27Template Usage Rate.png

Opening Thoughts

What’s up guys and gals, CaptainPlanet here to present the Overwatch Hero Tier List and Meta Report: World of Tanks Edition. Save us, Overawatch Open, you’re our only hope. The Meta is officially flooded with Tanks and barring the emergence of super secret strats at the Atlanta-based LAN it looks like we’re stuck with the NiP lineup for the forseeable future. As the Meta has evolved since Season 2’s release, so has the 3x3 strat – from a fringe lineup prior to Ana’s release to the many-headed juggernaut that it is today. This week, I will examine the history of the lineup, its current (and many!) forms, and what to watch for going into the weekend. Before we get to that though, let’s check out this week’s Tiers.

I’ve now begun to track all of my raw data in a much more database-friendly manner. You can now sort by any dates starting with Season Two, but keep posted for me eventually adding legacy data. This week’s data pre-sorted can be found HERE . The data is presented Row-wise for each Map Side. REMINDER: This Data was collected from the Lenovo Group Stages, and the Gosugamers EU and NA weekly tournaments that took place September 12-18th

27 new tier.png

THE TIERS

S Tier (>=95% Usage Rate*): No one!

A Tier (>80% Usage Rate): Lucio

B Tier (>50% Usage Rate): Zarya, Reinhardt, Ana

C Tier (>20% Usage Rate): Zenyatta, Roadhog, Winston, Tracer, Reaper, Genji, McCree

D Tier (>5% Usage Rate): Hanzo, Mercy

F Tier (<5 % Usage Rate): Mei, D.Va, Pharah, Torbjorn, Soldier 76, Junkrat, Widowmaker, Bastion, Symmetra

*What is usage rate? For every match, I record the time spent on each hero and divide it by that match’s total time duration. Each of these data points (a number from 0-1) are then summed across all sides of all matches, then divided by the total number of sides and converted to a percentage to produce a hero’s overall usage rate.

The tier ranges I’ve chosen** reflect different states of “balance” in the Meta.


S Tier “The Overpowered heroes”, and A Tier “The Core Heroes”

There’s something to be said about no one making the S Tier for the third week in a row, but for the second week in a row, Lucio was less than a percent from reaching the S Tier, so I’ll be addressing him here. Hidden in the hype and the hate of the 3x3 Meta, Lucio still reigns supreme. He fits all team compositions, all maps, and all strategies – he’s been Overwatch’s most consistent hero ever since I started writing these reports. Possibly, ever since he was released. Seriously, look at his usage:

2016-09-21_2103.png

At this point, I’m imagining a future for Overwatch where Lucio no longer exists – where the Overwatch devs decide to remove the hero entirely and bake his entire kit into the game as passive changes. Every hero would have their base movement speed increased, every hero would have a passive amount of healing, and his ultimate, wall-ride, and knock-back abilities would be introduced as abilities to future heroes. Would we really notice the difference? Sure, Lucio mains would be upset, but Blizzard would release a new short where Lucio sacrifices himself to save Tracer, D.Va, Mercy, Bastion’s bird friend, and the girl from the Soldier 76 short and everyone would feel sad, but accepting.

B Tier “The Favorites”

The B Tier better reads as “The heroes present in all Tank Meta lineups, regardless of map, map position, or game mode”. All variants of the NiP strat include a Reinhardt, an Ana, and a Zarya at minimum, so it’s not surprising to see them appear in the next tier below the ubiquitous Lucio. Reinhardt and Zarya have been in and out of the upper levels of usage of Overwatch, because whatever the Meta – you still need good tanks. Ana, however, more than doubled in usage from last week to this week, catapulting to a new high.

