Due to a security update, you may have to reset your password. Don’t panic, nothing has gone wrong and your password is safe. If you don’t have access to that email, send Tube a message at [email protected] More info here: https://status.vanillaforums.com/incidents/2zdqxf3bt7mj
Our new Indie Games subforum is now open for business in G&T. Go and check it out, you might land a code for a free game. If you're developing an indie game and want to post about it, follow these directions. If you don't, he'll break your legs! Hahaha! Seriously though.

[Overwatch] Ashe is now Live!

13567101

Posts

  • KreutzKreutz Registered User regular
    Flippy_D wrote: »
    Kreutz wrote: »
    I went 0-7 yesterday during my (dis)placement matches playing Mercy, including a 7-round slog on Anubis that I was just happy to be done with. Then I played some 6v6 randos, and found myself tearing it up with Moira so I tried her out for the remainder of my placement matches, because what's the worst that could happen, right? Well, I won the last three matches and placed gold for healing in all three. I think a lot of the difference is not having a giant bullseye on my character as soon as the wings come out, as well as being able to do something about gank attempts. Which isn't to say that Mercy's pistol doesn't work in certain situations, but compared to a tracking DoT with minor lifestealing capabilities it may as well be a peashooter. I don't think I mind giving up the damage boosting abilities in exchange for that.

    Am I the only one who thinks Moira is OP? Entirely possibly to get gold for healing and silver for damage without even being that good, plus autolock aim.

    x06k3vf6j092.jpg

    You know, I think I agree with you.

    (That probably reflects poorly on my group rather than well on me, though. They decided to defend on Hollywood without a tank. They lost miserably.)

    ChanceFlippy_D
  • Flippy_DFlippy_D Digital Conquistador LondonRegistered User regular
    edited September 2018
    Flippy_D wrote: »
    Okay, sooooOOOOooo I was waiting for the new thread to do this, but here are the results of a bunch of data crunching. There's a lot to say and this took several days to write, so uh, I apologise in advance.

    There does need to be context as to my motives, first: I'm really good at shooters and always have been. CounterStrike, TF2, Battlefield, anything, you name it, I was usually topping the leaderboard and in or around the semi-pro scene. I've won money playing FPS and sniper classes have always been my go-to.

    So Overwatch is released and I'm like, hell yeah Widow. But I play like a hundred hours with her and it just doesn't click. I find this weird but not particularly upsetting, in part because I'm getting to quite like the style of shooty boi Hanzo.

    Competitive begins and I go for it, and although nobody likes me playing Hanzo and the hero is generally regarded as a liability, I do reasonably well. 51% winrate. Rank 50 high. Basically the lowest end of Plat. That'll do me. Season 2 I get my historic high of 2531, again just on the edge of Plat, and even though I then have a rough patch and end up 2008, low gold, that feels like mostly bad luck. Highs and lows, etc.

    So to recap: always been pretty good at FPS but never super-pro; OW S1 and S2 positioned me as... pretty good, but not super-pro. That's cool! Reasonably happy about this. Having fun in-game and playing against, and with, pretty good people.

    The along comes Season 3, and everything starts going wrong. I don't feel like I'm playing that much differently, but I qualify much lower and start to sliiiiiiiide. My season high of 2280 felt like, okay, I'll just have to climb back up into low Plat. But instead I slipped all the way to 1935, high silver.

    What got me about this was less the losses themselves (annoying though they were), but more how the nature and impact of these losses felt totally arbitrary. Yes, leavers and idiots, but one had to assume that those were more or less equally distributed. It did feel that my more risky, skirmishing, strategic style of play was less effective when not backed up with a competent team, but it also felt like the way I went up and down after wins and losses was not a fair reflection of how I was playing. I was playing my arse off, but to no avail. I got discouraged, defeated and also just quite angry, which I disliked. I decided OW competitive - at least at this lower skill level - was too toxic and making me into someone unpleasant, so I quit the game entirely. At the time I also came to the conclusion that Hanzo's net utility was also probably not good enough, and that I was playing with the wrong deck.

    Time passes. I get suckered back in. I'm playing well and dominating QP, and so after initially hard avoiding competitive I try it again. But nothing has changed - in fact, it's got worse. Qualified low gold, which was bad enough, but after a senseless losing streak of 14 games more or less in a row, I'm actually laughing in disbelief at my rank of 1852. It seems to have absolutely nothing to do with how well I'm playing or not. The teams are dreadful. I always seem to be working with morons. The last loss had seen the first round absolutely crush us in just 1:50. These just aren't fair games. There's nothing I can do about it.

    Ultimately, I slid seriously hard, into the bottom ~7% of players at the lowest point. The entire time, I felt more or less powerless. My efforts, abilities and tactics seemed to make little difference to enemy teams who were often just better, net, than mine. My net winrate at times was as low as 38%.

    6bn2ykg3uboi.png

    But I'm a data guy and so I start to gather some. In fact, I gather 47 metrics each game, every game, for 166 games. I also gathered all the data the game shares with you at a rolled-up level; a further 30 datapoints. What I learned was... well.

    Just presenting the data would be a total splurge, so instead I'm going to position these as a series of hypotheses.

    1. I am getting worse

    An extremely reasonable take! You're losing skill, less effective in games, and so ranking down accordingly. Fair enough.

    But, doesn't look like it. For example, indexed data from my highest-played character, season-on-season (red bar is season 9, all benched vs earliest season I've played that hero):
    ajo46jb10yl5.png
    There are some ups and downs but as you'll see, some of the core 'skill' metrics are strong:
    • Accuracy up 35%
    • Crits up 53%
    • Damage per 10 mins up 59%(!)
    • Elims/10 still 5% off S1 but getting better season-on-season (On Fire and Final Blows are similar stories)
    • Obj time up 119%
    • Gold medals up 32%
    • Crit

    Other heroes are here: https://imgur.com/a/wJJIkMc.

    On most of these, there are meaningful improvements (red bar going right or higher than dark green bar). Ironically, the two heroes I've statistically improved least with - Junk and Bastion - also have my highest winrates¯\_(ツ)_/¯

    But it's hard to make a convincing case that I've really got worse over time.

    2. Everyone else has got better

    Doesn't matter if you're going at x acceleration if all other traffic is going at x2. The playerbase has probably also fallen off a bit, meaning fewer poor opponents and more good ones, proportionately. I personally reckon the second point is more salient, given that I do in fact meet some clearly good players in Silver.

    But again, data doesn't indicate such.

    35ofwipebd7n.png
    (this screenshot is contemporaneous to the end of season 9, so should include all of it - registering, for example, 65 games with Hanzo).

    Overbuff, which compares your data to others, puts me firmly in the top 50-10% for most of what would be thought of as 'key' metrics. This is data from competitive. I even die much less than other people. So it doesn't feel like I'm so terrible compared to this dataset, which is itself already skewed towards top players.

    3. I wasn't being a team player!

    I always try to be a team player but ultimately, DPS is DPS. This confess this was haunting me a bit (it's a team game!) so I deliberately stepped away from a more solo mindset and started to play healer, a servant to my team. I got quite a lot of compliments on my healing, for what it's worth.

    I guess you could say 'playing healer doesn't necessarily make you a team player', but I really poured myself into it. Again, looking at the Overbuff data in the prior post my healing output seemed to be absolutely top-end compared to most competitive. I also tried to more explicitly play the objective on all heroes (I always play the objective, but for example I would previously, when capping a point, move ahead of my team to try and repel/thwart counterattacks).

    So, I really did go for being a team player. I played majority healer in 40% of all matches and as a secondary function in 30% of matches.

    The unfortunate irony here is that I seem to have contributed more as a DPS, my natural home. When playing DPS, we won more often, and those wins were more decisive, than when I was healing:

    1kp46yzv4yef.png

    This was despite my actual performance being relatively comparable across both win and loss scenarios:

    39bg1engoyep.png

    This tallies with this Reddit post making a pretty convincing case that the best thing you can do for your team is play the hero you win the most with. I am a good healer relative to the playerbase, but a better DPS.

    So: massively upped my 'team player'ness, but ultimately the outputs seemed to indicate that this didn't really make a huge difference to winrate, and may have even slightly negatively affected it.

    4. I wasn't playing the right 'kind' of game

    This is still one of the hardest hypotheses, partly because it's vague. I don't really know how to prove it one way or the other. Included for sake of completeness, but will have to come back to it some day.

    Certainly my objective time massively increased, across the board (~39% more as healer, 81% more as DPS and 982% more as Bastion specifically!). I changed heroes more, always responded to team needs to fit 2-2-2 and so on, and communicated actively. I used to play raid leader in WoW so I know how to work in groups. I currently have a '4' on my teammate score, split ~25% sportsmanship, ~15% leadership and ~60% good play.

    Didn't help my winrate, though.

    With that said, one thing I did notice and certainly the first actual concrete takeaway was that I wasn't doing well on attack/defend maps (Hollywood, Lunar, Anubis, Volks., BlizzWorld, Hanamura):

    008f582751h0.png

    This is something I'm going to have to think about, although it's not lost on me that these are arguably the maps that require the most organisation and teamwork. Nor that my best successes come from tight, killzone-based maps.

    5. The -/+ system just isn't very good

    This was, to be honest, my chief suspicion. Yes, I admit it, the suspicion probably comes from wounded pride, but I just didn't feel like the dynamic range on how much SR is awarded to subtracted after a win or loss (-/+16 to -/+32) was consistently reflective of in-game quality. This wasn't helped by some of my worst point deductions (based from the minimum of -/+16) came from my closest games (note also that there were far more close games when winning than when losing!):

    5oe8acwfyv1z.png

    So, I made my own metric.

    Every single game, I recorded the data from the aftermath screens. Again, this was for 166 games.

    w1go9er7q44u.png

    From this, I made a score that indexed/benchmarked my performance on my own rolling averages. Without going too much into it, it also discriminated between healing and dmg (so I wasn't judging Mercy on kills, for instance) and weighted certain metrics more (e.g. so medals weren't over-important).

    I also allowed, from game 80 onward, for the addition of a modifier if I thought my scoring system wasn't quite working. I used this option 20 times (vs 86 games), mostly for healing wherein the range of influencing metrics was slimmer.

    The results of my custom score vs Overwatch's score were interesting.

    Turns out, the question of 'does the -/+ line up with my self-perceived performance?' is quite difficult to answer. Mainly because, whilst Overwatch always assigned me -16 to -28 on a negative for a loss, and +16 to +32 for a win, there were obviously games where https://youtube.com/watch?v=1TCX90yALsI]I performed well but was still on the losing side.

