28Template Usage Rate.png

Opening Thoughts

What’s up guys and gals, CaptainPlanet here to present the Overwatch Hero Tier List and Meta Report: Overwatch Open Deep Dives. I’ve missed you all! Between this report and last, I took a two week break to devote an entire piece to Overwatch’s largest tournament to date: Eleague’s Overwatch Open. Running over several months, this open tournament concluded in Atlanta on Friday and was broadcast live on television in the US, crowning Misfits as champions – and awarding them $100,000 USD of the $300,000 prize pool.

This report marks a great opportunity to what happens when the best of the best compete, especially because the teams present in the dataset have been noticeably absent from the past few reports due to their clandestine bootcamping in preparation for the tourney. Now that the Overwatch Open has concluded, we can shine some light on the “secret strats” that these teams have been cooking up behind the scenes and see where the pro meta currently stands. Using what ended up being a much larger dataset than usual, I can also perform some deep dives on trends that you, my readers, have suggested. As always, before we get to all that let’s take a look at the hero tiers from the largest Overwatch tournament to date.

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 Alienware Monthly Melee finals, and the Gosugamers EU and NA weekly tournaments that took place September 2-4th

28 new tier.png

THE TIERS

S Tier (>=95% Usage Rate*): Lucio

A Tier (>80% Usage Rate): Zarya

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

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

D Tier (>5% Usage Rate): Zenyatta

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

*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, A Tier, and B Tier: The “Beyblade Lineup” of the Overwatch Open

If someone were to ask me “Hey Captain, what was the ‘Meta’ lineup this week coming out of the Overwatch Open?” I would point them to these five heroes. These heroes make up the core of the current meta: Zarya and Reinhardt on tank, Lucio and Ana on support, and Deathblossom Reaper as DPS. Lucio and Zarya need little explanation – they have been perennial favorites regardless of meta due to Lucio’s speed boost and Zarya’s all-around solid kit. Ana, Reinhardt, and Reaper on the other hand were the defining element of the many strategies showcased throughout the tournament. Misfits in particular picked this core composition and practiced the hell out of it, a 5+1 lineup with a floating DPS for specific areas of specific maps.

Think of this lineup like some sort of sci-fi energy cannon. The Ana is the battery that powers the cannon: her ultimate charges extremely quickly in the opening fights and reaches full charge before any other hero. The Reinhardt is the primer: he receives the first Nanoboost and charges forward – cleaving a gigantic “V” of space for the rest of his team to move into and claim kills from advantageous positioning. After the cannon has been primed, it’s time to fire. The Reinhardt’s initial Nanoboost adventure helps to charge the Reaper’s Death Blossom and the Ana’s second Nanoboost back to full, and like a Beyblade* it’s time to “Let ‘er Rip!!”. Or more appropriately “Let ‘er Reap!”. The Reaper uses his teleport ability to get in behind or above the enemy, then drops in amongst them as Ana Nanoboosts his Death Blossom – releasing the charged cannon’s energy and securing the easy team wipe.

* thanks to Surefour for aptly naming the term

C Tier “The Balanced Heroes”

Then, there’s the rest. The heroes of the C Tier represent those heroes chosen to fill the +1 of the Beyblade lineup – every single of of these heroes saw usage alongside the five that make up the core. Tracer and Winston climbed their way to the top of the C Tier with 46% and 40% usage respectively because of their positive skew towards King of the Hill maps – which saw Tracer at 89% usage and Winston at 76% usage. Misfits’ Soon put the exclamation point on Tracer’s usage this week by securing their finals win with a tournament-defining Sticky Bomb to close out their final match.

Moving through the rest of the tier, Mei has now been tapped as the premier “Anti-Tank” hero and still remains an absolute pest on defending points – as I’m sure players of all skill levels can attest. Clever use of her ultimate, Ice Block, and Ice Wall can cause huge amounts of disruption both to lines of sight for Anas and to already-boosted Reapers and Reinhardts. Genji, McCree, and Roadhog fill in as the +1 to the 5+1 – with Roadhog functioning more as an assassin than a tank. Taimou even played Roadhog instead of his main McCree on King of the Hill – his hooks being the only thing standing between EnVyUs and Soon’s Tracer soloing their entire team*.

