• Log InLog In
  • Register
Liquid`
Team Liquid Liquipedia
EST 20:59
CET 02:59
KST 10:59
  • Home
  • Forum
  • Calendar
  • Streams
  • Liquipedia
  • Features
  • Store
  • EPT
  • TL+
  • StarCraft 2
  • Brood War
  • Smash
  • Heroes
  • Counter-Strike
  • Overwatch
  • Liquibet
  • Fantasy StarCraft
  • TLPD
  • StarCraft 2
  • Brood War
  • Blogs
Forum Sidebar
Events/Features
News
Featured News
RSL Season 3 - Playoffs Preview0RSL Season 3 - RO16 Groups C & D Preview0RSL Season 3 - RO16 Groups A & B Preview2TL.net Map Contest #21: Winners12Intel X Team Liquid Seoul event: Showmatches and Meet the Pros10
Community News
RSL Season 3: RO16 results & RO8 bracket12Weekly Cups (Nov 10-16): Reynor, Solar lead Zerg surge1[TLMC] Fall/Winter 2025 Ladder Map Rotation14Weekly Cups (Nov 3-9): Clem Conquers in Canada4SC: Evo Complete - Ranked Ladder OPEN ALPHA12
StarCraft 2
General
RSL Season 3: RO16 results & RO8 bracket SC: Evo Complete - Ranked Ladder OPEN ALPHA RSL Season 3 - Playoffs Preview Mech is the composition that needs teleportation t GM / Master map hacker and general hacking and cheating thread
Tourneys
RSL Revival: Season 3 $5,000+ WardiTV 2025 Championship StarCraft Evolution League (SC Evo Biweekly) Constellation Cup - Main Event - Stellar Fest 2025 RSL Offline Finals Dates + Ticket Sales!
Strategy
Custom Maps
Map Editor closed ?
External Content
Mutation # 500 Fright night Mutation # 499 Chilling Adaptation Mutation # 498 Wheel of Misfortune|Cradle of Death Mutation # 497 Battle Haredened
Brood War
General
Data analysis on 70 million replays A cwal.gg Extension - Easily keep track of anyone soO on: FanTaSy's Potential Return to StarCraft [ASL20] Ask the mapmakers — Drop your questions FlaSh on: Biggest Problem With SnOw's Playstyle
Tourneys
[Megathread] Daily Proleagues Small VOD Thread 2.0 [BSL21] GosuLeague T1 Ro16 - Tue & Thu 22:00 CET [BSL21] RO16 Tie Breaker - Group B - Sun 21:00 CET
Strategy
Current Meta Game Theory for Starcraft How to stay on top of macro? PvZ map balance
Other Games
General Games
Stormgate/Frost Giant Megathread EVE Corporation Path of Exile [Game] Osu! Should offensive tower rushing be viable in RTS games?
Dota 2
Official 'what is Dota anymore' discussion
League of Legends
Heroes of the Storm
Simple Questions, Simple Answers Heroes of the Storm 2.0
Hearthstone
Deck construction bug Heroes of StarCraft mini-set
TL Mafia
Mafia Game Mode Feedback/Ideas
Community
General
Russo-Ukrainian War Thread The Games Industry And ATVI US Politics Mega-thread Things Aren’t Peaceful in Palestine About SC2SEA.COM
Fan Clubs
White-Ra Fan Club The herO Fan Club!
Media & Entertainment
[Manga] One Piece Movie Discussion! Anime Discussion Thread Korean Music Discussion
Sports
2024 - 2026 Football Thread Formula 1 Discussion NBA General Discussion MLB/Baseball 2023 TeamLiquid Health and Fitness Initiative For 2023
World Cup 2022
Tech Support
TL Community
The Automated Ban List
Blogs
The Health Impact of Joining…
TrAiDoS
Dyadica Evangelium — Chapt…
Hildegard
Saturation point
Uldridge
DnB/metal remix FFO Mick Go…
ImbaTosS
Customize Sidebar...