    In fact, I put in above-average performances (always my personal average) in ~45% of losses and ~55% of wins. It's hard to know to what extent being beaten influences that - in other words, am I playing below average, contributing to a loss, or is being outmatched the equivalent of cutting 18% (or 10 pc pts) of my performance? For example in losing games I did 46 less damage per minute, but I can't DPS effectively if I'm receiving no support or if the enemy team is well-organised with shields etc, right? I think it's a bit of an impossible question to answer without more data, and we also won more when I played DPS, so... hard to say.

    Nonetheless, here is one way of looking at it:

    gm16a4y4fen4.png

    What this shows is not my absolute rank change, but rather the rank change expressed as a percentage of its deviation from average. In other words, if I lost -16 that was actually better than the average loss of -22 and so, therefore, actually measured as +27%. Equally, a win that only gave me 16 points was, although a positive score, realistically a poor reflection on the game's assessment of me, and worth -29%. This controls for the straight positive/negative dichotomy. My own score was similarly measured in the sense of deviation from the average.

    So just to be clear, this isn't wins on the left and losses on the right. This just performance vs the game's reward of my performance.

    ... and frustratingly, the trendlines are very similar. The deviation is likely mostly explained by 'performing' worse during losses (see above) and MoE.

    The reason why this is frustrating is because it is does suggest that my way of gauging how well I played is, in the macro, in line with the game's. This means in short - much as I don't like it - that the scoring is accurate when taken as a net. Another surprise was that Overwatch was just, slightly, giving out more for wins - average 22.97 vs average -22.19, so 0.46 more.

    Where it falls over, however, is on individual matches. As you can see, there are plenty of matches where my self-assessed performance outstripped the relative score from the game. Bear in mind this is mostly not manually done, but is a score automatically generated from my class metrics. If we prioritise my metrics instead, we can see what's up:

    ccyl9s578xgi.png

    Namely, it looks like the way Overwatch deals with wild variance in performance is to flatten it - yes, triumphs aren't rewarded as much, but losses are also not so harshly penalised. In fact, my system is marginally less forgiving, on paper.

    Still, although there's this data it still feels slightly inconclusive, and I still think that Overwatch does a generally poor job of responding to in-game actions. Presumably, the flattening is to account for that very discrepancy, which would explain why the trendlines in the first graph are so similar...

    ... but I'm still not entirely assuaged here. I fear I shall never be.

    ---

    So, what are the conclusions here?
    • I am getting worse - data doesn't support this, I'm actually improving
    • Everyone else is getting better - data doesn't support this, my relative metrics vs others are pretty favourable
    • I wasn't being a team player - no data, but I significantly adjusted my playstyle and ironically appear to still be contributing most as a DPS
    • I wasn't playing the right kind of game - somewhat supported! I am proportionately worse at attack/defend maps and need to figure this one out
    • The way Overwatch scores you isn't properly reflective of in-game performance - data inconclusive. In some ways it appears no, in other ways it seems yes. I grudgingly think it still gets it wrong, but have to admit it probably doesn't have as much an effect as I feel it does. Although as a bonus, I played above average in 68% of 'close' games.

    Where am I left after all these numbers and analyses? Somewhere obvious: teamwork is in fact more important than any individual metric or how well you perform, and you are basically therefore at the mercy of being in a good, or (if you're a leader type) attentive, team. I think it no co-incidence that two of my three 'climbs' were when I (by chance) found a decent team and stuck with it:

    72crn8ajxkfj.png

    To me that feels a bit... shit. For two reasons:

    1) It's still just the goddamn luck of the draw. no matter how well I play, or don't play, personal performance can only ever put the brakes on a bad team performance. I can't enforce comp, I can't really make people notice shit, and I can't 'carry'. That last point is, incidentally, where I wound up last time, without data: that the classes I play can't carry like an insane Doomfist, Genji, Moira, Brigitte, or to some extent Rein or Sombra, can. I don't happen to be practiced with those heroes, either.

    2) If that Redditor's analysis is correct, Overwatch could probably do better on distributing people into teams! It could easily figure out the propensity for people to play heroes with high winrates and calculate the projected winrate. From there, it could create fairer matchups. But I guess that will never happen, and so you'll always be subjected to stomps and (it never feels often enough) delivering stomps.

    Anyway, that's more than enough. I leave you with specific hero win data, which I'm now at least using to try and give myself the best chance of being useful on any game that I join.

    Because, I after all, I really, really want to win.

    hpeisq2sn3mn.png

    I hope nobody minds me 'bumping' this on the new page but goddamn I wrote like 2,700 words based on hundreds of hours hand-coded data, I would like engagement.

    I guess I just wanna know if a) anything thinks I've missed something or drawn the wrong conclusion, b) has an alternative view, or c) can tell me how to fix my apparent weird blackspot of attack/defend maps.

    Flippy_D on
    p8fnsZD.png
    soylenthChanceFencingsaxDevoutlyApatheticDyvim TvarUnluckyMusicoolH3Knuckles
  • TexiKenTexiKen 1, 2, 3, 4, 5, 6-Pack Yes! Registered User regular
    America the greatest as always.

    Q8ghZFx.png
  • -Loki--Loki- Don't pee in my mouth and tell me it's raining. Registered User regular
    Flippy_D wrote: »
    ...

    Where am I left after all these numbers and analyses? Somewhere obvious: teamwork is in fact more important than any individual metric or how well you perform, and you are basically therefore at the mercy of being in a good, or (if you're a leader type) attentive, team. I think it no co-incidence that two of my three 'climbs' were when I (by chance) found a decent team and stuck with it

    Yeah this is pretty spot on. I feel like this game is closer to a PvP RPG like Guild Wars than to a normal team shooter like Counter Strike or TF2. Each character has a role, above and beyond what class they’re in. If people know their role and build a team accordingly, you generally have close, fun matches even with minimal communication against another team doing the same.

    I find I can solo heal as Mercy pretty well and steer the engagements in our favour, but if the blocking tanks are trying to DPS or the DPS tanks are trying to be fast flankers, and the fast flankers are trying to be frontline tanks and the supports are trying to be DPS, or any combination of not understanding roles, and the other team is even slightly more on the ball, we’ll lose hard.

    If I pull my pistol out as Mercy to stop a flanker, someone didn’t do their job. Likewise if Reinhardt drops to a half hearted push, I didn’t do my job (probably because I had my pistol out dealing with a flanker). Both of those cause the team to suffer as a whole.

    People were piling on Reynolds last thread because he was saying he was dropping in position because he was having shit luck with teams and saying it’s not luck, but it kinda is. There’s only so much you can do to help the team win. If you got put on a team who don’t understand their roles, and the other team do understand their roles, at the same rank, that’s not skill, that’s luck. You can try to self improve at that point but you have to accept you’ll probably lose ranking because of it.

    Reynolds
  • The Escape GoatThe Escape Goat Brave Sir Robin Registered User regular
    Chance wrote: »
    Flippy_D wrote: »
    Kreutz wrote: »
    I went 0-7 yesterday during my (dis)placement matches playing Mercy, including a 7-round slog on Anubis that I was just happy to be done with. Then I played some 6v6 randos, and found myself tearing it up with Moira so I tried her out for the remainder of my placement matches, because what's the worst that could happen, right? Well, I won the last three matches and placed gold for healing in all three. I think a lot of the difference is not having a giant bullseye on my character as soon as the wings come out, as well as being able to do something about gank attempts. Which isn't to say that Mercy's pistol doesn't work in certain situations, but compared to a tracking DoT with minor lifestealing capabilities it may as well be a peashooter. I don't think I mind giving up the damage boosting abilities in exchange for that.

    Am I the only one who thinks Moira is OP? Entirely possibly to get gold for healing and silver for damage without even being that good, plus autolock aim.

    She can do a lot of healing or damage, sure, but it's more difficult for her to do impactful damage or healing. It's like the Torb turret problem.

    I dunno, I think Moira can be a pretty reliable problem for D.Va, Genji and Mercy in particular. Her primary aim is so forgiving that you can just keep it on a Mercy as she flits around, stopping her self-heal.

    D.Va being in that list is surprising, as I normally think of her as an outright counter to Moira. The beam really doesn't do that much to the mech if it's being healed at all, and a flash of DM eating an orb for almost no resource cost is a huge blowout.

    3clipse wrote: »
    God is dead and life has no purpose.
    ChanceKana
  • ChanceChance Registered User regular
    Chance wrote: »
    Flippy_D wrote: »
    Kreutz wrote: »
    I went 0-7 yesterday during my (dis)placement matches playing Mercy, including a 7-round slog on Anubis that I was just happy to be done with. Then I played some 6v6 randos, and found myself tearing it up with Moira so I tried her out for the remainder of my placement matches, because what's the worst that could happen, right? Well, I won the last three matches and placed gold for healing in all three. I think a lot of the difference is not having a giant bullseye on my character as soon as the wings come out, as well as being able to do something about gank attempts. Which isn't to say that Mercy's pistol doesn't work in certain situations, but compared to a tracking DoT with minor lifestealing capabilities it may as well be a peashooter. I don't think I mind giving up the damage boosting abilities in exchange for that.

    Am I the only one who thinks Moira is OP? Entirely possibly to get gold for healing and silver for damage without even being that good, plus autolock aim.

    She can do a lot of healing or damage, sure, but it's more difficult for her to do impactful damage or healing. It's like the Torb turret problem.

    I dunno, I think Moira can be a pretty reliable problem for D.Va, Genji and Mercy in particular. Her primary aim is so forgiving that you can just keep it on a Mercy as she flits around, stopping her self-heal.

    D.Va being in that list is surprising, as I normally think of her as an outright counter to Moira. The beam really doesn't do that much to the mech if it's being healed at all, and a flash of DM eating an orb for almost no resource cost is a huge blowout.

    My own theorycraft agrees with you - D.Va's ability to full counter a huge piece of Moira's kit is a massive blow, and D.Va's boost CD is comparable to Moira's fade, meaning she can easily chase down a scared Moira. Moira's damage is almost inconsequential without the orb, and her healing only becomes really powerful with it - it's very important.

    Buuut Matrix is no defense against Moira's primary and - most importantly - my D.Va-main brother complains about Moiras endlessly, and far more than any other hero. So it's quite anecdotal but to him there's nothing worse than a Moira on the enemy team.

    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
  • soylenthsoylenth Portland, ORRegistered User regular
    I had a moira in MH take me on as Bastion in a super-confident carry moira kind of way and ... I literally just sat there in turret mode healing and laughing (because her primary fire does not enough damage in turret mode to overcome self-healing) and then obliterated her once it got boring. I mean, if her team had started helping I would have been in trouble, but as it was ... that tickles.

