| Boardin March 21 2012 07:59. Posts 187 | Profile # |
I know that the MMR is working as intended, however it currently just helps people who mass games. The league would be a better place if better players were captains.
As it stands now many people who captain are under 50 percent win % but have 250+ games. They lose .1 pts when they lose a game and i lose 2 due to their terrible picks, ensuring that they will cap again the next game and make more bad decisions.
In a game where picks decide many games, its a vicious cycle that needs fixed. Thoughts? |
|
|
| Sprungjeezy United States. March 21 2012 08:12. Posts 1196 | Profile Blog # |
On March 21 2012 07:59 Boardin wrote: I know that the MMR is working as intended, however it currently just helps people who mass games. The league would be a better place if better players were captains.
As it stands now many people who captain are under 50 percent win % but have 250+ games. They lose .1 pts when they lose a game and i lose 2 due to their terrible picks, ensuring that they will cap again the next game and make more bad decisions.
In a game where picks decide many games, its a vicious cycle that needs fixed. Thoughts?
You probably ought not look to the TL bot for above average skilled games. Several people (like the majority of the Dota community) seem to think that they are far better than they are and because of that they are not willing or interested in hearing other opinions regardless of how accurate they may be. At one point the TLIH inhouses were a fun place to play with other people you recognize and at one point through team drafting there were some decently well played games, but now it is just people with high MMR, as you said, making poor decisions even though (and sometimes in spite of) other people make recommendations or suggestions. I don't know if changing the way the ELO system works would fix this because it mostly boils down to the captains in CM and how much they listen to their team mates. |
|

|
| Alur Denmark. March 21 2012 08:16. Posts 1546 | Profile Blog # |
| Because the outcome of the game doesn't depend on your own performance at all. Most of the current high rated players were "subjects" to "poor" captains (quotations because I believe you're exaggerating), but got to where they are due to their sheer impact on the game. And if you want to tell me that some picks are so bad it's impossible to win, I'll tell you that that's going to be a risk for a lot of players using the bot - which means that the losses for which you exclusively blame your captain, happen to others as well, so I don't understand how there is any form of vicious cycle. If anything it's a vicious temporary decrease in rating that will even out due to probability. |
| |
|
| Soulfire United States. March 21 2012 08:18. Posts 229 | Profile # |
This is true in a sense, but I personally know that every single person above 95% MMR atm is a perfectly capable drafter, and most if not all have a 60%ish W/L. Get high enough in MMR (30 TSR, which if you're as good as you say you are is incredibly easy) and only queue for pro, and you'll have a 95%+ every single time that can draft very well, or at least well enough for this league, because it rarely comes down to picks at that point.
If you're playing regular and you think you can cap better, try asking. I give up my cap slot all the time because I know there are people that are lower MMR that are way better than me. Don't be a dick when you ask like the majority do. |
| | ◕ http://kiwiclonearmy.sadlife.net/ ◕ |
|
|
| JeeJee Canada. March 21 2012 08:27. Posts 5330 | Profile Blog # |
On March 21 2012 08:16 Alur wrote: Because the outcome of the game doesn't depend on your own performance at all. Most of the current high rated players were "subjects" to "poor" captains (quotations because I believe you're exaggerating), but got to where they are due to their sheer impact on the game. And if you want to tell me that some picks are so bad it's impossible to win, I'll tell you that that's going to be a risk for a lot of players using the bot - which means that the losses for which you exclusively blame your captain, happen to others as well, so I don't understand how there is any form of vicious cycle. If anything it's a vicious temporary decrease in rating that will even out due to probability.
Indeed. There is no vicious cycle because if this is happening then it's happening to your opponents as much as it's happening to you. So if you are better, then your team will win more often than not, and your rating will increase. It's really that simple.
|
| (\o/) If you want it, you find a way. Otherwise you find excuses. No exceptions.
/_\ aka Shinbi (requesting a name change since 27/05/09 ☺) |
|