Website Feedback

Closed Threads



Active: 1664 users

Ranked matchmaking coming to Dota 2 - Page 114

Forum Index > Closed
2303 CommentsPost a Reply
Prev 1 112 113 114 115 116 Next
Please keep the QQ to a minimum if you do not like this update. We are happy to hear your reasoning for not liking a ranked system, but no "OMG VOLVO WHY" posts.
DougJDempsey
Profile Joined April 2010
747 Posts
January 21 2014 17:07 GMT
#2261
On January 21 2014 06:57 cecek wrote:
I think when people say "elo hell", what they mean is that they are much better than their teammates and enemies, but are still stuck with them in matchmaking. By definition, if the player doesn't outclass the other people in his match, he's not in elo hell.

Or... something like that. The "logic" behind elo hell is very confusing.


Well... Its highly likely 99% of the people crying about "elo hell" are just unable to face facts and are simply in denial... There really isnt a legit excuse that they could pull out to justify why somehow they are just stuck playing with "worst/terrible" players when they are oh so clearly better and all.... Except that they are at the exact same level and play equally as bad. Its easy to judge. Even a 2k player can see when dendi fucks up but they cant do even a tiny bit of what dendi can.
Vaelone
Profile Blog Joined February 2011
Finland4400 Posts
Last Edited: 2014-01-21 18:07:57
January 21 2014 17:28 GMT
#2262
Personally I did my 10 games when the system launched and no more.

Figured I'l play it the next time when I feel like I've improved as a player and watch my rating go up, instead of suffering from it going down from my initial rating which would be disheartening.

Ranked games didn't really seem better than regular anyway so I'l happily stick to regular queue for now and avoid any stress related to ranked. This was at only 3.9k rating though, and only 10 games.
Rhyme
Profile Joined May 2010
United States1069 Posts
January 21 2014 17:36 GMT
#2263
i see a big difference between my ranked and unranked games

people take ranked way more seriously imo; in unranked i get more frequent pub bullshit (no chicken, suboptimal team comp, poor communication, infighting, etc)

it's alright with me though, i think that's how it should be
dont ever say that
Plansix
Profile Blog Joined April 2011
United States60190 Posts
Last Edited: 2014-01-21 18:01:11
January 21 2014 17:59 GMT
#2264
From the bottom of the trench(2500-2900) I can say that ranked and Unranked feel about the same. As I have said before, I normally have pleasant games, so I offer these insights from my time in the trench:

- All pick is horrible at the bottom MMR. If you want to play with people who give no fucks, play that. No one will buy mek or wards. In my personal experience, it is where players who just want to play their hero go and do whatever they want.

- all the drafting modes on 2500-2900 are more passable. Although you will not get the beat drafts, anyone playing that mode has at least the understanding that they might not get to play the exact hero they want. This may be true at other mmrs, but in the deep trench, the players also lack any skills to back up their greed.

- people love leadership and team play. People respond well to anyone taking charge or trying to help. If you see a support that is behind a couple of level and you are lifestealer, just have them follow you around the jungle and get them a few level. Playing with warlock as another support, tell him you're getting mek and that he should rush scepter to, it will help.

- never be afraid to draft. At worst, just ask people what they want and go from there. I normally just draft CM and SS and fill in cores as needed if no one steps up. Be fearless and be friendly. But most of all, listen to your team and don't pick heroes they can't play.

- be nice, for the love god. Stop telling people how to play, or their build it bad. Never in the history of games has screaming at strangers made them play better. If you want to lose a game, yell at your team mate.

As I said before, I don't win every game, but I have a reasonable time and meet good people. Even at the bottom of the trench, you can find good games and people to play with.
I have the Honor to be your Obedient Servant, P.6
TL+ Member
cecek
Profile Blog Joined August 2011
Czech Republic18921 Posts
January 21 2014 18:14 GMT
#2265
Rant incoming:

Let's say you're 300 points below your level of play. It can happen... maybe you fell down by playing on tilt a lot, or you've been practicing meepo, or doing troll builds, maybe you played a lot of games outside of ranked and improved significantly. Now of course it's going to take a long time to climb to the improved MMR (and stabilize there). You can't expect to win 100% of the games suddenly, it's a 5v5 game (and even in 1v1, he better player doesn't win 100% of matches if the skill difference isn't massive), you probably just raise your probability of victory to 60% (a total estimate, obviously). Let's assume 60% improvement is correct and you gain or lose 25 MMR for every win and loss. It will then take you 60 games (36 wins, 24 losses) to erase that deficit. I don't see what's wrong with that. How else would you do it?

