|
|
| Author |
Thread Statistics | Show CCP posts - 1 post(s) |
Vitantur Nothus
Nos Nothi
1121
   |
Posted - 2014.12.11 14:26:00 -
[1] - Quote
Imp Smash wrote:Jaysyn Larrisen wrote:CCP Rattati actually directly addressed this on Biomasses Podcast this weekend. I asked him about it and he is iterating on an idea with the CPM but he's generally looking at changing the base squad size to 4x players and eventually trying to link the squads into a platoon for FW.
... Hence discussion now -- considering past history of 4 man squads being universally hated (and I mean universally -- the squad size increase was the ONLY time EVER we had a change to Dust where there were 0 dissenting opinions) -- do we want to return to 4 man groups? Has the meta and the in game connections matured to the point where that would have fewer downsides than it did way back in closed?
A few possible explanations ...
A. Limited Population
How many squads of six hardened vets are online at any one time? Of that sample, how many are playing on a common battle server? Of that reduced sample, how many are playing the same game mode? Of that further reduced sample, how many are presently waiting in queue?
B. Queue Times / Monotony
If matchmaking were instructed to only produce fair fights, it would need to consistently pit a full vet squad against another full vet squad. At any given time, there are a limited number of vet squads available. This translates to substantially longer queue times. Perhaps more importantly, it translates to fighting the same opponent over-and-over again.
[u]C. Statistical Anomalies[/u
There is a distinct possibility that 6-vet-stomp-squads maintain outlier-grade win:loss ratios. If this is the case, and matchmaking priorities are accounting for points A and B above, then squad size reduction is the only tenable solution. |
Vitantur Nothus
Nos Nothi
1143
|
Posted - 2014.12.12 17:35:00 -
[2] - Quote
Imp Smash wrote:Vitantur Nothus wrote:
A few possible explanations ...
[snip]
C. Statistical Anomalies
There is a distinct possibility that 6-vet-stomp-squads maintain outlier-grade win:loss ratios. If this is the case (and is considered to be a problem) then squad size reduction is an obvious and tenable solution. If the past is any indication, Rattati doesn't seem to care for outliers. Or stomps, for that matter; a comment by Rattati about soccer teams comes to mind (2014.07.09 19:03, "Hotfix - Episode Charlie: Scotty must die").
First off, let me just say that all of the above is what I was hoping for with this thread. If I may, a few questions to this. Could you explain C a bit for me? I don't follow and I would like to know how 4 man squads would allow for easier matchmaking.
Absolutely.
In a perfect matchmaking world, win:loss ratios would be normally distributed around a median of 1.00. As opposing forces would always be more-or-less equal, the odds of winning a given match would always more-or-less equal the odds of losing. Perfect matchmaking is impossible to achieve, but we can look to the properties of our win:loss ratio distribution to gauge how far we're removed from the ideal.
I hypothesize in Point C that 6 man super squads are presently and consistently able to generate win:loss ratios which are far removed from the norm ("outliers"). Unless a super squad is checked by like type, its odds of winning are extremely high. The odds of Scotty matching super squads against one another are directly related to the number super squads available at the time of matchmaking. Decreasing squad size from 6 to 4 increases this pool by roughly 50%.
Gû+ Squad Size == Gû¦Total Number of Squads
Gû¦ Total Number of Squads == Gû¦ Available Squads
Gû¦ Available Squads == Gû¦ Available Super Squads
Gû¦ Available Super Squads ==Gû¦Odds of Successful Matching
Successfully matching squads of like type against one another is key to yielding fair fights, and the frequency of fair fights (i.e. matchmaking health) can be measured by the distribution of win:loss ratios. A normal distribution prettily packed around 1.00 would indicate that Scotty is doing well; asymmetry might mean that Scotty is doing less well; the presence of outliers can only mean that Scotty is not doing well. If our goal is healthy matchmaking, then we must find ways to help Scotty counteract the forces responsible for outliers.
Gû¦Odds of Successful Matching == Gû¦Odds of a Fair Fight
Gû¦Odds of a Fair Fight == Gû+Odds of Stomp
Gû+Odds of Stomp == Gû+Odds of a Recurring Stomp
Gû+Odds of a Recurring Stomp == Gû+Number of Outliers[/i]
TL;DR: Decreasing squad size gives Scotty a bigger pool from which to draw like-types when matchmaking. |
Vitantur Nothus
Nos Nothi
1258
|
Posted - 2014.12.18 05:09:00 -
[3] - Quote
I suspect that Ambush matchmaking/competition would benefit substantially from reduced squad size. And to a lesser degree Dom and Skirm. I don't see FW or PC being largely affected. |
|
|
| |
|