• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Dota 2 |OT| To Hell and Back and Back to Hell

Status
Not open for further replies.
L

Lotto

Member
Finally got my 200th win tonight and with a semi-competent Meepo on my team to boot! A much need victory after having a terrible losing streak this week.
 
D

Dictator93

Member
SneakyStephan said:
Oh my fucking god, I just ran into a one in a million pub support while playing dragon knight.

A CM, he leveled his W
warded jungle entrance
carried clarities and teleport scrolls
pulled the jungle creeps, triple stacked 2 of the jungle camps and let me have the farm
When bounty was in our lane he bought sentries and didn't waste them by waiting for him to overextend.
He fucking BODY blocked a nighttime nightstalker (as cm with the pityful movement speed) for about 5seconds to save our nature's prophet till I could get in range for a stun.
He stopped attacking when enemies were super low to not take the farm.
He let me have last hits on towers
He denied as much as he could , he harrassed in lane while not getting his squishy butt owned by creep aggro like most pub players
He carried dust when we were pushing to deal with bounty hunter and put down really nice defensive sentries behind us every time we were taking down a tower.
And he was ALWAYS there for teamfights and pushes while doing all this.
Oh AND while playing hard support he found the gold to buy a mek , tranquil boots and double bracer on top of all those wards, clarities , dust and tele.


What a blessed experience, we wrecked their team so fucking hard.
I'm sad now as future pub games will look like crap compared to having this guy in your lane.
Click to expand...

So this is what meeting god would be like. Now you can die in peace my son.

Go back to your recent games list and friend this fellow
 
K

Kreunt

Banned
NBtoaster said:
I lust noticed axe illusions do full damage spins. Is manta axe a good joke build?
Click to expand...

not really, same amount of incoming hits, so same amount of chances that a spin triggers regardless of how many axe's there are.

edit: i suppose against something with a multi/split should it could slightly work.
 
J

JustinBB7

Member
I don't use mac/linux but everybody said the test client version ran like shit, did they fix that? Just wondering if it's good now.
 
A

Artanisix

Member
http://www.reddit.com/r/DotA2/comments/1ilxoq/psa_theres_a_valve_dev_actively_participating_in/

Really good post with Valve dev explaining games that seemed imbalanced. Shows how elos end up very close to even most of the time.

Yes, we use a Elo-type system. We have data that shows that it correlates well with a number of quality metrics.

Showing players their Elo has proven to cause a lot of unwanted behaviour. (The same problems exist in Chess.)

We know that there really is value in knowing how the match was formed, and even in knowing which players were the best, and also who was partied up with whom. It helps to analyze your own skill and try to understand how to get better. We are actively working to try to figure out what data we CAN provide to players (perhaps after the match) to help them do this analysis, without causing too many of the undesirable effects I mentioned. Right now, if there's a steamroll, people look at the only data available to them, which is totally understandable. Also, I believe that there have been problems in the past where it was not doing a good enough job segregating truly inexperienced players from experienced players. We've improved on that. However, our definition of "new player" probably doesn't match exactly with many players. (Their definition is often: "the other guy has fewer wins that I do.") And we have to deal with smurfs and players coming in from other games in the genre, where win count is definitely totally unrelated to the skill level of the player.
Click to expand...

Just one last comment on this. Elo is a TERRIBLE way to give players a sense of "progress." Many (most?) people reach a plateau, and their Elo stabilizes. It is simply not mathematically possible for Elo to keep increasing in general for players indefinitely as more and more games are played.

Given this reality, if players used Elo to measure "progress", we would constantly be reminding them that they are NOT making any. That would be really bad.
Click to expand...

If you post a match ID, I might be able to help explain why it decided to form the match. Either it will help us improve the tuning, or there's probably some piece of data that makes it make sense that we are not exposing.

We are looking for a way to provide information to players to help them understand why matches were formed in a way that doesn't lead to abuse of whoever the "worst" player is (according to whatever metric is preferable to the person giving the abuse). In the meantime, walking through some particular matches will help us better understand where it's still getting things wrong, or educate players about how it's making its decisions.