    Also, people on PS4 think that jumping in circles around me in turret form is just going to confuse and fright me and, I'm sorry, did you notice the massive chain gun?

  • ChanceChance Registered User regular
    soylenth wrote: »
    I had a moira in MH take me on as Bastion in a super-confident carry moira kind of way and ... I literally just sat there in turret mode healing and laughing (because her primary fire does not enough damage in turret mode to overcome self-healing) and then obliterated her once it got boring. I mean, if her team had started helping I would have been in trouble, but as it was ... that tickles.

    Also, people on PS4 think that jumping in circles around me in turret form is just going to confuse and fright me and, I'm sorry, did you notice the massive chain gun?

    In their defense, jumping in circles around someone while landing shots on them is generally pretty effective!

    I'm a terrible Bastion >.< using movement to fine-tune my aim is like hardwired into me, so I'm decent in recon mode and I can hit tanks in sentry and that's about it. And I'm always healing when I want to transform, and transforming when I want to heal - I'm just bad in every direction lol

    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
    Pailryder
  • soylenthsoylenth Portland, ORRegistered User regular
    Not as moira, she just doesn't do much to you if you stay calm. Tracer or Genji, yes, more worrisome. I'm not saying I don't get ultra-murdered on bastion some games, but generally he's one of my "enough of this bullshit" picks, especially if said bullshit is big and/or shieldy. Every since they upgraded his rifle he's decent in recon mode, especially at picking off pharahs and the like, and still great at tank busting. I just try to not being where they're expecting. If they know where I am it's time to move, unless I'm paired with my own shield tank in which case I just stick with them. Once they know where you are and that you can aim they will focus half the team on you which is generally not survivable, as you can't depend on rando teammates to support you. So get up and move!

    Whether you should stick it out in turret mode* and burn them down or get up and take your chances in recon is situation and hard to call sometimes though, I agree. Sometimes you're just boned either way! Contemplate the meaninglessness of the universe as you fall to pieces.

    *I am learning which encounters I can brute force and which I can't through trial and regrettable error. For instance, Zarya is very killable 1v1 in turret mode. You will charge her to full, yes, when she pops a bubble, but she won't have much time to make use of it as you immediately burn her down in a couple seconds.

  • ChanceChance Registered User regular
    soylenth wrote: »
    Not as moira.

    Well true lol

    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
  • miscellaneousinsanitymiscellaneousinsanity grass grows, birds fly, sun shines, and brother, i hurt peopleRegistered User regular
    Chance wrote: »
    Chance wrote: »
    Flippy_D wrote: »
    Kreutz wrote: »
    I went 0-7 yesterday during my (dis)placement matches playing Mercy, including a 7-round slog on Anubis that I was just happy to be done with. Then I played some 6v6 randos, and found myself tearing it up with Moira so I tried her out for the remainder of my placement matches, because what's the worst that could happen, right? Well, I won the last three matches and placed gold for healing in all three. I think a lot of the difference is not having a giant bullseye on my character as soon as the wings come out, as well as being able to do something about gank attempts. Which isn't to say that Mercy's pistol doesn't work in certain situations, but compared to a tracking DoT with minor lifestealing capabilities it may as well be a peashooter. I don't think I mind giving up the damage boosting abilities in exchange for that.

    Am I the only one who thinks Moira is OP? Entirely possibly to get gold for healing and silver for damage without even being that good, plus autolock aim.

    She can do a lot of healing or damage, sure, but it's more difficult for her to do impactful damage or healing. It's like the Torb turret problem.

    I dunno, I think Moira can be a pretty reliable problem for D.Va, Genji and Mercy in particular. Her primary aim is so forgiving that you can just keep it on a Mercy as she flits around, stopping her self-heal.

    D.Va being in that list is surprising, as I normally think of her as an outright counter to Moira. The beam really doesn't do that much to the mech if it's being healed at all, and a flash of DM eating an orb for almost no resource cost is a huge blowout.

    My own theorycraft agrees with you - D.Va's ability to full counter a huge piece of Moira's kit is a massive blow, and D.Va's boost CD is comparable to Moira's fade, meaning she can easily chase down a scared Moira. Moira's damage is almost inconsequential without the orb, and her healing only becomes really powerful with it - it's very important.

    Buuut Matrix is no defense against Moira's primary and - most importantly - my D.Va-main brother complains about Moiras endlessly, and far more than any other hero. So it's quite anecdotal but to him there's nothing worse than a Moira on the enemy team.

    it takes too long to be worth it, but moira wins the 1v1 against dva, since she can fade the micro missiles and dva has no protection against moira's tickle beam (and the mech's bigass hitbox means moira has reliable health regen the whole time)

    in the context of an actual teamfight, i don't think either upend the scales or threaten each other too much

  • BionicPenguinBionicPenguin Registered User regular
    Chance wrote: »
    Chance wrote: »
    Flippy_D wrote: »
    Kreutz wrote: »
    I went 0-7 yesterday during my (dis)placement matches playing Mercy, including a 7-round slog on Anubis that I was just happy to be done with. Then I played some 6v6 randos, and found myself tearing it up with Moira so I tried her out for the remainder of my placement matches, because what's the worst that could happen, right? Well, I won the last three matches and placed gold for healing in all three. I think a lot of the difference is not having a giant bullseye on my character as soon as the wings come out, as well as being able to do something about gank attempts. Which isn't to say that Mercy's pistol doesn't work in certain situations, but compared to a tracking DoT with minor lifestealing capabilities it may as well be a peashooter. I don't think I mind giving up the damage boosting abilities in exchange for that.

    Am I the only one who thinks Moira is OP? Entirely possibly to get gold for healing and silver for damage without even being that good, plus autolock aim.

    She can do a lot of healing or damage, sure, but it's more difficult for her to do impactful damage or healing. It's like the Torb turret problem.

    I dunno, I think Moira can be a pretty reliable problem for D.Va, Genji and Mercy in particular. Her primary aim is so forgiving that you can just keep it on a Mercy as she flits around, stopping her self-heal.

    D.Va being in that list is surprising, as I normally think of her as an outright counter to Moira. The beam really doesn't do that much to the mech if it's being healed at all, and a flash of DM eating an orb for almost no resource cost is a huge blowout.

    My own theorycraft agrees with you - D.Va's ability to full counter a huge piece of Moira's kit is a massive blow, and D.Va's boost CD is comparable to Moira's fade, meaning she can easily chase down a scared Moira. Moira's damage is almost inconsequential without the orb, and her healing only becomes really powerful with it - it's very important.

    Buuut Matrix is no defense against Moira's primary and - most importantly - my D.Va-main brother complains about Moiras endlessly, and far more than any other hero. So it's quite anecdotal but to him there's nothing worse than a Moira on the enemy team.

    it takes too long to be worth it, but moira wins the 1v1 against dva, since she can fade the micro missiles and dva has no protection against moira's tickle beam (and the mech's bigass hitbox means moira has reliable health regen the whole time)

    in the context of an actual teamfight, i don't think either upend the scales or threaten each other too much

    I don't know about that. I play a fair amount of D.Va these days and I've never hesitated to 1v1 a Moira. Generally, I'll lose her when she fades, but if I catch where she went, she's almost certainly dead.

    9MpLNPA.jpg
    Steam: MightyPotatoKing/King of the Impossible - Battle.net: PotatoKing#1893
    Chance
  • The Escape GoatThe Escape Goat Brave Sir Robin Registered User regular
    If their worst fear as D.Va isn't a competent Zarya I frankly don't trust them.

    3clipse wrote: »
    God is dead and life has no purpose.
    ChanceBionicPenguinDonnicton
  • miscellaneousinsanitymiscellaneousinsanity grass grows, birds fly, sun shines, and brother, i hurt peopleRegistered User regular
    i'm not sure i have ever felt threatened by dva when playing moira so, :shrug: ymmv

    The Escape GoatChanceDonnicton
  • BionicPenguinBionicPenguin Registered User regular
    If their worst fear as D.Va isn't a competent Zarya I frankly don't trust them.

    Mei's a close second.

    9MpLNPA.jpg
    Steam: MightyPotatoKing/King of the Impossible - Battle.net: PotatoKing#1893
    Donnicton
  • FremFrem Registered User regular
    If their worst fear as D.Va isn't a competent Zarya I frankly don't trust them.

    Mei's a close second.

    Surprised nobody has mentioned Brigitte. She shreds D.va in no time flat if she can get close enough.

  • EndaroEndaro Registered User regular
    Flippy_D wrote: »
    Flippy_D wrote: »
    Okay, sooooOOOOooo I was waiting for the new thread to do this, but here are the results of a bunch of data crunching. There's a lot to say and this took several days to write, so uh, I apologise in advance.

    There does need to be context as to my motives, first: I'm really good at shooters and always have been. CounterStrike, TF2, Battlefield, anything, you name it, I was usually topping the leaderboard and in or around the semi-pro scene. I've won money playing FPS and sniper classes have always been my go-to.

    So Overwatch is released and I'm like, hell yeah Widow. But I play like a hundred hours with her and it just doesn't click. I find this weird but not particularly upsetting, in part because I'm getting to quite like the style of shooty boi Hanzo.

    Competitive begins and I go for it, and although nobody likes me playing Hanzo and the hero is generally regarded as a liability, I do reasonably well. 51% winrate. Rank 50 high. Basically the lowest end of Plat. That'll do me. Season 2 I get my historic high of 2531, again just on the edge of Plat, and even though I then have a rough patch and end up 2008, low gold, that feels like mostly bad luck. Highs and lows, etc.

    So to recap: always been pretty good at FPS but never super-pro; OW S1 and S2 positioned me as... pretty good, but not super-pro. That's cool! Reasonably happy about this. Having fun in-game and playing against, and with, pretty good people.

    The along comes Season 3, and everything starts going wrong. I don't feel like I'm playing that much differently, but I qualify much lower and start to sliiiiiiiide. My season high of 2280 felt like, okay, I'll just have to climb back up into low Plat. But instead I slipped all the way to 1935, high silver.

    What got me about this was less the losses themselves (annoying though they were), but more how the nature and impact of these losses felt totally arbitrary. Yes, leavers and idiots, but one had to assume that those were more or less equally distributed. It did feel that my more risky, skirmishing, strategic style of play was less effective when not backed up with a competent team, but it also felt like the way I went up and down after wins and losses was not a fair reflection of how I was playing. I was playing my arse off, but to no avail. I got discouraged, defeated and also just quite angry, which I disliked. I decided OW competitive - at least at this lower skill level - was too toxic and making me into someone unpleasant, so I quit the game entirely. At the time I also came to the conclusion that Hanzo's net utility was also probably not good enough, and that I was playing with the wrong deck.

