Dota 2

[boozie]

No, it shouldn't. It has to be based on some sort of ELO algorithm. So, winning 5 games in a row would boost him some, but to be in VERY HIGH, which is like the top 5% of players I'd assume, it has to be using something other than wins.

It's called beating the people in the high bracket with high uncertainty. (Bayesian ELO or "TrueSkill" factor uncertainty) I have no facts to back this, but it seems completely reasonable to me. Much more reasonable than item or specific gamestat analysis.

 

[smackababy]

It's called beating the people in the high bracket with high uncertainty. (Bayesian ELO or "TrueSkill" factor uncertainty) I have no facts to back this, but it seems completely reasonable to me. Much more reasonable than item or specific gamestat analysis.

In 5 games he shouldn't have reached players of the ELO required to be in the high. Look at the chess ELO ratings. 1000 is considered a bright beginner and the median is 657. So, if we assume 1000 is what he started at, he couldn't get to what would be even the top 10% in 5 guys. Even if he played people of outrageously higher ELO. It has to consider some other things, such as performance. If I do awful and still win my first 5 games, there is no way I should be considered in the high. If it takes into account advanced "techniques" such as denying creeps, denying heroes when they are deniable (which is a high level move), getting good GPM for your role, being involved in team fights, having items helpful at the end of the game. Why shouldn't that be an indication of skill? You obviously know how to play and, therefore, are better than 75% of the players out there.

 

[boozie]

In 5 games he shouldn't have reached players of the ELO required to be in the high. Look at the chess ELO ratings. 1000 is considered a bright beginner and the median is 657. So, if we assume 1000 is what he started at, he couldn't get to what would be even the top 10% in 5 guys. Even if he played people of outrageously higher ELO. It has to consider some other things, such as performance. If I do awful and still win my first 5 games, there is no way I should be considered in the high. If it takes into account advanced "techniques" such as denying creeps, denying heroes when they are deniable (which is a high level move), getting good GPM for your role, being involved in team fights, having items helpful at the end of the game. Why shouldn't that be an indication of skill? You obviously know how to play and, therefore, are better than 75% of the players out there.

I'm not saying that CAN'T WORK as an indication of skill. I am saying that it is far more likely that it doesn't look at items you have. I will cede that it is entirely possible there is some sort of placement period where it looks at your individual (instead of team) performance. But I'm still inclined to stand by the idea that it's an accelerated metric for a new account and that someone with 500 games at 50% bracket couldn't duplicate those 5 games and wind up in Very High. It is important to have an uncertainty grace period. It helps make sure people (smurfs) go where they belong.

 

[smackababy]

I'm not saying that CAN'T WORK as an indication of skill. I am saying that it is far more likely that it doesn't look at items you have. I will cede that it is entirely possible there is some sort of placement period where it looks at your individual (instead of team) performance. But I'm still inclined to stand by the idea that it's an accelerated metric for a new account and that someone with 500 games at 50% bracket couldn't duplicate those 5 games and wind up in Very High. It is important to have an uncertainty grace period. It helps make sure people (smurfs) go where they belong.

If there is an accelerated metric for new accounts based just on win, it would put new players who get smurfs on their team in higher brackets, that would take longer to fall from and be fairly unenjoyable for them. It will, in time, balance out, but in a game like Dota the first 50 games are about learning. Constantly getting outplayed when you don't even know the items would be very discouraging.

 

[boozie]

If there is an accelerated metric for new accounts based just on win, it would put new players who get smurfs on their team in higher brackets, that would take longer to fall from and be fairly unenjoyable for them. It will, in time, balance out, but in a game like Dota the first 50 games are about learning. Constantly getting outplayed when you don't even know the items would be very discouraging.

Right, I just said it could look at individual performance. Regardless, one skewed game in the balance isn't that bad. (Not to mention if a smurf is leading the way on a stomp, the individual teammates will likely perform much better than they should anyways.)

 

[TheVrolok]

I'm not saying that CAN'T WORK as an indication of skill. I am saying that it is far more likely that it doesn't look at items you have. I will cede that it is entirely possible there is some sort of placement period where it looks at your individual (instead of team) performance. But I'm still inclined to stand by the idea that it's an accelerated metric for a new account and that someone with 500 games at 50% bracket couldn't duplicate those 5 games and wind up in Very High. It is important to have an uncertainty grace period. It helps make sure people (smurfs) go where they belong.

This is my guess. It certainly seems like these are all smurf games I'm playing. In a few there may be 1 or 2 true noobs, but it's generally far more experienced players who know what's going on. Seems like the smurf accounts got weeded out. What should be interesting is whether or not I remain in very high. My primary account queues between high/vh (occasionally normal) when I queue solo, but that is rare as I play 95% of my games with friends who have never gotten above normal.

 

[Arsinek]

Measuring of e-peens is always amusing. Especially when someone pulls the 'Let me 1v1 you' card to try to justify how right or awesome they are. Reminds me of those 12 year olds on xbox. And 1v1 sort of defeats the purpose of Dota 2, imo.

I do find it amusing all the smack he was talking but he wont 1v1 me. lol. Wheres the harm in a friendly 1v1 game?

 

[Dijeangenie]

I do find it amusing all the smack he was talking but he wont 1v1 me. lol. Wheres the harm in a friendly 1v1 game?

Am I the only one who doesn't understand 1v1? It's like playing 1v1 football ( soccer to most of you) it doesn't matter who wins because you play football with 11 players on a team...

