No matches

All about stats is a monthly column that discusses trends in Dota 2 based on recent patches.

It has nearly been a month and a half since the very awaited 7.20 patch was released for Dota 2. With a lot of map changes, a few new items and massive reworks to hero abilities, 7.20 changed the game in more ways than one. A lot of old heroes were forgotten, and the new guard replaced them. There haven’t been a lot of tournaments played on the new patch, but a few events like the ‘MegaFon Winter Clash’ and ‘I Can’t Believe It’s Not Summit’ along with the qualifiers for the Chongqing Major and the Bucharest Minor did take place on the 7.20 patch. So there is a good bit of data (350 professional games to be precise) to judge how the new patch has been going.


Most Picked in 7.20

7.20 definitely has been a breath of fresh air with regards to in meta heroes. If the 15 top picked heroes at TI8 are compared with the 15 top picked heroes in the Dota 2 7.20 patch, there isn’t a single common one! That is quite the change right there. Heroes like Vengeful Spirit, Tiny, Wraith King, Weaver, Necrophos and Clinkz which were main stay at The International 2018 are nowhere to be seen. Most of them have, if not all, have been nerfed a good deal to make sure they don’t raise their heads and get back into the popularity charts anytime soon.

Rubick leads the new charts with his improved Arcane Supremacy and negative status resistance Fade Bolt. The hero is being played as a mid laner as well now, a lot more regularly than before. Not that I think it’s a coincidence, but it’s a good time for Rubick to be popular with the new arcana coming out. However, the Grand Magus’ win rate is not that high and could do with a bit of raising up. Perhaps teams are still adjusting to the new mid Rubick. Sand King and Lich find themselves at the top as well, with their ability reworks. Sand King was the most popular hero of the 2017-2018 DPC season with over 1000 games. SK was picked throughout the year and was nerfed throughout the year as well, until it got to a point where the hero was just not good enough. But with the improved Sand Storm, SK has become a formidable foe in the laning stage with an invisibility that cannot be broken with a stun easily, as the opponents don’t have a clue where Sand King is within the Sand Storm. Lich might have a few mana problems with Sacrifice being sacrificed, but Frost Shield and Sinister Gaze have made a huge impact on games.  Grimstroke is in the mix as well, with the hero just being introduced to Captain’s Mode.

A hero that I am really glad to see in the most played list is Phantom Assassin. PA has always been popular in the pub scene, but never really so in the professional scene. IceFrog tried and failed at improving her in such a way that the hero would not be overpowered in pubs but would be a viable option in pro games at the same time. But I think he struck gold with 7.20! Phantom Assassin is either being picked or banned left, right and center and for a good reason as well. She has a win rate of nearly 57%, which is the second highest in the top 15 heroes. The active ability added to Blur has propelled her up the popularity charts as not only serve a way to farm faster without being damaged, it is also a safe way to know the whereabouts of enemies.

From 6.74 to 7.19: PA contest 1139 professional games out of 20848 (5.46%)
in 7.20: PA contested in 277 professional games out of 350 (77.94%)

The highest win rate goes to Magnus, who is reaching for a win rate of 69%! That is some crazy stat. Magnus was always a good offlane hero with the occasional mid Magnus showing up here and there. But with 7.20, even though all cleave damage became physical, the cleave distance and final width for all cleave items, abilities and talents got increased.

Since then, Magnus has become a lot more popular as a position 4 hero. His main job is to keep Empowering the position 1 hero, who is either a Phantom Assassin or Anti-Mage (or the likes) and try to hit the occasional Reverse Polarity, or RP as it’s more commonly known. Magnus and PA together have won 11 games out of 13 (84.62% win rate) while the combination of Magnus and AM have won 9 out of 12 games (75% win rate). It is like having a 4400 gold item (Battlefury) on your carry without having to spend the gold. It accelerates the farm exponentially and the opposition carry just cannot keep up. The most recent patch, 7.20e saw a sizable nerf to Empower with every level losing 10% of the bonus damage. However, the cleave is still intact and the upcoming tournaments will pass the verdict on whether or not Magnus is still considered a strong hero (I believe he is and will be contested quite a bit).

Most picked heroes in the 7.20 patch (stats taken from


A hero that does not make the top 15 but still has a staggering win rate is Anti-Mage. Anti-Mage has been picked in 49 games and has managed to win his team 34 of those (69.39% win rate)! For a hero that was totally ignored in 7.19 (picked in 19 games out of 912), this seems to be a good comeback patch. With the active Spell Shield, it is like having and Aghanim’s Scepter from the onset. Add to that Magnus’ Empower, that’s two items with a total cost of 8600 gold that Anti-Mage can have with abilities. With no Scepter required now, it also frees up a slot for AM which is very beneficial for the fast farming carry. The Chongqing Major is the first big event of 7.20 where all the big teams will be fighting it out on LAN on the new patch. That is where we will get to see AM’s true strength and potential with his new gifts.

