Can Blizzard employees exploit matchmaker insider information in order to get higher rankings and/or better quality matchups? And if so, does this suggest that there would be a bias among Blizzard employees against fixing such problems with the matchmaker?
For example, if you knew that the MMR calculator favored certain specific behaviors, then if you were a Blizzard employee playing Overwatch, presumably you would consciously try to match these behaviors, whereas those of us without the inside information would just be trying to help our team win. In many situations it’s not either-or: you can help the team win, via engaging in these specific behaviors. This would seem to mean that those with inside information could achieve a higher ranking and better quality matches than those without it – they would know they got this personal benefit, and would naturally be biased against fixing the problem.
they used to do this then stopped. around season 3 they use to derank you at the beginning of the season, just so you could climb. they admitted it was bad and removed it.
they could be fiddling with match maker, but orange posts suggest otherwise, but they never told us about the deranking thing, so who knows
My friend, this is 100% correct. The people who know how the matchmaker really truly work understand that it is performance base, and they know what metrics to pay attention to and can use this to rank up. I do not know all metrics but I know the important ones and I have used it to rank up to master and can get gm if I like but it is so boring to play for metrics.
PBSR is not a thing beyond Plat. And all it does is make climbing easier for people who are obviously higher than their rank so that they can get to their proper SR faster which results in more balanced games for everyone.
Can they? Possibly, depends on what the matchmaker and scoring systems do. If it’s a deep learning AI trained on the entirety of the player base, like I suspect, then probably not.
Even if they can, we’d never know.
If they can and do, something tells me Jeff would be higher than Platinum.
Why not? You’d know the inputs to the AI and have total liberty to analyze how the AI was processing those inputs. Seems like it should be quite straightforward to easily find a general idea of what the AI had deemed to be important and somewhat more complicated to get a precise idea, but certainly very possible.
I know, it seems that way - but it’s not so. I’m a professional software engineer, so I’m not speaking from ignorance, either. There are many documented instances where deep learning algorithms came up with solutions and analyses that are equal parts extremely counter-intuitive and extremely effective / accurate.
I’ve worked with AI for 20 years now. If you want to believe that you can reliably extrapolate behavior of a deep-learning algorithm based on its inputs, then more power to you.
How does what make me feel? I’m saying it makes sense that you work in the AI field, given the kind of misguided posts I’ve seen you make. You wag your credentials as if we should all be impressed, and maybe those who have only a passing familiarity with the field would be, but for someone like me, who has a lot more knowledge of the subject, I know better.
Also, don’t wag your credentials in front of people in lieu of making good arguments, it makes you look even worse than your bad argumentation already did make you look.
