Hero's winrate does not represent how OP he is. Even less if you're watching it under "all time" tab considering trends and heroes themselves are constantly changing.
Warlock and Zeus are both amazing mids with ultis people usually can't handle. Jakiro and Lich are often used supports (one is a lot harder to use than another though, ie Jakiro who is prolly more used in high+ mm)
i think centaur is game-breaking. check out my win-rate as him. all games were after 54 matches. I was concentrating with him for team sakes rather for myself and that stampede (ult) made excellent ganks and escapes that made a difference.
zues with a good team will always win. as u think u are about to escape he ults (with refresher)
---
drow is only pub hero. no one picks her in competitive cause easy to counter.
pl also good in pubs only. in competetive u would counter him and kill him before he gets farm.
magnus needs skill to pull of, he is not easy to play hero unless u got ahead in farm.
'pub win rates'
'83% of games are in the normal bracket'
there you go.
While we are at it, we should buff QoP, LD, SD and Wisp?
Top Win Rates for Diamond this week
----------------------------------------------
Tree 65.58%
Wisp 61.50%
Zues 59.79%
Troll 58.00%
SB 57.44%
Nyx 57.31%
WL 56.33%
Chen 55.50%
CK 55.39%
Visage 54.28%
Top Win Rates for Bronze this Week
------------------------------------
WL 60.68%
Jakiro 58.36%
Necro 57.96%
Zues 57.36%
Troll 56.39%
Cent 56.03%
SB 55.69%
Tree 54.73%
---------------------------------------
There is a big difference in how well the heroes perform depending on the skill of the players.
Wisp is 40.78% Bronze and 61.50% Diamond.
Chen is 43.15% Bronze and 55.50% Diamond.
Tree, Zues, Troll, WL, and SB do very well at all levels of pub dota.
basically what others have said. One, I would think these stats are misleading because it's only people who are signed up with this site and are having their stats being shared (i'm assuming, could be wrong) Also as stated some hero are easier to play and are more forgiving. Drow or Sniper is an easy hero to learn is going to have a higher winrate then say a hero like Meepoo which requires microing skills. Even looking at your link Meeps is dead last with 40%, so if 55%+ must be game breaking, does that mean poor meeps is broken?
Also, if you sort by amount of times the hero has been used you'll find that the WR is pretty much close to 50% +/- a couple points except for drow which is 55. The more games played the better the sample size of getting a real WR%. As I mentioned, it's misleading to compare a hero that's only been used 6,207,675 times to one that's been used 23,748,620
Wisp is my favorite example. Low tier, good chance you'll lose. High tier, good chance you'll win.
When the sample is 6 million vs 20 million games and both are 100% of the games played sample size has no impact.
Even if you have stats off Dotabuff still gets the information on which hero was used and what items. They just can't connect it to players. So all of the games are there.
Zues is a hero that dominates pubs at all levels. He is easy to gank and shut down early...but the fact is even very high pubs just don't do it. Zues will crush you in any teamfight he when he does not die before casting much because of static field. It does 11% dmg per cast to everyone in a 1000 aoe...zues hits like a very farmed gryocopter on anyone who is not magic immune, even if he just casts on creeps.
If zues gets off arc-bolt-ult-arc-arc-bolt-arc that takes 6-7 seconds for a highly skilled zues and will have pretty much killed the entire enemy team from full health himself.
Tree is actually OP. But he is taking forever for people to figure out how to use him. Living armor can let you jungle, lane, and tower dive in ways that just are not possible otherwise. Tree may as well skill stats for any other level than living armor its so powerful you can win the game with just living armor and some basic support items. I like to get the other tree skills over stats....but seriously you could win with nothing else.
Trolls ult is the most powerful siege tool in the game. If they don't nerf it before he goes into captains he will become first ban and first pick. With trolls ult any team that was losing can take your rax on a single teamfight win so fast if you don't have glyph defending is impossible.
Wisp, 85% chance to win, this hero is way too gd in pubs, recently training with this hero, winning non-stop
"tree is actually OP"
lol, yes living armor is good, but its the ONLY thing the hero has. Stop pretending that its roof vs nothing, its roof vs Leshrac, SD, KOTL, Rubick, WR, etc
actually if you want to see a more accurate win rate, take a look at the winrate of the more popular builds. some heroes, when played by people who actually know how to use them, have over 60-70% win rate on the most popular builds, compared to their flat 40 odd % win rate.
Setting a new benchmark for speaking out of ignorance, the casual gamer has not only never played a real game of tree, but has never had a tree on his team and has only played 2 games against trees that fed.