* slight exaggeration

D Tier “The Meta Dependent Heroes”

The D Tier is a strange tier this week, housing only Zenyatta. Zenyatta’s decline in usage has mirrored the increase in Ana’s since Season 2’s release as more and more pros find Zenyatta and the 3x3 support/tank lineups to be less optimal than the 5+1 Beyblade lineup. Zenyatta’s Discord Orb may have received the nerf it deserved, but it seems that the low output of Orb of Harmony’s healing and the strength of Ana’s ultimate charge rate leave the robot monk high and dry in a game that seems to gravitate to a maximum of two supports in its meta lineups.

F Tier “The Not-Tournament-Ready Heroes”

Behind the scenes, the teams competing in the Overwatch Open played against each other in scrims, optimized their lineups, and decided on strategies that they felt would give them the best chance at winning the grand prize. Due to all of this preparation, it’s not surprising to see a significant amount of heroes drop to negligible amounts of usage throughout the tournament. The Pros had time to determine that Torbjorn, Widowmaker, Junkrat and the rest of this cohort were just not going to win them any matches, and the chaff was removed.

But that’s not the whole story. As many of my readers are very quick to point out, if a Hero falls into the F Tier it does not necessarily mean that the Hero did not have impactful usage in specific scenarios, or specific map areas. Later in this report, I will examine the F Tier heroes and the situation that made them worth picking. Torbjorn, Bastion, and Widowmaker though – these are the true losers of the Overwatch Open. Torbjorn in particular had 0 picks the entire tournament, absolutely incredible considering the >7,300 hero picks and swaps that occurred during the week. Hard work….did not pay off :(

Once again, take the F Tier with a grain of salt…it only represent 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.


Overwatch Open Deep Dives

There are so many things I want to talk about regarding the meta and the Overwatch Open itself, and not enough time to make them all flow and make sense in a narrative way. Think of what follows as their own individual articles, and let’s GOOOOO!


Misfits and the 5+1 Beyblade Lineup: Counter-able, or not?

Leading into the Overwatch Open, we knew pretty much what to expect: lots of Ana, lots of Tanks, and lots of Nanoboosts. What came as a surprise was that the Meta had mutated once again: behind the scenes as the top teams prepped for the tourney. Reaper had been discovered, Mei had been mastered, and what arose was the 5+1 Beyblade Meta we see today. Many teams employed this lineup, but none so willingly devoted entire maps to the core Heroes than Misfits, the eventual champions. Usually, when a lineup becomes so ubiquitous and static, it becomes prey to counters – this is what I would call a dynamic meta. Think Rock, Paper, Scissors: Misfits became extremely good at rock while the rest of the teams were playing mostly rock, but sometimes paper and sometimes scissors depending on the map. The best map that demonstrates this phenomenon is Eichenwalde – Overwatch’s newest map and one that saw an uptick in play at the Open. Eichenwalde has a wide variety of map geometries for teams to take advantage of, but Misfits chose to stick with their core 5+1 in all of their matches – while other teams swapped to and from a rotating cast of different Heroes depending on the progression of the payload:

Misfits

2016-10-02_1904.png

All Teams

2016-10-02_1905.png

This begs the question: if Misfits was constantly playing rock – why didn’t teams play more paper? The answer lies in what paper actually represents, in terms of an Overwatch lineup that can counter the Beyblade composition. The kind of lineup that should do well against a core including a Reaper as its main damage dealer and Zarya as one of its main tanks would include a Pharah who can damage them from areas and distances that they cannot reach. Then, if you have a Pharah that also means you have a Mercy. Yes – the Pharah-Mercy combo has potential to do well against the Beyblade comp. However, there’s still the +1 of the 5+1 to consider….

EnVyUs actually tried this exact swap against Misfits in the finals, and EnVyUs also happens to feature arguably the best Pharah player in the world (Talespin) and one of the best Mercys (Chipshajen). As luck would have it though, Nevix (the player commonly playing the +1 Hero for Misfits) had just died on his Mei to secure a team wipe of EnVyUs and was smartly sitting in spawn waiting for them to choose their heroes. As soon as he saw Talespin and Chips swap to Pharah-Mercy, he swapped to McCree – countering two heroes simply by waiting a couple extra seconds. Misfits easily dealt with the new threat – especially since EnVyUs had to sacrifice significant ultimate charge to do so – and won the map.

