|
Brosnan posted:Then it ceases to be a skill evaluation and becomes a measure of grinding. You grind less/get farther if you're better, but that doesn't really matter (see: people in Quick Play getting bent out of shape about level differentials even though those levels don't indicate skill in any way). He's describing the Quick Play system. It's a joke. Not a very good one, because it's not actually that complicated to make your MMR visible without going "HEY YOU ONLY GOT THIS MUCH HAVE A FULLSCREEN ANIMATION SHOWING HOW LITTLE IT IS," but a joke nonetheless.
|
#
?
Jul 5, 2016 19:27
|
|
|
|
| # ? May 10, 2024 00:33 |
|
|
Well, fixing the leaver issues will solve, like, 90% of my complaints I think. I'm glad they got to saying something so quick, but Kaplan's been pretty good about that kind of thing. The only other personal issue I have is if supports are, in fact, getting less XP. I have no way of testing it, though, because I am always support.  It's also important to remember that this is still a brand new system for CM. I wasn't in the Closed Beta, but wasn't all of this completely overhauled from what was tested originally? It makes sense there's be a few kinks if all it got was a week long PTR phase.
|
|
#
?
Jul 5, 2016 19:35
|
|
|
I think it might help to just have a more granular scale than 1-100. Losing a whole level seems to be a lot worse than in Dota where you lose a game and go from 3583 MMR to 3560 or something.
|
|
#
?
Jul 5, 2016 19:36
|
|
|
Maybe the screen should show every player's MMR going up or down. People might stop believing it always decreases more than it increases when they see their opponents get a big boost. It would also clarify whether certain character picks are getting shafted in particular.
|
|
#
?
Jul 5, 2016 19:37
|
|
|
stop referring to the sliding rating bar as XP
|
|
#
?
Jul 5, 2016 19:39
|
|
|
Supercar Gautier posted:Maybe the screen should show every player's MMR going up or down. People might stop believing it always decreases more than it increases when they see their opponents get a big boost. It would also clarify whether certain character picks are getting shafted in particular. I can see the chat logs now and they are glorious.
|
|
#
?
Jul 5, 2016 19:39
|
|
|
If you leave in competitive and don't just outright quit playing for a bit, but rather queue into something else, the game should go "oh you're just being a bitch? well, enjoy -25% damage and half hp on everyone you play for the next week"
|
|
|
#
?
Jul 5, 2016 19:40
|
|
|
Zoness posted:stop referring to the sliding rating bar as XP It is XP Xtra Pointless
|
|
#
?
Jul 5, 2016 19:40
|
|
|
Axel Serenity posted:Huh? That's not what I wrote at all, dude. I said no sport uses straight win/loss to judge individual performance. Team sports don't use win/loss to judge individual performance precisely because they're team sports. The win/loss of the player would be identical to the win/loss of the team, minus games lost to benching and injury, and would therefore be useless for identifying individual contribution. The typical Overwatch player is playing in an wide variety of environments, which makes the quality of their wins and losses a useful measure of their individual performance. I'd also point out that there is a sport that uses straight win/loss to judge individual performance: pitchers in baseball. Whether or not the stat as-is is actually useful, it at least conceptually makes sense because the w/l of a subset of games that a pitcher plays might vary significantly from the team's overall w/l. There's also the issue of sheer volume. A 10 year career in the NFL will come out to under 200 games played total. Many Overwatch players can go through 200 games in a week. The comparison is a bit less stacked for the sports with long seasons, but even there you're going through a year's worth of games every week. Finally, if you're agitating for sabermetrics derived ratings, consider the scale of the problem. An NFL ratings system only has to account for a bit more than 32 people per offensive position. Even if you have a system that evaluates, say, every NBA player that's still a total of ~300 people. You're familiar with the performance of most of these players, and if you notice someone being under or overvalued you just put an asterisk on your report and adjust your ratings for next season. For Overwatch, your sabermetrics system would have to be close to perfect because it's being used to make tomorrow's schedule. It's also being used to determine player salaries, so they have an interest in actively sabotaging your efforts. It covers well over one hundred thousand players, so individual tracking is useless. Oh, and the reason the playerbase is so large is because your single system needs to perfectly rate little league, junior varsity, varsity, every collegiate level, minor league, and professional league play, and it's also responsible for promotions and demotions between those levels. The level of complexity to the problem is nightmarish, and if the support rating penalty is real than Blizzard has already screwed it up.
|
|
#
?
