This is an excellent idea!
The data is already available, on the each hero page. But its not easy to compare use of an item by different heroes.
Yeah they should be using percentage base data. Like Warlock builds Scepter on %56 of his games.
For instance top Aghs usage by raw numbers...the following heroes are not in correct order if you show the %.
Pudge 25%
Warlock 61%
Zues 55%
Invoker 45%
Juggernaut 32%
QoP 41%
Lich 47%
That shows a very different picture than the total usage, which is heavily biased by the popularity of the hero.
Other popular aghs heroes as significant as pudge or more did not make the list because of this distortion...
Enchantress 53%
Meepo 51%
Tiny 48%
Clockwerk 43%
Pugna 42%
Witch Doctor 39%
Ogre Magi 38%
Shadow Shaman 38%
Earth Shaker 37%
Chen 35%
Necrolyte 34%
Brewmaster 33%
Sand King 27%
Beastmaster 27%
AA 26%
Dark Seer 24%
Visage 24%
Lina 22%
Lion 21%
Doom 21%
Jakiro 20%
OD 18%
On several of these heroes aghs scepter is their most often purchased item...even ahead of the most popular boots. Its absolutely core, but you can't tell which ones from looking at the item page because the items real use is hidden by hero popularity.
Ooooo!
With a bit more calculation you could show the importance of the item even more dramatically....for instance.
Win Rate with Aghs vs Without
Warlock 74% with 35% without
Zues 71% with 36% without
Meepo 65% with 22% without
Brewmaster 68% with 39% without
This shows very clearly that people who know how to win with these heroes get aghs scepter. People who don't know how to play them don't get aghs, or fail to get it by the end of the game. Of course winrate with divine rapier is high for any hero...but win rate without it is basically identical to the average win rate. If you are only showing the impact on the heroes most likely to get the item this sort of distraction by outlier data will not be visible.
it just like if u have aghanim on support, it means that u have farm\kill and ur team are wining. Because of that aghanim has more winrate
No, that's not the same at all. That is like what I said about the DR. But aghs heroes are far more likely to lose without their core item. They are not far more likely to lose without some random expensive item...like dagon 5 or hex or heart.
Win Rate with/without Dagon 5 vs Average winrate
Warlock 77% with, 58% without, 58% avg
Zues 80% with, 55% without, 55% avg
Meepo 78% with, 43% without, 44% avg
Brewmaster 79% with, 48% without, 48% avg
Dagon 5 is nice if you happen to be way way ahead, but it does not win the game for these heroes. If they don't get dagon 5 they are just as likely to win. But if they don't get that aghs, their winrate drops way down. For warlock not getting aghs drops winrate by 23%...and what you don't even see there is that the games on the margin are won or lost based on how soon you got the aghs.
This is a big deal even if you are playing against people who really know how to counter warlock and pick up a quick diffusal on the carry. It's easy to purge away one golem, but the 2nd one still gets you. Diffusal has an 8 sec cd and in that time, the golem has already done its job.
I manage to build aghanim's in almost all of my meepo games(very few exceptions), but I doubt that I have 65% win rate with it.
It's because the averages are mostly based on very low-level pub play, where rarely a game will go longer than 20 minutes if there's a meepo in. And meepos manage to build scepter if they pwned early game and got fat. because most players get it after mek, vlads, treads.
Or I'm worse than average meepo player....
I hope to hear an official word from Dotabuff administration on this topic. I'm also going to send this feature request via their contact e-mail.
The best meepos typically are getting scepter as their 2nd item after boots and blink. Of course the best could win with about anything, but the typical build for good meepo players is not mech or vlads...those are too slow...used to be popular 2 years ago maybe.
Often they would vlads later, its best as a late game support item. They might go vlads first if losing. But the second you pick up aghs you just jumped a level on your ult...its almost like getting an instant 5 levels...but even better because stats sharing % goes up a huge amount...30% to 100%.
Just look at your winrate 52.9%, Vlads win rate 52.4% also....Aghs win rate? 65%
I bet if you start rushing aghs you are going to kick butt.
http://dotabuff.com/matches/272410810 <=== Neopets vs Dignitas.Universe
http://dotabuff.com/matches/272901273
check out Neopets games to see how to do it...neopets master meepo specialist, join his meepo guild
This guys games alone increase the winrate of meepos measurably...including dota 1 he must have like 4 k meepo games.
I disagree with "ooh this big item has higher winrate it is good!" mentality. Yes, a Crystal Maiden has a higher winrate with Aghanims, but that's because before she got the Aghanims she was already winning; otherwise, if early game was bad, she wouldn't even have the money to get Aghanims.
Basically, if you have money to get big items, you probably win. All big items have a high win rate.
I see you either did not read, or completely failed to comprehend it. You are of course entitled to your opinion that core items are indistinguishable from any expensive items...but there is no evidence or rational for that position.
Support, it should be % based and not total number of times used.
Also the % of games a certain item is built and it's winrate is something to look at and then think about, more in line with what relentless said. You don't blindly say oh X item have a massive win rate on Y hero I must get it, or go the other extreme and blindly ignore all the stats because all expensive items give a high win rate. Crystal maiden have an 89% win rate with a radiance, best item ever on CM must get it... That's not what the stats are there for.
side note: OD on that aghs list is an eye sore. Aghs is not even close to being "core" on OD.