Just to clarify, there was never any "win rate" calculation. Ever. It is true that a goal of matchmaking is to make even teams, so that the odds of the Radiant winning any given match is 50%. The matchmaker also will raise your Elo and try to put you in players of equivalent skill, which indirectly tries to get the win rate to 50%. However, it has never looked at your historical win rate and, for example, put you in a game where it knew that you were expected to lose, to end a winning streak, or given you a stomp to end a losing streak.
Click to expand...

"Noob" is a relative term. We don't consider a person with 150 games to be a "noob". We have some good data that by 75 wins (approx 150 games), Elo is pretty accurate, and so we rely on it almost exclusively at around that point. If you are getting matched with those players, it should be because your Elo is approximately the same. Parties can complicate things considerably. I might be able to provide further insight into why it thought the match would be a good one if you provide a MatchID.

Is 75 the wrong number? Should a person with 2000 wins complain about playing with you, since you are a "noob" with only 1000 wins?
Click to expand...
 
H

Hylian7

Member
Artanisix said:
http://www.reddit.com/r/DotA2/comments/1ilxoq/psa_theres_a_valve_dev_actively_participating_in/

Really good post with Valve dev explaining games that seemed imbalanced. Shows how elos end up very close to even most of the time.
Click to expand...

How do you feel about winrates though? People use that for some kind of analysis for where they stand or how good they are, and I kind of think that is innacurate too. I can't keep myself above 50%, and I do a good amount of both solo queue and parties. Right now I'm sitting at 49.61% with 966-981 in my W-L. Some people think that's good, others think that's garbage. I think it's kind of irrelevant, as you are only 1/5 of your team.

I know you "don't believe" in the trench or whatever, and I don't believe I am in the trench, but whether you are in the trench or not, you will get games where you are doing your job and can't pick up the slack for 3-4 people on your team, thus causing you to lose.

The fact that this is a team game makes it make since why there isn't any kind of visible ELO, and I'm not even sure there should be.

At the very least, maybe they should take off of the Norma/High/Very High idea and maybe have five levels that are (optionally) shown on your profile about your relative skill level. This shouldn't be purely based on wins and winrates, and other factors should come into play. Factor in KDR on some type of heroes, wards bought and placed as supports, average number of heroes stunned with AoEs such as Reverse Polarity or Ravage.

Bottom line, I feel like ELO/MMR/ladder/whatever, only really works in games where you everything is based on you and you alone. Dota isn't one of those games.

I feel like the matchmaking is fair in it's current state.
 
B

Boken

Banned
Hylian7 said:
Bottom line, I feel like ELO/MMR/ladder/whatever, only really works in games where you everything is based on you and you alone. Dota isn't one of those games.

I feel like the matchmaking is fair in it's current state.
Click to expand...

they do use an elo system though, with a k-factor modification for new accounts

i think what you mean is that you believe that elo is an indicator of skill only in solo games - thats fair enough

but as a matchmaking system, elo is perfectly fine.


Artanisix said:
http://www.reddit.com/r/DotA2/comments/1ilxoq/psa_theres_a_valve_dev_actively_participating_in/

Really good post with Valve dev explaining games that seemed imbalanced. Shows how elos end up very close to even most of the time.
Click to expand...

douchebag neogaffers never inv to guild

hail dreamsvisions
 
S

Sibylus

Banned
The only thing a win percentage tells you is if you're generally climbing up the ELO or sinking.

...but even then, that's averaged over every single game you've ever played. It's only demonstrative over a long period of time.
 
S

SneakyStephan

Banned
Idk what elo is but I get matched up with an extremely wide variety of skill levels of other players.

All the way down to autoattacking, creep aggro self harrassing supports who fight carries over farm and that are as green or clueless as they come and don't know you can deny creeps or towers or what enemy heroes' abilities are , and all the way up to really skilled players who are very efficient and do all of the meta and efficiency stuff you learn over time and who pretty much carry the entire match (not dota carry, but carry the other players).