Jul 5, 2016 19:45
|
|
|
I want to know what my Heals Above Replacement Lucio are.
|
|
#
?
Jul 5, 2016 19:48
|
|
|
https://twitter.com/C9_Adam/status/748045932911075330 I mean, it's anecdotal, but this was after they queued together for all their placements. Also like, hypothetically rating a support by just their healing output would be incredibly stupid, since damage boost and speed boost are essential to pushes. Zoness fucked around with this message at 19:54 on Jul 5, 2016 |
|
#
?
Jul 5, 2016 19:50
|
|
|
Wow, I understand it's anecdotal, but holy gently caress that would be aggravating.
|
|
#
?
Jul 5, 2016 19:53
|
|
|
Supercar Gautier posted:Maybe the screen should show every player's MMR going up or down. People might stop believing it always decreases more than it increases when they see their opponents get a big boost. It would also clarify whether certain character picks are getting shafted in particular. People would get TREMENDOUSLY salty seeing that lovely Hanzo on their team increase more/decrease less than they did.
|
|
#
?
Jul 5, 2016 19:56
|
|
|
Zoness posted:https://twitter.com/C9_Adam/status/748045932911075330 Would it be better if there was a per character mmr and then an average mmr?
|
|
#
?
Jul 5, 2016 19:56
|
|
|
Just showing straight mmr for everyone at the game end screen and yourself and your friends out of game would be so superior to any garbage they window dress it with. For the vast majority of dota players the mmr range in game is so tight only the truly salty ever make a deal about it (they are gonna bitch regardless so who cares).
|
|
#
?
Jul 5, 2016 19:58
|
|
|
I think the best is to just take skill rating with a grain of salt and understand what it does and doesn't mean.
|
|
#
?
Jul 5, 2016 19:59
|
|
|
ImPureAwesome posted:Would it be better if there was a per character mmr and then an average mmr? Per-character MMR is basically giving someone 20 smurf accounts for free. The costs outweight the benefits.
|
|
#
?
Jul 5, 2016 20:00
|
|
|
If anyone is interested, Cloud9's surefour solo queues on his alt, Tisumi, and last I saw, he was around 79-81 in competitive. Similarly, Cloud9's greggo solo queues on his alt, cuteboy420, and when I was watching, he was ranking up from around 61 to 65. The answer apparently to rank up in solo queue is to be good enough to carry your team. Surefour, for example, can solo carry on KOTH maps even against other rank 60 and 70 with his Tracer by just killing the other team by himself. It was honestly crazy. I don't think I've ever seen anyone dive onto a point and successfully 1 v 6 it and cap it. He ended up with like 100 kills or something. Greggo's alt seemed to do a bit more mixup between OP support (not surprising given his team role) and DPS carry (on Widowmaker while I was watching). I guess the game just expects someone who is really 10 or more lvls higher in skill rank to their teammates to be able to carry their team to a win. Edit: Sad for Adam, but how much individual healing are you getting done while playing with surefour or something? The top dps are probably just mowing down the opponents so the game can't get a good read on how good you are as a healer. Katreus fucked around with this message at 20:06 on Jul 5, 2016 |
|
#
?
Jul 5, 2016 20:00
|
|
|
Supercar Gautier posted:Maybe the screen should show every player's MMR going up or down. People might stop believing it always decreases more than it increases when they see their opponents get a big boost. It would also clarify whether certain character picks are getting shafted in particular. God this would be hilarious
|
|
#
?
Jul 5, 2016 20:00
|
|
|
also given that you don't queue up as a given character (because this would be an awful idea, despite this being how a lot of people would want to play the game) what would a character mmr even mean for the matchmaker
|
|
#
?
Jul 5, 2016 20:01
|
|
|
Papercut posted:People would get TREMENDOUSLY salty seeing that lovely Hanzo on their team increase more/decrease less than they did. Excellent.
|
|
#
?