Some guy suggested giving bonus points for win streaks to help climb back faster. You would have to do one of two things and I think both are bad. Either you would have to make people lose "bonus" points from losing streaks, which would obviously do nothing, because you would fall quicker and deeper, so climbing back up would take as long as now. Also, it would make the whole matchmaking more volatile, because people would jump around much more often. Also, knowing people, they would probably be whining about loss streaks way more than they would be happy about winning streaks, so I don't think that would work out.
The other way would be to just give the bonus points for winning streaks and not caring about losing streaks. This would probably make people happy, but it would also mean that the MMR given out is not a zero sum anymore (I assume it more or less is now), so the whole dota population would just be climbing up. Everyone will get a random winning streak here and there. I'm not sure, but it seems like it would be more random and screw with the matchmaking because of it. I think hearthstone does this, but they also do a complete reset of the ranking periodically, which is absolutely unacceptable for Dota for obvious reasons.

Another point is people suggesting that there should be a system that takes individual performance into consideration. Again, I believe this is impossible. There are so many ways you can affect a game positively and negatively that there is no way Valve could make a system to take everything into consideration. Let's say they would want to make such a system. How would it work? More kills = better? More wards bought = better? Creep kills? Denies? gmp? Checking the enemy inventories? How about what you write in the chat? It can have a huge impact if you tell a guy to get back when he's getting ganked. Should Valve develop a system that can understand what you're typing and saying in voice chat? It's just not possible. And the other huge problem is that any system of this sort would get gamed immediately. Many people will not play DotA, but play the system instead. Whore out kills, lasthits, gpm, etc. The only way to measure a player's impact on winning or losing a pub game is by observing the end result of the game, a win or a loss. Sure, it might not be fair for one individual game, there are plenty matches you deserve to win, but lose and vice versa. But over a large enough sample size, it is by far the most accurate and fair measurement.

rant over
super gg
HollowLord
Profile Blog Joined August 2010
United States3862 Posts
Last Edited: 2014-01-21 18:33:22
January 21 2014 18:31 GMT
#2266
The only way there could be a method that takes individual player performance into account would be if every player had a valve employee assigned to them, who watched all of their games and observed their performance to take into account.

99% of Dota is decision making. It is not enough to put a ward somewhere and judge that as something that was good. Was it safe to plant the ward? Did you need to put it there? Could it have been placed somewhere else? Did you waste too much time warding and dewarding? These are all things that are only learned through hours upon hours upon hours of experience and can only be analyzed by human intelligence.

Valve's matchmaking system is processed by computers. Computers can not make decisions or judgment calls. A computer receives input, and produces output. A computer can not determine whether something was or was not the right idea because the computer lacks a human mind that is thinking about every possible consequence or benefit to any action. The Ranked Matchmaking system currently in place is as close as one can get to an accurate representation of one's skill, because it takes into account a basic premise: Every player on a team has the ability to lead their team to victory. It may be with a massive K/D/A or it might be with intelligent ward control, but each player has the ability to contribute enough to their team that they can be the reason why they win. This is the premise regardless of what role you play, what hero you play, what language you speak or what country you are from.

And honestly, if you do not think you can lead your team to victory, why are you even playing?
dota 2 stream #noskill #feed #noob twitch.tv/dmcredgrave
zezamer
Profile Joined March 2011
Finland5701 Posts
Last Edited: 2014-01-21 18:39:23
January 21 2014 18:38 GMT
#2267
gaben just needs to create AI to fix matchmaking issues
sounds like a easy task for a mind that gave us hats
Sn0_Man
Profile Blog Joined October 2012
Tebellong44238 Posts
Last Edited: 2014-01-21 18:50:54
January 21 2014 18:41 GMT
#2268
Gradual MMR inflation doesn't seem that bad considering it would be a somewhat elegant way of having people who don't play for a while "decay" while simultaneously having people feel like they are improving and not really costing anybody anything.

I mean, its not like "people above 4K are gold and we want gold to be the middle 20% of the community" or whatever ridiculous garbage blizzard tried. 4K is just a number, and if 5K is the new 4K 2 years from now, so be it.

