At least I’m not rambling on about some number that has no meaning because I got beat in a game of HotS. 
2 Likes
So you’ve been working with statistics for 20+ years, and you don’t understand the significance of a statistic like winrates?
Or are you talking about the fact that I’m arguing about a videogame character, and not something more important or substantial? Well, news flash: You’re doing the same thing, buddy.
And funny thing: I don’t even find Deathwing that frustrating or difficult to play against. His kit is actually pretty well-designed, from my point of view; just needs more tuning. I simply know an overperforming character when I see (And play as) one.
2 Likes
“Winrates” means nothing independently and without context.
2 Likes
Well, it’s a good thing we have context, then.
Can you refute any of that? What more context do we need here?
1 Like
Is that why I was able to give her a ton of energy as Stitches?
completely made up nonsense.
1 Like
please just stop feeding this troll.
So, I just made up the fact that Deathwing has an extraordinarily high winrate, despite his high popularity?
I just made up the fact that a Hero’s winrate can go up or down, depending on how prevalent their counters are?
I just made up the fact that data-aggregate websites don’t have enough data for the first few days after a new Hero is released?
And I just made up the fact that Deathwing isn’t an unpopular niche character, isn’t situational to specific maps or team compositions, wasn’t just released a few days ago, and manages to thrive in the meta despite the presence of his counters?
For someone who managed to finish 2 science degrees and has 20+ years of experience with statistics, you don’t seem to be very articulate. I would think that after years of writing papers and engaging in scientific/scholarly debate, you’d be able to properly explain why my points are incorrect.
I would love to see you provide your evidence that proves this hypothesis.
See, here’s where your naivety shines through. Do you understand that “statistics” are a tool?
How old are you?
2 Likes
Or they can just permaban him, oh wait 
There were 1 week(?) since the nerfs and he loses his lead since it. With this rate it is possible he’ll be at 55% wr after another week without further changes.
But if he receives more nerfs (other than Heat Wave nerf), he’ll be the next Cho’Gall. A barely played Hero like almost every truly unique addition.
Or he could stay in the spot of Abathur: a highly banned oddness which isn’t op just really strong.
Around 8% are Diamond+. HotS has 330k followers on twitch (not everyone follows twitch who play HotS and following their twitch acc is pointless since HGC is gone, so the numbers are probably a lot higher, but let’s use this number).
8% is around 2.4k player. And after a week, from these 2.4k(+) ppl, we got less than 500 games.
I’m not satisfied with this sample size (not talking about DW’s 101% popularity… 
).
2 Likes
Well, different characters have different matchups. Tracer, for example, has favorable matchups against immobile and fragile characters that lack the tools to deal with her, and her winrates reflect that; Lt. Morales has only a 44% winrate against her, while a character like Fenix has a 62% winrate against her, due to Tracer’s inability to avoid Fenix’s main source of damage: autoattacks.
So, logically, in a meta where Tracer is highly popular, Lt. Morales’ winrate would go down, while the winrate for characters like Fenix would go up. And this was reflected in the winrates during the initial Tracer meta, before her nerfs.
Yes, they are a tool. A tool to try and learn the truth of a matter, rather than relying on mere anecdotes and opinions.
But you seem to be going off on a tangent, so I’ll ask you again: How are all of my points just “Completely made up nonsense”? If my points are wrong, then refute them.
Let me ask you something: What even is there to learn when playing against Deathwing? Dodge his skillshots, don’t burn your CC, use % damage, poke him down from a distance, be careful not to overextend in case he lands near you? All of those things apply to other characters too. There’s nothing fundamentally different about playing against Deathwing. So why do you keep arguing that people haven’t learned to play against him?
Also, why not look at Quick Match? You can’t ban anyone in Quick Match. So how does Deathwing have a staggering 65% winrate in Diamond+ MMR, with a 68% pickrate? That spells “overpowered” to me in big, bold letters.
No, he needs nerfs. And the longer we let this slide, the worse off the game will be. Unique heroes need to be balanced just like everyone else. People will eventually get tired of unique heroes if they appear in every match and can walk all over everyone else with minimal effort.
Your math is off. Only a small minority of HotS players play Storm League. The most-played mode is Versus A.I., the second most-played mode is Quick Match, and Storm League is only the third most-played mode.
So, Diamond+ is a small minority of players, within a small minority of players. So 500 games could easily be a pretty significant portion of all the Diamond+ games that took place in the past few days since the most recent patch, especially since Deathwing’s high banrate means he would only be played in 9% of all Diamond+ matches.
2 Likes
If he doesn’t take heat wave, I have no issue with the DW. He dies fast and barely contributes to his team that any other pick couldn’t have done way better.
When he does, it feels awful to play against. That’s the only fix I’m looking forward to, and hopefully it’s on the radar. He can have that power switched to something else as long as it’s not as awful to deal with, but HW is problematic.
Who can and who cannot 1v1 him and with what builds.
Baiting out Heat Wave so he can’t stack it up.
Not letting him use his flight or delaying it as much as possible.
Jumping on him when he lands.
When to focus and when to ignore him, because he can’t peel for his allies.
How to outmacro him, since his HS and (pseudo-)global Z are actually the same ability.