Even if you assume I'm somewhere in the middle, you often see those ultragreen players and godlike ones in the same matches, yesterday we had a guy who started playing literally the day before in the enemy team.

I should be getting matched with the green ones if I'm dogshit, or with the mediocre to great ones if I'm good, but I get a mix of both.
I feel like I've improved a lot over the past 200 games as I start to finally play each hero more than 4-5 times and learned how to counter pick and make better item builds each match, but one day I'll get a bunch of brilliant players and the next day I'm in the proverbial trench with the green players I described before , but usually there is a mix of dogshit and good players each game , with players that clearly do not belong in said match (too good or too hopeless compared to the rest)



Also the longer the matchmaking takes the wider that bar gets of the search range... so even if their elo system is good they seem to just ignore it by widening the search range after 3-4 minutes.
 
K

Kreunt

Banned
Hylian7 said:
How do you feel about winrates though? People use that for some kind of analysis for where they stand or how good they are, and I kind of think that is innacurate too. I can't keep myself above 50%, and I do a good amount of both solo queue and parties. Right now I'm sitting at 49.61% with 966-981 in my W-L. Some people think that's good, others think that's garbage. I think it's kind of irrelevant, as you are only 1/5 of your team.
Click to expand...

well in an elo system like he's talking about, you shouldn't ever be able to solo queue yourself very far above or below 50%
 
H

Hylian7

Member
I guess I forgot to mention one point in that: The IXDL used to have a rule that you had to have a 50% winrate. They dropped this rule, and rightly so, as I don't think it was that fair.
 
V

Vaporak

Member
Hylian7 said:
Bottom line, I feel like ELO/MMR/ladder/whatever, only really works in games where you everything is based on you and you alone. Dota isn't one of those games.

I feel like the matchmaking is fair in it's current state.
Click to expand...

These two opinions are incompatible, unless you think that who you are matched up with/against is unrelated to skill.
 
H

Hylian7

Member
Vaporak said:
These two opinions are incompatible, unless you think that who you are matched up with/against is unrelated to skill.
Click to expand...

I guess I meant that a VISIBLE MMR to measure your skill is a bad thing. Sure, MMR isn't accurate to skill, but it still forms good matches.
 
shira

shira

Member
The M.O.B said:
How is the best looking Drow bow a common?

256px-Jeweled_Longbow.png


Simple and Sleek
Click to expand...

Modelers think it needs to have sausages and dogs.
Reddit hates everything
Volvo jus wants $$$ and no lawsuits.
 
A

Aselith

Member
Speaking if ELO, had someone tell my team to let me take mid because being assigned pink meant I had highest ELO. Must have been an ex-HON guy.
 
Haunted

Haunted

Member
Late to the patch, love some of the new stuff.

Added automatic detection of GG calls in tournament and practice lobby matches, which then ends the game 10 seconds later.
Click to expand...
aw yeah.

I like that they're testing a surrender option, good job Valve.
 
I

Ikuu

Had his dog run over by Blizzard's CEO
Haunted said:
Late to the patch, love some of the new stuff.


aw yeah.

I like that they're testing a surrender option, good job Valve.
Click to expand...

They're not going to add it to the pub games, this is just letting pro games end without the admin having to enter a command.
 
Haunted

Haunted

Member
SneakyStephan said:
Idk what elo is but I get matched up with an extremely wide variety of skill levels of other players.

All the way down to autoattacking, creep aggro self harrassing supports who fight carries over farm and that are as green or clueless as they come and don't know you can deny creeps or towers or what enemy heroes' abilities are , and all the way up to really skilled players who are very efficient and do all of the meta and efficiency stuff you learn over time and who pretty much carry the entire match (not dota carry, but carry the other players).

Even if you assume I'm somewhere in the middle, you often see those ultragreen players and godlike ones in the same matches, yesterday we had a guy who started playing literally the day before in the enemy team.