This is quite apart from anything else, I'm just musing because somebody mentioned this and thought it would be bad.

E: I suppose there are drawbacks as well. For example, if you trained hard and got better, it would take longer for those results to show considering the fact that you would eventually need to gain 500 mmr to reach fair games instead of 300 mmr, which would be more games unless the system adjusted how much you won or lost dynamically which it doesn't seem to right now. Still, that seems minor and overcomeable.
LiquidDota StaffSCIENTISTS BAFFLED | 3275929302
HollowLord
Profile Blog Joined August 2010
United States3862 Posts
January 21 2014 18:50 GMT
#2269
On January 22 2014 03:41 Sn0_Man wrote:
Gradual MMR inflation doesn't seem that bad considering it would be a somewhat elegant way of having people who don't play for a while "decay" while simultaneously having people feel like they are improving and not really costing anybody anything.

I mean, its not like "people above 4K are gold and we want gold to be the middle 20% of the community" or whatever ridiculous garbage blizzard tried. 4K is just a number, and if 5K is the new 4K 2 years from now, so be it.

This is quite apart from anything else, I'm just musing because somebody mentioned this and thought it would be bad.


From a purely theoretical level, the idea behind the system is that it's a zero-sum game; for every point of MMR I gain, someone out there has to lose a correlating point of MMR. However because Dota has new players signing up every day, and because you can't really 'run out of MMR'(I'm not sure if it's possible to go beyond 0 MMR and into the negatives), this means that it actually does have inflation. More points of MMR are being introduced to the system every day.

Add in intentional inflation, and it's hard to see what the consequences of it could be. You could look at Starcraft 2 circa 2010: before the Masters league was implemented you were only good if you were a "2000 point Diamond". Two weeks later, you had to be "2100 point Diamond" to be considered good. Then you had to be "2200 point Diamond" another month later.
dota 2 stream #noskill #feed #noob twitch.tv/dmcredgrave
Sn0_Man
Profile Blog Joined October 2012
Tebellong44238 Posts
January 21 2014 18:57 GMT
#2270
Currently, zero sum still functions reasonably correctly. Adding players at the "average" mmr maintains that average mmr, even if it skews the exact distribution some.

Obviously inflation isn't super-desirable but going from "good is 2000 point diamond" to "good is 2200 point diamond" over the course of a year isn't terribly bothersome to me. Players who didn't play for 6 months would "lose" 100 rating while simultaneously players who played and didn't really improve (aka the vast majority) would still "gain" rating points for their troubles, which is kinda dumb but pretty popular/attractive to "the masses".
LiquidDota StaffSCIENTISTS BAFFLED | 3275929302
Veles
Profile Joined May 2011
United States3280 Posts
January 21 2014 18:59 GMT
#2271
On January 22 2014 03:14 cecek wrote:
Rant incoming:

Let's say you're 300 points below your level of play. It can happen... maybe you fell down by playing on tilt a lot, or you've been practicing meepo, or doing troll builds, maybe you played a lot of games outside of ranked and improved significantly. Now of course it's going to take a long time to climb to the improved MMR (and stabilize there). You can't expect to win 100% of the games suddenly, it's a 5v5 game (and even in 1v1, he better player doesn't win 100% of matches if the skill difference isn't massive), you probably just raise your probability of victory to 60% (a total estimate, obviously). Let's assume 60% improvement is correct and you gain or lose 25 MMR for every win and loss. It will then take you 60 games (36 wins, 24 losses) to erase that deficit. I don't see what's wrong with that. How else would you do it?

Some guy suggested giving bonus points for win streaks to help climb back faster. You would have to do one of two things and I think both are bad. Either you would have to make people lose "bonus" points from losing streaks, which would obviously do nothing, because you would fall quicker and deeper, so climbing back up would take as long as now. Also, it would make the whole matchmaking more volatile, because people would jump around much more often. Also, knowing people, they would probably be whining about loss streaks way more than they would be happy about winning streaks, so I don't think that would work out.
The other way would be to just give the bonus points for winning streaks and not caring about losing streaks. This would probably make people happy, but it would also mean that the MMR given out is not a zero sum anymore (I assume it more or less is now), so the whole dota population would just be climbing up. Everyone will get a random winning streak here and there. I'm not sure, but it seems like it would be more random and screw with the matchmaking because of it. I think hearthstone does this, but they also do a complete reset of the ranking periodically, which is absolutely unacceptable for Dota for obvious reasons.