Does this mean the lineup is immune to counters? I am not quite convinced. It is possible EnVyUs could have had better results if they had rolled out with Pharah-Mercy from the beginning instead of starting at a disadvantage in ultimate charge. It is also possible that there is a better shell of a lineup out there with which to surround the Pharah and the Mercy as well. Then again, Misfits could still make the same adjustments to deal with the combo, and Pharah-Mercy is very map-position dependent regardless of the current Meta. Only time – and the upcoming reduction in Ana’s ultimate charge rate – will tell if the Beyblade Meta is here to stay.


Will the nerf to Ana have any effect on the Meta?

On that note – will that nerf even do anything? Consider that most teams are choosing – outside of the initial usage on a Reinhardt – to use their Nanoboosts to combo with Reaper’s Death Blossom. If Ana players are forced to wait on Reaper’s ultimate to use their Nanoboost, is reducing her charge rate going to accomplish anything at all at the Pro level? The Reaper’s ultimate charge is the limiting factor in Nanoboost’s usage now, in that context Ana has received no change to her ultimate and a buff to her Biotic Grenade to boot!


Answering Questions

This week, I posted to Reddit asking for my readers to supply me with some questions to answer regarding the Overwatch Open, the Meta, and trends that they wanted to see explored this week. I too had questions I wanted to answer! What follows will be my attempts to answer these questions using some very basic data analysis, as well as a bit of speculation and context to explain the results.


What About F Tier Heroes?

Every week I get multiple comments about how F Tier Heroes aren’t necessarily bad. And they’re not! For further reassurance, I have plotted every single instance of “F Tier” Hero Usage including how long they were played, how much of the match share that represents, and who played them in one convenient chart. Unfortunately for Torbjorn, he was picked exactly 0 times so he doesn’t get a data point :(

F Tier Hero Usage

Looking at this chart, it does not take a lot of time to immediately toss out a couple heroes as beyond “niche” in their usage. Bastion, Widowmaker, Soldier 76, and Symmetra rarely saw significant usage even within their own game – not enough to really make an impact. Others, like Junkrat, were used for entire games or map sections, but usually only under extreme circumstances (like in the third or fourth side of a payload map, or second half of a 2CP map).

Heroes that saw less extremely low usage, but fell just short of escaping the F Tier because of their niche status include Hanzo, D.Va, Mercy, and Pharah. D.Va stands out because she appears in a lot of matches, but not for a significant amounts of time – pointing to her status as a last-point stall mechanism. Hanzo is niche both in map usage and player usage, he was played primarily by Buds, Zappis, and Tviq and even then primarily on Hollywood and King’s Row offense. Pharah was map-dependent as well, mainly used on Watchpoint: Gibraltar and Dorado offense, with sprinklings throughout Eichenwalde as well. Mercy had a similar spread as to be expected, but saw a boost from Rawkus in particular for his usage of the hero on Eichenwalde offense. Keep this in mind when you roll into ranked yourselves – you may want to think twice about shunning the player who’s playing Pharah on Eichenwalde, or Hanzo on Hollywood. Maybe they know something…


I was confused about how quickly teams dropped Zenyatta, and what the deal is with Mercy. Is she just bad?

I somewhat answered the question regarding Zenyatta above – Zenyatta has fallen out of favor in optimized Ana-based lineups because his main debuff was (fairly) nerfed, and his healing output cannot compete with Ana’s. However, I figured I would flesh out the rise and fall of heroes over time, starting with the first week of the qualifying rounds of the Overwatch open:

Heroes: Rising

Heroes: Falling

As a reminder, you can mouse over a hero to make their trend stand out.

We can see that over time the more recent “Risers” follow their leader, Ana, in their recent spike in usage. It is worth noting that Mei’s boost is due to her role as a counter to the tanks that are popular pairs with Ana, rather than as a boost to the Ana lineup itself. This is not to say that Mei does not see usage in the 5+1 strat, as Misfits’ Nevix played a great deal of Mei throughout the tournament as their +1 player.

Heroes that are on the decline can trace their fall either to direct pressures or indirect pressures forcing them out of their lineup slots. Zenyatta competes directly with Ana for his role as “the healer that is not Lucio”, while McCree, Roadhog, and Genji have been indirectly crowded out by the fact that only one of them can pair up with Reaper in the +1 slot, and that they also have to fight it out with Mei and each other for limited usage time.