    Time passes. I get suckered back in. I'm playing well and dominating QP, and so after initially hard avoiding competitive I try it again. But nothing has changed - in fact, it's got worse. Qualified low gold, which was bad enough, but after a senseless losing streak of 14 games more or less in a row, I'm actually laughing in disbelief at my rank of 1852. It seems to have absolutely nothing to do with how well I'm playing or not. The teams are dreadful. I always seem to be working with morons. The last loss had seen the first round absolutely crush us in just 1:50. These just aren't fair games. There's nothing I can do about it.

    Ultimately, I slid seriously hard, into the bottom ~7% of players at the lowest point. The entire time, I felt more or less powerless. My efforts, abilities and tactics seemed to make little difference to enemy teams who were often just better, net, than mine. My net winrate at times was as low as 38%.

    6bn2ykg3uboi.png

    But I'm a data guy and so I start to gather some. In fact, I gather 47 metrics each game, every game, for 166 games. I also gathered all the data the game shares with you at a rolled-up level; a further 30 datapoints. What I learned was... well.

    Just presenting the data would be a total splurge, so instead I'm going to position these as a series of hypotheses.

    1. I am getting worse

    An extremely reasonable take! You're losing skill, less effective in games, and so ranking down accordingly. Fair enough.

    But, doesn't look like it. For example, indexed data from my highest-played character, season-on-season (red bar is season 9, all benched vs earliest season I've played that hero):
    ajo46jb10yl5.png
    There are some ups and downs but as you'll see, some of the core 'skill' metrics are strong:
    • Accuracy up 35%
    • Crits up 53%
    • Damage per 10 mins up 59%(!)
    • Elims/10 still 5% off S1 but getting better season-on-season (On Fire and Final Blows are similar stories)
    • Obj time up 119%
    • Gold medals up 32%
    • Crit

    Other heroes are here: https://imgur.com/a/wJJIkMc.

    On most of these, there are meaningful improvements (red bar going right or higher than dark green bar). Ironically, the two heroes I've statistically improved least with - Junk and Bastion - also have my highest winrates¯\_(ツ)_/¯

    But it's hard to make a convincing case that I've really got worse over time.

    2. Everyone else has got better

    Doesn't matter if you're going at x acceleration if all other traffic is going at x2. The playerbase has probably also fallen off a bit, meaning fewer poor opponents and more good ones, proportionately. I personally reckon the second point is more salient, given that I do in fact meet some clearly good players in Silver.

    But again, data doesn't indicate such.

    35ofwipebd7n.png
    (this screenshot is contemporaneous to the end of season 9, so should include all of it - registering, for example, 65 games with Hanzo).

    Overbuff, which compares your data to others, puts me firmly in the top 50-10% for most of what would be thought of as 'key' metrics. This is data from competitive. I even die much less than other people. So it doesn't feel like I'm so terrible compared to this dataset, which is itself already skewed towards top players.

    3. I wasn't being a team player!

    I always try to be a team player but ultimately, DPS is DPS. This confess this was haunting me a bit (it's a team game!) so I deliberately stepped away from a more solo mindset and started to play healer, a servant to my team. I got quite a lot of compliments on my healing, for what it's worth.

    I guess you could say 'playing healer doesn't necessarily make you a team player', but I really poured myself into it. Again, looking at the Overbuff data in the prior post my healing output seemed to be absolutely top-end compared to most competitive. I also tried to more explicitly play the objective on all heroes (I always play the objective, but for example I would previously, when capping a point, move ahead of my team to try and repel/thwart counterattacks).

    So, I really did go for being a team player. I played majority healer in 40% of all matches and as a secondary function in 30% of matches.

    The unfortunate irony here is that I seem to have contributed more as a DPS, my natural home. When playing DPS, we won more often, and those wins were more decisive, than when I was healing:

    1kp46yzv4yef.png

    This was despite my actual performance being relatively comparable across both win and loss scenarios:

    39bg1engoyep.png

    This tallies with this Reddit post making a pretty convincing case that the best thing you can do for your team is play the hero you win the most with. I am a good healer relative to the playerbase, but a better DPS.

    So: massively upped my 'team player'ness, but ultimately the outputs seemed to indicate that this didn't really make a huge difference to winrate, and may have even slightly negatively affected it.

    4. I wasn't playing the right 'kind' of game

    This is still one of the hardest hypotheses, partly because it's vague. I don't really know how to prove it one way or the other. Included for sake of completeness, but will have to come back to it some day.

    Certainly my objective time massively increased, across the board (~39% more as healer, 81% more as DPS and 982% more as Bastion specifically!). I changed heroes more, always responded to team needs to fit 2-2-2 and so on, and communicated actively. I used to play raid leader in WoW so I know how to work in groups. I currently have a '4' on my teammate score, split ~25% sportsmanship, ~15% leadership and ~60% good play.

    Didn't help my winrate, though.

    With that said, one thing I did notice and certainly the first actual concrete takeaway was that I wasn't doing well on attack/defend maps (Hollywood, Lunar, Anubis, Volks., BlizzWorld, Hanamura):

    008f582751h0.png

    This is something I'm going to have to think about, although it's not lost on me that these are arguably the maps that require the most organisation and teamwork. Nor that my best successes come from tight, killzone-based maps.

    5. The -/+ system just isn't very good

    This was, to be honest, my chief suspicion. Yes, I admit it, the suspicion probably comes from wounded pride, but I just didn't feel like the dynamic range on how much SR is awarded to subtracted after a win or loss (-/+16 to -/+32) was consistently reflective of in-game quality. This wasn't helped by some of my worst point deductions (based from the minimum of -/+16) came from my closest games (note also that there were far more close games when winning than when losing!):

    5oe8acwfyv1z.png

    So, I made my own metric.

    Every single game, I recorded the data from the aftermath screens. Again, this was for 166 games.

    w1go9er7q44u.png

    From this, I made a score that indexed/benchmarked my performance on my own rolling averages. Without going too much into it, it also discriminated between healing and dmg (so I wasn't judging Mercy on kills, for instance) and weighted certain metrics more (e.g. so medals weren't over-important).

    I also allowed, from game 80 onward, for the addition of a modifier if I thought my scoring system wasn't quite working. I used this option 20 times (vs 86 games), mostly for healing wherein the range of influencing metrics was slimmer.

    The results of my custom score vs Overwatch's score were interesting.

    Turns out, the question of 'does the -/+ line up with my self-perceived performance?' is quite difficult to answer. Mainly because, whilst Overwatch always assigned me -16 to -28 on a negative for a loss, and +16 to +32 for a win, there were obviously games where https://youtube.com/watch?v=1TCX90yALsI]I performed well but was still on the losing side.

    In fact, I put in above-average performances (always my personal average) in ~45% of losses and ~55% of wins. It's hard to know to what extent being beaten influences that - in other words, am I playing below average, contributing to a loss, or is being outmatched the equivalent of cutting 18% (or 10 pc pts) of my performance? For example in losing games I did 46 less damage per minute, but I can't DPS effectively if I'm receiving no support or if the enemy team is well-organised with shields etc, right? I think it's a bit of an impossible question to answer without more data, and we also won more when I played DPS, so... hard to say.

    Nonetheless, here is one way of looking at it:

    gm16a4y4fen4.png

    What this shows is not my absolute rank change, but rather the rank change expressed as a percentage of its deviation from average. In other words, if I lost -16 that was actually better than the average loss of -22 and so, therefore, actually measured as +27%. Equally, a win that only gave me 16 points was, although a positive score, realistically a poor reflection on the game's assessment of me, and worth -29%. This controls for the straight positive/negative dichotomy. My own score was similarly measured in the sense of deviation from the average.

    So just to be clear, this isn't wins on the left and losses on the right. This just performance vs the game's reward of my performance.

    ... and frustratingly, the trendlines are very similar. The deviation is likely mostly explained by 'performing' worse during losses (see above) and MoE.

    The reason why this is frustrating is because it is does suggest that my way of gauging how well I played is, in the macro, in line with the game's. This means in short - much as I don't like it - that the scoring is accurate when taken as a net. Another surprise was that Overwatch was just, slightly, giving out more for wins - average 22.97 vs average -22.19, so 0.46 more.

    Where it falls over, however, is on individual matches. As you can see, there are plenty of matches where my self-assessed performance outstripped the relative score from the game. Bear in mind this is mostly not manually done, but is a score automatically generated from my class metrics. If we prioritise my metrics instead, we can see what's up:

    ccyl9s578xgi.png

    Namely, it looks like the way Overwatch deals with wild variance in performance is to flatten it - yes, triumphs aren't rewarded as much, but losses are also not so harshly penalised. In fact, my system is marginally less forgiving, on paper.

    Still, although there's this data it still feels slightly inconclusive, and I still think that Overwatch does a generally poor job of responding to in-game actions. Presumably, the flattening is to account for that very discrepancy, which would explain why the trendlines in the first graph are so similar...

    ... but I'm still not entirely assuaged here. I fear I shall never be.

    ---

    So, what are the conclusions here?
    • I am getting worse - data doesn't support this, I'm actually improving
    • Everyone else is getting better - data doesn't support this, my relative metrics vs others are pretty favourable
    • I wasn't being a team player - no data, but I significantly adjusted my playstyle and ironically appear to still be contributing most as a DPS
    • I wasn't playing the right kind of game - somewhat supported! I am proportionately worse at attack/defend maps and need to figure this one out
    • The way Overwatch scores you isn't properly reflective of in-game performance - data inconclusive. In some ways it appears no, in other ways it seems yes. I grudgingly think it still gets it wrong, but have to admit it probably doesn't have as much an effect as I feel it does. Although as a bonus, I played above average in 68% of 'close' games.

    Where am I left after all these numbers and analyses? Somewhere obvious: teamwork is in fact more important than any individual metric or how well you perform, and you are basically therefore at the mercy of being in a good, or (if you're a leader type) attentive, team. I think it no co-incidence that two of my three 'climbs' were when I (by chance) found a decent team and stuck with it:

    72crn8ajxkfj.png

    To me that feels a bit... shit. For two reasons:

    1) It's still just the goddamn luck of the draw. no matter how well I play, or don't play, personal performance can only ever put the brakes on a bad team performance. I can't enforce comp, I can't really make people notice shit, and I can't 'carry'. That last point is, incidentally, where I wound up last time, without data: that the classes I play can't carry like an insane Doomfist, Genji, Moira, Brigitte, or to some extent Rein or Sombra, can. I don't happen to be practiced with those heroes, either.

