The more you lose, the higher the chances of winning the next game, which is why it is extremely rare to lose more than 10 times in a row unless you purposely make mistakes. Calibration does not just take into account your win rate, it also looks at the pentagon in your stats, as well as how often you play the game. If you reached high skill, then chances are you will be in legend, if not it is archon. Actually dont bother too much about your rank. If you are really good then you will naturally climb.
the more you lose the higher chance of winning. Check my recent games @dreadnought325 then :D
6 losses - 1 Win - 6 losses....I don't really like the "The more you lose, the higher the chances of winning the next game". xD
it is true though, the games get easier because your rating gets lower so the skill level of everyone (including your enemies) gets lower. assuming your play is consistent, lower mmr average makes the game easier, regardless of whether you're on the higher or lower rated team.
individual streaks are bound to happen in a large enough sample size, but when you look at the big picture they are meaningless and unnoticable.
if you flip a coin a million times you'd probably expect one side landing 10 times in a row somewhere, and still end up close to 50% at the end of it. to expect it to alternate heads/tails (win - loss in this case) every single time would be less likely.
back on topic, unranked games still have mmr, it's just hidden whereas ranked it is displayed. during calibration your skill uncertainty is higher which means the mmr change from the winning or losing is higher than in normal games in order to help you reach your skill level in fewer games.
in regular ranked games a game gives +/- 30 mmr if you're solo and +/- 20 mmr if you're in a party. during calibration these values are higher to make it quicker to adjust to your real rank, but i believe gets lower with each game based on relative performance (and other factors that valve doesn't want to disclose), so that if games are stomps that usually means the skill disparity is high so someone isn't in the correct skill group and if your performance is expected for other players of that level then that should mean you're close to your actual rating
If I am wrong then quit dota. Since the more you lose, the higher chance of you losing the next game, then you must either improve your skills exponentially or drop all the way to archon.
@Bambooni, in your centaur game https://www.dotabuff.com/matches/5543624734 you could have won. I am not sure if you are pos3 or 4, but the item build focused too heavily on mobility. You should have gotten some strength item for you skill set and survivability. In case you did not read centaur's skills, 3 of his skills require strength to deal damage. Force staff and blink will not save you from PL, but a halberd or even an armlet will. Your team actually have sufficient nukes to deal with PL, so being tanky is key to giving your teammates enough time to land their spells. You are supposed to tank but simply did not.
Please sign in to post comments.
Ive had this account for a number of years now as just a casual account.
Always used to play really casual and bad in earlier years.
Also worth noting, never really got into ranked since there is no ranked servers in my country. Id classify myself as a sub decent player. (at least archon 5+) not really some divine player. But I have an atrocious win rate, (like 49%)
I do however get quite abit of high skill games, even when solo queing unranked. Id say atleast 60% of my games are high skill (when I dont party Q with my Herald mates from school) in the past month or so.
If I were to try and calibrate. How does dota sense your skill group? Do they take your all time skill and pop it in accordingly. Or do they take your recent skill level?
thanks in advance