News Realms Remove Ads - Go Premium! Sign In or Register

Bluetracker

Tracks Blizzard employees across various accounts.

This mornings meta snapshot. 10 games, 8 Demon Hunters. Can't recommend playing right now 😢 https://t.co/kGdrqRgtmZ

This mornings meta snapshot. 10 games, 8 Demon Hunters. Can't recommend playing right now 😢 https://t.co/kGdrqRgtmZ

  • Savjz
    Posted 3 years, 9 months ago (Source)
    This mornings meta snapshot. 10 games, 8 Demon Hunters. Can't recommend playing right now 😢 https://t.co/kGdrqRgtmZ

    • Savjz
      Posted 3 years, 9 months ago (Source)
      Yes obviously a small sample size and not an accurate representation etc. That is obvious. Just showing this *can* happen, and there for real are *MANY* DHs. for more accurate data I recommend @HSReplayNet

      • RegisKillbin
        Posted 3 years, 9 months ago (Source)
        @Savjz @HSReplayNet Anyone citing "small sample size" against you doesn't understand people experience games as individuals, not as data sets. When these bad experiences happen to an everyday player, they don't go to HSReplay and review the data first. They close the app and don't play for a month.

        • Chadd Nervigg
          Posted 3 years, 9 months ago (Source)
          @RegisKillbin @Savjz @HSReplayNet Well, I think the point of the small sample size comment is that it may not be an *avoidable* problem. I don't think anyone argues that facing 8/10 DHs in a row is good, just that with a large playerbase, even with a perfect class popularity split, this will happen.

          • Chadd Nervigg
            Posted 3 years, 9 months ago (Source)
            @RegisKillbin @Savjz @HSReplayNet But I'd be curious what people think about... what if it *was* avoidable? What if we made some rule in the matchmaker that if you've played against the same class twice+ in a row*, start decreasing the odds of facing that class? (Not in development, merely a thought experiment.)

            • Chadd Nervigg
              Posted 3 years, 9 months ago (Source)
              @RegisKillbin @Savjz @HSReplayNet *and you haven't changed your deck, so you can't try to take advantage of this by counterqueueing. (Of course, if you're playing the counter to it, you might WANT to play against that class 8/10 times, and this is a downgrade for you.)

            • jcandance
              Posted 3 years, 9 months ago (Source)
              @Celestalon @RegisKillbin @Savjz @HSReplayNet would this be an opt in experience, or part of the algorithm for everyone? It probably works better then the "hey you get to ban a class", and I'd be for it. It'd be even better if it wasn't just one game, but say you had to see two or three different classes before repeat

              • Chadd Nervigg
                Posted 3 years, 9 months ago (Source)
                @jcandance @RegisKillbin @Savjz @HSReplayNet Maybe, who knows, it's just an idle thought at this point. The thing I dislike about this is if there ends up being any sort of metagaming where you should counterqueue or play differently based on what you've played against in the last few games.

            • Danny_Jackpot
              Posted 3 years, 9 months ago (Source)
              @Celestalon @RegisKillbin @Savjz @HSReplayNet I would not be a fan of this. Especially at high legend.

              • Chadd Nervigg
                Posted 3 years, 9 months ago (Source)
                @Danny_Jackpot @RegisKillbin @Savjz @HSReplayNet Yeah, it's possible that this is a bad thing for Legend players where the sanctity of random matchmaking is important, and a good thing for everyone else.

            • RickMKick
              Posted 3 years, 9 months ago (Source)
              @Celestalon @RegisKillbin @Savjz @HSReplayNet That would add some nice variance. In conjunction with that, increasing the odds of facing an underrepresented class would add a bit more. Would that hurt the players of the popular class though? I imagine that would increase the odds of a mirror match. Might be too frequent.

              • Chadd Nervigg
                Posted 3 years, 9 months ago (Source)
                @RickMKick @RegisKillbin @Savjz @HSReplayNet Yeah, I'm not sure. I think that ultimately the same number of players of each class end up queued up and matchmade, so over a long duration, your % of opponents by class stays the same. This would just smooth the streaks out, I think? Matchmaking is super complex, not positive.
Tweet