Aug 20, 2018 Llarion Insights

Ladies and Gentlemen, welcome to my first post on the HeroesLounge blog.  I’ll be diving deep into the stats of the Division 5 games played up to  week 3, bringing you a snapshot of the battlegrounds, hero bans, hero  picks, compositions and player awards, all based on the games you’ve played so far and the numbers that were achieved in them.


Battlegrounds

First up, we’ll be looking at the most popular map choices for our teams from Division 5. Unsurprisingly, Infernal Shrines sits at the top with a total of 18 selections so far, closely followed by our favourite spider extermination grounds of Tomb with 16 picks. Volskaya Foundry comes in third with 12 picks, and after that it’s a fairly wide split of all other maps with 6 to 9 picks.

The shortest game so far was concluded in 10:39 minutes between Sköna Lirare and CEP, the longest slugfest award goes to Klondike Express and CoB Antimeta for a game on Tomb that dragged on for 33:13 mins.

One interesting outlier in terms of stats is that on Infernal Shrines, the first-pick team has a win rate of 66.7%, making it the most successful map for first-pick teams. On the opposite end of the spectrum, first-pick advantage has led to success in only 29% of cases on Dragon Shire, however that number is to be taken with a grain of salt, as there have only been 7 overall games played on that map.


Hero bans

Moving on to the draft meta, we’ll be looking into hero bans first of all. This is where I expect Division 5 to deviate a bit from other divisions, as personal comfort picks might have a higher overall value than the strictly meta heroes, but let’s see.

Leading the charge of the ban brigade is everyone’s favourite space marshal Jimmy Raynor, who clocks in at 47 times on the sidelines. Following him is, largely because weeks one through three were still played before her balance patch, our resident unfun-to-play-against gnome dragon Chromie with 46 bans. After that we have a bit of a drop-off in numbers, with Garrosh, Genji, and Yrel coming in at 37, 31, and 29 bans, respectively. On the support side, Stukov and Deckard follow with 22 and 21.


Hero picks

On the other side of the Petman-faced coin, the most-picked hero in the first three weeks of Division 5 play was, leading by a fairly large margin, Johanna with 45 overall picks, followed by fellow main tank Muradin with 38. Those two were the overwhelming most popular choice at main tank, with a runner-up Diablo only coming in at 24 picks. However, Johanna’s win rate sits at only 38%, begging the question whether she was really the most fitting pick for the situation. In terms of tank win rate ETC is actually the overwhelming leader of the charts with a staggering 85% of his 13 games won.


Following the main tanks in absolute numbers are, unsurprisingly, the other category of heroes that is the least populated in the game, support. Rehgar wins out over Deckard by a slim margin, 37 to 36 times chosen overall, while Malfurion still takes a respectable third place at 29 picks. The wolf shaman is haunted by the same problems as our most-picked tank though and sits at an equally bad win rate of 38%, while Deckard and Malfurion are doing much better at 58% and 48%, respectively. The lesser played supports Ana, Brightwing, Lucio, and Auriel find themselves at the top of the win rate charts, once again begging the question whether meta is always better. However, none of them have a significantly high enough number of games to make any definitive answer to that, we’ll have to wait and see how things develop throughout the season.


Jaina the Frost Mage seems to be by far the most popular assassin so far, with the likes of Fenix, Hanzo, Maiev or Genji nowhere to be found. Jaina leads her fellow mages Li-Ming and Kael’thas with 35 compared to 25 and 24 picks, respectively. She is also the most-picked hero early in the draft, with 20 of her 35 selections coming in the first round of draft. Surprisingly, compared to Johanna and Rehgar, this tendency for early picks has not lead to a decline in her win rate, where she sits at 57%.

The most-picked auto-attackers then finally come in at 23 picks, and they, quite surprisingly, feature everyone’s favourite Gilnean protector and part-time doggo, Greymane, joined by Jimmy and Fenix.

The assassin with the highest win rate over a meaningful number of games is Gul’dan, who ended up taking 71% of his 24 games.

