[Megathread] DotA 2

[PizzaSHARK]

icefrog is still in charge of balance decisions

valve doesnt have "interns" that do balance

Do we have any proof of this? The Icefrog thing, not the intern thing, that's mostly a joke. I'm just assuming that it's a group thing now, even if Icefrog is nominal head of it. Honestly, I figured it's always been a group thing - DotA just seems to have too many variables for it to be a truly one-man job.

[Charge me Doctor]

Welcome to people in charge of DotA knowing absolutely nothing about balancing games. Or it's probably more accurate to say they just don't care about balance anymore, just making money.

Or they want more Linas to be on mid to discourage melee mid laners with bottle that already got so fucking boring to see

- - - Updated - - -

Honestly, last time I played War3 DotA was like back in... 2006 or something. It felt like you were damned near always updating a new version, even playing games through Clan TNA and other "official" sources.

What? I remember not playing freshly updated maps for months before X.YZa version or something because fuck the day alchemist was released, it was a fucking disaster for dota, this OPed as fuck hero shouldn't even be implemented in his state.

Then again, it should be pretty obvious that doing something like increasing WR's Powershot range by almost 50% in a single patch, or giving heroes like Troll Warlord effectively an additional 100 IAS (this also applies to WR's Focus Fire, another reason why she's insanely strong now) is pants on head retarded.

I remember a long time ago people said that DotA and similar games were balanced incrementally. That is, you see fairly frequent patches that bring a variety of relatively minor changes, slowly and sometimes subtly pushing things in one direction or another. A lot of patches still have those kinds of changes, but there are also a lot of extremely huge changes (Naga losing 12 base damage, Ogre Magi going from 25% multicast to 40%, Powershot from 1800 range to 2600 range, etc) that just make me think that either they had an intern handle that, or they deliberately decided "we want/don't want to see this hero getting played a lot so let's do some silly shit to guarantee they get played/don't get played a lot."

Why is it retarded? And they make changes for exact reasons you stated here, they want some heroes to be more popular, and others less popular, that's why you see Windrunner Powershot range increase, to "counter" current meta. Small changes are never fun, if the game isn't patched every single week, changes like "Powershot now 1900 range" only to change it to 2000 in next month is shitty change. It's not league of legends to be patched every week or so, that's why changes are more "strong" here

I can't wrap my head around these "incremental" changes, if you look at any allstars changelog there is nothing "incremental" in them, there is no such a thing. If consider Seeker having his bloodbath buffed to get health from assist, and getting increased range on "assist range" for it in next patch, it's not incremental. Dota always was known for its patches that fuck up meta and make you to relearn things you already knew.

Also Wyvern is OP as fuck, plz don't buff

- - - Updated - - -

Win rate means less as a data point than the nitrogen proportion of my farts. And pub win rates mean even less.

Sniper no longer has easy counters. .

Sniper is kinda easily countered by Bane, Spectre, Lina or Windrunner from my experience, but he shrapnel should be nerfed, it at least shouldn't block daggers

[Planetarism]

Lol.

DotA development & balancing has stopped being a 1man job since 2005-2006. It is evidently implied in the way IceFrog wrote changelog notes as far back as then

- - - - - - - - - - - - - -

Add http://steamcommunity.com/id/MetaPlanetarism/ -&- http://steamcommunity.com/id/Planetarism/

Either experienced shadowshaman/broodmother/NP/warlock/leshrac/kotl/veno/drow splitpushers -OR- decent players highly motivated to learn this playstyle^ (I'll teach in botgame practice w/ you)

End games very fast, b4 25-30min - NecroBook|'Fresher|Aghanim|Drums|BoT - 2 Towers down by ~7:00, 1-2 Barracks by ~15:00.

[PizzaSHARK]

Sniper is kinda easily countered by Bane, Spectre, Lina or Windrunner from my experience, but he shrapnel should be nerfed, it at least shouldn't block daggers

tbh clockwerk gives me more headaches than anyone else if he's any good. PA can be brutal if they have a draft that can work with her, too. Honestly though I'm fine with literally anything fucking over blink dagger because that item is just way too fucking good for its cost. So many heroes completely ignore major balancing factors in their abilities and base stats once they get blink. I'd say just remove the goddamn thing entirely if not for the fact that some heroes would basically become unusable at competitive level without it (which makes me wonder why you'd even design a hero like that in the first place.)