I should be getting matched with the green ones if I'm dogshit, or with the mediocre to great ones if I'm good, but I get a mix of both.
I feel like I've improved a lot over the past 200 games as I start to finally play each hero more than 4-5 times and learned how to counter pick and make better item builds each match, but one day I'll get a bunch of brilliant players and the next day I'm in the proverbial trench with the green players I described before , but usually there is a mix of dogshit and good players each game , with players that clearly do not belong in said match (too good or too hopeless compared to the rest)

Also the longer the matchmaking takes the wider that bar gets of the search range... so even if their elo system is good they seem to just ignore it by widening the search range after 3-4 minutes.
Click to expand...
Did you play solo queue? Because otherwise, it was probably one of those really good players being in a stack with one of those really new players.


Ikuu said:
They're not going to add it to the pub games, this is just letting pro games end without the admin having to enter a command.
Click to expand...
Definitely a good change for tournament and practice games.
 
G

george01

Banned
the odds of the Radiant winning any given match is 50%. The matchmaker also will raise your Elo and try to put you in players of equivalent skill......
 
B

BraXzy

Member
The problem with matchmaking sometimes is, you get a few people with less than 100 wins, their team all has 300, yeah sure wins aren't everything but the people with more wins have more experience of the game by far, even if they might not be the best. It seems to have improved though.
 
A

Aselith

Member
BraXzy said:
The problem with matchmaking sometimes is, you get a few people with less than 100 wins, their team all has 300, yeah sure wins aren't everything but the people with more wins have more experience of the game by far, even if they might not be the best. It seems to have improved though.
Click to expand...

Or they're smurfs
 
K

KingKong

Member
Hylian7 said:
At the very least, maybe they should take off of the Norma/High/Very High idea and maybe have five levels that are (optionally) shown on your profile about your relative skill level. This shouldn't be purely based on wins and winrates, and other factors should come into play. Factor in KDR on some type of heroes, wards bought and placed as supports, average number of heroes stunned with AoEs such as Reverse Polarity or Ravage.
Click to expand...

CS:GO has 18 skill groups that it puts people in and tries to match them up and I think that works very well, and you can see yourself going up or down and see people's skill at the end of the game. http://blog.counter-strike.net/index.php/2012/10/5565/

I'm sure they're both using some kind of ELO matchmaking but just displaying a skill group to track and check your progress would be a big improvement imo.
 
RubxQub

RubxQub

φίλω ἐξεχέγλουτον καί ψευδολόγον οὖκ εἰπόν
BraXzy said:
The problem with matchmaking sometimes is, you get a few people with less than 100 wins, their team all has 300, yeah sure wins aren't everything but the people with more wins have more experience of the game by far, even if they might not be the best. It seems to have improved though.
Click to expand...
Wins mean nothing at all...like at all. OK, it means they've played the game a certain amount of time, but that's it.

With smurf accounts, and people that are apparently awful regardless of how many times they've played, someone's win-count is an extremely faulty metric to use to determine skill at all.

I've learned this the hard way a few too many times.
 
D

Drkirby

Corporate Apologist
george01 said:
the odds of the Radiant winning any given match is 50%. The matchmaker also will raise your Elo and try to put you in players of equivalent skill......
Click to expand...
Actually radiant win more games in pubs due to map layout, dire doesn't balance out until you hit about 40 minutes long. At least it is what one set of stats a few months back showed. The only hero who wins more on dire then radiant was Ursa.
 
Status
Not open for further replies.

Similar threads

The Verge: We played Valve’s secret new shooter: Deadlock
Opinion News 2 3 4
199
11K
near replied
Deadlock |OT|
|OT| Community 2
61
3K
MrSmooth replied
Coming Soon to Game Pass: Ara: History Untold, Frostpunk 2, and Wargroove 2
News Hype
42
1K
Jesb replied
The Legend Of Heroes: Trails Through Daybreak | Review Thread
Analysis Review 2 3 4 5 6
267
14K
Sept Terrion Of Nothing replied
Eiyuden Chronicle: Hundred Heroes | Review Thread
Analysis Review 2 3
143
11K
BadBurger replied
Top Bottom