The Dota 2 Competitive Hero Tier List for November 2018 includes data from 7.20 to 7.20c, for a total of 234 games. All heroes are split in five groups.
All heroes are arranged in five tiers based on their draft rate. The tiers are viewed as clusters of data, and each hero is placed in the one they are closer the most.
The requisites for each tier never change. The goal is offering a neutral comparison that shows how the meta evolves over time.
The win rate has no effect on the tiers, but is included for further info. The tiers depend only on drafts, and aren't affected either by the win rate of a hero, or by their performance in pub matches.
This tier list includes all matches played during version 7.20, from November 19th to 30th. Even if the period of time considered is quite limited, we still have lots of data thanks to The Chongqing Major Qualifiers.
More: All the Dota 2 Tier Lists
Phantom Assassin’s Blur - a few fatal mistakes that could be resolved easily
Phantom Assassin is the hero with the highest win rate (above 53%) among the top 10 most popular picks at the Divine rank. The situation is even more critical in competitive play where she achieved an astonishing 67% win rate with a 71% draft rate.
I do not remember the last time a hero obtained such stats. Even when our games were plagued by the offlaner Omniknight, the win rate of the hero wasn’t so high.
Blur has been reworked. Probably, IceFrog intended it as a sort of personal smoke as PA can freely move in the area controlled by the enemy’s wards without being detected, which fits an assassin.
Passive evasion: 20%/30%/40%/50%
Vanish dispel radius: 600
Vanish dispel delay: 0.75
Vanish duration: 25
Mana cost: 20. Increased to 30 in 7.20c
Turns Phantom Assassin invisible. Does not dispel on attack or spell cast, and cannot be revealed with true sight. Only dispels when getting within range of heroes or buildings, with a small delay before getting dispelled.
The average player can use it to farm the ancients receiving no damage, move from a side to another of the map without being detected by wards, and push the lanes while remaining invisible.
When creating this ability, they made two huge balancing mistakes, which - in my opinion - show that the hero has not been tested properly as everyone (even lower-ranked pub players) noticed the potential of the skill after a few games.
You can activate the skill at any time, even when the enemies are inside the vanish dispel radius of 600.
There is a vanish dispel delay of 0.75 seconds.
The skill has been created as an utility tool, but in practice an experienced player is able to use it in team fight to disjoint spells - even when the enemy hero is next to you!
Valve gave PA an ability that makes her a sort of hybrid Riky, but you can detect the latter with the Dust of Appearance if he tries to escape, while PA’s invisibility is unbreakable, so it can’t be detected by any form of true sight. Void’s Chronosphere is the only skill in the entire game able to break this state.
To better understand the value of the actual Blur, we can compare it to Puck’s Phase Shift. Also the latter disjoints projectiles if used at the right time, but that’s all: you can’t move, you have no extra benefits, and you must try to escape using the Blink Dagger or the Illusory Orb once the effect ends. Of course, the comparison is not perfect because Phase Shift also hides Puck from the fight for 3.25 seconds, which has its own uses.
Phantom Assassin can disjoint projectiles, smoke at will, move while the effect is active, attack lane, neutrals, and ancient creeps without being detected, and has also a 50% passive evasion (upgradable to 75% with the level 25 talent). I believe the actual ability does too much because it can be used as an utility, defensive, and even offensive tool.
There are two easy solutions to the problem: Valve can either reduce the dispel delay, or prevent the player from using the ability when engaged in combat or if there are enemy heroes in the dispel radius. Both these nerfs would absolutely not hinder the hero, who would remain pretty strong.
The new match duration fits perfectly PA’s timings
Another factor contributing to the success of Phantom Assassin is the shorter match duration: she is very powerful during and after the mid game, but can’t win easily against the late game carries.
Considering the pro games played during 7.19, 37 minutes is equal to the 51.40 percentile of the matches, and 47 minutes to the 81.37.
In 7.20 so far, 33 minutes is equal to the 52.53 percentile, and 41 to the 83.41.
Without going into details, just based on the data above we can roughly tell that the average duration of the pro matches decreased by at least 5 minutes (and probably more).
I do not have proper data on pub games, but from both personal experience and the results shared on socials and forums, I believe the situation is pretty much the same.
The new armor formula boosts PA’s snowballing capabilities
Changed the Armor formula to: ( 0.052 * Armor ) / ( 0.9+ 0.048 * Armor)
Updated various armor values to closely align with the new formula:
Tier 1 Towers: 17 to 12
Tier 2 Towers: 19 to 14
Tier 3 Towers: 19 to 14
Tier 4 Towers: 29 to 21
Melee Barracks: 18 to 13
Ranged Barracks: 12 to 9
Ancient: 18 to 13
Fillers: 14 to 10
Shrines: 24 to 17
The following graph, shared on Twitter by PimpmuckL, shows the old and the new formula.