Bane is worthless in the current meta. Troll gets too much damage too fast for Enfeeble to matter, Juggy can just faceroll with Omnislash, and Sniper can partly ignore Enfeeble with Headshot damage (plus he's usually too far away to easily Enfeeble anyway.) Might work paired with Lina or potm for setup, but I'd honestly rather have Shadow Demon for that and I can't see using a slot on potm when so many very strong or outright overpowered heroes that can lane almost anywhere exist. Bane is extremely good on paper, looks ideal to counter current meta heroes, but in practice he's incredibly easy to zone out or just blow up. Almost reliant on getting blinker before he can land Nightmare reliably and I'd just rather have Lion, VS, etc instead of Bane.

- - - Updated - - -

Crazy idea to deal with games being 70+ min stalemates:

Suspend all regeneration effects (except Living Armor and other outside influences) on all buildings, including backdoor regen (damage reduction bonuses remain intact), starting at 45 mins. Towers also begin slowly losing health, starting from the outward in, tier by tier. Towers can die to this health loss and will be marked as a deny. The health loss increases at 50 mins and every 5 mins thereafter, and the Ancient will begin losing health starting at 60 min, even t3 or t4 towers remain intact.

UT2k4 had a system like this in place for its Onslaught game mode, where the power core in each team's base would begin losing health once the game went beyond a certain specified time limit, to prevent boring, mindless stalemates. I think it could work in DotA. Removing or severely nerfing buyback would just make teams play even more conservatively. The idea is to force teams to fight each other and leave their base if necessary. Hell, buyback's nerfs could be reverted if it worked out alright. Could maybe couple this with some nerfs to the rubberband mechanics.

It'd probably encourage ratting, though. Not sure if I really want to see Profit make a return or not :-/

[Charge me Doctor]

The thing with Bane is that you can nullify Snipers farm by Enfeebling him on lane, and continuously CC his supports or himself, and since he lost his mini-stun on headshot, he can't end up in a situation when he attacks, you ult and attack procs headshot interrupting your ult. Also, iirc, you can juke Snipers ult with Nightmare.

If Bane have issues with connecting his ult when he has dagger/force staff then it's either inexperienced player, or unfavorable match up. Off course you won't pick Bane in every single occasion when you see Sniper, if you want just CC bot to shut Sniper down (which often is not enough, hugely depends on team composition and game flow) you can get Rattletrap or Lion.

I can't see how current Troll is any different from Void or pre-bash-nerf-Troll. Yeah, when fed or lucky (for opportunity to ult) he is good, but it's not impossible to counter him and/or shut him down. I personally love playing Wyvern when enemy team has Troll, it's too awesome to ult someone on their team when Troll uses his own ult. Or using force staff on cursed target to ram him into enemy backline hard hitters like Sniper. Or using force staff to save your frozen allies from enemy hard hitters like Sniper

I would also not want blink to be removed, it would favor 60 minutes games with Dusas, Sylas and shit all over again. The nerf of some sort is indeed needed, but it shouldn't be mana cost or huge CD increase.

- - - Updated - - -

Crazy idea to deal with games being 70+ min stalemates:

Suspend all regeneration effects (except Living Armor and other outside influences) on all buildings, including backdoor regen (damage reduction bonuses remain intact), starting at 45 mins. Towers also begin slowly losing health, starting from the outward in, tier by tier. Towers can die to this health loss and will be marked as a deny. The health loss increases at 50 mins and every 5 mins thereafter, and the Ancient will begin losing health starting at 60 min, even t3 or t4 towers remain intact.

UT2k4 had a system like this in place for its Onslaught game mode, where the power core in each team's base would begin losing health once the game went beyond a certain specified time limit, to prevent boring, mindless stalemates. I think it could work in DotA. Removing or severely nerfing buyback would just make teams play even more conservatively. The idea is to force teams to fight each other and leave their base if necessary. Hell, buyback's nerfs could be reverted if it worked out alright. Could maybe couple this with some nerfs to the rubberband mechanics.

It'd probably encourage ratting, though. Not sure if I really want to see Profit make a return or not :-/

Profits, Sylas, Tanky-crazy-siege-damage-dealers-like-CK-or-DK all over the place. Stale game up until 45 minute mark, rat your way to the TI5. Your idea doesn't force teams to fight each other, it forces them to pick Naga, farm for 45 minutes, leave Naga at base with refresher to defend and rat their way to the enemy antient ignoring all creeps and towers. Honestly, things like these would probably work in LoL (some sort of "after 3rd baron buff enemy team buildings melt to the ground") or HotS (where it actually already exist and called map mechanics), but not for dota.

