Paul Inouye, on 04 May 2020 - 12:52 PM, said:
Hey MechWarriors,
As I mentioned last week, I'd be playing with a few of the match making numbers on the live servers throughout the weekend and I'd share info that was discovered in doing so.
[...]
It became very clear that there was a heavy trade-off in wait times doing even minimal adjustments to the release valves and hard limits on Tier/Weight Class. Essentially it boils down to the same principle of economics:
Basically we are putting a point (set of MM values/restrictions) somewhere in the above triangle. Unless the point is dead center, there's a give and take effect happening. If that point leans toward any given corner of the triangle, the other two properties will take a hit in balance.
[...]
In the image above, I tried to find a happy medium where time to get a match, Tier/PSR matching and Weight Class matching all had their place. The combination of Tier and Weight matching is additive so no matter what the settings are set to, the global wait time is going to increase. It now comes down to how much time is acceptable to get the Tier balancing and Weight balancing having a positive impact.
What interests me is how altering the MM preference actually influences the quality of matches.
As we know, the value of each slot on a team is a function of the pilot's skill, the effectiveness of the mech (of which tonnage is only a small component), and the combination of the pilot in that given mech.
If you tweak the MM to prefer evenly matched tonnage across teams, how
much does that improve the evenness of the matches? You can look at how even the matches were when they started as you did in the OP, but what I really want to know is how evenly they ended. Or rather, how much does tonnage imbalance influence the result of a match? How strong is that influence? If tonnage imbalance to a certain degree can be shown to have little effect on the actual winner of the match, then it is not worth sacrificing MM queue time in order to achieve a perceived tonnage balance, because other factors are having a stronger influence.
Same argument for Tier balancing. When there is a tier imbalance, how much does the tier-favoured team actually come out as the victor? Is it a very strong correlation? By answering this question, you can decide how valuable it is to spend MM queue time to ensure more even Tier balancing.
What I suspect, is that within certain limits, these correlations are not extremely strong - really only readily apparent in more extreme imbalance. Match results depend on so many factors, that even with perfect tonnage and tier balancing, there is a lot of variance in whether the players perceive the matches to be even themselves, so striving for better weight and tier balance would lengthen queue times for very little benefit.
This is the type of research I began here, with limited resources available:
https://mwomercs.com...is-of-the-12-0/
Ie., are bad results (matches that ended imbalanced) correlated to some basic starting conditions?
For instance, I found that tonnage actually had a negative correlation. When a match was a stomp, it was
very slightly more likely that the
lighter tonned team was the victor. Now, my research was limited in scope, but that this result even occurred *at all* I think is significant, and begs a more thorough investigation.
And the strongest correlation I noticed was predicted by W/L ratio. When there was a stomp, it was most likely that the team who won, was already comprised of players who averaged a higher W/L ratio than the losing team. This component is not measured by PSR in its current form. Nor is a player's average match score (which is roughly a continuous record of the direct value of a player). Sure, these components influence the rate of change in a player's PSR, but PSR ultimately does not accurately measure how valuable a player for the purposes of fair matchmaking. And PSR, not being a zero-sum system, has a cap.
So among the upper echelon of players, Tier 1, there is a
humongous variance in the actual skill that these players wield, and PSR does not currently account for that variance
in any way whatsoever.
So how valuable is it, truly, to sacrifice any MM expediency in favour of more balanced PSR matching, when so many Tier 1 players are really Tier 2 players in actual skill... and so many Tier 2 players are simply people who haven't grinded enough PSR to inevitably reach Tier 1 yet?
Before we can achieve a proper balance between MM time, and quality of matches, we need to improve both our means of assessing the actual quality of a match, and our means of measuring the value of an individual player (how influential they routinely are on the outcome of a match). Until then, I daresay it would be safer to simply lean towards quicker matches.
6