    2) If that Redditor's analysis is correct, Overwatch could probably do better on distributing people into teams! It could easily figure out the propensity for people to play heroes with high winrates and calculate the projected winrate. From there, it could create fairer matchups. But I guess that will never happen, and so you'll always be subjected to stomps and (it never feels often enough) delivering stomps.

    Anyway, that's more than enough. I leave you with specific hero win data, which I'm now at least using to try and give myself the best chance of being useful on any game that I join.

    Because, I after all, I really, really want to win.

    hpeisq2sn3mn.png

    I hope nobody minds me 'bumping' this on the new page but goddamn I wrote like 2,700 words based on hundreds of hours hand-coded data, I would like engagement.

    I guess I just wanna know if a) anything thinks I've missed something or drawn the wrong conclusion, b) has an alternative view, or c) can tell me how to fix my apparent weird blackspot of attack/defend maps.

    I want to type out a reply, but I hurt my hand and can't type much. I read it all, and will get back to you later!

    FencingsaxH3Knuckles
  • EnlongEnlong Registered User regular
    edited September 2018
    I'm totally going to fail to get the Dva rewards, because of my delightful blend of procrastination and incompetence. Lovely.

    edit: Dammit. If I'm going to just be like this all the time, I should just remove the game from my computer again, because I clearly cannot be trusted to be a good person in the game or while discussing it. So I have.

    Enlong on
    YggiDee wrote: »
    Having teenaged RPG leads is really cool until you stop being a teenager yourself. Do you remember being seventeen? You're a dumbass at seventeen! I wanna be saved by the guy who's twenty-seven. He's at least payed taxes. He knows how to do shit. He can drive.
  • Phoenix-DPhoenix-D Registered User regular
    Comp game just now:

    I'm Mercy. Solo healing. I am on the cart, everyone else is 100 feet thataway toward the end.

    I die, and they respond by trying to spawn camp the enemy team.

    We get a leaver 3 feet from the end.

    We win anyway. o.O

    ChanceKreutzMrVyngaardFrem
  • -Loki--Loki- Don't pee in my mouth and tell me it's raining. Registered User regular
    edited September 2018
    Sometimes leavers are because of connections. I left a Dawn of War match I was about 5 second away from winning because I stretched and kicked the power cord out of the router.

    -Loki- on
    Raw Concrete
  • djmitchelladjmitchella Registered User regular
    tried some quickplay -- first three real games, lost them all, died a lot

    but

    in two of those three games, the 'play of the game' involved someone killing me, so I guess at least I am bad in an entertaining-to-watch way?

  • KanaKana Registered User regular
    edited September 2018
    Flippy_D wrote: »
    Flippy_D wrote: »
    Okay, sooooOOOOooo I was waiting for the new thread to do this, but here are the results of a bunch of data crunching. There's a lot to say and this took several days to write, so uh, I apologise in advance.

    There does need to be context as to my motives, first: I'm really good at shooters and always have been. CounterStrike, TF2, Battlefield, anything, you name it, I was usually topping the leaderboard and in or around the semi-pro scene. I've won money playing FPS and sniper classes have always been my go-to.

    So Overwatch is released and I'm like, hell yeah Widow. But I play like a hundred hours with her and it just doesn't click. I find this weird but not particularly upsetting, in part because I'm getting to quite like the style of shooty boi Hanzo.

    Competitive begins and I go for it, and although nobody likes me playing Hanzo and the hero is generally regarded as a liability, I do reasonably well. 51% winrate. Rank 50 high. Basically the lowest end of Plat. That'll do me. Season 2 I get my historic high of 2531, again just on the edge of Plat, and even though I then have a rough patch and end up 2008, low gold, that feels like mostly bad luck. Highs and lows, etc.

    So to recap: always been pretty good at FPS but never super-pro; OW S1 and S2 positioned me as... pretty good, but not super-pro. That's cool! Reasonably happy about this. Having fun in-game and playing against, and with, pretty good people.

    The along comes Season 3, and everything starts going wrong. I don't feel like I'm playing that much differently, but I qualify much lower and start to sliiiiiiiide. My season high of 2280 felt like, okay, I'll just have to climb back up into low Plat. But instead I slipped all the way to 1935, high silver.

    What got me about this was less the losses themselves (annoying though they were), but more how the nature and impact of these losses felt totally arbitrary. Yes, leavers and idiots, but one had to assume that those were more or less equally distributed. It did feel that my more risky, skirmishing, strategic style of play was less effective when not backed up with a competent team, but it also felt like the way I went up and down after wins and losses was not a fair reflection of how I was playing. I was playing my arse off, but to no avail. I got discouraged, defeated and also just quite angry, which I disliked. I decided OW competitive - at least at this lower skill level - was too toxic and making me into someone unpleasant, so I quit the game entirely. At the time I also came to the conclusion that Hanzo's net utility was also probably not good enough, and that I was playing with the wrong deck.

    Time passes. I get suckered back in. I'm playing well and dominating QP, and so after initially hard avoiding competitive I try it again. But nothing has changed - in fact, it's got worse. Qualified low gold, which was bad enough, but after a senseless losing streak of 14 games more or less in a row, I'm actually laughing in disbelief at my rank of 1852. It seems to have absolutely nothing to do with how well I'm playing or not. The teams are dreadful. I always seem to be working with morons. The last loss had seen the first round absolutely crush us in just 1:50. These just aren't fair games. There's nothing I can do about it.

    Ultimately, I slid seriously hard, into the bottom ~7% of players at the lowest point. The entire time, I felt more or less powerless. My efforts, abilities and tactics seemed to make little difference to enemy teams who were often just better, net, than mine. My net winrate at times was as low as 38%.

    6bn2ykg3uboi.png

    But I'm a data guy and so I start to gather some. In fact, I gather 47 metrics each game, every game, for 166 games. I also gathered all the data the game shares with you at a rolled-up level; a further 30 datapoints. What I learned was... well.

    Just presenting the data would be a total splurge, so instead I'm going to position these as a series of hypotheses.

    1. I am getting worse

    An extremely reasonable take! You're losing skill, less effective in games, and so ranking down accordingly. Fair enough.

    But, doesn't look like it. For example, indexed data from my highest-played character, season-on-season (red bar is season 9, all benched vs earliest season I've played that hero):
    ajo46jb10yl5.png
    There are some ups and downs but as you'll see, some of the core 'skill' metrics are strong:
    • Accuracy up 35%
    • Crits up 53%
    • Damage per 10 mins up 59%(!)
    • Elims/10 still 5% off S1 but getting better season-on-season (On Fire and Final Blows are similar stories)
    • Obj time up 119%
    • Gold medals up 32%
    • Crit

    Other heroes are here: https://imgur.com/a/wJJIkMc.

    On most of these, there are meaningful improvements (red bar going right or higher than dark green bar). Ironically, the two heroes I've statistically improved least with - Junk and Bastion - also have my highest winrates¯\_(ツ)_/¯

    But it's hard to make a convincing case that I've really got worse over time.

    2. Everyone else has got better

    Doesn't matter if you're going at x acceleration if all other traffic is going at x2. The playerbase has probably also fallen off a bit, meaning fewer poor opponents and more good ones, proportionately. I personally reckon the second point is more salient, given that I do in fact meet some clearly good players in Silver.

    But again, data doesn't indicate such.

    35ofwipebd7n.png
    (this screenshot is contemporaneous to the end of season 9, so should include all of it - registering, for example, 65 games with Hanzo).

    Overbuff, which compares your data to others, puts me firmly in the top 50-10% for most of what would be thought of as 'key' metrics. This is data from competitive. I even die much less than other people. So it doesn't feel like I'm so terrible compared to this dataset, which is itself already skewed towards top players.

    3. I wasn't being a team player!

    I always try to be a team player but ultimately, DPS is DPS. This confess this was haunting me a bit (it's a team game!) so I deliberately stepped away from a more solo mindset and started to play healer, a servant to my team. I got quite a lot of compliments on my healing, for what it's worth.

    I guess you could say 'playing healer doesn't necessarily make you a team player', but I really poured myself into it. Again, looking at the Overbuff data in the prior post my healing output seemed to be absolutely top-end compared to most competitive. I also tried to more explicitly play the objective on all heroes (I always play the objective, but for example I would previously, when capping a point, move ahead of my team to try and repel/thwart counterattacks).

    So, I really did go for being a team player. I played majority healer in 40% of all matches and as a secondary function in 30% of matches.

    The unfortunate irony here is that I seem to have contributed more as a DPS, my natural home. When playing DPS, we won more often, and those wins were more decisive, than when I was healing:

    1kp46yzv4yef.png

    This was despite my actual performance being relatively comparable across both win and loss scenarios:

    39bg1engoyep.png

    This tallies with this Reddit post making a pretty convincing case that the best thing you can do for your team is play the hero you win the most with. I am a good healer relative to the playerbase, but a better DPS.

    So: massively upped my 'team player'ness, but ultimately the outputs seemed to indicate that this didn't really make a huge difference to winrate, and may have even slightly negatively affected it.

    4. I wasn't playing the right 'kind' of game

    This is still one of the hardest hypotheses, partly because it's vague. I don't really know how to prove it one way or the other. Included for sake of completeness, but will have to come back to it some day.

    Certainly my objective time massively increased, across the board (~39% more as healer, 81% more as DPS and 982% more as Bastion specifically!). I changed heroes more, always responded to team needs to fit 2-2-2 and so on, and communicated actively. I used to play raid leader in WoW so I know how to work in groups. I currently have a '4' on my teammate score, split ~25% sportsmanship, ~15% leadership and ~60% good play.

    Didn't help my winrate, though.

    With that said, one thing I did notice and certainly the first actual concrete takeaway was that I wasn't doing well on attack/defend maps (Hollywood, Lunar, Anubis, Volks., BlizzWorld, Hanamura):

    008f582751h0.png

    This is something I'm going to have to think about, although it's not lost on me that these are arguably the maps that require the most organisation and teamwork. Nor that my best successes come from tight, killzone-based maps.

    5. The -/+ system just isn't very good

    This was, to be honest, my chief suspicion. Yes, I admit it, the suspicion probably comes from wounded pride, but I just didn't feel like the dynamic range on how much SR is awarded to subtracted after a win or loss (-/+16 to -/+32) was consistently reflective of in-game quality. This wasn't helped by some of my worst point deductions (based from the minimum of -/+16) came from my closest games (note also that there were far more close games when winning than when losing!):

    5oe8acwfyv1z.png

    So, I made my own metric.