|
| Durak Canada. March 21 2012 08:35. Posts 3221 | Profile Blog # |
I think the Trueskill system is fine. Your conclusion that mass gamers have an advantage because they "lose less points" is false. The system simply has a better idea of their true skill rating so they have less variance than you. Conversely, you also get more points for a win (also takes in account how fair the teams are/who you're playing against) than the mass gamer.
Focus on getting better rather than systems that have been researched quite well. It's not worth your time. |
|
|
| Cel.erity United States. March 21 2012 08:36. Posts 3595 | Profile Blog # |
| It's worth noting also that play skill has almost nothing to do with how well you pick. Below the pro level, there are very few high skill players who can pick worth shit. You're going to have to deal with this in any league. |
| | We found Dove in a soapless place. |
|
|
| Boardin March 21 2012 09:03. Posts 187 | Profile # |
On March 21 2012 08:35 Durak wrote: I think the Trueskill system is fine. Your conclusion that mass gamers have an advantage because they "lose less points" is false. The system simply has a better idea of their true skill rating so they have less variance than you. Conversely, you also get more points for a win (also takes in account how fair the teams are/who you're playing against) than the mass gamer.
Focus on getting better rather than systems that have been researched quite well. It's not worth your time.
With the current TSR, anyone over 90% will never be able to go lower.
This is flawed |
|
|
| Boardin March 21 2012 09:03. Posts 187 | Profile # |
On March 21 2012 08:18 Soulfire wrote: This is true in a sense, but I personally know that every single person above 95% MMR atm is a perfectly capable drafter, and most if not all have a 60%ish W/L. Get high enough in MMR (30 TSR, which if you're as good as you say you are is incredibly easy) and only queue for pro, and you'll have a 95%+ every single time that can draft very well, or at least well enough for this league, because it rarely comes down to picks at that point.
If you're playing regular and you think you can cap better, try asking. I give up my cap slot all the time because I know there are people that are lower MMR that are way better than me. Don't be a dick when you ask like the majority do.
They aren't all perfectly capable drafters, but you are right, I should just sign pro |
|
|
| Boardin March 21 2012 09:11. Posts 187 | Profile # |
| I'd also like to add that while trueskill is a "proven" formula. It's implementation is very different on a service like xbox live with millions of users and hundreds around your skill level. the amount of games until the rating system converges on your "true skill" should vary based on the size of the player pool. I believe that the certainty should be far less games and perhaps should never get to the point where a player can lose just .1 pts per game. |
|
|
| Durak Canada. March 21 2012 09:12. Posts 3221 | Profile Blog # |
On March 21 2012 09:03 Boardin wrote: Show nested quote +On March 21 2012 08:35 Durak wrote: I think the Trueskill system is fine. Your conclusion that mass gamers have an advantage because they "lose less points" is false. The system simply has a better idea of their true skill rating so they have less variance than you. Conversely, you also get more points for a win (also takes in account how fair the teams are/who you're playing against) than the mass gamer.
Focus on getting better rather than systems that have been researched quite well. It's not worth your time.
With the current TSR, anyone over 90% will never be able to go lower. This is flawed
They're better than 90% of players. If new players sign up who are better than them, and play more than a couple games, they'll get pushed down. It's just a relative measure so I don't see how it is flawed. |
|
|
| FarmI3oy United States. March 21 2012 09:40. Posts 98 | Profile # |
I think the MMR system is fine currently, I just started playing a few days ago in this league. The only major flaw I see is that a lot of high MMR people are elitist tools that automatically assume your shit before you even play. Normally I'd say "Well this is dota after all why would it be different in this league".... Team liquid is a respected name because of good manners at least that's what I always thought.
Sadly the TLIH is becoming more and more elitist by the day. I know some admins are trying to cure this while others seem to encourage it. Raging and Cursing your teammates obviously will happen but at least try to teach people what they are doing wrong instead of bashing them for the entirety of the game. |
|
|
| Boardin March 21 2012 09:45. Posts 187 | Profile # |
On March 21 2012 09:12 Durak wrote: Show nested quote +On March 21 2012 09:03 Boardin wrote: On March 21 2012 08:35 Durak wrote: I think the Trueskill system is fine. Your conclusion that mass gamers have an advantage because they "lose less points" is false. The system simply has a better idea of their true skill rating so they have less variance than you. Conversely, you also get more points for a win (also takes in account how fair the teams are/who you're playing against) than the mass gamer.
Focus on getting better rather than systems that have been researched quite well. It's not worth your time.
With the current TSR, anyone over 90% will never be able to go lower. This is flawed
They're better than 90% of players. If new players sign up who are better than them, and play more than a couple games, they'll get pushed down. It's just a relative measure so I don't see how it is flawed.
see my post before this. the system is "too certain". they can never lose enough points to move down substantially. I don't want to name names as its not appropriate but if there were a reset I suspect they would not reach the same rating |
|