I am pointing out these changes because now having a lot more or a lot less armor has a bigger impact on the damage mitigated and received. As a critical-based hero, Phantom Assassin has always benefited from armor reduction, and in fact the Desolator remains her most bought core item, with an average purchase time of 20.04 in 7.20.
In this context, her level 15 talent -4 Armor Corruption has more value than in the past, and in fact Valve decided to adjust it to -3 Armor Corruption in 7.20c.
Even with the nerf to Cleave, the other level 15 talent +30% Cleave is still useful if you need to catch up with farm or you require the AoE damage (against illusion-based heroes), but the armor corruption is generally the best choice, especially if your team will fight a lot.
I have checked manually the games played at The Chongqing Major Qualifiers, and the great majority of the pro players selected the -3 Armor Corruption talent.
Stacking armor is stronger in this patch. -3 armor might not feel a lot, but it does stack with the 7 armor reduction provided by the Desolator, and with any other hero’s item and ability, such as Dazzle’s Bad Juju, which is one of the most popular supports of 7.20.
Interactions between Phantom Assassin and Grimstroke
Grimstroke is the most drafted hero in the game, and his abilities work well in conjunction with a carry such as PA.
He slows the enemies with Stroke of Fate, silences them with Phantom’s Embrace, and increases the movement speed of an allied hero with Ink Swell. PA is a melee auto-attacker, so she greatly benefits from all these disables and the movement speed bonus. She can can jump on the target with Blink Strike, attack it, and proc a max stun-duration Ink Swell.
Furthermore, Grimstroke’s Soulbind is very powerful if combined with PA’s Stifling Dagger and her level 20 talent Triple Strike Stifling Dagger.
Soulbind: “Any unit-targeted ability that either bound hero receives also gets cast on the other hero”.
I tested the interaction to find out how it works exactly. If you place two dummy targets next to each other, and then cast Soulbind and Stifling Dagger on them, they will both be attacked by a total of three daggers each.
The potential damage is huge. In the demo, I was hurting them for 186x3=558 each with no items. Add on top of this all the extra damage via items, any item effect, and your critical chance - the combo is very easy to reproduce and has a great killing potential.
Phantom Strike buffed
I will conclude this analysis on PA pointing out that also Phantom Strike has been buffed. This is the list of changes:
No longer has an attack limit of 4 attacks
Potentially, it opens also to item builds with more attack speed.
No longer gets dispelled when attacking a different unit than the cast target
It increases the versatility of the skill as you can jump on a creep to reduce the distance, and afterwards engage the enemy hero.
Rescaled attack speed bonus from 130 on each level to 100/125/150/175
Overall, it is a buff that boosts her late game. The reduced attack speed at level 1 and 2 won’t hinder her performance in the lane phase, especially considering Valve also reduced the cooldown of the spell at early levels.
Reduced duration from 4 to 2.5
Reduced cast time from 0.3 to 0.25
Reduced mana cost from 50 on each level to 35/40/45/50
Reduced cooldown from 14/11/8/5 to 11/9/7/5
A few hours before publishing this article, Valve released version 7.20d. I will have a quick look at PA’s changes.
Recipe cost increased from 220 to 240
20 gold more won’t change the popularity of this item and the triple Wraith Band strategy on many agility-based cores. A heavier nerf is needed, such as reducing the agility bonus from +8 to 5.
Phantom Strike attack speed duration reduced from 2.5 to 2.25
Blur now has a 0.4 second cast point
We talked a lot about Blur and the benefits it bring to PA during a team fight. With 7.20d, it lost a lot of defensive potential because she can’t vanish instantly anymore.
This nerf doesn’t resolve the problem completely, but it helps if you find PA out of position and disable her before she uses the skill.
Coup de Grace damage reduced from 230/340/450% to 210/330/450%
The damage of this passive in the mid and late game matters the most, and it is basically unchanged.
Overall, expect more nerfs :P.
I have published a series of articles and analyses on the 7.20 update. They cover most aspects of the game, general and specific game changes, new and old items.
This competitive tier list is the first article in which I talked about a hero. I decided to wait a few weeks to learn the pro meta before starting any hero guide or analysis. In December, I will update my meta series to 7.20.
|Keeper of the Light||40||17.1%||21||9.0%||19||8.1%||38.1%|
In tier 4 and 5 we find all the situational heroes. They are strong when paired with others, or against specific lineups. These heroes aren't bad, they just don't work in every game, so the pick and ban rates aren't outstanding.
|Queen of Pain||5||2.1%||4||1.7%||1||0.4%||25.0%|