And the main reason behind it not working is dota being very complicated game where changes like these shift meta completely, you have to be very careful with what you change, because it can decrease pool of reliable heroes to 10-15.

I personally don't think that there should be any kind of "sudden death" mechanic in dota, games do not often last 200 minutes, and seeing one or two of them is fine in my book. 40-60 minute long games are fine too, because, well, turtleing is a strategy too. Look at SC2 LotV, with beta changes to economy they completely fucked up all turtleing strategies and community is outraged (jk gaem iz ded). Removing one strategy because game take too long is not an option, they should provide something else then to fill removed strategy spot

[Sorcereria]

I didnt understan whole troll bandwagon becouse when they released all hero challange I failed troll in 9 games. Trully, I played it plain WRONG.

That hero is so overpowered once I did some games now after having all that experience from pro games.
Axes too f*** strong. 1vs3 with 10% HP, I think only PA could pull it off with enough rng.

That hero needs a BASE movement speed nerf SO BADLY by minimum 30-40, it is a joke, you can escape most gank attempts by simply outrunning enemy heroes. Kiting? Forget.

[Longview]

Do we have any proof of this? The Icefrog thing, not the intern thing, that's mostly a joke. I'm just assuming that it's a group thing now, even if Icefrog is nominal head of it. Honestly, I figured it's always been a group thing - DotA just seems to have too many variables for it to be a truly one-man job.

email him and he will tell you the same thing - icefrog at gmail

Lol.

DotA development & balancing has stopped being a 1man job since 2005-2006. It is evidently implied in the way IceFrog wrote changelog notes as far back as then

thats not what i said

what i mean is that he has final say in balance decisions

there's a list of core testers somewhere on pd

[StrawberryZebra]

http://www.dotabuff.com/heroes/winni...=current_patch

You'd be hard pressed to call Sniper, Troll, or Axe OP. Especially Axe. Those 3 heroes all have counters, and don't have winrates that cause them to be insta-locks by everyone. For the record, I have played numerous games as Sniper and Troll. And I can tell you they are still plenty killable.

All of these are very poor metrics to measure balance in my opinion. A much better way to look at it is how many other Heroes they're forcing out not because they counter them, but because they're just so good at what they do that its not worth picking anyone else. This is why those kinds of strong Hero end up being first pick or ban material. Like Lycan at TI last year, he was so good you either had to pick him, ban him or struggle to deal with him and of those, picking or banning was much easier to do.

Its also worth looking at it from the other side too. Omniknight has had a win rate consistently above 50% (and usually even over 55%!) on Dotabuff for as long as I can remember, is he overpowered? I'd say no, because he exists in a space that allows other Heroes like Dazzle, Witch Doctor or Abaddon, who can all fill a defensive support role along side him, to be good picks independent of Omniknights presence. At the same time, Omni doesn't crowd out other support picks like Lich or Venge even should he suddenly be overpowered, nor would Lich or Venge force Omni out should they become overpowered either. There are too many incomparable abilities in their skill sets which make it impossible for them to force each other out.

Why is it retarded? And they make changes for exact reasons you stated here, they want some heroes to be more popular, and others less popular, that's why you see Windrunner Powershot range increase, to "counter" current meta. Small changes are never fun, if the game isn't patched every single week, changes like "Powershot now 1900 range" only to change it to 2000 in next month is shitty change. It's not league of legends to be patched every week or so, that's why changes are more "strong" here

I can't wrap my head around these "incremental" changes, if you look at any allstars changelog there is nothing "incremental" in them, there is no such a thing. If consider Seeker having his bloodbath buffed to get health from assist, and getting increased range on "assist range" for it in next patch, it's not incremental. Dota always was known for its patches that fuck up meta and make you to relearn things you already knew.

Doing small frequent changes allows them to see what effect they're going to have in the short term, as well as see how its going to change trends in the long term. This is a good way to slowly nudge some heroes upwards in power, or tone them down as needed without totally messing up the meta and forcing people to relearn everything they already knew. Its also far less likely to end up with Heroes buffed to the point of stupidity or nerfed to the ground, and in some cases six feet under. An extra point or two of Str here or a slight cost adjustment on an item there aren't going to cause sudden and dramatic shifts the way patches do currently.

