Gorefiend, Ryan, Brooklyn Kurtz, Eater Ghost, Great Deep Soul, lol I love google translation.
^wo* not 'wor'.
Actually I'm kinda interested - what translator actually told you to type 'wor'?
EDIT: Holy shit, if the stuff that site provides is legit, I'm impressed! Can someone who can read Chinese tell us how they got skill tier information? I saw that they related MMR brackets to skill tiers, but if that was how they got skill tiers, how did they obtain MMRs?
No, it shows skill tiers of all recent matches. Maybe they really do search every so often matches by skill tier and record every single match? It's the only way I can think of.
At least now I know my high wr is 52% and my normal and low wr are below 40% now I understand what's going on...
Q: 什么是Normal, High, Very High的比赛？Dotamax如何得到的数据？
A: 每个dota2 的账号都隐藏了一个匹配值Match Making Rating(MMR)，后来给出天梯分基本就是MMR的一个直接体现，当玩家选择个人(组队)寻找比赛时，dota2会根据你(你队伍)的MMR(平均MMR)匹配其他正在寻找比赛的玩家(队伍)。Skill等级: Normal(普通), High(高水平), Very High(非常高), 是大致放映MMR区间的一种表示方式，在9月的一次更新后Value屏蔽了使用玩家数字ID得到Skill等级的办法，但并没有取消Skill的获得和计算，Dotamax通过不断的轮询所有的比赛而不使用用户的数字ID仍然可以得到Skill,也就是说Dotamax得数据来自官方API,对于有些玩家的质疑，我们表示理解并支持给我们指出错误，但是也不希望因为少数人的误解让大家怀疑我们数据的真实性，希望大家可以自己去分辨对错而不是人云亦云。事实胜于雄辩，其实大家可以对照下你得SKILL等级和天梯分区间的关系。。SKill等级是在比赛前就确定的，目前的经验判断 Normal:大约是天梯分(0 ~ 3200), High: 大约是天梯(3200 ~ 3800), Very High: 大约是天梯分( >3800)。
SHUUUURE
Q: What are the Normal, High, Very High matches? How does Dotamax get these statistics?
A: Every dota2 match id hides a MMR, which formed the basis for Ranked Matchmaking. When a player queues for a match, Dota 2 matches the game according to his MMR or his party's average MMR. The various brackets are Normal, High and Very High, and was a way of showing the average MMR of a match. Although Valve eventually released Ranked Matchmaking, they did not remove the calculations bla bla bla
basically api still holds data for the skill brackets, normal 0-3200 high 3200-3800 very high over 3800 on average
they APPROXIMATED normal/high/very high by mmr .normal 0-3200 high 3200-3800 very high over 3800
the cutoff are arbitary. not from valve. THEY MADE IT THAT WAY. not the valve normal/high/veryhigh.
I thought they said the skill tiers are approximately those brackets, not that they 'derived' skill tiers from those cutoffs. I've started going back after every game and checking the skill tiers and they seem to be accurate for me? Maybe Relentless was right and they decreased the threshold requirement for Very High.
I think Valve's Very high was ~3900. I had solo games on Very High and I'm pretty sure I've never been a 4500 player. Currently I'm ~3500, so a 3900-4k spike would sound a bit more realistic.
Normal game (including dropped) (8.38%)
Matches: 134 Wins: 48.51% KDA: 3.0
High game (38.09%)
Matches: 609 Wins: 56.98% KDA: 3.2
Very High game (52.28%)
Matches: 836 Wins: 60.05% KDA: 3.1
Professional game (1.25%)
Matches: 20 Wins: 85.00% KDA: 4.8
Ladder Competition
Matches: 139 Wins: 60.43% KDA: 3.3
Normal ladder match (0.00%)
Matches: 0 Wins: 0.00% KDA: 0.0
High ladder match (6.47%)
Screenings: 9 Wins: 66.67% KDA: 3.2
Very High ladder match (93.53%)
Matches: 130 Wins: 60.00% KDA: 3.3
Streak: 20 games
Losing streak: 7 games
very cool, but what's ladder match? it doesn't look like it means ranked
lol. I am more likely to win high/veryhigh than normal. I should show this to my noobish stack buddies.
Normal 天梯比赛 (36.84%)
场次: 105 胜率: 42.86% KDA: 3.1
High 天梯比赛 (39.30%)
场次: 112 胜率: 56.25% KDA: 3.0
Very High 天梯比赛 (23.86%)
场次: 68 胜率: 50.00% KDA: 2.4
hi everyone, next time ask me to translate any chinese for you.
google translate is unreliable like dota match making.
I am your friendly chinese neighbor. lol
How do they know the average mmr of each ranked game? Or how do they know if a game was normal, high or very high? Apparently they are quite accurate.
Zaloguj się aby dodawać komentarze.
Chinese make a wesite, DOTAMAX.com
i think U guys can learn something from it