Another point is people suggesting that there should be a system that takes individual performance into consideration. Again, I believe this is impossible. There are so many ways you can affect a game positively and negatively that there is no way Valve could make a system to take everything into consideration. Let's say they would want to make such a system. How would it work? More kills = better? More wards bought = better? Creep kills? Denies? gmp? Checking the enemy inventories? How about what you write in the chat? It can have a huge impact if you tell a guy to get back when he's getting ganked. Should Valve develop a system that can understand what you're typing and saying in voice chat? It's just not possible. And the other huge problem is that any system of this sort would get gamed immediately. Many people will not play DotA, but play the system instead. Whore out kills, lasthits, gpm, etc. The only way to measure a player's impact on winning or losing a pub game is by observing the end result of the game, a win or a loss. Sure, it might not be fair for one individual game, there are plenty matches you deserve to win, but lose and vice versa. But over a large enough sample size, it is by far the most accurate and fair measurement.

rant over


IMO the major flaw in these arguments is the assumption that 300 rating is significant. I know that I have been matched with players ~1400 points above me and I wouldn't be surprised to learn that I have been matched with players ~800 below me. Maybe at low mmr matchmaking isn't quite as liberal, but if the variance that you describe exists, then skill shouldn't change significantly over 300 mmr.

I think that rating does generally correlate with skill. However, especially at the higher level, players who are 500 rating above you might be playing the same games. To state it explicitly: even if you had higher mmr, your games would be exactly the same.
HollowLord
Profile Blog Joined August 2010
United States3862 Posts
January 21 2014 19:01 GMT
#2272
On January 22 2014 03:57 Sn0_Man wrote:
Obviously inflation isn't super-desirable but going from "good is 2000 point diamond" to "good is 2200 point diamond" over the course of a year isn't terribly bothersome to me. Players who didn't play for 6 months would "lose" 100 rating while simultaneously players who played and didn't really improve (aka the vast majority) would still "gain" rating points for their troubles, which is kinda dumb but pretty popular/attractive to "the masses".


Well, that's where the debate lies in whether it's a good or a bad thing. You can see a similar effect in a lot of places in the videogame media. How many people take a break from an MMO for a month or two only to come back and be denied spots in groups because "omg this noob doesnt even have tier 8 gear wtf".

After all, when he stopped playing it only went up to tier 5.
dota 2 stream #noskill #feed #noob twitch.tv/dmcredgrave
Sn0_Man
Profile Blog Joined October 2012
Tebellong44238 Posts
January 21 2014 19:15 GMT
#2273
Okay but why raid with a guy who only has tier 5 gear?
If he's really that legit he'll get to tier 8.

MMR would be kinda like that. Why play with somebody who was good when 6k was good? If he's still good he'll get to 6.5K in no time if he's really legit.

Please understand that I HATE FAST INFLATION BULLSHIT like wow pulled off where all your shit from last year is worse than low level garbage this year. Its a dirty hook developers use to make people play games that they can't be bothered to improve so they just artificially produce "progression". I hate that, and I'll admit that thats kinda what MMR inflation smells like. I'm not necessarily advocating for this inflation, but I'm pointing out that its not ALL bad and that it has some real convenient attributes. Dunno. Food for thought.
LiquidDota StaffSCIENTISTS BAFFLED | 3275929302
Ryndika
Profile Blog Joined August 2010
1489 Posts
Last Edited: 2014-01-21 19:41:02
January 21 2014 19:35 GMT
#2274
On January 22 2014 02:59 Plansix wrote:
From the bottom of the trench(2500-2900) I can say that ranked and Unranked feel about the same. As I have said before, I normally have pleasant games, so I offer these insights from my time in the trench:

- All pick is horrible at the bottom MMR. If you want to play with people who give no fucks, play that. No one will buy mek or wards. In my personal experience, it is where players who just want to play their hero go and do whatever they want.