Jul 5, 2016 20:02
|
|
|
Efexeye posted:seems like that's a legitimate complaint, no? I think a big problem with competitive solo queue is that a support can't "carry" a team. If your dps is bad (that includes tanks, as they do kill people in this game), a support won't suddenly make them kill things. Don't get me wrong, a good support is amazing to have around and can definitely be the difference between a win and loss, but they can't make a bad player hit their shots (or not be stupid and run in alone). So tl;dr is that I think good support solo-queuers are just purely at the mercy of the RNG of getting a few decent team members. I think your luck has just been poo poo and isn't indicative of the system as whole 
|
|
#
?
Jul 5, 2016 20:04
|
|
|
i've decided that all i care about is how many comp points i get towards my gold grenade launcher and that's let me feel a lot more chill about the whole thing it also helps to realize that even good players can get on a bad losing streak and drop 10 ranks in a weekend. it's okay! it's gonna happen. it's probably not your fault. 
|
|
#
?
Jul 5, 2016 20:05
|
|
|
Zoness posted:also given that you don't queue up as a given character (because this would be an awful idea, despite this being how a lot of people would want to play the game) what would a character mmr even mean for the matchmaker Well maybe the matchmaker would use the current mmr, but you could see the per character mmr so support mains would feel better?
|
|
#
?
Jul 5, 2016 20:05
|
|
|
It's likely they go with these obscure ranks and poo poo to hide the terrible matchmaking or mmr system in the back end.
|
|
#
?
Jul 5, 2016 20:08
|
|
|
subx posted:I think a big problem with competitive solo queue is that a support can't "carry" a team. If your dps is bad (that includes tanks, as they do kill people in this game), a support won't suddenly make them kill things. Yeah. I already have 14 hours in CM as Lucio alone, and it can be rough. I can hop around a payload all day, but if the team is bad, all I've done is made the other team win slower in the end. I'm trying to play other characters, though, and thankfully people are more willing to go Support the past day or so than when it opened. Apparently if I really want to carry a team, Master Overwatch says I should always insta-lock Torb. I am somehow ok with this.
|
|
#
?
Jul 5, 2016 20:09
|
|
|
subx posted:So tl;dr is that I think good support solo-queuers are just purely at the mercy of the RNG of getting a few decent team members. I think your luck has just been poo poo and isn't indicative of the system as whole  Anecdotally I've been as low as 39 as a support main but have no problems keeping up with my teammates in rank 56 matches v  v. It makes sense - when we do lucio-mercy and I pocket a good pharah on hollywood or route 66 streets that pharah is going to devastate our way through the second point, and I'm going to accomplish a lot more as symmetra when my team actually knows where to hold on dorado or gibraltar instead of overextending to the offense spawn.
Zoness fucked around with this message at 20:15 on Jul 5, 2016 |
|
#
?
Jul 5, 2016 20:11
|
|
|
It should be obvious if you think you are better than your team mates to pick someone other than the aoe heal bot with a weak rear end gun.
|
|
#
?
Jul 5, 2016 20:13
|
|
|
Zoness posted:
Honestly ranking supports has to be a really difficult task to give an algorithm to accurately measure performance. You can't just base it purely off "amount of healing" or anything like that. I think just giving everyone on the team the same amount of "ranked xp" would probably end up being more fair to everyone.
|
|
#
?
Jul 5, 2016 20:16
|
|
|
flashman posted:It should be obvious if you think you are better than your team mates to pick someone other than the aoe heal bot with a weak rear end gun. I have a couple friends that are amazing supports that have great game awareness and ability to stay alive that I can't match, but can't aim all that well and actually enjoy supporting.
|
|
#
?
Jul 5, 2016 20:20
|
|
|
flashman posted:It should be obvious if you think you are better than your team mates to pick someone other than the aoe heal bot with a weak rear end gun. The problem is that the teammates who are obviously bad (like who insta-lock Hanzo immediately at the start of every round and map) aren't going to switch even if the team has no support.
|
|
#
?
Jul 5, 2016 20:23
|
|
|
Supporting is fun. I like supporting. I feel I am at least reasonably good at supporting. Most importantly if I don't support odds are someone else isn't going to or will only do it grudgingly and unhappily which rarely leads to good play. If I lose a game I rarely felt like I could have changed it as an additional Soldier 76 with no support at all compared to as a Lucio or Mercy where at minimum I can revive a dead team or allow for a strong push to help carry weaker players over a hump.
|
|
#
?