    Every single game, I recorded the data from the aftermath screens. Again, this was for 166 games.

    w1go9er7q44u.png

    From this, I made a score that indexed/benchmarked my performance on my own rolling averages. Without going too much into it, it also discriminated between healing and dmg (so I wasn't judging Mercy on kills, for instance) and weighted certain metrics more (e.g. so medals weren't over-important).

    I also allowed, from game 80 onward, for the addition of a modifier if I thought my scoring system wasn't quite working. I used this option 20 times (vs 86 games), mostly for healing wherein the range of influencing metrics was slimmer.

    The results of my custom score vs Overwatch's score were interesting.

    Turns out, the question of 'does the -/+ line up with my self-perceived performance?' is quite difficult to answer. Mainly because, whilst Overwatch always assigned me -16 to -28 on a negative for a loss, and +16 to +32 for a win, there were obviously games where https://youtube.com/watch?v=1TCX90yALsI]I performed well but was still on the losing side.

    In fact, I put in above-average performances (always my personal average) in ~45% of losses and ~55% of wins. It's hard to know to what extent being beaten influences that - in other words, am I playing below average, contributing to a loss, or is being outmatched the equivalent of cutting 18% (or 10 pc pts) of my performance? For example in losing games I did 46 less damage per minute, but I can't DPS effectively if I'm receiving no support or if the enemy team is well-organised with shields etc, right? I think it's a bit of an impossible question to answer without more data, and we also won more when I played DPS, so... hard to say.

    Nonetheless, here is one way of looking at it:

    gm16a4y4fen4.png

    What this shows is not my absolute rank change, but rather the rank change expressed as a percentage of its deviation from average. In other words, if I lost -16 that was actually better than the average loss of -22 and so, therefore, actually measured as +27%. Equally, a win that only gave me 16 points was, although a positive score, realistically a poor reflection on the game's assessment of me, and worth -29%. This controls for the straight positive/negative dichotomy. My own score was similarly measured in the sense of deviation from the average.

    So just to be clear, this isn't wins on the left and losses on the right. This just performance vs the game's reward of my performance.

    ... and frustratingly, the trendlines are very similar. The deviation is likely mostly explained by 'performing' worse during losses (see above) and MoE.

    The reason why this is frustrating is because it is does suggest that my way of gauging how well I played is, in the macro, in line with the game's. This means in short - much as I don't like it - that the scoring is accurate when taken as a net. Another surprise was that Overwatch was just, slightly, giving out more for wins - average 22.97 vs average -22.19, so 0.46 more.

    Where it falls over, however, is on individual matches. As you can see, there are plenty of matches where my self-assessed performance outstripped the relative score from the game. Bear in mind this is mostly not manually done, but is a score automatically generated from my class metrics. If we prioritise my metrics instead, we can see what's up:

    ccyl9s578xgi.png

    Namely, it looks like the way Overwatch deals with wild variance in performance is to flatten it - yes, triumphs aren't rewarded as much, but losses are also not so harshly penalised. In fact, my system is marginally less forgiving, on paper.

    Still, although there's this data it still feels slightly inconclusive, and I still think that Overwatch does a generally poor job of responding to in-game actions. Presumably, the flattening is to account for that very discrepancy, which would explain why the trendlines in the first graph are so similar...

    ... but I'm still not entirely assuaged here. I fear I shall never be.

    ---

    So, what are the conclusions here?
    • I am getting worse - data doesn't support this, I'm actually improving
    • Everyone else is getting better - data doesn't support this, my relative metrics vs others are pretty favourable
    • I wasn't being a team player - no data, but I significantly adjusted my playstyle and ironically appear to still be contributing most as a DPS
    • I wasn't playing the right kind of game - somewhat supported! I am proportionately worse at attack/defend maps and need to figure this one out
    • The way Overwatch scores you isn't properly reflective of in-game performance - data inconclusive. In some ways it appears no, in other ways it seems yes. I grudgingly think it still gets it wrong, but have to admit it probably doesn't have as much an effect as I feel it does. Although as a bonus, I played above average in 68% of 'close' games.

    Where am I left after all these numbers and analyses? Somewhere obvious: teamwork is in fact more important than any individual metric or how well you perform, and you are basically therefore at the mercy of being in a good, or (if you're a leader type) attentive, team. I think it no co-incidence that two of my three 'climbs' were when I (by chance) found a decent team and stuck with it:

    72crn8ajxkfj.png

    To me that feels a bit... shit. For two reasons:

    1) It's still just the goddamn luck of the draw. no matter how well I play, or don't play, personal performance can only ever put the brakes on a bad team performance. I can't enforce comp, I can't really make people notice shit, and I can't 'carry'. That last point is, incidentally, where I wound up last time, without data: that the classes I play can't carry like an insane Doomfist, Genji, Moira, Brigitte, or to some extent Rein or Sombra, can. I don't happen to be practiced with those heroes, either.

    2) If that Redditor's analysis is correct, Overwatch could probably do better on distributing people into teams! It could easily figure out the propensity for people to play heroes with high winrates and calculate the projected winrate. From there, it could create fairer matchups. But I guess that will never happen, and so you'll always be subjected to stomps and (it never feels often enough) delivering stomps.

    Anyway, that's more than enough. I leave you with specific hero win data, which I'm now at least using to try and give myself the best chance of being useful on any game that I join.

    Because, I after all, I really, really want to win.

    hpeisq2sn3mn.png

    I hope nobody minds me 'bumping' this on the new page but goddamn I wrote like 2,700 words based on hundreds of hours hand-coded data, I would like engagement.

    I guess I just wanna know if a) anything thinks I've missed something or drawn the wrong conclusion, b) has an alternative view, or c) can tell me how to fix my apparent weird blackspot of attack/defend maps.

    I think my short form answer to his whole thing isn't going to be very satisfying for you, but... Who cares what your stats are?

    Like just as with the whole convo with Moira's stats earlier... Moira is not OP, Moira is just good at racking up certain stats that happen to show up on the leaderboard. She is very frequently a worse selection than other heroes who will nevertheless end up with less tracked stats.

    You're talking about being useful in games that you join, but if you're collecting More Stats while neglecting basic game principles like ult management and team engagements, then your stats really don't matter much. There's a lot of non-intuitive game strategies that are really important to Overwatch, and don't actually show up that much in other shooters.

    Like this is just based on a vague recollection of playing with you a while ago, but I remember you playing with the PA group a couple times in the past and honestly kinda getting beat on. Not because of your mechanics, I know your mechanics are just fine. But because you didn't have game awareness, you went off on your own and got punished, engaged in fights after the fight was already over, staggered yourself. I dunno, maybe I'm remembering it wrong, in which case my apologies and just treat it as a hypothetical scenario. You can have good mechanics and collect good stats and not actually be contributing nearly as much to your team as a player with much worse stats and mechanics.

    Like, I feel like I'm pretty solid on D.Va. Whether or not I got good stats on D.Va or not in a round is pretty much entirely inconsequential to how much game impact I had. If the enemy team runs a pocketed Widow the whole game and I'm constantly in her face and not letting her take shots? I don't give a shit that I never killed her, that's not my win condition. The point is I stopped two enemy players from doing what they wanted at the cost of 1 ally, and therefore gave my team a consistent man advantage in every fight. Ain't no stat for "gravs eaten" but that can change the entire course of a match. And so on.

    Not to be a downer, but I think the best advice would be to take those stats and push them to the side and not look at them anymore, and instead watch some Jayne vids and do some tape review of your own play.

    Kana on
    A trap is for fish: when you've got the fish, you can forget the trap. A snare is for rabbits: when you've got the rabbit, you can forget the snare. Words are for meaning: when you've got the meaning, you can forget the words.
    Unlucky
  • BaidolBaidol I will hold him off Escape while you canRegistered User regular
    I don't usually post my clipshows here, but I think this one is particularly good.

    Steam Overwatch: Baidol#1957
    H3KnucklesChanceEndaro
  • MusicoolMusicool Registered User regular
    edited September 2018
    Hey guys, I have a friend of mine who was Plat in like, Season 3, then never again. He was really down about it for a long time.

    But he only just recently cracked Plat, and then he stormed all his way to 2751! He was so unreservedly happy, it was infectious. I dunno, I guess I just wanted to say that these numbers make us glum for so long, but sometimes - sometimes - after a great deal of work and listening to others (we were throwing a lot of advice around) they can make you happy too!

    The crazy thing is, he used to be one of those Genji players, but after months of polite suggestions, he tried new things! He spent a few seasons on Mercy during peak Mercy. He moved to Reinhardt and Brig, and now he's a terrifying Reinhardt. And I gotta say, I really appreciate not being our team's main tank all the time. More Hamster and Zarya for me!

    Musicool on
    Burtletoy wrote: »
    I disagree completely.

    hAmmONd IsnT A mAin TAnk
    unbelievablejugsphp.png
    Chance
  • -Loki--Loki- Don't pee in my mouth and tell me it's raining. Registered User regular
    The difference between an average Reinhardt and a good Reinhardt is amazing.

    MusicoolChanceKanaNobodyKreutzPhoenix-DMrVyngaardExtreaminatusReynolds
  • MusicoolMusicool Registered User regular
    edited September 2018
    -Loki- wrote: »
    The difference between an average Reinhardt and a good Reinhardt is amazing.

    Fo sho. It's the difference between "we may as well have an Orisa if you're going to do nothing or just feed with charge, which are basically the same thing" and "our main tank just got a double: let's walk onto point guys".

    And I've noticed he's more of the latter much more often.

    Musicool on
    Burtletoy wrote: »
    I disagree completely.

    hAmmONd IsnT A mAin TAnk
    unbelievablejugsphp.png
    Chance
  • ChanceChance Registered User regular
    Baidol wrote: »
    I don't usually post my clipshows here, but I think this one is particularly good.


    It is ^.^

    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
    H3Knuckles
  • nexuscrawlernexuscrawler Registered User regular
    Moira is a hard counter to baby DVA

  • The Escape GoatThe Escape Goat Brave Sir Robin Registered User regular
    edited September 2018
    -Loki- wrote: »
    The difference between an average Reinhardt and a good Reinhardt is amazing.

    Recently I recoiled away when someone on my work team said that I was a good Reinhardt, and this is exactly why. If I was a good Reinhardt I'd be actively winning the Rein mirror and landing more than a third of a my shatters, and shattering in places where my team could actually follow up; as it is I'm just someone who's willing to play Rein when no-one else will.

    Edit: The one thing I'll say I do well as Rein is that I build shatters quickly, mostly because I play very aggressively. But that has its own downsides.

    The Escape Goat on
    3clipse wrote: »
    God is dead and life has no purpose.
    Chance
  • RoeRoe Farming Greater Rifts Registered User regular
    edited September 2018
    I've seen some bad Reinhardt players. Most like to charge half way across the map to pin someone down.