- all the drafting modes on 2500-2900 are more passable. Although you will not get the beat drafts, anyone playing that mode has at least the understanding that they might not get to play the exact hero they want. This may be true at other mmrs, but in the deep trench, the players also lack any skills to back up their greed.

- people love leadership and team play. People respond well to anyone taking charge or trying to help. If you see a support that is behind a couple of level and you are lifestealer, just have them follow you around the jungle and get them a few level. Playing with warlock as another support, tell him you're getting mek and that he should rush scepter to, it will help.

- never be afraid to draft. At worst, just ask people what they want and go from there. I normally just draft CM and SS and fill in cores as needed if no one steps up. Be fearless and be friendly. But most of all, listen to your team and don't pick heroes they can't play.

- be nice, for the love god. Stop telling people how to play, or their build it bad. Never in the history of games has screaming at strangers made them play better. If you want to lose a game, yell at your team mate.

As I said before, I don't win every game, but I have a reasonable time and meet good people. Even at the bottom of the trench, you can find good games and people to play with.

4.3k players have to play against 2.9k players in ranked. Most of games are as stomp as in nonranked. (or more as stacks are even better).
as useful as teasalt
Kazeyonoma
Profile Blog Joined April 2010
United States2912 Posts
January 21 2014 20:06 GMT
#2275
On January 22 2014 04:35 Ryndika wrote:
Show nested quote +
On January 22 2014 02:59 Plansix wrote:
From the bottom of the trench(2500-2900) I can say that ranked and Unranked feel about the same. As I have said before, I normally have pleasant games, so I offer these insights from my time in the trench:

- All pick is horrible at the bottom MMR. If you want to play with people who give no fucks, play that. No one will buy mek or wards. In my personal experience, it is where players who just want to play their hero go and do whatever they want.

- all the drafting modes on 2500-2900 are more passable. Although you will not get the beat drafts, anyone playing that mode has at least the understanding that they might not get to play the exact hero they want. This may be true at other mmrs, but in the deep trench, the players also lack any skills to back up their greed.

- people love leadership and team play. People respond well to anyone taking charge or trying to help. If you see a support that is behind a couple of level and you are lifestealer, just have them follow you around the jungle and get them a few level. Playing with warlock as another support, tell him you're getting mek and that he should rush scepter to, it will help.

- never be afraid to draft. At worst, just ask people what they want and go from there. I normally just draft CM and SS and fill in cores as needed if no one steps up. Be fearless and be friendly. But most of all, listen to your team and don't pick heroes they can't play.

- be nice, for the love god. Stop telling people how to play, or their build it bad. Never in the history of games has screaming at strangers made them play better. If you want to lose a game, yell at your team mate.

As I said before, I don't win every game, but I have a reasonable time and meet good people. Even at the bottom of the trench, you can find good games and people to play with.

4.3k players have to play against 2.9k players in ranked. Most of games are as stomp as in nonranked. (or more as stacks are even better).


Proof please? I've played at least 50 ranked games around high 3k low 4k and i've never been paired with anyone +/- 500 range. If you're going up against 2.9k players, it's because they were probably partied and had teammates that countered his low rating + the 'innate teamwork' bonus Valve gives in MM algorithm. I've waited in queue for over 20 minutes even at my level and have never hit someone that low before. either on my team or against. I get it if someone is at 5k+ rating who gets paired with a 4.5k or something but that's because tehre's not that many 5k+ rating people and realistically, while no one will doubt a 5k+ rating pro/semi pro (merlini, actual pros, etc) are better than everyone else, the difference between a 4.5k rating player and a 5k+ rating player, should be miniscule enough to still make for a fair 50/50 chance game. a 4.3k vs 2.9k player game does not warrant a 50/50 chance unless the 2.9k player had a 5k player with him. I think a lot of people are just making a lot of blind assumptions here.
I now have autographs of both BoxeR and NaDa. I can die happy. Lim Yo Hwan and Lee Yun Yeol FIGHTING forever!
govie
Profile Blog Joined November 2012
9334 Posts
January 21 2014 20:16 GMT
#2276
On January 22 2014 05:06 Kazeyonoma wrote:
Show nested quote +
On January 22 2014 04:35 Ryndika wrote:
On January 22 2014 02:59 Plansix wrote:
From the bottom of the trench(2500-2900) I can say that ranked and Unranked feel about the same. As I have said before, I normally have pleasant games, so I offer these insights from my time in the trench:

- All pick is horrible at the bottom MMR. If you want to play with people who give no fucks, play that. No one will buy mek or wards. In my personal experience, it is where players who just want to play their hero go and do whatever they want.

