1) dunno
2) pretty accurate tbh as it takes your non-ranked MMR as base
Jerax actually went over 6k (6.2k i think) but went down to 5k and sth when he started playing other heroes. If you feel like picking invoker every ranked game i dont see why u cant as lots of ppl play only one hero even in rmm
2) ranked MMR is a mix of your stats before the ranked and 10 "tryhard games" to calibrate it, this makes it fairly accurate, the more games you play the more accurate it gets...
for the side note, I have a friend who only played Death prophet
@atum
According to your statement, then we can conclude that ten through twenty ranked MM games is not a fair representation of one's MMR. If it was accurate, Jerax would not have falled from 6.2k down to 5K.
This is especially true for a player that does not adapt his/her hero pick. If you are lucky, your hero works great with the team and even counter the other team. If you are unlucky, man your game is not winnable.
@Grimorum
noo.. he didnt start at 6.2. he got there by ONLY playing earth spirit, a hero he was really skilled with (checking his profile right now - he played over 200 ES games in a row, only playing ES, it was a HUGE boost to his MMR). when he started playing other heroes, with which he cant dominate every single game like he did with ES, he started losing and, thus, dropping in MMR. It had nothing to do with the "calibration" phase.
1. 85% solo mm q with Invoker But that doesn't really mean anything does it
2. I think it's pretty balanced seeing as how the average okay-ish person is 4200 after 10 games, and decent pro player gets around 5000 after 10 games so I think it's legit.
Main point is you can get as high as points as you want if you spam your hero, but that still does not mean that you are actually a player of that rating. Rating will never effectively tell how skilled you are unlike starcraft ladder, but it does give you a very vague general sense of how skilled you are, if you're 3000+ you're average, if you're 4000+ you're a decent player but you can still suck at times, if you're 5000+ you're pretty good, and if you're 6000+ you pro-player level.
4500 sounds pretty good for only invoker since you did not adjust to counter pick, etc.
However, I just realized there are more variables. Atum mentioned that our non ranked games are used as base.
So unless someone plays all ranked and non ranked games with only invoker, the MMR is still inaccurate.
u can get ur base mmr MINUS 0-1500 to PLUS 0-1500
for the 10 games. its mostly averaging at 750. which means if u rly do good and win 10 games u can get +750 from ur base mmr. if u happen to have unfair games where enemy overall mmr is way higher than urs and u win, u could get +150 a game.
I'm running an invo smurf, solo only, ap/euw, no lpq abuse or anything. Instalock of course, I only left when s1 picked it faster or called mid. Reached level 13 and started ranked. 5400 after calibration (6-4), now 5448. Aiming for 6k. That one guy in dotabuff friendlist was totally random, player pool is small up there in non ranked.
http://dotabuff.com/players/126943640
http://i.imgur.com/j7anh0E.jpg
1) Has anyone done every solo queue ranked mm game with only Invoker?'
i got like 300 mmr by picking sf only
2) How reliable is ranked MMR with only ten games? According to the law of large numbers, ten does not seem adequate at all.
your rating gets imported from unranked after those ten games
@wave
Valve has never told players the details, but as I understand it, when a new player makes and account and starts stomping hard, smurf detection kicks in and then shoots the new account extremely high up in ranking, after which most players normally come crashing down. Mathismight however, kept winning, and so he stayed high.
There also might be some differences with normal matchmakings MMR system. It is less strict, and may reward or detract more than the average 25 points per game.
so where a player of my skill level would probably cap off out somewhere on the lower end of the old "high" rating, A player like you or math would stay at the top of the graph.
Following that up. The "ten games" of calibration is completely arbitrary from what i understand. It is simply close to the number of games it will take for you unpredictability variable in the MMR variable to shrink from its originally inflated variable. There will always be an predictability variable for each match, but when you start your first ten matches, that value is arbitrarily inflated by about 100 points and slowly decreases.
Valve is basically using a modified version of Microsoft's trueskill system http://en.wikipedia.org/wiki/TrueSkill
Here is a better explanation someone else gave at joindota:
"I won all 10 of my initial ranked games and my MMR is 3000 while my friend lost all his games and has a MMR of 4000. WTF Volvo???"
Your 10 calibration matches don't actually determine your MMR as valve has lead you to believe. Your initial MMR is already pre-determined by your mean skill in unranked play. This is why you were not matched against completely new players in your first game. Your rating does not start from 0 and jump up until it hits your true rating.
You start where you left off in unranked but your uncertainty rating is reset to very high. Your rating changes more than it normally would in unranked but not so much after 10 games that it would make a big difference to your rating. If you were a 2500 rating player and you won your 10 matches, you would be 3000 but your friend may have been a 4500 rating player so he dropped to 4000 after losing 10 matches.
