[QUOTE=KnightLight;50508408]
[img]http://i0.kym-cdn.com/photos/images/original/001/134/819/cd8.jpg[/img][/QUOTE]
You belong in a garbage caaaan.
[QUOTE=Worstcase;50508389]Actually, being placed against significantly better teams and going on loss streaks is perfectly reasonable. If your MM skill rating was higher than your true skill rating than you'll naturally be placed against better teams where you will promptly lose more often than you win. That's by definition how the matchmaking system should work. Now whether or not such a case applies to you, I don't know. If you play in a large group where you get a boosted rating, its totally possible. Not to mention win/loss streaks are simply a natural occurrence just like a coin flip is likely to have head and tail streaks.[/QUOTE]
If your MM skill rating is significantly higher or lower than your true skill level that in and of itself suggests that the matchmaking is flawed.
New dump of aggressive Reinhardt :v:
[media]https://youtu.be/J-6x-ikVKIc[/media]
[media]https://youtu.be/i3gz9wAELPc[/media]
[media]https://youtu.be/ie4wcCFrFQQ[/media]
[media]https://youtu.be/171aKQPCFaU[/media]
tbh I feel like the steamroll streaks people are talking about is in a big way confirmation bias. Like, try and be fair about it, and record on a piece of paper whether each match was a steamroll, reasonably balanced, or a really good, close game. See how the results look, I'd bet a majority end up actually falling under category 2, and you're just remembering any category 1's more clearly because they stand out. I mean I could be totally wrong and you guys could be just having some really shitty luck in comparison to me or the matchmaking could even be rigged to encourage steamrolls, but I think before you put on the tin foil hats you ought to at least try a semi-objective way of checking. 50% winrate is statistically likely given a large sample size and players near your own skill level, that's not evidence of the game forcing steamrolls. It'd be interesting to see people's results regardless.
[QUOTE=Rufia;50508459]If your MM skill rating is significantly higher or lower than your true skill level that in and of itself suggests that the matchmaking is flawed.[/QUOTE]
It doesn't imply anything other than what you quoted me saying. Artificially high (or low) ratings happen all the time in games like DotA 2 or League of Legends. Both of which utilize comparably similar matchmaking systems. As far as the developers are concerned, it is perfectly legal to play with your significantly better friends even if it does skew your rating. Do you propose that that shouldn't be allowed because it causes a skewed rating and thus is 'flawed'?
At any rate, matchmaking systems never claimed to be perfect. They are inherently flawed. That much should be obvious.
Lost the last 6 games in a row. Haven't won a single game all day today. I wish this game had unranked play.
[QUOTE=Mort Stroodle;50508546]tbh I feel like the steamroll streaks people are talking about is in a big way confirmation bias. Like, try and be fair about it, and record on a piece of paper whether each match was a steamroll, reasonably balanced, or a really good, close game. See how the results look, I'd bet a majority end up actually falling under category 2, and you're just remembering any category 1's more clearly because they stand out. I mean I could be totally wrong and you guys could be just having some really shitty luck in comparison to me or the matchmaking could even be rigged to encourage steamrolls, but I think before you put on the tin foil hats you ought to at least try a semi-objective way of checking. 50% winrate is statistically likely given a large sample size and players near your own skill level, that's not evidence of the game forcing steamrolls. It'd be interesting to see people's results regardless.[/QUOTE]
Even if landslide victories happen at a fairly high rate, I don't understand why people would assume there is some forced winrate as opposed to the game just lending itself to landslide victories. Most competitive games have playerbases that are plagued with consistency issues.
The other issue is that landslides are often due to team composition rather than individual skill, which can probably screw with matchmaking.
I feel like it'll give me easy fights when I'm below 50% and when I'm above 50% I get matched against a McCree like plays like fucking Django and some autistic wizard sniper.
GODDAMN IT
[t]http://i.imgur.com/PEH6wpq.jpg[/t]
well let's open that loootbox
[t]http://i.imgur.com/rHfUz5Y.jpg[/t]
oh
[QUOTE=Mort Stroodle;50508546]tbh I feel like the steamroll streaks people are talking about is in a big way confirmation bias. Like, try and be fair about it, and record on a piece of paper whether each match was a steamroll, reasonably balanced, or a really good, close game. See how the results look, I'd bet a majority end up actually falling under category 2, and you're just remembering any category 1's more clearly because they stand out. I mean I could be totally wrong and you guys could be just having some really shitty luck in comparison to me or the matchmaking could even be rigged to encourage steamrolls, but I think before you put on the tin foil hats you ought to at least try a semi-objective way of checking. 50% winrate is statistically likely given a large sample size and players near your own skill level, that's not evidence of the game forcing steamrolls. It'd be interesting to see people's results regardless.[/QUOTE]
Yes, even in a completely random system you would have win/lose streaks.
I think the true issue is that steamrolls happen statistically too often regardless of matchmaking.
[QUOTE=Talishmar;50508764]I think the true issue is that steamrolls happen statistically too often regardless of matchmaking.[/QUOTE]
The point is that people are saying steamrolls are happening too often but they're just going off their memory, aren't backing it up with gathered stats.
[editline]13th June 2016[/editline]
and what's considered 'too often' anyway
I actually did a study on matchmaking at university, did some statistical analysis.
Assuming blizzard's matchmaker works how I think it does (pull 12 players of a similar skill rating into a group, then balance the teams) then the steamroll scenario makes a lot of sense to me.
