A Post-6.78 Hero Draft Analysis

Posted by Huy "Chronosphere" Nguyen at 30 June 2013 21:45

Which heroes are trending up? Which are fading into forgotten land? GosuGamers Crew takes a look at how the current metagame is reflected through the drafting phase in the past two months of competitive DotA 2.


With major changes that came with the release of the 6.78b patch, we decide to use June 4th - the patch's release date, as the cut-off point for our statistical analysis. Using data meticulously compiled by our friends at datDotA.com, we take an in-depth look at how team captains fielded their lineups, before and after 6.78. 



1. The Regulars



Top 10 picked/banned


Pre-6.78 (475 games) Post-6.78 (497 games)
Heroes Picked/Banned Picked/Banned (%) Heroes Picked/Banned Picked/Banned (%)
 Batrider 474 99.8  Batrider 495 99.6
 Lifestealer 473 99.6  Dark Seer 464 93.4
 Lone Druid 428 90.1  Gyrocopter 445 89.5
 Shadow Demon 426 89.7 Io 431 86.7
 Nyx Assassin 418 88.0  Lifestealer 409 82.3
 Gyrocopter 414 87.2  Shadow Demon 355 71.4
 Dark Seer 410 84.2 Prophet 353 71.0
 Magnus 400 84.2  Nyx Assassin 337 67.8
Io 379 79.8  Lone Druid 321 64.6


376 79.2  Visage 317 63.8


Even with some slight nerfs (decreased Firefly flying vision and increased manacost for Flamebreak), Batrider remains a no-brainer ban/pick in virtually every single game. The rest of the Top 10 saw some shakeups following the installment of the patch. Lifestealer, used to be an omnipotent and versatile carry, lost a lot of its lane control ability with a signifcant range nerf to Open Wounds (600 -> 200/300/400/500). Now forced to sacrifice early skill points in Rage in favor of maxing out Open Wounds, Lifestealer dropped from being a consensus no-brainer ban/pick before the patch to being selected in 82.2% of all the drafts. Still a strong hero, but no longer dominant. The same can be said for Lone Druid, with its popularity taken a significant hit from the nerfs (no Armlet, 300 gold bounty on Spirit Bear), dropping from 90.1% to 64.6%. Meanwhile, Magnus has completely fallen out of the Top 10 (more on that later).

On the other side of the spectrum, Dark Seer made its return as a top-tiered selection following the patch, either banned or picked in 93.4% of games. That mark is second only to Batrider. Visage stormed into the Top 10 despite receiving no buffs, perhaps due to a combination of an increase in Chinese DotA being played this June as well as the nerfs to other popular support heroes.


2. The Trends  

The Hot


Heroes Pre-patch Picked/Banned (%) Post-patch Picked/Banned (%) Differential
 Treant Protector 11.2 53.4 +42.2
 Naga Siren 2.7 31.3 +28.6
 Visage 38.1 64.0 +25.9
 Outworld Devourer         7.6 28.6 +21.0
 Dragon Knight 27.8 43.6 +15.8
Prophet 58.5 71.0 +12.5
 Kunkka 19.6 30.0 +10.4
 Spectre  8.4 17.5 +9.1
 Lich 0.2 8.7 +8.5
 Rubick 63.0 70.7 +7.7


It took teams quite a while to take notice of Treant Protector's potential. Once they realized how powerful this hero can be, its popularity has risen exponentially. The patch itself did give Treant a slight buff (Leech Seed's slow 24%->28%), but it was clearly not the reason why this hero has been picked so often recently. Quite frankly, Living Armor is arguably the most powerful early game spell in the game right now. It can almost certainly guarantee mid lane advantage now that bottle crowing is no longer a viable option. It also shuts down early aggression from the enemy's team. Tower-healing is just the icing on the cake. Just like Keeper of the Light was a year ago, Treant Protector went from a forgotten hero to a top-tier pick overnight despite receiving no significant buffs; it's just a domino effect. Once teams began to realize the hero's potential, they brought it to the metagame for good. 

Also of note is the emergence of new mid-lane options. As previously mentioned, the bottle-crow nerf has led to teams going for heroes with good rune-control abilities or are not bottle-dependant as their mid-laners. OD has always been known for being a great solo hero. With good Astral Imprisonment timing, OD can completely zone out the opposition and dominate the lane. It's a great counter to popular solo-mid Intelligence heroes such as Queen of Pain or Puck. In regards to the melee heroes, Kunkka can always push the lane with his cleave and control runes. Naga can do the same with Rip Tide and can even use illusions to scout runes. Dragon Knight can spam Breath of Fire to push lanes and also has built-in regeneration.

Mid lane is still not quite back to the old-fashioned full-on aggression way it used to be played. That said, this bottle-crow nerf is definitely bringing some fresh new faces and different approaches to the lane. It's also safe to say that the boring days of Magnus vs. Puck both bottle-crowing and each minding their own business in mid are behind us. 