[Charge me Doctor]

Doing small frequent changes allows them to see what effect they're going to have in the short term, as well as see how its going to change trends in the long term. This is a good way to slowly nudge some heroes upwards in power, or tone them down as needed without totally messing up the meta and forcing people to relearn everything they already knew. Its also far less likely to end up with Heroes buffed to the point of stupidity or nerfed to the ground, and in some cases six feet under. An extra point or two of Str here or a slight cost adjustment on an item there aren't going to cause sudden and dramatic shifts the way patches do currently.

In case of games that rely on feedback from public beta tests - yes, but dota gets its feedback mainly from closed beta tests afaik, from people who dedicate themselves to test the changes made by other people who analyze balance. Dota don't take it "slowly", or maybe they do take it slowly, but under the curtain, and we see final result of whole balancing process that is then (afterwards) analyzed again and they do, what they think is appropriate changes. And if you follow dota changelogs history, they do these meta "scrambles" very often, while also doing small tweaks for some heroes.

I agree that doing small, frequent changes is the way to go, but it won't work for dota. It just doesn't work for some games, and some developers do not want to "share" what they are doing with game and how they approach balancing process. Frankly, seeing LoL patch notes, i start to think "do they just change things a little and see at response from community?", and if it is true, i find it amazingly irresponsibly, there should be people dedicated to testing balance, public beta tests should be kept for glitches, bugs and stability, not to balance.

My comment in quoted post about not being able to understand "incremental" changes was aimed at dota, not on overall approach of balancing the game.

I also almost always hear complaints about heroes being nerfed to the ground or buffed too much. Well, that just happens, and i'm pretty sure that they have all their rights to overnerf of overbuff whoever they like, it's their game after all, but i can't recall any character being too powerful (except freshly released characteres, thanks that dota 2 is saved from that nightmare) that it leaves zero ways to handle him

[Crackleslap]

Despite the hero being very awful in the meta game. Lifestealer (with farm) is a very good counter too oracle.

Oracle ults high hp hero -> do more damage for longer -> orb bursts -> watch enemy hero blow up.

[StrawberryZebra]

In case of games that rely on feedback from public beta tests - yes, but dota gets its feedback mainly from closed beta tests afaik, from people who dedicate themselves to test the changes made by other people who analyze balance. Dota don't take it "slowly", or maybe they do take it slowly, but under the curtain, and we see final result of whole balancing process that is then (afterwards) analyzed again and they do, what they think is appropriate changes. And if you follow dota changelogs history, they do these meta "scrambles" very often, while also doing small tweaks for some heroes.

I imagine that the Dota player base probably has more played time with Hero within an hour of a patch than a team of testers could manage with a year of testing. You could get a lot more feedback a lot faster with small changes than you probably ever could with a testing team. You also have to consider the skill level of the testers too, some heroes require a lot more time to get at least competent with. Finally, testing out changes for a Hero they're not at all familiar with is going result in poor feedback.

Releasing small regular changes probably generates more useful feedback for Devs than any other method. Unfortunately it does make the game feel very much like its in a constant state of beta. It's a no-win situation.

As things stand at the moment I think they like going overboard with both buffs and nerfs to intentionally shake things up. Its almost a fail safe way to make sure that a hero previously underplayed in Pro matches ends up being picked while the previous round of strong picks vanish almost entirely from the scene until its time for them to get buffed again.

[Drunkenvalley]

So what do people here think about Spectre+MoM anyhow? I forgot what game it was I saw it, but it seemed like an interesting build. Tried it in a few pubs, it seemed to do a lot towards making you pretty scary early on with Desolate.

[PizzaSHARK]

6.84 is allegedly going to arrive after Starladder ends. I'll probably throw things if blink dagger doesn't get nerfed :-/

[Charge me Doctor]

It's probably characters that benefit from blink dagger will be nerfed

- - - Updated - - -

I imagine that the Dota player base probably has more played time with Hero within an hour of a patch than a team of testers could manage with a year of testing. You could get a lot more feedback a lot faster with small changes than you probably ever could with a testing team. You also have to consider the skill level of the testers too, some heroes require a lot more time to get at least competent with. Finally, testing out changes for a Hero they're not at all familiar with is going result in poor feedback.

Releasing small regular changes probably generates more useful feedback for Devs than any other method. Unfortunately it does make the game feel very much like its in a constant state of beta. It's a no-win situation.

As things stand at the moment I think they like going overboard with both buffs and nerfs to intentionally shake things up. Its almost a fail safe way to make sure that a hero previously underplayed in Pro matches ends up being picked while the previous round of strong picks vanish almost entirely from the scene until its time for them to get buffed again.