|
Kipsate Netherlands. March 21 2012 10:33. Posts 18114 | Profile Blog # | |
| | Park Ji Yeon/Ye Eun/Kang Min Kyung, T-ara/Wonder Girls/Davichi | |
|
|
| TheWarbler United States. March 21 2012 10:51. Posts 1241 | Profile Blog # |
On March 21 2012 10:33 Kipsate wrote: Got a better idea?
Keep it the same but have seasons. Every couple months TSR resets.Last edit: 2012-03-21 10:52:05 |
| | if you can believe you can concieve |
|
|
| JBright Canada. March 21 2012 10:52. Posts 3772 | Profile # |
On March 21 2012 09:40 FarmI3oy wrote: I think the MMR system is fine currently, I just started playing a few days ago in this league. The only major flaw I see is that a lot of high MMR people are elitist tools that automatically assume your shit before you even play. Normally I'd say "Well this is dota after all why would it be different in this league".... Team liquid is a respected name because of good manners at least that's what I always thought.
Sadly the TLIH is becoming more and more elitist by the day. I know some admins are trying to cure this while others seem to encourage it. Raging and Cursing your teammates obviously will happen but at least try to teach people what they are doing wrong instead of bashing them for the entirety of the game.
I haven't played in a TLIH but I don't really see the problem of assuming that you're bad until you can show that you're good. Everyone needs to prove themselves one way or another so people will get to know your ability after you consistently perform well. The assumption part is ok - how they act towards people who they think are bad is not. |
| | The good and the wise lead quiet lives. Neo's #1 Frenemy and nightmare. | |
|

|
| TheyCallMePops United States. March 21 2012 11:21. Posts 68 | Profile # |
ELO rankings in team games have always been and always will be complete crock. Applies to DotA/LoL and everything else. The only argument that can be made for it is in the case of pre-arranged teams. Otherwise you're going to be playing with friends that are much worse than you and losing because you're raising the average ELO of your team to disproportionate levels or lose because they're lowering your ELO and you're getting terrible players. Can this swing both ways? Of course it can. Does it always even out? Definitely not, especially when the higher ELO player isn't a dedicated carry player.
I'm not saying MMR/ELO is a terrible system for DotA, I think it works well enough because we dont have some dedicated ladder like LoL and it does an okay job of matching players, but I think it's a load of shit for people to base their skill level on their win % because of the shortcomings of ELO in team games.
I personally have two accounts, one that is for playing with my team (whether everyone is on or we are just duo/tri-queueing) and another for playing with the multitude of terrible friends that I have that are new to the game/genre. It's amazing to see the disparity of skill levels between the two accounts even though they both have similar win rate % (~55-60%) |
|

|
| Boardin March 21 2012 12:41. Posts 187 | Profile # |
On March 21 2012 10:33 Kipsate wrote: Got a better idea?
Yes. A new ELO system (a quick google search suggests a few, I would be more than happy to research and help implement, I'm a IT consultant who majored in compsci/math), a variation of the current Trueskill (there are configurable variables that would help to keep people from being "stuck"), as well as incorporating seasons.
I'm not just here to complain, I'd love to help |
|
|
blahz0r March 21 2012 18:22. Posts 2138 | Profile # |
The dxd bots were like perfect for dota/dota 2 =x
The problem with trueskill is that it requires a really large amount of players. |
| |
|
| Velr Switzerland. March 21 2012 20:29. Posts 5110 | Profile Blog # |
You should just stop playing CM and go with SD or AP and have fun.
There ya go, i fixed TLIH for you. |
|
|
| 1 2 Next All |
|