The Not


Heroes Pre-patch Picked/Banned (%) Post-patch Picked/Banned (%) Differential
 Magnus 84.3 24.3 -60.0
 Keeper of the Light 79.2 52.3 -26.9
 Leshrac 52.0 25.6 -26.4
 Lone Druid 90.1 64.6 -25.5
 Juggernaut 50.3 28.0 -22.3
 Nyx Assassin 88.0 67.8 -20.2
 Lifestealer 99.6 82.3 -17.3
 Anti Mage 38.1 22.5 -15.6
 Lina 31.4 19.9 -12.5
 Clockwerk  68.0 55.7 -12.3


Poor Magnus... This new patch has taken him from a top tier pick to merely a situational option. 50 damage from Reverse Polarity at level 1 means that there's no threat of him solo-killing any hero mid. But the main problem is, again, the bottle-crow nerf. Without constant regen to spam Shockwave and get farm, there are just so many different ways Magnus can be shut down in the mid lane. The hero itself is still strong, but it's more of a hit-or-miss with an over-reliance on RP to win teamfights. Frankly, there are safer and better options to send mid than taking the risk with Magnus. 

Lone Druid also saw a steady decline in popularity following the patch. No Armlet on Spirit Bear means that if Lone Druid is sent to the off-lane and gets zoned out early, it will be extremely difficult for him to farm his way back. Also, 300 is a lot of gold to give away, meaning Lone Druid players have to be extra careful on the Spirit Bear.  

Past top-tier picks such as Nyx and Keeper of the Light, while still seen regularly in drafts, are no longer first-pick first-ban heroes, mainly because teams are so used to playing against these two that they have found ways to effectively neutralize them.        



3. Effectiveness


Post-6.78 Top 10 Highest Winrate (Min. 30 games picked)


Heroes Games Wins Losses Winrate (%)
 Disruptor 45 31 14 68.9
 Treant Protector 133 83 50 62.4
 Lich 39 24 15 61.5
Io 70 43 27 61.4
 Nyx Assassin 163 97 66 59.5
 Puck 149 88 61 59.1
 Chen 109 63 46 57.8
 Mirana 35 20 15 57.1
 Naga Siren 90 50 40 55.6
 Weaver 122 67 55 54.9


Interesting to see Disruptor topping the list with an impressive 68.9% winrate, although it might have just been a fluke considering he's only been played 45 times in a total of 497 games recorded. On the other hand, the recent surge in popularity of Treant Protector is justified by his 62.4% winrate. Lich is an interesting case; he's been on the rise lately and rightfully so. An offlane Lich with Sacrifice spam can be absolutely devastating to the opposing carry hero; and 10 Chain Frost bounces can deal a ton of damage with a proper setup. Mirana is still not picked frequently enough to make anything out of her high winrate. 

On the other hand, it's not really surprising to see the regulars Io, Nyx, and Puck and Chen atop the winrate list. Of course, just because a hero's winrate is high doesn't mean it can win an entire game on its own. Other factors such as team composition, skill level are also involved. Nevertheless, having one of these heroes in the lineup certainly won't hurt any team's chances. 

Post-6.78 Top 10 Lowest Winrate (Min. 30 games picked)


Heroes Games Wins Losses Winrate (%)
 Lina 83 30 53 36.1
 Luna 48 19 29 39.6
 Leshrac 95 38 57 40.0
 Clockwerk  172 71 101 41.3
 Chaos Knight 41 17 24 41.4
 Slark 42 18 24 42.9
 Juggernaut 73 32 41 43.8
 Storm Spirit 57 25 32 43.9
 Doom 34 15 19 44.1
 Magnus 54 24 30 44.4


Usually when a team picks Lina, it either means that they want to be aggressive early on or they just run out of supports to pick. Lina's strength is that all her nukes, including her ultimate, deal decent amount of damage and have short cooldowns. But she needs a setup to be effective in the early game and doesn't offer much the longer the game goes. She's very effective in small ganks when she can instantly bursts down a target but fades away when teams start to group up. With the current metagame favoring 5-man DotA, Lina is just not really a good option as a support, evidenced by an underwhelming 36.1% winrate.  

Other than Lina, all the other heroes on this list have their winrates hovering around the 40-45% mark, which obviously is not very good, but not oh-my-god-these-heroes-suck either. Again, just because a hero's winrate is low doesn't mean that picking that hero will compromise your team's chances. There are many other factors contributing to a loss. Teams still pick these heroes and win with them.   



4. General Thoughts


The new patch has not even been in effect for a full month, so obviously we're working with a relatively small sample size in this analysis. Teams are still tinkering with new lineups and strategies. That said, numbers never lie; we are definitely seeing quite a few notable changes in the metagame, especially in terms of hero choices. The new drafting system also results in a larger hero pool; and that's always a good thing for us DotA fans. Will we see another huge evolution of the metagame in the near future? Only time will tell. 

We would like to say a special thanks to DatDota for helping us in the retrieval of information.