Yeah, that's good way to get statistics (win rate, build paths, power spikes and stuff) but it has its downsides and statistics is not only criteria for balancing heroes, i just think that dota devs do not want this public approach to balancing their game, and i personally see no issues here.

I think we put different meaning in "feedback", small regular changes work fine with statistics, but statistic is a huge lie (in LoL, any nerfed character drops in popularity even if nerf was justified or unnoticable), feedback of sorts like long-well-made-posts-about-character-i-always-loved-but-now-you-ruined-it doesn't help much either, it's just public opinion. While having people (say, three hundreds) daily working with numerous amount of patches on heroes that have to make reports on how every change affected gameplay (say, you see 50% increase in powershot range in changelog, but when they were working on patch, the tested all ranges fron 1% to 100% increase and made in clear that 50% increase suits their goals better, we can't see that and (even if i want to) we shouldn't because it would "force" devs to cater to public pressure)

And you are right, they love to shake everything up, there is buffs like 25 to 14 CD on Centaurs double edge, there is also buffs like +10 damage on Searing arrows of Clinkz, nerfs like -0,4 int gain on Tini, and -300 death coil range on Abba.

[Lysah]

It's probably characters that benefit from blink dagger will be nerfed

So, all of them? The game has definitely degenerated to some degree. People get blink on Sniper so that they can't be caught out, and people get blink on Troll so that they can jump on blink dagger Snipers. It's become a game where everyone gets blink to escape everyone and everyone else gets blink to catch everyone in the first place.

[PizzaSHARK]

So, all of them? The game has definitely degenerated to some degree. People get blink on Sniper so that they can't be caught out, and people get blink on Troll so that they can jump on blink dagger Snipers. It's become a game where everyone gets blink to escape everyone and everyone else gets blink to catch everyone in the first place.

Pretty much. It really is funny if you look at people discussing item builds/skill builds etc for heroes, blink dagger always comes up like that. "You need blink to position because Lion has blink to instahex you and sniper has blink to get away from your blink and you need to blink so you can blink to stop their blink before they can blink and..."

I'd think that should be a sign there's something wrong.

- - - Updated - - -

(in LoL, any nerfed character drops in popularity even if nerf was justified or unnoticable)

Happens all the time here. How many times have we seen characters become popular because they got, like, +2 STR starting (omg, hero was shit before, but now that they have 38 more starting HP, they're totally pro stfu nab they were shit b4) or suddenly disappear because they lost 3 base damage? It's almost like the perception of nerfed/buffed matters more than the actual nerf or buff.

[Drunkenvalley]

Blink dagger is my primary reason for enjoying Storm Spirit mid- and lategame. Ruined so many initiations just balling through them. Although it really feels like Dota 2 has just become "initiate before they can blink", and it really feels like Blink Dagger has little real counterplay. If you can't hurt them for the next 3 seconds they're gone.

[PizzaSHARK]

Blink dagger is my primary reason for enjoying Storm Spirit mid- and lategame. Ruined so many initiations just balling through them. Although it really feels like Dota 2 has just become "initiate before they can blink", and it really feels like Blink Dagger has little real counterplay. If you can't hurt them for the next 3 seconds they're gone.

Wouldn't mind blink dagger's cooldown being reset entirely if you take damage from heroes or towers. Make it clearly a mobility or initiation item, not a get out of jail free card.

That'd be a substantial nerf to Puck, though :-/

[Charge me Doctor]

Wouldn't mind blink dagger's cooldown being reset entirely if you take damage from heroes or towers. Make it clearly a mobility or initiation item, not a get out of jail free card.

That'd be a substantial nerf to Puck, though :-/

Oh no, a hero with a make-a-choice ability to either do damage, initiate fight or escape losing a free get-out-of-jail card via blink dagger nerf, now he will have to make actual choice and not yolo in fights and get away with it. The her ois ruined

[Drunkenvalley]

I'd argue for a nerf to the CD of Blink Dagger itself perhaps. A 12 second cooldown is pretty much nothing, very often enabling not just initiation for the fight but also an easy means of exiting it. Something I see a lot is players engaging into a short fight with blink, then back off and just blink out again. Worse yet, the reverse often happens as well, with a hero fleeing from the scene using blink, only to be able to re-initiate very shortly after.

The fact that the cooldown is short enough to be able to do that is pretty nasty.