2016-09-21_2108.png

Ana’s ability to generate ultimate charge faster than any other hero makes her into the engine that makes the lineup go. Even if her own charge rate were reduced by a reasonable amount, this strategy could likely still function well simply on the back of the strength of tank ultimates and how well tanks synergize with Nanoboost in general. Don’t even get me started on the Biotic Grenade + Amp it Up! synergy…

C Tier “The Balanced Heroes”

Once again, the C Tier has sorted itself out into heroes that “fit” into the Tank Meta, Tracer, and heroes that “fit” into more traditional lineups. Depending on the map position, player style, and opponent lineup teams will swap between Roadhog and Winston freely – reducing their overall usage but landing them at similar numbers. It is worth noting that Winston tends to have much higher usage in king of the hill maps – so Roadhog by comparison fills in more often on payload/hybrid/assault maps. Zenyatta just barely missed the cutoff to the B Tier (< 3%), as he’s sometimes subbed out for a single dps in the form of Tracer, Hanzo, McCree, or Genji. The aforementioned dps heroes fill out the rest of this tier for their usage in traditional and occasionally Tank Meta lineups alike – although Tracer has been carried into the tier in part due to her high usage on king of the hill maps.

D Tier “The Meta Dependent heroes”

The D Tier was a short list this week – documenting the continued decline of Mercy and rise of Hanzo. Mercy is presently only being used by stubborn teams who refuse to conform to the Tank Meta, or who are still attempting Pharah-Mercy duos for the surprise factor. Unfortunately for Mercy’s (and Pharah’s) usage, these occurrences are becoming more and more rare. Hanzo has solidified a permanent seat in the D Tier because of his offensive strength on hybrid maps – subbing for Zenyatta to abuse the early pick potential before either side builds their ultimates.

F Tier “The ‘Even Season 2 cannot save you’ Heroes”

As the Meta crystallizes, so does the F Tier grow. Rather than harp on the Defense Heroes, I’d like to express how surprised I am about Mei and D.Va appearing in the F Tier this week. Mei is now a constant complaint not only among less-skilled players, but also in the Pro scene. Mei – and D.Va to a lesser extent – have emerged as beastly final point protectors. When played correctly, this duo can turn an already-lost team fight into a multiple-minute stall by staggering spawns. D.Va can rush to the point, absorb all projectiles with her Defense Matrix, fly in a spiral upwards to dodge more, and then dance around the point with her tiny, hard-to-hit body. As the attackers hunt down the D.Va, Mei can roll up to the point, find a corner, and wall herself off for 4.5 seconds, then pop her Ice Block for an additional 4 seconds – leaving only 1.5 seconds until the Wall comes back off cooldown and the process can be repeated. Combine the two, and you’re looking at a real headache when capping a contested point – and that’s before we even consider their gigantic area-denying ultimates.

This isn’t to say that this strategy will work every time, but we can expect every pro team to at least try it at Eleague. D.Va and Mei may be stuck in the F Tier this week, but I predict that their usage will spike much higher this weekend in Atlanta.

Once again, take the F Tier with a grain of salt…it only represents present usage at the top levels of Overwatch, it is not meant to tell you that your favorite Hero is garbage. (although, if you play Symmetra on offense I have no sympathy for you)

** I do not chose the placement of heroes in a Tier, only the Range which defines the Tier. By determining Usage Rate directly from hero Time Played in Tournament Matches, my data is Objectively determined, and not subjective at all.


Historical Tracking

2016-09-21_2130.png

For a more interactive chart (also less messy looking):

Click this link

Hero Picks and Success Rates, Present Week

2016-09-21_2131.png

For a more interactive chart (also less messy looking):

Click this link


World of Tanks

The Evolution of the NiP Strat

The NiP strat always existed on the fringes of viability, a strategy born out of necessity for an SG-1 team with extremely strong tank players. For example, Hymzi is known as the “OG Hog” – the master of the secret hooker arts before Harbleu or MoonMoon ever picked up the piggy. SG-1 leveraged their team’s strengths to create an unusual but not quite top-tier strategy – a strategy that was somewhat forgotten among the murk of 1 hero limit’s implementation, nerfs to Zenyatta, Widowmaker, and McCree, and the release of Ana. It was Ana’s release, however, that brought the tank-based lineup back to the forefront. The strategy was now much stronger: with Nanoboost a tank could do more damage than a DPS and tank more damage than a tank, effectively soloing as 2+ lineup slots for a limited amount of time. The Triple Support Lineup as it was called in the pre-season 2 days was even tried by NRG a few times, but they – along with SG-1 under their new moniker NiP - could not break through in a Meta dominated by Zenyatta’s 50% Discord Orbs. Ana and Nanoboosts’ strength could not be denied, but she was simply crowded out by a more overpowered mechanic.

