[Megathread] DotA 2

[EqualWin]

Razor is probably my favourite for the event. Has good amounts of early teamfight and can snowball well. At the same time he turns into pushing/objective taking monster once he gets fat.

can you drain pet dmg?

[Lysah]

can you drain pet dmg?

Yes, and it's very easy to get a rampage at 5 minutes against an unsuspecting team. Get 300 damage and go two shot people while you totally ignore the beast.

I've probably had closer and more intense games in this mode than I've had in regular matches in months. The addition of sets makes people try way harder and never give up or lose hope.

[Deleted]

If you have someone to upgrade the beast to level 3, takes the support route and pick zeus, 100% win if you get the beast to level 3 without getting 100% stomped.

[Lysah]

The amount of people who click random and then refuse to listen to people who have actually done this event before makes me want to just give up on humanity. Clicking random is already an AWFUL idea for year beast, but getting a hero like SA when the enemy has a free 3k hp gem of true sight and not listening to your team is just the pinnacle of dumb.

We did beat a team with two tier 3 upgrades last night, though it wasn't easy and they weren't smart enough to get speed path for the scepter necro ult spam.

[Sorcereria]

I hope the idiot whos idea was a timed windows to play beast dies in a horrific car crash where he isnt found for 2 days after accident so he can suffer more.

Why? Becouse spanish living style grownups working partido, in the morning and afternoon with siesta time, basicly CANT PLAY it.
How f**** awsome is it to actually buy CM arcana and not be able to even use the 5k points that you bought with arcana.
- I come back from morning work, NICE! Next event in 3hours, I just missed it. Great, I have to leave to work again by the time I could finish next game.
- Uff, afternoon work done, now I can jump into event. Nope. I am 10min late, next one if 3 hours again so either I can sleep decent 8h or sleep 6h just to play the beast ONCE.

For once, I will try to reimburse money from the arcana just to screw up with valve. Glad I am not the only one doing it.

Last year beast was awsome. Just FYI.

They should open and close valve HQ toilets like in the beast event. Suffer b***es.

[Deleted]

Not to mention that MM isn't even working for the event most of the time. I got in 2 out of 6 events, other times it would just be finding match for 20+ minutes until I cancel it, or dota network crash after 10 minutes.

gg volvo.

[EqualWin]

I hope the idiot whos idea was a timed windows to play beast dies in a horrific car crash where he isnt found for 2 days after accident so he can suffer more.

Why? Becouse spanish living style grownups working partido, in the morning and afternoon with siesta time, basicly CANT PLAY it.
How f**** awsome is it to actually buy CM arcana and not be able to even use the 5k points that you bought with arcana.
- I come back from morning work, NICE! Next event in 3hours, I just missed it. Great, I have to leave to work again by the time I could finish next game.
- Uff, afternoon work done, now I can jump into event. Nope. I am 10min late, next one if 3 hours again so either I can sleep decent 8h or sleep 6h just to play the beast ONCE.

For once, I will try to reimburse money from the arcana just to screw up with valve. Glad I am not the only one doing it.

Last year beast was awsome. Just FYI.

They should open and close valve HQ toilets like in the beast event. Suffer b***es.

ROFLMAO

1ch

[PizzaSHARK]

http://www.datdota.com/blog/?p=1110

So this is pretty neat. It's a useful way of gauging the relative strength of heroes inside each patch's competitive bubble. It doesn't do much for heroes that aren't seeing much use, or are being used poorly, but since the game is balanced around the competitive bubble, I guess those data don't matter much, anyway.

Now we can just spam this URL anytime some fuckwit starts talking about win rates as if they matter for shit

[Lysah]

Pretty sure I've argued this exact thing a million times before and not been taken seriously, but some guy writes a blog post about it and it's all over the internet.

His metric is still extremely poor and says very little about a hero, but it's a slight step in the right direction. Pointing out the obvious is lost on a lot of people. Alchemist is only 40% winrate must be shit tier hero pls buff already gaben noob! Or, maybe it's the fact that low skill 1-2k MMR pubs have no idea how to play him and feed a lot when they get him and lose every time?

CM winrate has gone down since she got an arcana. Is it because she is a bad hero and deserves it, or is it because more idiots are picking her just to show off their shiny toy when they have no idea how to play her?

Comparing ELO delta is better than nothing, but it still doesn't account for people who don't know how to draft, people who draft against the meta game, etc. A lot of very underused heroes get suddenly picked by tier 2 teams trying to beat tier 1 teams, if they lose it isn't necessarily because the hero isn't as good. A big enough sample size would alleviate this issue in competitive, but we are nowhere near that, maybe in another 200 years of DotA =p

[StrawberryZebra]

http://www.datdota.com/blog/?p=1110

An interesting read, certainly. Relevant and useful information? Probably not, assigning each hero an ELO value is all good and well, but its very important to remember that ELO also tends to correlate strongly with Win Rates. It gives us a sum of how much ELO each Hero has earned or lost Pro's over the patch which is something most, if not all, Dota players are at least familiar with as a measurement of a players skill (or lack of). Does it do what it sets out to do and give us a metric to score how "good" a Hero is? Objectively no, it does not. Its just performing analysis on Win Rates and ELO data, and while it may be used as a tool to help us identify strong Heroes, it is really no better at doing so than Win Rates or ELO are.

There are also hundreds of other variables that aren't taken into account. I doubt many of them can actually be quantified in any meaningful way so getting a true measurement of a Heroes strength is practically impossible.

Pretty sure I've argued this exact thing a million times before and not been taken seriously, but some guy writes a blog post about it and it's all over the internet.

