Found this interview about the release with Dustin Browder on Eurogamers.
Some highlights: * 3-4 week hero release cycle (good start with Johanna already I guess) * ~2 month patch cycle * Enhanced MMR taking in gameplay into account * Apparently the unit is already sustainable with the current playerbase / spending levels (but it's hard to tell if that's exactly what he meant)
kudos to the interviewer for having the balls to ask "so how long will this game last?".
i've heard Force and a few other of these gaming journalist luminaries comment that the player/viewer base is very small... so it was a tough question that had to be asked.
this ain't EA we're dealing with here though..if Blizz says they're committed long term then the chance they're packing it in within a year is zero.
The mention of enhanced MMR taking gameplay into account is a concern for me. I really don't know what objective metric you can use to determine player performance. Damage, Healing, EXP, deaths, etc. are all contextual in nature. I fear a situation where the MMR system incentivizes players to play the game suboptimally, which I think would quickly become the case if gameplay MMR adjustments are implemented poorly.
On June 04 2015 23:37 Darthsanta13 wrote: The mention of enhanced MMR taking gameplay into account is a concern for me. I really don't know what objective metric you can use to determine player performance. Damage, Healing, EXP, deaths, etc. are all contextual in nature. I fear a situation where the MMR system incentivizes players to play the game suboptimally, which I think would quickly become the case if gameplay MMR adjustments are implemented poorly.
I have a hard time believe the metrics used will be anything you see on the score screen (not even takedowns since they seem to like the idea of split pushing heroes, but it's unlikely to be anything made public due to the incentivization issue you mentioned, so we'll just have to speculating.
What I found interesting is that it's kind of a tautological task isn't it? We want to identify skill beyond win-loss so we're going to look for common metrics shared by high skilled players... which we can only identify by win-loss. In any case I'm sure they'll be iterating on it for a while.
On June 04 2015 23:37 Darthsanta13 wrote: The mention of enhanced MMR taking gameplay into account is a concern for me. I really don't know what objective metric you can use to determine player performance. Damage, Healing, EXP, deaths, etc. are all contextual in nature. I fear a situation where the MMR system incentivizes players to play the game suboptimally, which I think would quickly become the case if gameplay MMR adjustments are implemented poorly.
Yeah like as Raynor I can play super safe and augment my hero damage by just staying on the back lines and punching bullets into the tank but ideally I need to try and take some risks and put some damage into their damage dealers. This results in more deaths and lower hero damage but can ultimately decide if the game is a win or lose. It would be a shame if the game decided that since I didn't die a bunch and I had good hero damage this loss must not have been my fault and awards me less negative MMR than my team mates. Or similarly if I get carried in the same fashion and I gain more MMR despite having an actual negligible effect on the outcome.
They really should just make it binary. You either won the game or you lost the game. None of this in between stuff.
The problem is that if you want a meaningful individual ranking then you need to take account individual skill. Right now, one person alone may be the best player ever but if his team mates are shit he's going to lose, in similar fashion you may be average but if you queue with another 4 good players your rating will be above average. So right now the ranking could or could not reflect your actual skill, and that makes it meaningless to me.
But I know taking into account specific individual metric can be problematic.
On June 04 2015 23:37 Darthsanta13 wrote: The mention of enhanced MMR taking gameplay into account is a concern for me. I really don't know what objective metric you can use to determine player performance. Damage, Healing, EXP, deaths, etc. are all contextual in nature. I fear a situation where the MMR system incentivizes players to play the game suboptimally, which I think would quickly become the case if gameplay MMR adjustments are implemented poorly.
I have a hard time believe the metrics used will be anything you see on the score screen (not even takedowns since they seem to like the idea of split pushing heroes, but it's unlikely to be anything made public due to the incentivization issue you mentioned, so we'll just have to speculating.
What I found interesting is that it's kind of a tautological task isn't it? We want to identify skill beyond win-loss so we're going to look for common metrics shared by high skilled players... which we can only identify by win-loss. In any case I'm sure they'll be iterating on it for a while.
That's probably true. I just honestly can't think of a single objective metric that would accurately be able to suss out individual contributions. In any case, I wonder how big of an issue this really is. On a long enough timescale, everyone is going to get pretty close to their "true" MMR no matter what.
On June 05 2015 04:53 [Phantom] wrote: The problem is that if you want a meaningful individual ranking then you need to take account individual skill. Right now, one person alone may be the best player ever but if his team mates are shit he's going to lose, in similar fashion you may be average but if you queue with another 4 good players your rating will be above average. So right now the ranking could or could not reflect your actual skill, and that makes it meaningless to me.
But I know taking into account specific individual metric can be problematic.
You're right that over a sample of a few games, win rate isn't an accurate indicator of personal skill. However, over a large enough sample of games, the only true constant is you (and anyone you choose to play with) and just by the law of large numbers, your unlucky games are going to be supplemented by games where you are lucky to have won. The problem is that people are less likely to notice games where their team is particularly good, or the opponent's team is particularly bad, and so many people have this conception that they're only being held down by their teammates.
Note that I'm not implying that people don't get matched with terrible people and lose games they should be winning, just that that happens in equal measure to people and their opponents. In any case, that's more an issue with the match making system itself than individual MMRs.