I just won my game, have the highest GPM and most kills in the game, why did my MMR go down? This is so broken!!!!
Firstly, GPM and kills will not have any influence on rating. The system only determines whether you win or lose not by how much.
MMR can drop after a win or increase after a loss and here's why. If you remember how MMR is calculated from your mean skill rating: MMR = mean - (3 x uncertainty). Mean always goes up after a win and goes down after a loss. In a game where the system heavily favors you to win and you do win, your uncertainty drops very little and you gain very little points. If you recall, the game adds some uncertainty to your rating before each game to stop ratings from stagnating. In such games, the amount your uncertainty drops after the game is not enough to offset the extra uncertainty the system applies to you before the game resulting in your uncertainty actually going up and in conjunction with only a very slight increase in mean rating, ultimately results in a net decrease in MMR.
Let's look at an example. My mean rating is 6000 and my uncertainty is 500 so my MMR would be 4500. I just finished a game where I am heavily favored to win. My team gains 300 points to be distributed between our players. The system adds 50 uncertainty before the game and my uncertainty drops to 480 so I am at 530. My team's total uncertainty is 2000 so I get (530/2000)x300=about 80 points. My new MMR would be 6080 - (3x 530) = 4490. A net decrease of 10 MMR.
The same can be said for an increase in MMR after a loss. When the system thinks you have a high chance to lose the game, your mean loss is small and if you get a large decrease in uncertainty, it can offset that small mean loss and even result in a net gain. This generally only occurs when you first start ranked while uncertainty is high and is essentially a failsafe against new players getting stomped and losing a bunch of their rating points."
Later, he goes on to speculate that individual stats may play a role, but there is no way to confirm it.
no thats how smurf detection worked before ranked mm was released
there were constantly 1-2 win players on front page first game and they had 100~ KDA
@ Neah
1- I don't pretend to know the details behind the smurf protection algorithm. Valve keeps that under wraps and abstract explanations and estimations are the best available.
2- Mathismight didn't play "2-3 games with shitty stats" He played 265 or so games in normal and retained a 68%+ win rate.
@Wave
Why do you believe anti-smurf has changed since ranked was released? A new player will still have to play normal non ranked games until they reach the level minimum for ranked. Anti-smurf will take effect long before a real smurf hits ranked. Because his starting ranked MMR is based off his normally MMR with an inflated predictability variable, anti-smurf should have played a role. Did valve announce any changes in their anti-smurf algorithm?
pick invoker
get ganked by 5 ppls minute 1
get ganked by 3 ppls minute 3
perma 5 mid minute 7
ur team comes mid and feeds to
game over at minute 15
this is how reliable invoker in rmm
ops i meant it still works like it did before ranked, however i think kda is based on the early stages of rating
The matching system was actually changed when they added ranked matching. They are clearly using an ELO type system now, not TSR like before.
You can tell that it is ELO for ranked matches now because you do not gain or lose accelerating amounts of MMR for win streaks like you did way back when the skill bracket was public information. Now that they are using ELO any deviation from +/- 25 points per game indicates that the average MMR for the teams was not balanced. Usually this only happens at the highest and lowest MMRs.
TSR is now only used for anti-smurf when you begin an account and the calibration when you begin ranked.
@Mathismight
I am very impressed! You did all of this while ignoring team composition and laning. Since the release of v6.80 patch,
did your MMR increase or decrease for the solo invoker account?
Sadly, the other popular heroes I enjoy were underwhelming like Huskar 6.77, Medusa, and Razor. Embarrassingly, I have to admit that my OD was complete crap too because I was following a 6.77 OD guide and I never questioned or thought carefully about the spells.
Fortunately, Huskar received a big buff and I have been doing 2x better with him recently. I played Medusa and Huskar far too many times. For now, I am 5000 and expecting to climb higher in solo queue only.
Grimorum unless you update your profile and let us see where you really stand, on every of your profiles, you are full of shit for me. And i also bet your trash disciple gave up on you once you taught him all that he wanted to know... But you didn't see that coming, did you...? Newb, amateur coach...
@Don Jon aka iinvoke
You need to stop harassing/cussing/disrespecting me and other members. It is quite obnoxious and despicable.
Calm down Grim, that's not even the same Don John you're thinking about lol. Not everybody makes 20 different accounts.
კომენტარის დასაწერად გაიარეთ ავტორიზაცია.
1) Has anyone done every solo queue ranked mm game with only Invoker?
2) How reliable is ranked MMR with only ten games? According to the law of large numbers, ten does not seem adequate at all.
I think 100 ranked MMR would be fairly accurate.
On a side note, has anyone done every solo queue ranked MM game with only one hero? I heard Jerax did it with earth spirit and end up being nearly 6000 MMR.