IP ban south america and SEA( i rather get 4 peruvians than a single asian) from US servers
Reveal MMR during picking stage.
Taper the amount of MMR gained/loss throughout team. Blue/Pink (higher ranked) player should receive less MMR than Orange/Brown (lowest ranked) player.
Yeah, the color thing was speculation of the highest stupidity based on a few random occurrences where in solo queue the highest would be the blue player.
I know the color system doesn't match up with the MMR, but I'm saying it should so teams can see who the best player generally is.
^and then they pull a HoN and the lowest mmr MUST play support even if they are better at a different role?
There is a reason why you can only see mmr AFTER this game
1.) Reveal MMR to teammates (I once had a 4.3k MMR player insist on going mid over a 6k MMR player. The 5.7K MMR player on the other team just fed off of him)
2.) Make MMR more stable (have an uncertainty and gradually take the +/- to 1 point a game as your local win rate approaches 50/50, increase +/- as you start having streaks. This effect would be on top of the current MMR judging system where games are worth more/less depending on the "chance" of winning). It's really ridiculous for a player who is typically around some level to have a few days where he is -500 or -300 from that level just because he had a bad couple of days full of losses only to grind right back up to where he was.
3.) Reward support players something like 10% more MMR in a game they win if they play support and buy plenty of support items.
4.) Take away the mute system. Stop being little pussies and reporting people who are "doing the wrong thing" or who are "raging on mic". I understand a mute for someone who is actually griefing with his microphone like loud or disturbing noises, but rage is part of gaming. Use the local mute for non-offensive ragers.
5.) Have reports manually processed with stronger punishment.
6.) At least tell people the average MMR of a game so they can know what to expect. (e.g. say you're a 5.2k MMR player. Sometimes you get 4 other 5k mmr players, so supporting is fine. Other times, you get a bunch of 4.5k MMR players. You support, and you'll probably lose.)
7.) Code up a bot that searches match text for "gg just end" "i won't defend" "it's over" or "fuck mid" and instantly ban them permanently from the dota network.
Oh please it did, 80+/92 solo queue games as Blue, that wasn't random. It was based on queue times, which was in turn highly co-related with MMR and only relevant above 5k. C'mon Havoc, I expected you to be better than this... (the 5 games I weren't blue I got Winter/Chains/Meracle etc on my team). As a bonus when I tried re-starting the queue every time it exceeded 2 minutes I no longer got Blue when the system was still working that way (my average queue time is around 4+ minutes).
@Vandal - for me a big part of solo queuing pubs is checking the live game pages. Page 1 is on average 5.2-5.3k MMR and above, page 9-10 tends to be around 5k MMR and page 15 around 4.8k... you do the math. I've taken to checking the leaderboards for players I meet to know whether or not I should support or carry.
I think mmr can be great if:
-First wined game --> +10
-2nd --> +15
-3rd --> +22
-4th --> +31
-5th --> +43
-6th --> +61
-7th --> +75
-8th --> +90
-9th --> +90
bis
Same with lose streak but without limits. (No more buyed acc)
Report the intentional feeders, I intentional feed like a 4 ranked games and no 1 get a fucking report. And if u get a intentional feeder get back the 25 points when the report be succesfull.
If a player pick always slark or tb or whatever get bened this heros for a week or something similar.
And detection to see suport builds and carry builds (in 1-4.5 mmr bracket), if u play always carry get less mmr than the player who change the role to play.
Try to put your options to improve the mmr, i will put it on the steam petitions, i know that it will not succed but....
I KNOW MY ENGLISH SUCKS
and this is why valve does not listen to the community, because they are fucking braindead.
one very simple and small change:
remove the impact of factors such as kda, hd and gpm/xpm during the calibration proces (make it win/loss dependant only)
there is no reason as to why calibration matches should be different than normal ranked matches
and it heavily punishes support players
support players have = impact on the match as carry/mid players. Giving support players a bonus is a retarded idea. The problem is most people try and copy what support players do in pro matches which do not translate to pub play so it looks like they have 0 impact
@just dont feed
http://dotabuff.com/matches/597108361
when you have a 5.6k support and a 5k carry
abaddon hero damage: 12.4k
lifestealer hero damage: 5.4k
supports fucking win games if you have one or two competent teammates who can capitalize off it (in this game my mid/off).
