GAC Matchup Discrepency

"DarjeloSalas;c-1997944" wrote:

"BeralCator;c-1997899" wrote:

"DarjeloSalas;c-1997881" wrote:

"BeralCator;c-1997872" wrote:

"DarjeloSalas;c-1997846" wrote:

"dimi4a;c-1997839" wrote:

"Fanatic;c-1997806" wrote:

"dimi4a;c-1997792" wrote:

"Legend91;c-1997790" wrote:
The final round is where the (mostly) successful people are matched against each other so it's harder to win there.
And I also thought we've passed the "oh my god my opponent has xxxk more GP" time. GP is meaningless.

Well tell that to the Devs!

It's meaningless in terms of whether player A can beat player B. And it's meaningless in terms of the total GP differences players always point out, because GAC isn't based on total GP, it is based on top X toons in a players roster.

For the devs they need SOME number by which to calculate who is 'equivalent' to rank together. That is what GP is. Personally I'd like to see them move to a ladder ranking like most games use that is based on win/loss ratios (and which bracket you are in). But they do need to use a number. They chose GP. They could have chosen total zeta's applied, or total g12 characters, or any other arbitrary number. None of those would be meaningfully better than the current top X GP. They could try and weight different characters differently, but that is also a rather arbitrary assignment with which the player base (as a whole) would not agree with - and those character weights would need to be re-evaluated with every new character release and how they interact with existing toons.

But it's meaningless as in not a balanced and fair way to pit players against each other. GP doesn't take into account character abilities and synergies.

And what would you suggest as an alternative?

Without entirely blowing up the current GP calculation (which is somewhat impractical as other game modes are based on it)?

You could add an aging algorithm to each character that devalues their GP by an exponential amount for every quarter since their kit was last touched. The function could largely filter out launch toons like CUP, Ugnaught, Tuskens, etc. while preserving most of the value of anything after the introduction of zetas. I think we'd all agree that new characters have more utility and synergy, and that the adjusted health and protection and G13 finisher stats for older characters didn't really alter the landscape much.

There are only a handful of untouched kits that still play OK: Clone Sergeant, Ewok Elder, First Order Officer, Snowtrooper, and Resistance Trooper. They are already low GP anyway because they have only 3 abilities and generally don't show up in people's top X.

When was Thrawn’s kit last touched? Nest? They’re pretty good and are exactly the same as they were when released.

An adaptive matchmaking algorithm would be incredible, but let’s be honest with ourselves: there is absolutely no way the devs are up to the task of implementing and maintaining such a system effectively.

Thrawn was 6/13/2017, and Nest was 6/1/2018. Zetas were introduced near the end of 2016. I'm thinking of a curve where anything from 1/1/2017 forward is >75% value and approaching 100% as you move towards today, and anything before 1/1/2017 is approaching 50% or less as you get to the start date of the game.

There are obviously outliers like Nest and Thrawn, but I was trying to quickly suggest something that could be a single equation and would not yield nonsense results. There's never going to be a perfect system, but it's depressing that many of us can come up with a superior solution in under 20 minutes.

Another issue with your proposal is that the introduction of a new toon can often make an older, untouched toon suddenly become much more valuable. Think of the impact Rangetrooper has on an IT team.

Another, more niche, example is that some people are getting mileage out of Clone Wars Chewy with Vandor Chewy, as his lead has a good chance to buff Vandor without him doing anything, making it very hard for him to not be Prepared.

I know you came up with it quickly, and I appreciate that someone chiming in with these “what ifs” must be annoying, but I’m not convinced we can categorically state that this is a superior solution to GP alone just yet.

Also, I’ve not seen much, if any evidence, of “so many of us” coming up with any solution at all, never mind a superior one.

Clone Wars Chewbacca and Veers have had their kit adjusted with zeta additions. Only Snowtrooper and possibly Shoretrooper would predate my point of exponential devaluing (Starck and Death Trooper came later, both well into 2017) and as 3-ability characters they are less impacted by an aging algorithm.

Look, our current system states that a G8 Bodhi Rook has half the value of a G12 zzzRevan. I'd argue that virtually any idea is a superior one at this point.

I realize any solutions that are posited aren't perfect and that they may have edge cases and outliers, but that doesn't mean they aren't better the current matchmaking; it just means that they may require subsequent revision as new flaws are discovered. Good matchmaking is an iterative process and requires lots of A/B testing to get to an ideal state; doing nothing until you have the perfect solution isn't how science works.

I feel like the CG solution is doing nothing at all, and to be fair things may be satisfactory as they are; I don't have the data detailing how many matches are competitive. It may be that 95% of them are good, and that fixing the remaining 5% isn't worth the time or effort involved in devising a more adaptive and complicated system.