-------------------------------------------------------------------------------------------------------
Zenoth makes an interesting point. But if you just take games where Alchemist got MJ, sure he wins 66.1% compared to his usual 45.5%...but in part he got the item because he was winning and in part he was winning because he had a good build and played well.
To disentangle it, consider how well Alch does with MJ vs the average hero who gets the item or vs another typical MJ carrier.
Naix with MJ wins 74.3%
Sniper with MJ wins 73.0%
FVoid with MJ wins 74.4%
DK with MJ wins 76.5%
So while Alchemist does a lot better if he can get MJ...he is really still quite a bit less effective using it than other MJ carriers.
In fact the average win rate for any hero getting MJ in a game is 72.6% significantly higher than Alchemists 66.1% So Alchemist is still weaker in pub dota than most carries even with correct items, even in games where he farmed well.
He was talking about skill build, but nice try.
Really living up to the american stereotype, ignorant and dumb.
I hope you are wrong again Casual. I had assumed he was making an intelligent statement. But I see you assumed that he was not. If he was referring to skill builds, rather than item builds I'm sorry to say its not useful information.
The most popular skill builds are a tiny % of games of any hero you pick...they are the longest games, the easiest games due to match making variation...they are really a very bad sample that does not represent what the hero can do at all.
The skill builds shown have high win rates because they are inherently biased toward winning games where the hero reached level 18. This is why the numbers never come close to 100% in total. Its just a very incomplete picture.
One thing you can see with DotabuffPlus however, is how the popularity of a particular build varies with the DBR of the players. For instance Diamond players are far more likely to gravitate to the best skill builds. You can compare the build popularity at different levels of play to just which one is relatively more successful.
But its not useful for comparing one hero to another, because the sample is extremely biased toward winning games.
Whoa, Casual, pls be polite.
Also all your thoughts are interesting, I just forgot to think about skill brackets and skill/item builds.
Still, can you agree with a thought that some heroes may be over-, and some underpowered or over/underestimated ?
Take Treant for example, before trying him I thought "Oh, he's so bad, we're gonna lose" how wrong I was. Was on hard lane, went full heal mode, no one ever died, none towers taken even I think, all eveno before the armor buff, so I underestimated a hero that is actually good and could lose because of that.
@Relentless, we don't see item builds, only items. There is no other thing he could be talking about referring to more popular builds when skill builds are the only information that we are provided with. What they mean or if they are significant is another issue entirely. Pretty cute how keep trying to go in-depth to show the insignificance of various data, which could very easily be applied to almost all the data except for out-liers (ie. 70% CW). Normal bracket provides pathetic data for obvious reasons, and in high level games its mainly 5 stack vs 5 stack (at the very top end), and once more, with the shitty match making combined with the numerous players playing in stacks of 55-75% win rate players constantly playing against 52-55% WR players, how much of the data is relevant at all? Also with reference to Alchemist, take a look at his 3rd ability, its not surprising that he has a lower win rate with MJ than other heroes.
Tree is a hero that is good, for the first 10 minutes, and only for heroes that he isn't in the lane with. Why? Because any other standard support hero in that tri-lane is going to be far better offensively, and defensively too. The passive regen becomes useless later on, and lineups with big burst damage or able to siege relatively effectively are going to ensure that Treant is a complete non-factor after 10 minutes.
Kind of like a SWM or Pudge in the sense that you do get an advantage early, but with Treant its a much shorter duration, and its main effect is passive defense, rather than a ridiculous offense that can create an advantage. He may not be a "weak hero", although I think he is, but he is certainly not a strong one, not right now.
Pub win rates don't mean shit, one of the most OP heroes in the game has a 43% win rate.
SWM is actually weak in pubs becouse people don't know how to use him. Remember when troll just relased, he had 50%+ WR. Thats becouse he is matter only on pressing R for him and that also benefits his team a lot. SWM is not like that. I saw so many people maxing his slow first.. really? Than in pubs people don't really aware of their positioning (speaking for my normal fucking bracket im in). And with SWM the positioning is the most important thing ever. Its like playing a ranged carry - owning from distance and not in the middle of the fight and die first.
Well, everyone who knows dota well would max his axes first, ie. the slow, miss chance and dmg is rly gd early and later on, gives the high dmg/miss chance when melee, slow when range, wut it is more perfect than that?
Yes, casual your zero games of tree, and zero games even with tree on your team have clearly informed your ideas on when what tree can do and how much it matters.