- all the drafting modes on 2500-2900 are more passable. Although you will not get the beat drafts, anyone playing that mode has at least the understanding that they might not get to play the exact hero they want. This may be true at other mmrs, but in the deep trench, the players also lack any skills to back up their greed.

- people love leadership and team play. People respond well to anyone taking charge or trying to help. If you see a support that is behind a couple of level and you are lifestealer, just have them follow you around the jungle and get them a few level. Playing with warlock as another support, tell him you're getting mek and that he should rush scepter to, it will help.

- never be afraid to draft. At worst, just ask people what they want and go from there. I normally just draft CM and SS and fill in cores as needed if no one steps up. Be fearless and be friendly. But most of all, listen to your team and don't pick heroes they can't play.

- be nice, for the love god. Stop telling people how to play, or their build it bad. Never in the history of games has screaming at strangers made them play better. If you want to lose a game, yell at your team mate.

As I said before, I don't win every game, but I have a reasonable time and meet good people. Even at the bottom of the trench, you can find good games and people to play with.

4.3k players have to play against 2.9k players in ranked. Most of games are as stomp as in nonranked. (or more as stacks are even better).


Proof please? I've played at least 50 ranked games around high 3k low 4k and i've never been paired with anyone +/- 500 range. If you're going up against 2.9k players, it's because they were probably partied and had teammates that countered his low rating + the 'innate teamwork' bonus Valve gives in MM algorithm. I've waited in queue for over 20 minutes even at my level and have never hit someone that low before. either on my team or against. I get it if someone is at 5k+ rating who gets paired with a 4.5k or something but that's because tehre's not that many 5k+ rating people and realistically, while no one will doubt a 5k+ rating pro/semi pro (merlini, actual pros, etc) are better than everyone else, the difference between a 4.5k rating player and a 5k+ rating player, should be miniscule enough to still make for a fair 50/50 chance game. a 4.3k vs 2.9k player game does not warrant a 50/50 chance unless the 2.9k player had a 5k player with him. I think a lot of people are just making a lot of blind assumptions here.


Maybe a stack? would explain it maybe.
The two NBA teams in states with legal weed are called the Nuggets and the Blazers...
Sn0_Man
Profile Blog Joined October 2012
Tebellong44238 Posts
January 21 2014 20:18 GMT
#2277
Thats exactly what he said... :/
LiquidDota StaffSCIENTISTS BAFFLED | 3275929302
SkelA
Profile Blog Joined January 2007
Macedonia13069 Posts
January 21 2014 20:19 GMT
#2278
Whatever your rank you gonna play with players -+500 to 800 in most games if you are playing solo and no stacks are involved. At least that was my experience when we could see dota rank.

Now if you are in a stack for example you could end up playing even +- 1k to 1.5k because of friends having a high rank or others having really low rank.

For example when i was 3.5k ive seen players as high as 4.8k range and as low as 2.2k.
Stork and KHAN fan till 2012 ...
Sokrates
Profile Joined May 2012
738 Posts
January 21 2014 20:22 GMT
#2279
I play 4.3k mmr and i ve had players with more than 1k points above me and 1k points below me in one team. THAT is the ONLY point i rant about because having such a huge variance within your team makes the game snowball out of control pretty hard pretty fast. BUt that doesnt happen too often, most of the times the players sit at an equal level and you get nice games. The mmr is making the average mmr of the team even but i just dont like it when sometimes the variance is really huge within a team. Sometimes you cannot win a game when you got a player with you that fucks up your laning phase and starts feeding but you got to deal with it. It is just anoying when you are unlucky and you get such people 2-3x in a row which makes you tilt.
Sn0_Man
Profile Blog Joined October 2012
Tebellong44238 Posts
January 21 2014 20:23 GMT
#2280
I bet if ur 2kmmr and playing with like all modes all languages all regions you wouldn't experience much more than maybe 200mmr spread in a game very often.