Put 4 stacks back in ranked so I don't have to play with/againced stupid super tryhard 4 stacks in unranked
@feed
"Report the intentional feeders, I intentional feed like a 4 ranked games and no 1 get a fucking report. And if u get a intentional feeder get back the 25 points when the report be succesfull.
If a player pick always slark or tb or whatever get bened this heros for a week or something similar.
And detection to see suport builds and carry builds (in 1-4.5 mmr bracket), if u play always carry get less mmr than the player who change the role to play."
Banning people for playing a specific hero is just dumb, so it makes you feel better that there are people who play Treant Protector for 1000 games? Or is it for having a high win rate with that hero? Ban me for playing TA in ranked???
Getting back 25 points when reporting an intentional feeder? Welcome to stacks where everybody agrees to report one of the solo queuers to get back their rating.
Carry players getting less MMR? Support players getting more? Implying what exactly. This system makes no sense whatsoever - so someone who is good at carries should support? When I know I'm the 5.6k rated guy in a team of 4.5k players I should pick support just because you make the system punish me for playing carries?
@sampson
The 14 march update where they fixed the bug to always show MMR after a match, they removed colour ordering by queue time.
http://dotabuff.com/players/68379658/matches?date=&duration=&faction=&game_mode=&hero=terrorblade&lobby_type=ranked_matchmaking&page=2
The first 5 games on this page were after the patch, the other 15 games before.
All 15 games were solo queue. I was the first player in 13/15. Of the two players who were put above me (in both cases I was second by colour), one was anonymous, and the other one is this player http://dotabuff.com/players/101450083
I can't be assed to find all my game records as I play more party than solo which isn't affected by colour, but this one page is enough to give an idea.
@Zenoth no, ban this picked hero, no the player, the hero for this player.
I cant explain it in english like i want.
I didnt say punish the carry pickers, just play two roles in low brackets, cause the people in 2k-4k play always carry.
So if u play 20 games playing as support get less points
if u play 20 games playing only carry get less points
if u play 12 carry - 8 suppor u get normal points
Then how does the system detect support or carry? The definition is far too loose. Is Alch a support or a carry? Is Naga a support or a carry? Or will you go by GPM alone? Then a roaming PotM who scores kills and snowballs despite buying wards is going to be deemed as a carry?
If u make urn, buy wards, mekansm, forcstaff, sentrys. Support.
If u make sange and yash, bf, midas or something like that Carry.
IDK its hard to detect, but its possible.
Nah that's bullshit... so an OD mid who buys Mek and Force is a support? A Pudge who buys Urn and Force is a support?
A jungling CM who gets first blood and snowballs off a couple of kills, buying a Midas is a carry?
People will just game the system, and you have a minor annoyance which does no good for anybody.
Furthermore, why should someone who knows that he is better than the rest of his team by a large margin play support?
why would a lower MM player who is not good at support play the role? Having a support who constantly kills the creep equilibrium can kill a carries farm and have the offlaner snowball hard
I rather a 4.5k shit support screwing up the safe lane than the same 4.5k player going against a 5.6k mid and getting shat on with 0-5 by 5 minutes.
dude so what
burning is playing pubs but he will have to play wisp cause if he doesnt he will get punished because he plays carry
WAT
@Zenoth
somehow that that same 4.5k MM player feeding mid 0-5 is somehow not feeding anywhere else. Win mid = win game correct?
He may play a bad support, but his negative impact on the game is decreased.
It is far easier to win a game where your support goes 0-5 than when your mid goes 0-5.
