South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
How does one accurately measure performance in a game like dota?Thought it was based on performance. Same as in primary school if your friends fail you can still pass...
I think the standard is 25MMR for a win actually. I think it may be like the elo system from csgo and chess where the points you receive are based on the relative skill levels between the 2 teams. (i.e. you lose less and win more in games against teams with superior average MMR and vice versa)Gpm, experience, kills, deaths, Ward purchasing, Tower damage, hero damage... Dota is basically a world of statistics for each player. How do they work out MMR based on a win? Dire wins everyone on Dire gets 3 more MMR points?
Fair enough. I think the answer to your question then is that it is simply too complicated to create a system like that. The metrics you've mentioned are either not reliably measurable or they don't apply to all roles in a dota team.I didn't propose anything. I asked. You then said how would I work i said there are measurable elements.
I still really dig playing with you because you don't rage.For the most part during calibration, winning isn't everything, you can lose all your 10 games and still get a decent MMR.
After that you only gain MMR by winning, ranging between 15-30 MMR per match (those are the figures I've gotten in the past) But mostly it's 24-25MMR
No idea how mmr gets calibrated, couldn't really care either.
Amped to play some dokta this weekend, will probably catch a few of you in-game since we all pretty much have similar MMR.
For those that don't know yet, my dota handle is boerEsports (fighting hamster pic)
I still really dig playing with you because you don't rage.
Now with my rubbish MMR when we party up it isn't going to throw us into those High Skill games any longer where I feel bad for you when you struggle (I've been in those games where you know you're just way out-skilled by your opponents and it isn't fun).(although you've kicked butt before in those games too which is great
)
But as I said we can avoid that now in ranked because I'm down in the trenches now as well![]()
Actually when I first started dota I used to play with a mate in the US Navy. I did all my party MMR calibration games on US servers around the time I did the majority of my solo calibration games back when I first started dota. My party MMR is also 1900 and some change. So even there I'm in mashed potato bracketHaha yeah man, I very rarely rage as it serves no purpose, rather change your strategy and try finding a solution once things go south.
Would be fun getting some party mmr as I'm sure most of us still require those 10 calibration games.
P4L ... the clan... the legend...Potatoes-4-life unite![]()
Sweet Jesus. No wonder I'm at the arse end of the trench.Read something interesting about mmr calibration.
'if you want max mmr calib and have decided to reach 5k+ one day don't calibrate until you have at least good 10 games into the very high skill bracket. if you look at your matches on dotabuff they will each be labeled with normal, high or very high skill, normal is something like less than 3.2k base mmr, and very high is everything over 3.8 mmr and high is everything in between. once you get high base mmr calibration will just add or remove mmr from it based on your 10 calibration games. stats that most influence your mmr before calibration is over are hero damage per minute>hero healing per minute>k:d:a>structure dmg>gpm/xpm>last hits and denies, the system compares them to average hero picked stats and average team stats in that game. max mmr you can calibrate at is 4999 and lowest is 1.'
Other things I read is get over 240 creep kills per game and over 530gpm.
During your calibration (first 10 games) you can gain up to 500mmr per game.