The issue is like 3.9K mmr is top 5% of dota players or whatever, then factor in language choice, ranked vs unranked, modes, server choices, etc. The pool gets pretty small pretty fast.
LiquidDota StaffSCIENTISTS BAFFLED | 3275929302
Prev 1 112 113 114 115 116 Next
Please log in or register to reply.
Live Events Refresh
Next event in 5h 31m
[ Submit Event ]
Live Streams
Refresh
StarCraft 2
White-Ra 217
RuFF_SC2 115
StarCraft: Brood War
Leta 28
NaDa 13
Sexy 4
Dota 2
NeuroSwarm88
LuMiX0
League of Legends
JimRising 273
Super Smash Bros
C9.Mang0253
Other Games
summit1g15561
fl0m731
WinterStarcraft348
ViBE162
Trikslyr71
ToD15
Organizations
Other Games
gamesdonequick742
StarCraft 2
Blizzard YouTube
StarCraft: Brood War
BSLTrovo
sctven
[ Show 17 non-featured ]
StarCraft 2
• Hupsaiya 102
• davetesta39
• IndyKCrew
• sooper7s
• AfreecaTV YouTube
• Migwel
• intothetv
• LaughNgamezSOOP
• Kozan
StarCraft: Brood War
• Azhi_Dahaki23
• STPLYoutube
• ZZZeroYoutube
• BSLYoutube
Dota 2
• masondota22481
League of Legends
• Doublelift5110
• Rush710
Other Games
• Scarra1185
Upcoming Events
RSL Revival
5h 31m
Classic vs MaxPax
SHIN vs Reynor
herO vs Maru
WardiTV Korean Royale
10h 1m
SC Evo League
10h 31m
IPSL
15h 1m
Julia vs Artosis
JDConan vs DragOn
OSC
15h 1m
BSL 21
18h 1m
TerrOr vs Aeternum
HBO vs Kyrie
RSL Revival
1d 5h
Wardi Open
1d 12h
IPSL
1d 18h
StRyKeR vs OldBoy
Sziky vs Tarson
BSL 21
1d 18h
StRyKeR vs Artosis
OyAji vs KameZerg
[ Show More ]
OSC
1d 21h
OSC
2 days
Monday Night Weeklies
2 days
OSC
2 days
Wardi Open
3 days
Replay Cast
4 days
Wardi Open
4 days
Tenacious Turtle Tussle
4 days
The PondCast
5 days
Replay Cast
5 days
LAN Event
6 days
Replay Cast
6 days
Liquipedia Results

Completed

Proleague 2025-11-16
Stellar Fest: Constellation Cup
Eternal Conflict S1

Ongoing

C-Race Season 1
IPSL Winter 2025-26
KCM Race Survival 2025 Season 4
SOOP Univ League 2025
YSL S2
BSL Season 21
CSCL: Masked Kings S3
SLON Tour Season 2
RSL Revival: Season 3
META Madness #9
BLAST Rivals Fall 2025
IEM Chengdu 2025
PGL Masters Bucharest 2025
Thunderpick World Champ.
CS Asia Championships 2025
ESL Pro League S22
StarSeries Fall 2025
FISSURE Playground #2

Upcoming

BSL 21 Non-Korean Championship
Acropolis #4
IPSL Spring 2026
HSC XXVIII
RSL Offline Finals
WardiTV 2025
IEM Kraków 2026
BLAST Bounty Winter 2026
BLAST Bounty Winter 2026: Closed Qualifier
eXTREMESLAND 2025
ESL Impact League Season 8
SL Budapest Major 2025
TLPD

1. ByuN
2. TY
3. Dark
4. Solar
5. Stats
6. Nerchio
7. sOs
8. soO
9. INnoVation
10. Elazer
1. Rain
2. Flash
3. EffOrt
4. Last
5. Bisu
6. Soulkey
7. Mini
8. Sharp
Sidebar Settings...

Advertising | Privacy Policy | Terms Of Use | Contact Us

Original banner artwork: Jim Warren
The contents of this webpage are copyright © 2025 TLnet. All Rights Reserved.