Of course I'm generalizing here alot, and I know it doesn't apply to everybody/all situations, but for general cases it usually holds true.
Case Study: zerr3x of playdota forums, a 3.5k player who buys a 4.7k account, which means every game he plays he is practically the worst player on both teams. However he manages to keep his mmr from dropping by playing treant practically every game and minimizing his negative impact on the game (even though his treant is still godawful).
Every game is about maximizing and minimizing impact. There is an average skill level for every game, and players above it should maximize their impact on the game in general while players below it should minimizing any negative impact they may have.
Of course this is not saying that lower rated players cannot beat higher-rated players or anything, but on average this is the case.
dont fucking do this shit in my games valve
[IMG]http://puu.sh/841F1.jpg[/IMG] LUCKILY IM BOOSTING THIS ACC BUT THIS SHOULDNT HAPPEN
theres a huge fucking gap between 4.4k and 5.2k ITS UNREAL yes i was 5.2k
Supports have similar positive impacts on the game to cores. However, supports have less negative impact on the games than cores.
In other words a good player impacts the game equally well on supports and cores (assuming similar levels of competency), but a bad player pulls down a game more when on a core rather than a support.
Sampson, you're just a moron, man. I think you're that guy who's very arrogant yet makes subpar grades, doesn't do anything well, etc. The reason? "I wasn't trying". Okay, but. The world doesn't give a shit what you're able to do, it cares what you're doing.
you caught me, In my 3rd period math class yesterday I was suppose to be doing my 6 times tables but I was like 'NIGGA PLZ IM GOING TO TI4'
What zenoth said is totally true, shit happened when we playing as a support and our core barely have any impact to the game.
Related on topic, maybe revealing most played +winrate someone hero when we draft is good because we know that they really can handle the hero. It will be usefun even on cm/cd.
But i think its hard to implemented it.
I say make each player pick their character and then they have to pick a role (hard carry, carry, offlane, semi-support, hard support), and then base their points on how well they did their role.
Example: Alch could be listed in all 5 slots, but you choose what he is going to be. Sure you might start as a support and end up being the carry, but as long as you are fulfilling your support role, you will still get the points earned.
I realize that this system is flawed as I write this, but i mean to just use it as a starting point.
Why would someone like sampson he is a fucking asshole
btw even with low mmr some people just cant play support.
I have 3.5 k mmr, if you put me in the support role in a 4.5 k game i am going to fuck up a lot more than if you put me in mid or safe lane.
The reason is pretty simple, i DONT KNOW how to play support, i would feed a lot.
If you dont believe me ask sampson.
Maybe what you are saying is true for high brackets where everyone knows at least the basics of support.
1. Region lock, anyone above 250 ping is not alllowed to play (CS GO HAVE THIS, WTF).
2. Return solo queue.
If offlane dies more than 3 times in 5 minutes, you don't lose any points.
@zenoth the color theory is partially correct based on solo + long queue times but it wasn't a constant thing. People would instantly assume being blue was being the highest MMR....this theory was probably true 10% of the time but not enough to make it a constant factor.
Wave, that's cute.
Tell me more about this huge gap you´re talking about.
https://twitter.com/WagaGaming/status/449252144739135488
Obviously the system is not perfect, but people hating on it more than they should tbh, it's hard to factor in all the random drunks or ppl who are emotional when playing. No system can accurately predict the human factor in these games. Of course, it would be neat if I stopped getting people sub5k MMR in my games only because it would take long to find a higher one...
Some option to the queuing could be nice (Strict/Loose searchsettings)
@waga
Valve made a decision to cater to the masses instead of the top, it would be interesting to see how long the ques would be for 6k+ people if they limited the search range
Hey my friend i dont always play carry, i am just bad at supporting and offlaning probably, i can play those roles but if i do that our team is probably going to lose, not because i dont trust our carry but because i am just bad at those roles.
Моля, впишете се, за да публикувате коментар.
...............