What I learned from my analysis was if you manually balance the teams, you end up with more even matches, but to do that you need to swap players from team to team. If you can't do that, the matchmaker will give you teams of similar skill, but it can only do so well.
A quick example is, say you're solo queuing and you end up in a match with this spread:
(assume skill levels range from 1-10)
1,1,2,2,3,4,5,6,6,6,6,9
You can build 2 teams out of this that are fairly even:
9,1,6,5,2,2 (26)
vs
1,6,6,4,6,3 (26)
Stunningly even actually, my stats showed that manually balancing teams gave way better results than simply pulling players from the same skill bracket.
The problem happens when you have a premade team, instead of having all these numbers you can slide around, you get stuck with a single skill value for a team, which makes it *much* harder to give an even match, since you can't swap players you can't even out a disparity.
In a solo queue, a slightly unbalanced matchup can often be tweaked into a fair one, but in a team queue you have no recourse, so you end up with one-sided matches far more often.
[editline]13th June 2016[/editline]
I'm currently writing a simulator for the loot box system, does anyone know of a source where I can find drop rates and also the number of items in the game? Google just comes up with fucking terrible clickbait articles on 'how to best level up'
When you start losing or are getting steamrolled, spectate/observe what your teammates are doing. After I started doing this, I began to understand how/why we were losing so bad. Before I was just kind of focusing on what I was doing wrong and I feel that's what a lot of people are doing when they complain about the forced 50.
I'm rather curious as to why Hanzo of all people has an interaction with Junkrat about where he's hiding whatever treasure he found.
just had a 2 minute and 12 second game with 2 reins, mei, junkrat, hanzo and mercy on anubis offense, gg
[QUOTE=Furnost;50509032]just had a 2 minute and 12 second game with 2 reins, mei, junkrat, hanzo and mercy on anubis offense, gg[/QUOTE]
I think my anubis record is 1:47
I feel like you either storm your way in or bash your head against a heavily fortified wall for ten minutes
[QUOTE=Mellowbloom;50509083]I think my anubis record is 1:47
I feel like you either storm your way in or bash your head against a heavily fortified wall for ten minutes[/QUOTE]
I got 1:09 on Anubis once.
I love the game and it's very enjoyable to play. But I think there is something wrong with the map design on maps like Hanamura, Valskaya Industries and Temple of Anubis. There is something very frustrating that lies beneath these designs. One of those things are the very, [I]very[/I] close respawn rooms of the opposing team. Another is the impossibility to create a more diverse strategy on how to achieve a goal, like on Hanamura.
I think these maps need some redesigns.
[QUOTE=Antimuffin;50509126]I love the game and it's very enjoyable to play. But I think there is something wrong with the map design on maps like Hanamura, Valskaya Industries and Temple of Anubis. There is something very frustrating that lies beneath these designs. One of those things are the very, [I]very[/I] close respawn rooms of the opposing team. Another is the impossibility to create a more diverse strategy on how to achieve a goal, like on Hanamura.
I think these maps need some redesigns.[/QUOTE]
they seriously need to look to TF2 when it comes to designing maps. i dont know if its just because they have never really made an FPS game like overwatch before, but when it comes to map design in places like hanamura they shit the bed. the first point being a huge choke, with the only real "flank" being 2 inches to the left, and high enough that only heroes with the movement ability are able to utilize it, but it still isnt really a flank because theyll easily see you using that flank.
[t]http://i.imgur.com/hzWozrr.jpg[/t]
That's a pleasant surprise- and for 2 heroes I actually play, too.
[QUOTE=Furnost;50509181]they seriously need to look to TF2 when it comes to designing maps. i dont know if its just because they have never really made an FPS game like overwatch before, but when it comes to map design in places like hanamura they shit the bed. the first point being a huge choke, with the only real "flank" being 2 inches to the left, and high enough that only heroes with the movement ability are able to utilize it, but it still isnt really a flank because theyll easily see you using that flank.[/QUOTE]this. so much this. i'll be trying to look for ways to get around the defenders on hanamura/anubis/volskaya/basically any map and i just run into dead ends or walls and it's infuriating
with a 6-man premade you might be able to push thru those chokes well but in pub games (aka the majority of gameplay) it's pretty hard to
not to mention that the majority of the """flanks""" on those maps are small ass rooms right next to the entrance to the choke that you can't even get into without passing through the choke
[IMG]https://dl.dropboxusercontent.com/s/frgqcu69wbrpg8m/2016-06-13_12-33-30.png?dl=0[/IMG]
and my 2 most played heroes:
[IMG]https://dl.dropboxusercontent.com/s/2p02f8m7ozf914y/2016-06-13_12-35-09.png?dl=0[/IMG]
Does Reaper even have a purpose beyond hero limits with McCree as he is?
Still no Junkrat skin at all. I have at least one purple or legendary skin for every character now, apart from Junkrat, who main now pretty much.
[QUOTE=greeley;50509279]Still no Junkrat skin at all. I have at least one purple or legendary skin for every character now, apart from Junkrat, who main now pretty much.[/QUOTE]
I got the scarecrow junkrat in the first box I opened, but I stopped using it - it hides his face and that's where all junkrats character is
[QUOTE=Mellowbloom;50509293]I got the scarecrow junkrat in the first box I opened, but I stopped using it - it hides his face and that's where all junkrats character is[/QUOTE]
Jester skin fo life baybee
Sorry, you need to Log In to post a reply to this thread.