Then came Season 2.

Season 2 nerfed Discord Orb, left Ana unchanged, and blew the door wide open for new strategies to be theorycrafted. Or in NiP’s case: revisited. NiP kept doing what they were doing, but started to steamroll teams in their scrims – and those teams began to take notice. In the first week of Season 2, only three teams were seen to adopt the “3x3” strategy for most of their matches: Reason Gaming, Luxurywatch, and NiP. Reason Gaming was crushed by a superior Rogue team, but Luxurywatch won their first Gosugamers NA weekly on the back of the 3x3 strat – and showed the world what was to come. Let’s take a look at the usage and success of the Tank Meta over time, as more teams began to adopt the Heroes associated with the lineup:

2016-09-21_2139.png

A quick cheat sheet for this graphic:

  • Larger squares = more absolute play time
  • More Blue = more chance to complete a map successfully
  • More Orange = less chance to complete a map successfully
  • Keep an eye on Ana – since she’s used almost exclusively in the strategy and no others and will serve as our indicator of the presence of tank-based lineups

  • you can play around with the non-image version and check absolute stats by mousing over the squares at this Direct Link

For this analysis I chose the eight heroes known to fit into the Tank Meta: Ana, Hanzo, Lucio, Reinhardt, Roadhog, Winston, Zarya, and Zenyatta. In the first week of Season 2 it’s easy to see that the strategy was niche at best – having only minor success on Dorado defense and Hollywood offense. This is likely due to Reason Gaming attempting to run the strategy against Rogue on King’s Row and Hollywood Offense, leading to the red Ana indicators demonstrating their failure to break the superior team. As we move into week 2, there’s a massive jump in usage for the heroes – the strategy begins to be seen on maps like Watchpoint: Gibraltar, Numbani, and Eichenwalde as well as Volskaya. We can also see indicators of where the strategy may actually have weaknesses – namely Hollywood offense and King’s Row defense. Hanzo begins to appear on offense as the strategy mutates – turning into a 4+2 lineup of Ana, Lucio, Reinhardt, Zarya, one of Winston/Roadhog, and a situational filler. Hanzo’s usage continues through week 3, as the 3x3/4+2 lineups start to define what is now the Tank Meta. Some clear trends appear to have emerged in week 3, but given the (somewhat lesser) strength of the teams these could simply be artifacts in the data. For example, it looks like the 3x3 strat was significantly unsuccessful on Hollywood offense, but a significant amount of those data points came from Oppadanawa running into Complexity – a much better team – and losing. Other trends that have emerged (but that you should take with a grain of salt include:

Bias Towards Success:

  • Hollywood defense
  • Numbani offense
  • Watchpoint: Gibraltar offense
  • Temple of Anubis offense

Bias Towards Failure:

  • Hollywood offense
  • Numbani defense
  • Volskaya defense
  • Temple of Anubis defense

I will be very interested to see how these trends shake out with data from “higher quality” sources – that is to say the teams who’ve qualified for the Overwatch Open.

 

Is the Tank Meta a bad thing?

“But Captain”, you might be thinking, “this is the most balanced we’ve ever seen the Meta! The supports (except Mercy) are all seeing decent levels of play, and we haven’t had any S Tier nerf-alert heroes since Season 2 came out! Why are the pros so upset?”