    Roe on
    oHw5R0V.jpg
  • CoinageCoinage The Naming of Cats is a difficult matter Registered User regular
    Reinhardt is my most played hero and I always feel like I made the wrong decision. Weird how it can be such a combination of boring, frustrating, and intensity

    s586cu2r93hr.gif
  • KupiKupi Registered User regular
    The worst feeling I experience as Reinhardt is attacking three or more enemies at the same time and missing five consecutive hammer swings.

    You think it isn't possible to aim that badly and you are wrong.

    My favorite musical instrument is the air-raid siren.
    The Escape GoatKreutz
  • KreutzKreutz Registered User regular
    The worst feeling as Reinhardt is going for the medium-range pin, the target sidesteps, you overestimate your turning radius and fly off of the edge of the map, and somehow it all gets captured in the PotG.

    Chance
  • Brutal JBrutal J Sorry! Sorry, I'm sorry. Sorry. Registered User regular
    Kana wrote: »
    Flippy_D wrote: »
    Flippy_D wrote: »
    Okay, sooooOOOOooo I was waiting for the new thread to do this, but here are the results of a bunch of data crunching. There's a lot to say and this took several days to write, so uh, I apologise in advance.

    There does need to be context as to my motives, first: I'm really good at shooters and always have been. CounterStrike, TF2, Battlefield, anything, you name it, I was usually topping the leaderboard and in or around the semi-pro scene. I've won money playing FPS and sniper classes have always been my go-to.

    So Overwatch is released and I'm like, hell yeah Widow. But I play like a hundred hours with her and it just doesn't click. I find this weird but not particularly upsetting, in part because I'm getting to quite like the style of shooty boi Hanzo.

    Competitive begins and I go for it, and although nobody likes me playing Hanzo and the hero is generally regarded as a liability, I do reasonably well. 51% winrate. Rank 50 high. Basically the lowest end of Plat. That'll do me. Season 2 I get my historic high of 2531, again just on the edge of Plat, and even though I then have a rough patch and end up 2008, low gold, that feels like mostly bad luck. Highs and lows, etc.

    So to recap: always been pretty good at FPS but never super-pro; OW S1 and S2 positioned me as... pretty good, but not super-pro. That's cool! Reasonably happy about this. Having fun in-game and playing against, and with, pretty good people.

    The along comes Season 3, and everything starts going wrong. I don't feel like I'm playing that much differently, but I qualify much lower and start to sliiiiiiiide. My season high of 2280 felt like, okay, I'll just have to climb back up into low Plat. But instead I slipped all the way to 1935, high silver.

    What got me about this was less the losses themselves (annoying though they were), but more how the nature and impact of these losses felt totally arbitrary. Yes, leavers and idiots, but one had to assume that those were more or less equally distributed. It did feel that my more risky, skirmishing, strategic style of play was less effective when not backed up with a competent team, but it also felt like the way I went up and down after wins and losses was not a fair reflection of how I was playing. I was playing my arse off, but to no avail. I got discouraged, defeated and also just quite angry, which I disliked. I decided OW competitive - at least at this lower skill level - was too toxic and making me into someone unpleasant, so I quit the game entirely. At the time I also came to the conclusion that Hanzo's net utility was also probably not good enough, and that I was playing with the wrong deck.

    Time passes. I get suckered back in. I'm playing well and dominating QP, and so after initially hard avoiding competitive I try it again. But nothing has changed - in fact, it's got worse. Qualified low gold, which was bad enough, but after a senseless losing streak of 14 games more or less in a row, I'm actually laughing in disbelief at my rank of 1852. It seems to have absolutely nothing to do with how well I'm playing or not. The teams are dreadful. I always seem to be working with morons. The last loss had seen the first round absolutely crush us in just 1:50. These just aren't fair games. There's nothing I can do about it.

    Ultimately, I slid seriously hard, into the bottom ~7% of players at the lowest point. The entire time, I felt more or less powerless. My efforts, abilities and tactics seemed to make little difference to enemy teams who were often just better, net, than mine. My net winrate at times was as low as 38%.

    6bn2ykg3uboi.png

    But I'm a data guy and so I start to gather some. In fact, I gather 47 metrics each game, every game, for 166 games. I also gathered all the data the game shares with you at a rolled-up level; a further 30 datapoints. What I learned was... well.

    Just presenting the data would be a total splurge, so instead I'm going to position these as a series of hypotheses.

    1. I am getting worse

    An extremely reasonable take! You're losing skill, less effective in games, and so ranking down accordingly. Fair enough.

    But, doesn't look like it. For example, indexed data from my highest-played character, season-on-season (red bar is season 9, all benched vs earliest season I've played that hero):
    ajo46jb10yl5.png
    There are some ups and downs but as you'll see, some of the core 'skill' metrics are strong:
    • Accuracy up 35%
    • Crits up 53%
    • Damage per 10 mins up 59%(!)
    • Elims/10 still 5% off S1 but getting better season-on-season (On Fire and Final Blows are similar stories)
    • Obj time up 119%
    • Gold medals up 32%
    • Crit

    Other heroes are here: https://imgur.com/a/wJJIkMc.

    On most of these, there are meaningful improvements (red bar going right or higher than dark green bar). Ironically, the two heroes I've statistically improved least with - Junk and Bastion - also have my highest winrates¯\_(ツ)_/¯

    But it's hard to make a convincing case that I've really got worse over time.

    2. Everyone else has got better

    Doesn't matter if you're going at x acceleration if all other traffic is going at x2. The playerbase has probably also fallen off a bit, meaning fewer poor opponents and more good ones, proportionately. I personally reckon the second point is more salient, given that I do in fact meet some clearly good players in Silver.

    But again, data doesn't indicate such.

    35ofwipebd7n.png
    (this screenshot is contemporaneous to the end of season 9, so should include all of it - registering, for example, 65 games with Hanzo).

    Overbuff, which compares your data to others, puts me firmly in the top 50-10% for most of what would be thought of as 'key' metrics. This is data from competitive. I even die much less than other people. So it doesn't feel like I'm so terrible compared to this dataset, which is itself already skewed towards top players.

    3. I wasn't being a team player!

    I always try to be a team player but ultimately, DPS is DPS. This confess this was haunting me a bit (it's a team game!) so I deliberately stepped away from a more solo mindset and started to play healer, a servant to my team. I got quite a lot of compliments on my healing, for what it's worth.

    I guess you could say 'playing healer doesn't necessarily make you a team player', but I really poured myself into it. Again, looking at the Overbuff data in the prior post my healing output seemed to be absolutely top-end compared to most competitive. I also tried to more explicitly play the objective on all heroes (I always play the objective, but for example I would previously, when capping a point, move ahead of my team to try and repel/thwart counterattacks).

    So, I really did go for being a team player. I played majority healer in 40% of all matches and as a secondary function in 30% of matches.

    The unfortunate irony here is that I seem to have contributed more as a DPS, my natural home. When playing DPS, we won more often, and those wins were more decisive, than when I was healing:

    1kp46yzv4yef.png

    This was despite my actual performance being relatively comparable across both win and loss scenarios:

    39bg1engoyep.png

    This tallies with this Reddit post making a pretty convincing case that the best thing you can do for your team is play the hero you win the most with. I am a good healer relative to the playerbase, but a better DPS.

    So: massively upped my 'team player'ness, but ultimately the outputs seemed to indicate that this didn't really make a huge difference to winrate, and may have even slightly negatively affected it.

    4. I wasn't playing the right 'kind' of game

    This is still one of the hardest hypotheses, partly because it's vague. I don't really know how to prove it one way or the other. Included for sake of completeness, but will have to come back to it some day.

    Certainly my objective time massively increased, across the board (~39% more as healer, 81% more as DPS and 982% more as Bastion specifically!). I changed heroes more, always responded to team needs to fit 2-2-2 and so on, and communicated actively. I used to play raid leader in WoW so I know how to work in groups. I currently have a '4' on my teammate score, split ~25% sportsmanship, ~15% leadership and ~60% good play.

    Didn't help my winrate, though.

    With that said, one thing I did notice and certainly the first actual concrete takeaway was that I wasn't doing well on attack/defend maps (Hollywood, Lunar, Anubis, Volks., BlizzWorld, Hanamura):

    008f582751h0.png

    This is something I'm going to have to think about, although it's not lost on me that these are arguably the maps that require the most organisation and teamwork. Nor that my best successes come from tight, killzone-based maps.

    5. The -/+ system just isn't very good

    This was, to be honest, my chief suspicion. Yes, I admit it, the suspicion probably comes from wounded pride, but I just didn't feel like the dynamic range on how much SR is awarded to subtracted after a win or loss (-/+16 to -/+32) was consistently reflective of in-game quality. This wasn't helped by some of my worst point deductions (based from the minimum of -/+16) came from my closest games (note also that there were far more close games when winning than when losing!):

    5oe8acwfyv1z.png

    So, I made my own metric.

    Every single game, I recorded the data from the aftermath screens. Again, this was for 166 games.

    w1go9er7q44u.png

    From this, I made a score that indexed/benchmarked my performance on my own rolling averages. Without going too much into it, it also discriminated between healing and dmg (so I wasn't judging Mercy on kills, for instance) and weighted certain metrics more (e.g. so medals weren't over-important).

    I also allowed, from game 80 onward, for the addition of a modifier if I thought my scoring system wasn't quite working. I used this option 20 times (vs 86 games), mostly for healing wherein the range of influencing metrics was slimmer.

    The results of my custom score vs Overwatch's score were interesting.

    Turns out, the question of 'does the -/+ line up with my self-perceived performance?' is quite difficult to answer. Mainly because, whilst Overwatch always assigned me -16 to -28 on a negative for a loss, and +16 to +32 for a win, there were obviously games where https://youtube.com/watch?v=1TCX90yALsI]I performed well but was still on the losing side.

    In fact, I put in above-average performances (always my personal average) in ~45% of losses and ~55% of wins. It's hard to know to what extent being beaten influences that - in other words, am I playing below average, contributing to a loss, or is being outmatched the equivalent of cutting 18% (or 10 pc pts) of my performance? For example in losing games I did 46 less damage per minute, but I can't DPS effectively if I'm receiving no support or if the enemy team is well-organised with shields etc, right? I think it's a bit of an impossible question to answer without more data, and we also won more when I played DPS, so... hard to say.