How to draft against him.
Some might seems “common” knowledge to you, because you might need to do/learn the same with other Heroes, but just because cooking XY and cooking ZY is all cooking doesn’t mean that you can cook ZY just because you cooked XY before. You still need to learn the same processes with the new Hero.
Because (my loved) QM is clownfiesta. It has no real barring on balance.
Agree.
Show me yours then
You say SL is not that populated. But I already used a shrinked sample size to start with, since twitch numbers are a lot smaller than the actual playerbase (now that HGC is gone).
And I didn’t even pointed out that my estimated 2.4k diamond+ players would create over 16k games if the average player plays 1 match/day, which is incredibly low (and the nerfs hit 10 days ago, not 7, so the real match numbers are even higher).
500 games are nothing compared to that. Even if you say my (directly lowly estimated) numbers are too “high”, we still need a ton more less players and games to make 500 at least 10% of all matches.
And you want to use that as some kind of proof… You say SL is barely played, that Diamond+ are rare among SL players and we know HeroesProfile doesn’t have all the matches and from all the HP-matches we have 9% with DW to estimate his “true strength”? Nonsense…
How can you consider that as a reliable sample size for balance (and pls, don’t start with QM again)?..
3 Likes
Well, for one thing, bear in mind that Deathwing players will also get better at playing Deathwing as time goes on. They’ll make it more difficult for you to jump on him when he lands, they won’t miss Heat Waves, they’ll get better at 1v1s, they’ll get better at bodyblocking for allies, they’ll get better at macro play, and they’ll learn to adapt against their counters.
Also, it’s worth noting that Gazlowe is literally the only character with a winrate higher than 50% against Deathwing. An unpopular, niche hero that’s easily countered in draft.
Sure. We’ll assume that the total number of HotS players is 330k (Probably not that exact number). Since Storm League is the third most-played mode, I’ll estimate that 20% of all players play it.
Now, Diamond+ is equal to 8% of all Storm League players, and since Deathwing has a 91% ban rate and 9% pick rate in Diamond+, only 9% of all games in Diamond+ will have Deathwing in them.
So, that’s 330,000 * 0.2 * 0.08 = 5,328 people. So if there are only 5,328 people in Diamond+, and if Deathwing only gets picked in 9% of the matches they play (Due to his high ban rate), then 500 games could plausibly be a large portion of the total number of matches played with Deathwing during the past week and a half, depending on how many games people play per day in Diamond+.
I beg to differ. It’s the most-played gamemode outside of Versus AI, so it needs to be taken into consideration for balancing. And it’s also a good way to see how well a permabanned hero performs when placed in an environment where they can’t be banned.
Because there are likely only a few thousand people in SL Diamond+ to begin with. A sample size of 500 games would actually be a large portion of the total number of matches played with Deathwing in Diamond+, during the past week and a half.
And the only alternative to these statistics is using nothing but anecdotes and opinions, and I probably don’t need to tell you that we’ll get nowhere fast if we rely on that.
my god, you are wasting your time with this karabars dude.
he already said that he prefers his own biased opinions over the only reliable source of stats that we players have. let him live in his delusions. he is garbage 2500 mmr anyway. wtf does a platinum below average noob has to say about balance in master+ games? absolutely nothing.
The same people in this thread saying DW is fine said the same thing before the devs blanket nerfed him. Stats don’t matter. Opinions and anecdotes are the only things that matter.
1 Like
So you think he has a high skillceiling?
I don’t, therefore I can only disagree with your attempt for a counterargument.
Based on the site which says DW is played/banned more games than they have… (101% popularity, that’s ridiculous.)
20% based on what?
By a guess that half the playerbase plays the most popular gamemode (vsAI)?
But fine, for the sake of simplicity let’s go with this made up number and the twitch numbers (which we know is smaller than the real playerbase).
5k ppl plays 1 match/day. That’s 50k games since the nerfs (pretty low estimation).
500 games is 1% of those total games.
So HP -by your generous hypothetical scenario’s numbers- has around 1/10 of all games (probably a ton less actually).
And you call that a solid samplesize. <1/10.
Not talking about how low 500 games are in general if we want to talk about overall balance. We need more games.
2 Likes
Compared to most sample sizes, that’s more than enough to make a judgment. The sample size for a test involving a new prototype drug, for example, might only involve about 300 test subjects. But they might be planning to distribute the drug to literal millions of people across the country.
Do they abort the testing or discard the results because they could only get a sample size of a few hundred people, and that’s a teeny, tiny fraction of the whole population? No.
A sample size of several hundred games is more than enough to make a judgment. If you’re just going to dismiss valid stats collected from hundreds of real games, and say we should rely only on opinions and anecdotes, then this conversation is going to go nowhere.
Speaking of which, have you even played any ranked games against Deathwing? Have you actually put any of his supposed counters into practice?
The goalposts will keep moving. “you need 50 games on DW”, “you need 100 games”, “you need 300 games”, “500 games aren’t enough”. No matter what statistics or evidence you bring to the discussion it wont matter.
2 Likes
He only wins in qm when matchmaker favor him with 100% win teams vs 100% lose teams. Dw is a joke without a main tank.
2 Likes