Relentless
I know that the best meepo players rush aghanim's right away. but I mean that average/most players don't rush it. they probably buy it after treads,meks,vlads, popular items on pubs, and they need to be in lead in order to be able to buy aghanim's after this. IMO any item that average/most players do not plan to build early/don't rush them, have high win rates. but if you rush those items that doesn't mean you are bad.
It's like tinker having 45% win rate with boots of travel whilst any other hero has 60-65%+
You are right aghs is not very good on OD, but its not terrible. Anyway that's why its only used in 18% of games...its not the best choice.
-------------------------------------
@Luxon Meepo is hard to play well, but if you can play it pretty well you really should try getting aghs first and see the difference it makes. See if you can do it. Its so good if you can, that you should at least try to do it.
@relentress
You don't understand me :d. I mean that I always rush aghanim's first right after tranquils and that's why I have lower than average win rate with it. I never build vlads and almost never build mek. And I mean that this feature doesn't decide how good your build is unless you follow the most common one.
If the "without" winrate and the average winrate are identical or very close, that means there is a minimal amount of games being played with that item. Of course with your Dagon 5 example it seems Aghanim's is better, but Dagon 5 is an item that just doesn't really appear. I'd suggest using other items like Eul's or Heaven's Halberd to compare. (Not doing it myself cause match starting and I'm lazy)
Right, you are lazy at investigating and lazy at thinking and that is why....you have a poor understanding of these things. You are not dumb...you are just lazy.
Euls and Halberd are not good for meepo. That is why no one gets them except as a troll build. And that is why like Dagon 5, also a bad item for meepo not getting them has no impact on winrate...Not gettting Core items however, like Aghs drops win rate by a huge amount.
Lets consider an example that is easy to understand. Chen's most core item is Mechanism. The combination of his ult and mech give him a huge teamfight advantage to push towers early.
Chen winrate with Mech 52%, without 9%...average 45%
When Chen doesn't get mech he loses. Sure Chen could get whatever expensive item if he is way ahead and win because he is way ahead...lets say Eblade, but that does not fullfill his role on the team. Chen does not need Eblade to win.
Chen wins 81% with Eblade, but 45% without it, 45% average.
Hmm. That makes sense. Although, given the statistics, Chen basically never gets Eblade. Mek isn't all that expensive so it's quite plausible to be present on a losing team. Of course heroes without their core items are going to suffer, and the statistics given support that, but if those luxury items you listed are hardly present in any match - their winrate without the item being (almost) identical to the one across the board is indicative of this.
Expensive inappropriate items are almost never used not merely because they cost a lot...but also because they are bad on the hero.
There are limited resources in a dota game. But unless the game is complete disaster you can afford at least 1 core item if you manage your resources correctly. If you don't use that precious first 3k gold to get your core item and get a stupid dagon instead your hero will not function correctly and you will lose the game the vast majority of the time. Or if you save up forever trying to get a Tier 6 item without getting your core....you lose the game before you ever get it. That is why such expensive items are almost never purchased on inappropriate heroes....anyone who ever tries fails so hard they don't get the item before the end of the game.
The fact that Divine Rapier, or Eblade is almost never present on Chen is not luck or some fad or a coincidence. It is because trying to get Divine on Chen is so terrible that you can only do it in a game where it was won for you 4 v 5 while you were AFK jungle farming. Of course if you do get it...then the game was won 4v5 because anything less and it would be over before you finished DR...so if you happen to get DR you won the game. But this almost never happens and so not having DR has no impact on winrate.
This is why I prefer playing with two supports with one warding with focus on small items giving a better team support who would benefit from the Agha's upgrade the chance to get it.
E.G. I play CM with someone else keeping cash on the Warlock or Rhasta with gold staying on a lich or something of the kind.
That's what pro games are like (or what I think they are like). There's one guy that has nothing but boots and wards wardswardswardwardwardwrd and then there's the other guy that gets mek, and maybe pipe or mana boots.
However, in pubs I doubt people would be glad to hand all of their gold to wards. I did a game like this once, and it does not look good according to what most people say (i.e. omg noob 2 KDA and 400 GPM go away feeder). Even though the game was won, heh.
Basically in pubs even supports are going to try get their own special Scepter or Scythe because that's just how people work in pubs. I've heard "if your team doesn't commend you then they don't know where their minds are" to a carry that carried, but appreciation for supports is much rarer. People just go "wtf PT at 30 mins lol" and ignore the wards on the map most of the time. I'm glad to say that at least some people appreciate supports, but they're definitely a minority in the current system with everyone looking at KDA and GPM and such.
@Relentless I literally agree completely with you I don't know why we were arguing lol.
Zaloguj się aby dodawać komentarze.
Overall amount of times item purchased on a hero doesn't objectively show it's popularity on a hero. Instead, i want to see how many times item is purchased on a hero per 10 games, this would actually show how popular this item is.
For example, http://dotabuff.com/items/aghanims-scepter says Scepter is built 30k more times on a Pudge than on a Warlock. It may seem that Scepter on a Pudge is more popular than a Scepter on a Warlock. However, pudge matches played this month is 4kk and Warlock ones are only 1,6kk. It means Pudge purchases ~2,5 Scepters per 10 games and Warlock does every second game!