Casual you really need to stop insulting yourself, since you are a Normal Bracket player. People who play most of their games in Normal bracket are not pathetic and stupid. Many of them are smart and play hard. Your disdain for the vast majority of dota players (and insane arrogance since you are like them) is distasteful. The sample size of the dotabuff data is all of the pub games its more relevant to discussing the pub performance than any other data could be.
----------------------------------------------------------------------------------------------------
SWM is a very high skill cap hero. Even after people generally understand how to play him, he will still tend to lose in pubs because SWM is so difficult to use to full potential.
Tree is very easy to use if you just understand how to do it.
Whatever people do with troll, if their team groups to push at some point...all they need to do to win is time their ults when people are attacking in teamfights and hitting towers. Many skills in dota can be considered based on their roughly equivalent value in items. Trolls ult used correctly is better than 3 hyperstones on every hero...that is over a 30k gold advantage...its truly OP.
Troll wasn't so much OP before, I remember it had a higher cd and 4 sec effect on allies instead of the complete 8
well after getting tree banned in their first two games DD finally able to pick him. looking forward to seeing what Goblak does with him
DD doing a good job of showing the power of living armor this game. You can use heroes and tactics that just won't work any other way, survive more aggressive positions, get more farm...it has a huge game impact.
@Relentless keep talking mate, you do realise that the search function for other players isn't accurate? I can look up the highest ranked DBR players and it will have no results when I search for their matches for Very High games. My more recent matches have all been in the High or Very High bracket, and as I said earlier I dislike the match making system, and hence I only play IH matches (watching pro games, my replays and my friend's)
But keep talking like you understand the game, you have a negative KD in 15 of your last 20 matches, and your highest average GPM is 439, don't pretend that you are good at the game.
You are an idiot (which I think is a well established fact) if you believe that the average player is good at the game. The moment you learn to last hit, you are already better than 60% of players in the game. 1800 DBR is around the point when you start finding decent players, ones that actually know how to play the game. They aren't anything exceptional, but they do know what they are doing, at least most of the time. The sad thing is, this represents something like the top 10% of players, probably even less. How does that not indicate that the vast majority of players in this game are trash? Yes there are plenty of good players, I'm not denying that, but there is still a massive pool of extremely bad players in this game, sure some are learning, that doesn't remove them from that skill group at this point in time. ~60% of the total DotA player pool are fucking rubbish, and it doesn't take much to figure that out. You can go and spectate a random match, or you can go make a smurf account and experience it yourself first hand. If you want to include data like that as the sole reason to determine the strength of heroes, then you have to be mental. Even in higher level games its much more about the skill of the players, or the stack vs the other stack. Players that average 70% win rates are going to have a very different effect on data than those who average 50%.
SWM is a high skill cap hero? In comparison to which hero? Higher skill cap than Invoker? Puck? Magnus? SWM would be quite low in terms of skill cap, heroes which rely on well placed and well timed spells have a significantly higher skill cap than others, SWM doesn't even come close.
Who said Tree is difficult to use? Are you fucking kidding? You spam your shit on your allies as they need it, as well as on yourself. That's 90% of what playing Tree is about. And trying to conclude something from a single match is silly, Tree has been useless in many matches where the team is getting burst down straight away, or playing against heroes that they can't go super aggressive on. And when it comes to towers? Sure you will stop the passive heroes that take down towers after 10 pushes, but you aren't doing shit once any solid siege hero (Chen, Leshrac, Clinkz, LD, etc) decides to take it down. He offers far too little past the laning stage, where if you haven't created a big advantage already, its almost like playing 4v5 for the rest of the match. You can't create or help out too much in team fights, and are barely useful in both pushing and counter pushing, you don't have a stun, you don't have some game breaking ultimate. Compare that to the presence that other offlane heroes can have past the laning stage. I don't think I need to go any detail about what it is that they do: DS, Clock, LD, Tide, WR, Jugg
Troll is an incredibly powerful hero even without his ultimate. Spammable nukes equating to 350 damage, one with a 900 range vision giving throw, the other with a 60% miss chance attached to it, all the while having a base movement speed of 320/330, while having one of the highest EHPs in the early game.
Please sign in to post comments.
Let's look at the stats for all time http://dotabuff.com/heroes/winning?date=
We see, that it is not PL, not Drow(though close), not ,for example even hated by me Gyro, Magnus and Furion, not other hated heroes, but Centaur, Warlock , Jakiro, Lich and Zeus are actually above 55 % winrate.
It is either that most of all who play them are skilled people (which is impossible) or that something must be wrong with these heroes, that their abilities are game breaking in some sorts. Any thoughts on whether am I right or not, and why do you think so ? looking forward to productive discussions.