 

[Arsinek]

Am I the only one who doesn't understand 1v1? It's like playing 1v1 football ( soccer to most of you) it doesn't matter who wins because you play football with 11 players on a team...

Well, if you play 1v1 football youre playing football just like if you play 11v11 youre playing football.

 

[smackababy]

He fails to realize I said he sucks in Dota, and he wouldn't 1v1 me IN DOTA. He wants to play LoL. That is like being called bad in soccer then saying "oh well 1v1 me in rugby!"

 

[Rhezuss]

Never bought it but it's showing as "Owned" on the game store page...
Might as well install it

 

[LucJoe]

Never bought it but it's showing as "Owned" on the game store page...
Might as well install it

f2p...

 

[Rhezuss]

f2p...

Ok but what's the $30 for the Early access?

 

[Arsinek]

He fails to realize I said he sucks in Dota, and he wouldn't 1v1 me IN DOTA. He wants to play LoL. That is like being called bad in soccer then saying "oh well 1v1 me in rugby!"

Just by your post, I can tell you are not good at LoL...

:sneaky:

Lol, whatever dude. I just find it funny. Ive never even 1v1'd in LoL before but with how full of yourself you are I think it would be interesting.

 

[smackababy]

Ok but what's the $30 for the Early access?

It is still in beta. And valve gives out tons of keys to send to friends, but also charges $30 for people who don't have any.

 

[Arsinek]

You throw out your experienced old age with 3+ "fanboy" accusations and a 1v1 BROWDOWN challenge. We get it, you like LoL more than Dota 2. That is your prerogative.

Can you at least find the edit key instead of constant double/triple posting?



Yes you can do key modifiers. Alt works as a key modifier. The rest can be made modifiers through console/autoexec.cfg modification. Complain as needed about having to use Autoexec, I am merely saying it is possible for your/anyone else's knowledge.

Whats that you say?

 

[Arsinek]

You throw out your experienced old age with 3+ "fanboy" accusations and a 1v1 BROWDOWN challenge. We get it, you like LoL more than Dota 2. That is your prerogative.

Can you at least find the edit key instead of constant double/triple posting?



Yes you can do key modifiers. Alt works as a key modifier. The rest can be made modifiers through console/autoexec.cfg modification. Complain as needed about having to use Autoexec, I am merely saying it is possible for your/anyone else's knowledge.

I cant hear you.



Stop being an ass.
-ViRGE

 

[Zeze]

Settle down kids. Skywrath Mage is released!

Patch notes:

Added Skywrath Mage!
COMMUNITY
Added stricter system implementing communication (voice and text) bans for abusive players.
Combined voice and text reports into a single 'communication' report.
Reduced reports per week to 4.
Removed ability for spectators to report players in-game.
Reports now allow single selection of behavior type only.
GAMEPLAY
Beastmaster: Fixed Wild Axes losing vision after the second cast.
Doom: Scorched Earth now shows its duration on the buff icon.
Lich: Fixed Frost Armor auto-casting while Lich is channeling.
Magnus: Fixed a rare crash involving Skewer.
Slardar: Fixed Slithereen Crush not slowing attack speed.
Spectre: Fixed Haunt illusions moving while out of the game.
Timbersaw: Fixed Timberchain travelling too fast.
Timbersaw: Fixed Timberchain not latching quite far enough.
Tusk: Fixed Tusk not issuing an attack order on the snowballed unit after the snowball crashes.
Enabled Drow and Tusk to Captain's Mode (Tournament Version next week)
UI
Added tab to the Watch section showing downloaded replay files.
VISUALS
Riki: Added Smoke Screen and Blink Strike animations.
Sniper: Added victory animations and more aggressive posing of idle, run and attack when enemy is in range.
BOTS
Bots now know to avoid standing in Macropyre and Ice Path.
WORKSHOP
Added a button to the preview that allows you to test your imported model in-game, in a local server.

 

[LucJoe]

It is still in beta. And valve gives out tons of keys to send to friends, but also charges $30 for people who don't have any.

Technically, the $30 is for a bundle of in-game items. With the added bonus of a beta invite if you happen to be living under a rock.

 

[Clinkster]

Settle down kids. Skywrath Mage is released!

Patch notes:

Added Skywrath Mage!

Timbersaw: Fixed Timberchain not latching quite far enough.

Nice, for those of you like me who do not know about Skywrath Mage...

http://www.dota2wiki.com/wiki/Skywrath_Mage

Also, I had a feeling Timbersaw's chain was actually shorter than it should've been. Really hoping this is really fixed...

 

[Zeze]

He is the ultimate mana hungry glass cannon.

He's worse than Timbersaw for mana requirement. His ult is 800 mana at every 20 cd. LOL.

He can barely cast it 2x with Bloodstone + Arcane, forget other 3 spells.

 

[Mr. Pedantic]

Can't see it. Do I have to restart Steam or anything?

 

[Clinkster]

Can't see it. Do I have to restart Steam or anything?

This is a general FYI:

The notes our fellow ATer posted were the DotA 2 TEST patch notes, which are released only for the seperate test client of DotA 2 on Wednesdays.

Valve basically takes a day to iron out game breaking bugs and will release the patch on Thursday (to the normal DotA client).

So you have to either dl the test client or just wait until tomorrow.

 

[Zeze]

Can't see it. Do I have to restart Steam or anything?

Above is right. You don't have to download test client if you have the game already. It'll configure itself.

 

[Mr. Pedantic]

Above is right. You don't have to download test client if you have the game already. It'll configure itself.

O I see.