to Fix Matchmaking