    Nonetheless, here is one way of looking at it:

    gm16a4y4fen4.png

    What this shows is not my absolute rank change, but rather the rank change expressed as a percentage of its deviation from average. In other words, if I lost -16 that was actually better than the average loss of -22 and so, therefore, actually measured as +27%. Equally, a win that only gave me 16 points was, although a positive score, realistically a poor reflection on the game's assessment of me, and worth -29%. This controls for the straight positive/negative dichotomy. My own score was similarly measured in the sense of deviation from the average.

    So just to be clear, this isn't wins on the left and losses on the right. This just performance vs the game's reward of my performance.

    ... and frustratingly, the trendlines are very similar. The deviation is likely mostly explained by 'performing' worse during losses (see above) and MoE.

    The reason why this is frustrating is because it is does suggest that my way of gauging how well I played is, in the macro, in line with the game's. This means in short - much as I don't like it - that the scoring is accurate when taken as a net. Another surprise was that Overwatch was just, slightly, giving out more for wins - average 22.97 vs average -22.19, so 0.46 more.

    Where it falls over, however, is on individual matches. As you can see, there are plenty of matches where my self-assessed performance outstripped the relative score from the game. Bear in mind this is mostly not manually done, but is a score automatically generated from my class metrics. If we prioritise my metrics instead, we can see what's up:

    ccyl9s578xgi.png

    Namely, it looks like the way Overwatch deals with wild variance in performance is to flatten it - yes, triumphs aren't rewarded as much, but losses are also not so harshly penalised. In fact, my system is marginally less forgiving, on paper.

    Still, although there's this data it still feels slightly inconclusive, and I still think that Overwatch does a generally poor job of responding to in-game actions. Presumably, the flattening is to account for that very discrepancy, which would explain why the trendlines in the first graph are so similar...

    ... but I'm still not entirely assuaged here. I fear I shall never be.

    ---

    So, what are the conclusions here?
    • I am getting worse - data doesn't support this, I'm actually improving
    • Everyone else is getting better - data doesn't support this, my relative metrics vs others are pretty favourable
    • I wasn't being a team player - no data, but I significantly adjusted my playstyle and ironically appear to still be contributing most as a DPS
    • I wasn't playing the right kind of game - somewhat supported! I am proportionately worse at attack/defend maps and need to figure this one out
    • The way Overwatch scores you isn't properly reflective of in-game performance - data inconclusive. In some ways it appears no, in other ways it seems yes. I grudgingly think it still gets it wrong, but have to admit it probably doesn't have as much an effect as I feel it does. Although as a bonus, I played above average in 68% of 'close' games.

    Where am I left after all these numbers and analyses? Somewhere obvious: teamwork is in fact more important than any individual metric or how well you perform, and you are basically therefore at the mercy of being in a good, or (if you're a leader type) attentive, team. I think it no co-incidence that two of my three 'climbs' were when I (by chance) found a decent team and stuck with it:

    72crn8ajxkfj.png

    To me that feels a bit... shit. For two reasons:

    1) It's still just the goddamn luck of the draw. no matter how well I play, or don't play, personal performance can only ever put the brakes on a bad team performance. I can't enforce comp, I can't really make people notice shit, and I can't 'carry'. That last point is, incidentally, where I wound up last time, without data: that the classes I play can't carry like an insane Doomfist, Genji, Moira, Brigitte, or to some extent Rein or Sombra, can. I don't happen to be practiced with those heroes, either.

    2) If that Redditor's analysis is correct, Overwatch could probably do better on distributing people into teams! It could easily figure out the propensity for people to play heroes with high winrates and calculate the projected winrate. From there, it could create fairer matchups. But I guess that will never happen, and so you'll always be subjected to stomps and (it never feels often enough) delivering stomps.

    Anyway, that's more than enough. I leave you with specific hero win data, which I'm now at least using to try and give myself the best chance of being useful on any game that I join.

    Because, I after all, I really, really want to win.

    hpeisq2sn3mn.png

    I hope nobody minds me 'bumping' this on the new page but goddamn I wrote like 2,700 words based on hundreds of hours hand-coded data, I would like engagement.

    I guess I just wanna know if a) anything thinks I've missed something or drawn the wrong conclusion, b) has an alternative view, or c) can tell me how to fix my apparent weird blackspot of attack/defend maps.

    I think my short form answer to his whole thing isn't going to be very satisfying for you, but... Who cares what your stats are?

    Like just as with the whole convo with Moira's stats earlier... Moira is not OP, Moira is just good at racking up certain stats that happen to show up on the leaderboard. She is very frequently a worse selection than other heroes who will nevertheless end up with less tracked stats.

    You're talking about being useful in games that you join, but if you're collecting More Stats while neglecting basic game principles like ult management and team engagements, then your stats really don't matter much. There's a lot of non-intuitive game strategies that are really important to Overwatch, and don't actually show up that much in other shooters.

    Like this is just based on a vague recollection of playing with you a while ago, but I remember you playing with the PA group a couple times in the past and honestly kinda getting beat on. Not because of your mechanics, I know your mechanics are just fine. But because you didn't have game awareness, you went off on your own and got punished, engaged in fights after the fight was already over, staggered yourself. I dunno, maybe I'm remembering it wrong, in which case my apologies and just treat it as a hypothetical scenario. You can have good mechanics and collect good stats and not actually be contributing nearly as much to your team as a player with much worse stats and mechanics.

    Like, I feel like I'm pretty solid on D.Va. Whether or not I got good stats on D.Va or not in a round is pretty much entirely inconsequential to how much game impact I had. If the enemy team runs a pocketed Widow the whole game and I'm constantly in her face and not letting her take shots? I don't give a shit that I never killed her, that's not my win condition. The point is I stopped two enemy players from doing what they wanted at the cost of 1 ally, and therefore gave my team a consistent man advantage in every fight. Ain't no stat for "gravs eaten" but that can change the entire course of a match. And so on.

    Not to be a downer, but I think the best advice would be to take those stats and push them to the side and not look at them anymore, and instead watch some Jayne vids and do some tape review of your own play.

    Said it before, I'll say it again. Anyone looking to improve, go ahead and post a vod of game you've lost, and I'll be more than happy to look at it, and give tips for improving.

    Stats can tell you some things, like yes, if your aim is far below average on a character, then you need to improve it. If you meet the average, then your aim could be better, but it's not really holding you back. And if it's above average, then something else is definitely holding you back, but who knows what.

    Overwatch has a lot of depth to it, and people like to simplify it as, kill enemy, get wins. It's why everyone insta-locks DPS despite if you ask any pro, they will tell you DPS is the least important role. Have you ever won a game, and at the end the cards came up, and your team only had 1 card? Or no cards? That means the enemy won on stats. Did more damage, healed more, killed more, but they still lost the game. Getting a 4-man with dragon blade may show up as play of the game, but those kills don't matter if your whole team was already dead, and you die to the remaining 2 enemies.

    Stats require context, and are going to be very limited in what they can tell you.

    I made a smurf account to play DPS in comp which isn't something I typically get to do without feeling a ton of pressure that I'm going to drag the team down. I just did my placements and got gold, mostly playing with characters I don't have a lot of time on (Tracer and Doomfist), played a little Mei though here and there. And even before I looked at some profiles to confirm, I could tell I was playing gold placement matches. One thing you should consider when you think you're playing well is the context of what that statement means. Like, good for a gold? Play at the rank you want not the rank you are. Good for gold is still gold. And if you want to be plat you need to have the same understanding of the game as a plat player, this continues as you go up in ranks.

    Also realize that this can be difficult to achieve on your own. You would learn a lot faster if you were playing higher rank games (and putting the time/effort to understand differences), but you don't have the luxury. So it'll take longer if you just keep grinding away.

    Again, post a vod, and I, along with others, can help you improve on things that matter way more than just your aim.

    Kana
  • Brutal JBrutal J Sorry! Sorry, I'm sorry. Sorry. Registered User regular
    Played a comp game on the smurf as Mei, got King's Row, was on defense first, full hold, no deaths (on either side), gained 78 SR. Now, while my aim on Mei is quite good, it should be noted this shut-out was not due to that aim, but more position, wall usage, knowing when and when not to take fights, and good ult usage.

    ChanceH3Knuckles
  • ChanceChance Registered User regular
    78 sr lol.

    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
  • Brutal JBrutal J Sorry! Sorry, I'm sorry. Sorry. Registered User regular
    Chance wrote: »
    78 sr lol.

    though to be fair, it seems new accounts my have some volatile SR. Played another game as Pharah won that and it was plus 40something, then lose a game and lost 40something, and didn't do that bad in that game.

  • ChanceChance Registered User regular
    edited September 2018
    I find that kinda' weird. Like I'd assumed the 78 was because the system was like "this person needs to be 1500 SR higher than they are right now! SR! ALLL THE SR!!!" But to drop you 40 after a loss - that's kinda' like "the system has no fucking idea where to put me."

    Right now I think I'm where I'm supposed to be - 2450-2500. Won 2, lost 1, won 1, lost 1, won 1, lost 1 - at least I'm not fallin'! And the games are mostly good hard fights, so I'm very much in the sweet spot.

    My brother keeps buggin' me to make a smurf, but maybe I'm scared of what the SR system'll tell me...

    Edit: also I just noticed the quote under your avatar, @Brutal J - that voice line has been my text/email/general phone alert for like the past year. It usurped the Kim Possible ringtone and the MGS Codec sound.

    Chance on
    OuFrYk0.jpg
    'Chance, you are the best kind of whore.' -Henroid
    H3Knuckles
  • Houk the NamebringerHouk the Namebringer Nipples The EchidnaRegistered User regular
    Chance wrote: »
    I find that kinda' weird. Like I'd assumed the 78 was because the system was like "this person needs to be 1500 SR higher than they are right now! SR! ALLL THE SR!!!" But to drop you 40 after a loss - that's kinda' like "the system has no fucking idea where to put me."

    Right now I think I'm where I'm supposed to be - 2450-2500. Won 2, lost 1, won 1, lost 1, won 1, lost 1 - at least I'm not fallin'! And the games are mostly good hard fights, so I'm very much in the sweet spot.

    My brother keeps buggin' me to make a smurf, but maybe I'm scared of what the SR system'll tell me...

    Same thing happened to me when I bought my second account and jumped into comp right after hitting level 25. There isn't enough data to go on so you get these huge wild swings in both directions.

    The Escape Goat
  • RonaldoTheGypsyRonaldoTheGypsy Or just "Ron" Registered User regular
    As a Rein I think the worst feeling is when I go with my whole team of no dive heroes and then we are at the front and my whole team runs in front of the barrier and dies and then the enemy team just walks over me and I am like why do I do this to myself, I should not be the last one to die.

    Beyond NormalPhoenix-DFrem
This discussion has been closed.