That leaves us with solo-laners, where Sonya takes the most-picked position with 28 overall selections, followed by Blaze with 24 and Yrel with 23 picks. Sonya also leads solo laners in win rate, where she sits at a comfortable 64% of her games won. Leoric and the reworked Artanis follow closely behind, both with also more than 60% of their games ending with the victory screen.

As a bonus, here's the bubble diagram for specialists, as my analysis software can't differentiate between main tanks and solo laners, so they are contained within the tanks and assassin charts. Overall, these types of charts are unfortunately only useful at identifying outliers, apologies for the mess that happens in the overpopulated middle regions.

If you were expecting Li Li, Zagara, Butcher and win-blades Varian, I’ll unfortunately have to disappoint, the overall Division 5 meta sees relatively few eccentric hero picks. The most noteworthy thing may be that the Shimada brothers failed to show up as much as they do at higher levels of play.


Unpicked heroes

But which heroes were completely neglected by our Div 5 teams? Overall, 71 out of the 80 available heroes were played in the first three weeks of play, Whitemane obviously not being released in that timeframe yet and still banned for Lounge play anyway. The nine unplayed heroes were Chen, Illidan, Murky, Nova, Misha with her pet ogre Rexxar, Samuro, the Vikings, Tyrande, and Xul.


Compositions


So now that we’ve established the hero picks, let’s look at the combination of them and what proved the most successful so far. In terms of compositions, the vast majority of games was completed using the “standard” double tank, double assassin support comp, which was picked 65% of the time and achieved a win rate of 57%. Single frontline triple assassin comps with heroes like Fenix taking the solo lane were picked in 10% of games, but fell short in terms of win rate significantly, only managing to take 18% of their games. Replacing one of the assassins with a specialist in either of these two compositions were the next popular alternative at 6% pick rate each, both times coming close to an even win rate at 45%.

My most interesting comp award goes to team Southgate Boiz for their quadruple tank plus support comp of Muradin, Arthas, Artanis, Sonya, and Alexstrasza, that netted them the win in their week 1 match on Infernal Shrines.


Player awards

Finally, let’s look at individual performances across all roles and try really badly at making out an MVP for each role simply based on average stats. It can’t possibly be worse than the ingame MVP system, can it? Can it? Either way, only take this too seriously at your own risk!

For the tank role, I went with damage taken per death and total stun time. Jaldren of Earthworm Defence Force takes that first category despite not even playing main tank (remember not to focus the Dehaka guys!) and Uroroboros of Duet R is our stun king thanks to his Anub’arak play.

Because I’m a firm believer in not dying over doing maximum damage, I have selected total damage done per death for our damage dealer category as well. Here, Leverous of FAT Goblin leads MarineLorD from team PLagueMakers by a fairly slim margin.

In terms of overall healing done, swatata from FAT Goblin leads the field playing exclusively Alexstrasza and Rehgar.

Overall, the most deadly players are strangemas from FAT Bugbear and once again MarineLorD with 6.4 kills per game on average, the highest KDA was achieved by Numani of EDF on support.

And as a completely useless bonus stat, the most sprays and Bsteps award goes to FiL of Southgate Boiz, coming in at over 580 sprays used across his 6 games. That has to either be a world record or a bug with my parsing software.


So there you have it, the meta snapshot as of week 3 for Division 5. Look out for more to come as the season progresses. If anyone is interested in data behind this post, contact me or check out the condensed version spreadsheet for some raw numbers.

Special thanks to ebshimizu and his amazing HotS replay analysis and database software Stats of the Storm, without which this analysis would not have been possible. If you have any questions, comments or criticism, you can reach me on Discord as Llarion#0127!

Previous

#RIGGED - your week 5 preview

Aug 14, 2018 #Rigged
A look at the halfway point highlights of the season!

Next

Outside Lounge; Inside Open Division

Aug 19, 2018 Amateur League EU Inside Lounge Interview
An interview with several of the Heroes Lounge teams that participate in Blizzard's Open Division!