6.79 - 6.80: Evolution of the Mid Lane
Having previously written articles discussing data on support trends and carry trends early into 6.80, and having recently published a piece summarily discussing the changing popularity of heroes in 6.80, it occurred to me that I've not written anything completely focused on the mid lane in this patch. For this post, I chose to do something different to my usual analysis, though. After several people providing feedback (Thanks!) that drawing data about the competitive metagame from all official matches might actually be problematic, given the large range of tournaments that comprise the data parsed on datdota.com, I decided that I'd instead focus on data specific to some of the leading mid players. What follows is an analysis of the top 5 most played heroes of 6.79 compared to the top 5 most played heroes of 6.80 of Dendi, S4, Resolut1on, Ferrari, Mushi, Super, and Arteezy.
Starting with Dendi, the most striking thing for me is that he has played substantially less Storm Spirit in 6.80 than he did in 6.79. Storm Spirit was his most played, and most successful, hero in 6.79. The truly bizarre thing here is that Storm Spirit, overall, has been picked a lot more in 6.80 than it was in 6.79, having an overall pick-or-ban percentage of 37% in 6.79 which is now sitting at an improved 51%. Perhaps we can understand Dendi getting less Storm Spirit in 6.80 based on the fact that his win rate on the hero is a lot lower so far in those games he has had in 6.80. But why the change? Surely he didn't just forget how to play the hero overnight? My instinct is to say that in 6.79 Na'Vi normally picked Storm Spirit together with Lifestealer - possibly one of the main reasons for why they won so much with the hero. In 6.80, Puppey has picked XBOCT Lifestealer only 8 times so far, compared to the 22 Lifestealers he played over the course of 6.79. This theory is corroborated by the reduction in the number of Puck's Dendi has been playing recently, given that Puck is yet another hero that Na'Vi historically picked a lot to combine with Lifestealer.
Most of the other changes on Dendi's top 5 list are not too surprising. Invoker is, by all reports, the best hero in Dota at the moment and thus we should expect to see all mid players playing more of it than they did before. Meanwhile, Shadowfiend and Ember Spirit are both flavour of the month, especially given the increasing effect that EG's Arteezy has been having on the metagame. After yet another nerf in 6.80, OD has fallen largely out of favour. Lastly, Batrider did initially fall out of favour in 6.80, but has recently returned as a major pick. It is probably just a matter of time before we see Dendi playing more Batrider again.
While 6.80 has had Dendi playing less of his signature Storm, s4 is yet to have a single official match with Magnus in this version. That said, s4's number of Storm Spirit games has also decreased. I was happy to notice that my above hypothesis about Dendi's decrease in Storm Spirit appearances can also explain s4's similar decrease. Indeed, Loda has only played 10 Lifestealers in 6,80 so far, as opposed to his total of 25 in the previous patch. This also explains the reduction in appearances of s4's Puck. Elsewhere, Dragon Knight, one of s4's main heroes in 6.79, has seen an overall decrease in popularity worldwide, moving from a pick-or-ban percentage of 34% to this version's meagre 20%.
Notably, s4 has pounced on the opportunity to play Batrider again with the return of this hero to popularity. This is unsurprising given his preference for the hero in the previous version. For me, the most surprising thing about s4's data is his 100% win rate on Invoker in the current version. s4 is not exactly hailed for his Invoker - and in a context where Invoker players are the talk of the town, it is odd that the one who seems to always win with it is not talked about more often in the Invoker conversation. I know I will personally play closer attention to this particular Invoker player in the future.
Resolut1on's Empire have been the talk of the town since the start of the year, starting the year on better form than any other 'Western' team, and later maintaining an incredible run of wins until very recently. During the early parts of the year, Resolut1on would almost never be seen playing anything but Invoker or Storm Spirit. In fact, despite not yet being at the Superstar status of Dendi or s4, it is probably fair to say that Resolut1on's mid-laning served as a powerful force in shaping the current metagame when it comes to mid heroes. All mid players play a lot of Invoker at the moment, and most of them play quite a lot of Storm too. Interestingly enough, Silent has played roughly the same amount of Lifestealer in 6.80 as he did in 6.79 (12 -> 10). However, this is not to say that Empire are picking the hero a lot more than other teams at the moment. Rather, what this suggests is that Resolut1on's Storm might be a bit less dependant on Naix bombs than some others are in order to succeed.
Recently, Resolut1on has also started showing a lot of flair as Mirana in the mid-lane, a role that previously only C9's SingSing had had consistent success on. As someone who did not pay that much attention to this player before this year, I would love to see how his Nightstalker would fare in the current metagame, given his supposed preference for this hero in the past. Obviously, the 6.79 data here provides a substantially smaller sample for this player, and deals almost exclusively with him playing in a different team - so comparatively speaking, the data is a lot less valuable than is ideal, if useful at all.