Jul 5, 2016 20:23
|
|
|
Papercut posted:The problem is that the teammates who are obviously bad (like who insta-lock Hanzo immediately at the start of every round and map) aren't going to switch even if the team has no support. Also the Mei that blocks your ults. Like what do you even say aside from frantically yelling 'drop it drop it' into team comms when you hammer down 4-5 enemy players and your Mei ice walls them into safety?
|
|
#
?
Jul 5, 2016 20:26
|
|
|
Blizz needs to put Avoid Player back in for comp mode.
|
|
#
?
Jul 5, 2016 20:27
|
|
|
Essentially what the problem seems to be is that good teams, which steam roll the enemy, don't actually give healers a lot to work with. They might take scattered damage here and there, but for complete stomps it's at a point where the enemy are dying a lot, so the DPS is doing their job, damage isn't coming through, so the tanks are doing their job, and there's only a scattered few people who need to be topped up, so healers don't really get to do their job. Or rather, Mercy doesn't get to do her job since people aren't taking damage, she's not really healing that much, damage boost isn't that great, and she's not actually getting any eliminations/damage done herself. So the ideal situation for Mercy is a long, drawn out battle where your team struggle but ultimately win rather than you rolling over them. Lucio/Zenyatta are probably in a better position than Mercy as they can contribute meaningfully to the fight aswell as heal.
|
|
#
?
Jul 5, 2016 20:32
|
|
|
Kerrrrrrr posted:i've decided that all i care about is how many comp points i get towards my gold grenade launcher and that's let me feel a lot more chill about the whole thing After a few nights in the 45-55 range, this is truly the best attitude. Play comp over QP because comp gives out >0 comp points. After a few nights, I've found very little difference in the chill levels, team comps, or winrate between comp and pre-comp QP. Also, every QP game I've played since ranked came out has been absolute garbage. There's a tryhard queue and a practice/warm-up queue now. That's fine, but being as I'm not prone to Big Emotions I'm not sure why anyone would play QP for fun now. It's just worse games for less reward.
|
|
#
?
Jul 5, 2016 20:35
|
|
|
I really wonder how many people care about XP and rankings as they play the game Then I wonder if directly showing people how much XP they've forfeited as a stat on their player profile would help reduce the amount of ragequitting or quitting just because their team is doing bad
|
|
#
?
Jul 5, 2016 20:35
|
|
|
Zoness posted:Also the Mei that blocks your ults. Like what do you even say aside from frantically yelling 'drop it drop it' into team comms when you hammer down 4-5 enemy players and your Mei ice walls them into safety? I got invited to this group yesterday out of solo queue, so it wasn't clear who knew who there. This one member is like 5 ranks lower than the rest of us, and over the course of about 8 games he insta-locks Mei every time, regardless of scenario or team comp. I'm trying so hard to hold my tongue about it because i don't know if these guys are friends or whatever, when finally another member of the group is like "dude do you ever play anything other than Mei?" The worst part about it was that in like 3 of those games an enemy Mei was terrorizing us and the obvious team comp change was for our Mei to switch to Pharah or Junkrat but nope.
|
|
#
?
Jul 5, 2016 20:36
|
|
|
Ddraig posted:Mercy doesn't get to do her job since people aren't taking damage, she's not really healing that much, damage boost isn't that great, and she's not actually getting any eliminations/damage done herself. Damage boosting can make a lot more characters one shot people, that's kinda huge.
|
|
#
?
Jul 5, 2016 20:37
|
|
|
|
| # ? May 10, 2024 00:33 |
|
|
Lant posted:Damage boosting can make a lot more characters one shot people, that's kinda huge. It's great. but it's not necessarily great when it comes to Blizzard's weighting for contribution to teams, it seems. A lucio doing chip damage to a bunch of people with his gun/knockback/punch or Zenyatta slamming his balls at people seems to contribute a lot more to whatever algorithm Blizzard is using to determine how well you're playing than just pressing M2 on a DPS character. e: I wouldn't entirely be surprised if Mercy getting lower contributions than other people is an unintentional bug by them 'fixing' whatever it is that caused her ult to charge when damage boosting someone shooting shields. The way Blizzard have handled this entire competitive update makes me wonder if this is not the case, since they've hosed up everything else. Rush Limbo fucked around with this message at 20:43 on Jul 5, 2016 |
|
#
?
Jul 5, 2016 20:40
|
|