I’ll answer this question with a little story about closed beta. In closed beta, Blizzard was still experimenting with a great many things, one of which was seeking to answer the question of “how does ultimate charge get generated?”. The devs decided that heroes should generate ultimate charge by doing the sorts of things you’d expect them to: healers would generate it by healing, damage dealers would generate it by dealing damage, but this left tanks in a strange spot. Tanks could deal damage, sure, but their main role was not to be a damage dealer – it was to tank incoming damage. To remedy this, Blizzard decided to grant all heroes ultimate charge whenever they took damage. This helped tanks generate ultimate charge by doing what they were supposed to do, but it led to an interesting set of problems. For one, DPS players were dis-incentivized from dealing damage to their opponents, lest they feed their opponents’ much stronger tank ultimates. On the flip-side, tank players started to turn into suicidal maniacs – simply dying on Reinhardt was worth more than 25% ultimate charge given his health pool. This made for incredibly un-fun gameplay. Luckily, the devs were quick to change the system to ward off the unhealthy playstyle – removing ultimate charge gain from damage and adding a passive gain instead.

Fast forward a few months, to Ana’s release.

Ana can charge her ultimate much faster than most heroes, especially within a lineup surrounded by big beefy meat shields that can readily take damage without getting one-shot. Ana’s ultimate charges so fast that she flips the idea of ultimate batteries on their head, and has thrown the Meta back into the closed-beta, damage-charging-ultimate days. Roadhog used to be a liability because he’s an easy target to use to build damage-based ultimate charge. However, if there’s an Ana healing him suddenly the Roadhog becomes Ana’s battery. Apply this principle to the rest of the tanks in the 3x3 lineup, and suddenly the enemy team recoils from dealing any damage at all to your tanks, lest they bring your Ana to Nanoboost faster. The end result is that by releasing Ana with her present ultimate charge rate, the Overwatch devs have indirectly (and probably accidentally) made the same mistake they made in early closed beta – promoting a game state where players are afraid to actually play the game.


Deep Dives with the Captain

This is a new segment I’m going to start doing to help teach myself more about using Tableau, so I can make even better infographics and data-based analyses for you all. This week, I’ll be answering a question a reader proposed to me: how often do teams who take control of the capture point of a king of the hill map first go on to win the game?

2016-09-21_2201.png

Direct Link

While this is a fairly easy question to plot, working with only three week’s worth of data is a bit of a hindrance. As usual, the size of the indicators are weighted by total time played – then the first point capture rate runs along the x axis and the conversion rate of these first captures of the point into wins runs along the y axis. The conclusive answer to this question is that…..there is no conclusive answer. There are some strange outliers who populate the top of the graph with a 100% conversion rate of first captures into wins – but these are likely due to the limited dataset. Luminosity stands out as an outlier of outliers: their total time played is much higher than most teams and they sport a 100% winrate when they capture the point first, but they’ve only done so 5% of the time. In fact, this inspired me to dig directly down into my raw data: Luminosity has only won the point first once since Season 2 began, but that led to a win of the map. Maybe they should try it more often!

Reunited’s big(ish) circle provides our only other significant datapoint, and it tells the story of their continued dominance on king of the hill. Reunited has long been masters of this map type, and it looks like at least some of their success can be attributed to their taking the point first. Reunited captures first 64% of the time, yet converts >75% of these captures into wins. If I were gameplanning a king of the hill match against Reunited, I’d definitely try my best to take the point before they could set up to defend – the numbers don’t lie. This graphic will become a lot more telling as it fills up with data, so I’ll be sure to re-vist it in the future!


Final thoughts and shoutouts

Shoutout Cloud9 for letting me creep on their team practice this week after a squirrel bit through the power at Logitech HQ and they had to stop at my local gamer bar. As it turns out, Surefour is actually an anime character in real life, Reaver can crush my hand with his beefy forearms, and Grego was very nice and shared his salsa with me. Also shoutout to all of the Overwatch fanart that’s coming out of the woodwork lately! I’m running out of thumbnails to use for my meta reports, so if you would like your Overwatch fanart featured in future report thumbnails hit me up! I am more than happy and willing to provide Twitter shoutouts and links to your deviantart/pixiv/instagram/[name your platform]. You can send submissions to ben@overbuff.com, and please make sure that they are >1920x1080 resolution. Finally, Overbuff is hiring! If you want to work with me and the Overbuff content team, head on over to our job posting blog for more information.

Until next time,

CaptainPlanet