Then maybe its time for you to start your own blog? Apparently its the only platform in which one can state the obvious and have people listen.

[Deleted]

Win rates show how good a hero is in pub games (most of the time), best example would be Spirit Breaker is the best hero in game vs a team with bad teamwork, but against a decent team he is 100% trash.

[Lysah]

Then maybe its time for you to start your own blog? Apparently its the only platform in which one can state the obvious and have people listen.

I was doing that for a while, and people did seem to like it. But, I needed help making pretty gifs because I have zero artistic knowledge or talent and my support quit on me so I gave up

[EqualWin]

Win rates show how good a hero is in pub games (most of the time), best example would be Spirit Breaker is the best hero in game vs a team with bad teamwork, but against a decent team he is 100% trash.

thats why space cow isn't picked at all in pro games right?




idk rating /wr are irrelevant most of the time, except for alchemist, his wr show how trash the hero is.

[PizzaSHARK]

Pretty sure I've argued this exact thing a million times before and not been taken seriously, but some guy writes a blog post about it and it's all over the internet.

Welcome to how the internet works

His metric is still extremely poor and says very little about a hero, but it's a slight step in the right direction. Pointing out the obvious is lost on a lot of people. Alchemist is only 40% winrate must be shit tier hero pls buff already gaben noob! Or, maybe it's the fact that low skill 1-2k MMR pubs have no idea how to play him and feed a lot when they get him and lose every time?

Funny thing, someone was recently asking about buffs for Alchemist and I pointed out that Alchemist is still a strong hero, just a bit niche. He's kinda poopy as a hard carry but works great as a sort of utility core, and his stun and acid spray are still super good. His winrate crashed because he went from being a hero you could not possibly fuck up with that had a guaranteed 5 sec stun for 500 damage to a hero that required extremely strong positioning and timing to get that really good stun on, and it happened basically overnight. How many comp games did we watch where the pros were consistently stunning themselves instead of their targets? Hero wasn't broken, hero just went from Viper levels of "hero plays itself" to something a little closer to, like, Meepo or Invoker.

Comparing ELO delta is better than nothing, but it still doesn't account for people who don't know how to draft, people who draft against the meta game, etc. A lot of very underused heroes get suddenly picked by tier 2 teams trying to beat tier 1 teams, if they lose it isn't necessarily because the hero isn't as good. A big enough sample size would alleviate this issue in competitive, but we are nowhere near that, maybe in another 200 years of DotA =p

Yup, I've tried to explain that to people but... nope, Sniper is a shit tier hero that's only good because I'm a trench baddie (never mind that "trench" never actually gets defined... it seems to mean "disagrees with me") until the pros start using him and suddenly Sniper is a very good hero in certain situations, etc.

I think the fact that pro teams are completely unstable is also a huge issue towards accumulating useful data. A team that had played together for three years would probably have pretty consistent performance, making them a good source for data. But teams are constantly shuffling rosters, even at the top tier, and the lower tiers are damned near playing musical chairs with their players.

- - - Updated - - -

I was doing that for a while, and people did seem to like it. But, I needed help making pretty gifs because I have zero artistic knowledge or talent and my support quit on me so I gave up

Why not learn how?

thats why space cow isn't picked at all in pro games right?




idk rating /wr are irrelevant most of the time, except for alchemist, his wr show how trash the hero is.

Alchemist is fine.

[Longview]

I was doing that for a while, and people did seem to like it. But, I needed help making pretty gifs because I have zero artistic knowledge or talent and my support quit on me so I gave up

http://blog.bahraniapps.com/gifcam/

OnT: i think alchemist is a pretty good support hero if you play in a stack. go medallion after boots and then alternate between killing people and stacking camps, farming them whenever. blink+mek in whichever order is a pretty good build. alchemist also scales really well and can quickly fill item slots if you start pulling ahead in gold/xp.

[Lysah]

Why not learn how?

Not worth it.

http://blog.bahraniapps.com/gifcam/

OnT: i think alchemist is a pretty good support hero if you play in a stack. go medallion after boots and then alternate between killing people and stacking camps, farming them whenever. blink+mek in whichever order is a pretty good build. alchemist also scales really well and can quickly fill item slots if you start pulling ahead in gold/xp.

Yeah I've heard about things like this, but I assume they make massive files and I'm working with a strict 5MB limit (which is tiny for a gif) to be able to upload to imgur so I can embed. All of this could probably be fixed if I started my own actual blog but that is DEFINITELY not worth it. Maybe when people start paying me :P

[PizzaSHARK]

Man, watching some of these DAC games and they're playing Axe so poorly. Tranquils into Vanguard, completely pointless - Axe already has highest HP regen in the game, you only need one regen item, not two. Completely skipping Battle Hunger despite it being god-tier now. Blade Mail against teams without an orb hero (basically Silencer and OD.)

Bleh.

[Deleted]

Problem with alch is he has 900 hp at 11 as a strg hero, really easy to kill and skillset when compared to other heroes like him is pretty bad. Alch starting as a support works but i don't know why you would when you could have a wk do the same job a million times better.

If they want alch to be a good carry again then they should :

Buff his strg gain

Buff agi gain

and suddenly he isn't paper anymore.

[Sorcereria]

Maybe but it doesnt change the fact that untill you actually start picking tanking items, its useless.

I would say ulti+stat gains already offsets greedeness for dmg items.

[Deleted]

the str and agi gain are too low, 1100 hp and 4 armor at level 16 is a joke really, you need bkb, basher and something with attack speed to be useful in a team fight, otherwise you might as well just get blink and have the same usefulness.