Moving over to the Eastern Scene, earlier sentiments about Dragon Knight are echoed. This hero was a top pick for most mid players in the East before 6.80. Now, it's become something of a fringe pick. In Ferrari's case, the same could be said for Kunkka too. For me, the most notable thing about Ferrari's data, though, is just how little of it it there is for 6.80. This is a familiar concern when discussing the Eastern, and especially Chinese, Dota scene. If it wasn't already obvious from general observation, the data makes it pretty clear that the top Chinese teams are being quite starved for practice at the moment, relative to the Western teams. This also makes it more difficult to draw anything meaningful from this data.
That said, one thing definitely stands out - Ferrari is the only tier 1 Chinese mid player to have quickly picked up Ember Spirit as one of his core heroes in this patch. This is especially significant given that Ember Spirit was not even in Captains Mode for the entire 6.80 patch. While iG do still seem to be a bit behind DK in the East, it was this sort of picking-ahead-of-the-curve which proved to be an early sign of the 2012 iG's rise to power, ultimately ending in their victory at TI2.
However, DK's Mushi has also started to play some new heroes in 6.80 ahead of any other Chinese teams, particularly Zeus and Centaur Warrrunner. Zeus has not been picked much by any mid players worldwide while Centaur has been picked almost exclusively as an offlaner. For a very good advert for Mushi's Centaur, check out this VOD of DK vs Newbee game 1 during Starladder playoffs, a game in which he showed that his mid lane Centaur could laugh off a super-farmed enemy Spectre in the lategame. (Edit: This is a mistake - I'm blaming lack of sleep - the games where Mushi has played Centaur so far, he's been on the offlane. I still highly recommend the VOD I linked though.)
One obvious talking point when it comes to Mushi's hero pool is the lack of Invoker on his list. However, followers of the scene will recall that Mushi simply does not play the hero at all. Of course, it's interesting that a team can retain such a high level of dominance when it's mid player does not ever play what is supposed to be the best mid hero. While some might criticize Mushi for this, I think it speaks to his credit that he manages to have the same amount of impact as other players, without using what most would say is simply a 'stronger hero'.
Rounding off the Eastern Scene, I chose to look at Super, given VG's incredible success in 6.79 which has been followed by a lot less consistency in 6.80. As with the above two players, the sample is really small and thus there's really not that much to say here. DK, Pugna, and Kunkka have fallen off quite a lot as competitive picks and thus Super, like everyone else discussed above, has had to adapt and begin to focus on different heroes. This, with the single exception of Alchemist. In some ways, I think of Super as the quintessential Alchemist player these days. Following several nerfs, the hero has seen far less play as a support or a carry. While the nerfs hardly impacted the hero as a mid-laner, it seems that most teams have still opted to pick it less even for this lane. Not VG. Super's Fast-Mek Alchemist continues to have huge success into 6.80, showing that not everything has to change with a new patch.
I decided to discuss Arteezy last because I wanted to do something slightly different here. Firstly, Arteezy was not very active in 6.79. Secondly, where he was active, the data includes matches in which he was playing for different teams to the one he is now in. But most importantly, if Resolut1on influenced the mid-lane metagame early in 2014, it's Arteezy who is making the most waves as we move into April. I've included more than just his top 5 most played heroes for 6.80, just to emphasize how different his style of mid is to everyone else's. Even when he subbed for Speed Gaming, Arteezy specialized in heavy-farming mid laners like OD and TA. You might say he's just stuck to his guns with his new team - only now he is allowed more freedom to pursue his own philosophy. Arteezy's mid heroes will always farm hard, often more than Fear (or Mason) does, and will very often go for Hand of Midas. A few days ago, Arteezy defended his philosophy in his 20-20 for Teamliquid:
"In my opinion, it is just the better way to play the game. Having a carry like Naga Siren or Morphling middle ensures that they will go that extra mile to put you down (ex: placing wards in your medium camps/constantly pressuring via support rotation/constant support rune control) which evidently allows for other lanes to farm more and later on make up for the amount of space you provided for them. Heroes like MOrph and Naga are very easy to come back with as well due to their almost-invulnerability status (Song and morphings' entire skill set) and AoE nukes. I am still capable of playing utility mids that aren't centered on farming but I will only play them when I feel like it is the correct game."
It's very difficult to look at the mid lane in isolation. While most mid players will play a lot of whatever the most popular mid is at the given time, even here there are exceptions - e.g. Arteezy and Mushi not playing Invoker. Meanwhile, some mid selections will depend a lot on what synergy they can have with a given strategy. Super's mid Alchemist might work a lot better for VG than it would for any other team, given the pushing mentality that is so often behind VG's strategies. Meanwhile, teams who pick certain mid heroes to combo with heroes like Lifestealer will pick those heroes less often where they are picking Lifestealer less often. Lastly, the case of Arteezy shows that even fundamental underlying philosophies of mid players can differ. This highlights something truly amazing about Dota 2 as a game - in a very meaningful way, players get to decide how the game should be played - not always the other way around, as might be naturally expected.