Two heroes that are surprisingly in the top 15 (who I was hoping wouldn’t be there) are Centaur Warrunner and Terrorblade. Towards the end of 7.19 (7.19d), Centaur and TB were two of the most common picks and got contested in nearly every game. IceFrog revamped Centaur in 7.20, but the new Centaur Warruner, as it appears, is quite powerful as well. Terrorblade hardly got changed in 7.20, which comes as a big surprise seeing as to everyone in the community was fed up of seeing the demon. But somehow, he escaped the hell that was the huge nerf hammer and continues to be picked quite regularly post the release of the 7.20 patch. Even though their win rates are not too impressive, I hope to see some more nerfs come along the way of these two heroes to keep them away for a few months at least. After all, they have had their time in the limelight.


Most Contested

Coming to the most contested picks, it’s no surprise that Io is in the mix. The hero is banned in most games, irrespective of the patch. Grimstroke and Magnus deserve a special prize for pipping Io in the race for the most contested hero. But the stats on the first eight heroes are staggering.  A contention percentage of more than 70% for all eight of those means at least for now, the meta is converging on those few heroes. Rubick and Dazzle have found more appeal in the mid position to add to their support picks. Teams like J.Storm also use Dazzle as a position 3 hero, so the versatility is endless.  Lone Druid is another hero who sticks out in this list. LD isn’t included in the most played heroes as the hero gets banned out in a lot of games. With the range build coming back in business, LD has become a popular hero in both, the mid lane and the safe lane. A lot of these stats are from before the nerfs from 7.20e, so there is a possibility they might skew out of favor for some of these heroes.

Most contested heroes in the 7.20 patch (stats taken from


The Banished

The list below shows the heroes that were picked less than 5 games since 7.20 was introduced to the world. There are a lot of expected ones like Clinkz and Wraith King, who got nerfed after TI8. But quite a few surprises in there as well.  Bristleback, Chaos Knight, Night Stalker, Sven and Troll are all heroes who got decent buffs in 7.20 but have not been favored in the professional meta at all. Sven might have been buffed a bit in the follow-up patches, but the removal of Warcry’s armor has really crippled him in being a safe laner. A support Sven may be more common now, but that’s just a little sad. From one of the hardest carries in the game to a potential support is not really a promotion of any kind. I really hope that Sven is given back his old Warcry and is bolstered in other departments to make him viable as a carry again. Troll Warlord being under picked to this extent was not expected at all, but I guess even though Berserker’s Rage is a lot stronger now, the high cooldown makes it less lucrative. Also, if carries like CK and Troll are to be picked more frequently, there is a need to nerf PA and AM to an extent and nerf TB a lot. Terrorblade has been a part of the meta for a long time (except the month of TI8) and it is high time the demon is sent to hell. Abaddon, Underlord and Viper are a few of the others who haven’t been seen for a while and could use some IceFrog loving. The meta is still young and for all you know, in this long break we have right now, teams might figure out a bit more about what hero is better than he/she appears. The next big patch will possibly come after the Chongqing Major and will possibly depend on the picks in that tournament.

Heroes picked in less than 5 games since the release of the Dota 2 7.20 patch (stats taken from


Time Distribution of Games

Comparison of game distribution based on time for 7.20 and 7.19

A lot of hero picks and strategies depend on how long the games usually go. 7.20 made the game faster because of a lot of attack speed based buffs and talents. There were predictions from analysts that the game might become faster. and that is exactly what has happened. Comparing the average game distribution based on time for 7.20 to that of 7.19, the percentage of games that go past 40 minutes has gone down drastically (all three slots) while the percentage of games shorter than 30 minutes has gone up. A higher percentage of games now end up in the 20 to 30 minute slot than before. Dota 2 has become a lot more fast-paced as compared to 7.19 and it is clearly showing. The one slot that has nearly remained unchanged is 30 to 40 minutes. But either sides of that slot, there has been a complete reversal.


Usage of New Items

Three new large items (more than 2000 gold) were introduced to the game when Dota 2 7.20 was released. Those three were: ‘Holy Locket’, ‘Kaya and Sange’ and ‘Yasha and Kaya’. But how are the new items faring? Have they gained popularity? Let’s take a look at what the data says (taken from regarding the number of times each item was picked up from the 350 professional games played.

Holy Locket: 15
Kaya and Sange: 8
Yasha and Kaya: 55

Holy Locket and Kaya and Sange have an atrocious record of being picked up. Both have been picked up in less than 5% games. These two items clearly need some improvement so as to make players go for them. Yasha and Kaya, on the other hand, has done pretty well for itself. One of the major reasons for that is Outworld Devourer. Yasha and Kaya suits OD perfectly and is usually made as a first item on the hero.



The meta is still fresh and a lot needs to be deciphered, which is no surprise as it is just a month and a half old. Games have become faster with a propensity to finish between 20 to 40 minutes (68% of the games). Attention has been focused on a few specific heroes while others who have been buffed a lot have been completely ignored. The new items need some buffs to make them seem not just good, but worthy enough to replace a current item in an orthodox build. Teams and players are yet to totally understand the usefulness of a lot of the reworked skills, but once they do, it promises to be a whole new ball game. That is the reason why the Bucharest Minor and the Chongqing Major have the potential to be interesting tournaments. January 2019 will a month filled with a lot of entertainment for the Dota 2 community!

Share on FacebookShare on Twitter