Back to the Mercy question: as we can see right after Season 2’s release Mercy saw a spike in usage – probably to test out the re-balancing to her ultimate charge rate. However, the Pro community quickly determined that she could not even compete with Zenyatta, much less Lucio and Ana for the 2-3 support slots available. Mercy, unfortunately for her fans, has been out of the meta for quite some time now. Again, it would be misleading to say that she does not see usage, however, so please see the above question regarding niche hero usage from this tournament.


What’s the story with Mei? Is she better on certain maps?

I was curious about this question, and while answering it I noticed some other heroes had some interesting overall trends as well. I have chosen to spotlight these Heroes in the chart below:

Hero Successes and Failures: Hero Spotlight

In this chart, more blue = more likely a hero results in a win, more red = more likely a hero results in a loss, and grey is 5050. Surprisingly, Mei was the only hero out of 22 that was decidedly more blue than red – meaning that teams that picked or swapped to Mei generally won map sides more often than not. In terms of answering the posed question, Mei was very good both on Hollywood (offense and defense) as well as Eichenwalde defense. If we click into each of these data points, we can see that Kyky and Nevix are primarily to blame for Mei’s success on Hollywood

Focused Search: Hollywood

If we perform a similar search, we can find that a multitude of players make up the successes on Eichenwalde.

Focused Search: Eichenwalde

Both Hollywood and Eichenwalde have cramped close quarter final points which may explain Mei’s dominance on these maps, which seems to imply that King’s Row and Dorado would also see success with Mei defenses. As I will discuss in the next section, there is a reason behind this.

Moving on to the other heroes – Genji, McCree, Reaper, Tracer, and Winston all saw some significant negative trends on certain maps. Luckily, we can discount Tracer almost immediately: Tracer tends to be picked as a last resort similarly to D.Va by teams that are going to lose a map side regardless, in order to stall for precious seconds in the time bank.

Reaper stood out to me because he had poor success only in a specific area: King of the Hill maps. If we click into these data points, we can see that most of these losses could be attributed to teams that did not make the finals – that is to say teams not named EnVyUs or Misfits. Perhaps these teams were trying to punch above their weight by relying on their individual fragging skills on king of the hill, but ended up losing to the eventual group winners anyway. These teams – for whatever reason – tended to use Reaper in their lineups – and perhaps this was a mistake.

Genji, McCree, and Winson were the most interesting of the bunch: apparently they were poor picks on almost every map. This makes sense, however, since these heroes are used in “out of meta” compositions which tended to lose to compositions similar to Misfits’, who ended up winning the whole thing.


Are there any egregious imbalances in map win/loss rates? A lot of people seem to think Eichenwalde is hard to attack; do the statistics reflect this, or are pro players seeing something that “normals” aren’t?

The answer to this question is yes! And I’ve created a chart to demonstrate it (see the pattern yet?):

Hero Successes and Failures: Map Spotlight

Several maps had a noticeable bias towards offense or defense, and I have listed the ones that stood out to me here. In no particular order:

  • Hanamura is biased against attackers
  • Temple of Anubis is biased against defenders
  • Volskaya Industries is biased against defenders
  • Dorado is biased against defenders
  • Watchpoint: Gibraltar is slightly biased against defenders
  • Eichenwalde is biased against attackers (but not towards defenders in the data, I’ll explain why)
  • King’s row is biased against defenders
  • Numbani is biased against defenders

Regarding Eichenwalde: this map was and is biased so much against attackers that teams tended to complete the map during overtime – if they completed it at all. The way that I track “successes” and “failures” is dependent on map completion, which means that when time bank comes into play, teams are much more likely to go 5050 on defense, since the map rarely can be completed in that amount of time.

Also, addressing why Mei does not fair as well on King’s Row and Dorado defense as Eichenwalde and Hollywood: these maps are skewed overall against defenders in the first place which seems to affect Mei as well.


What were the key differences between the EU and the NA Meta?

Again, here’s a graphic!

Hero Successes and Failures: EU vs. NA

The EU vs NA meta was less about heroes, and more about maps. NA teams were much more likely to draft and ban into king of the hill maps compared to their EU counterparts, picking Reaper on these maps. Unfortunately, this Reaper lean was often to their detriment as previously discussed. NA teams both did not pick as much Genji as EU, and tended to have less success on him anyway. Perhaps EU’s Genji players are stronger as a whole – players like Kyb, Shadowburn, Linkzr, Tseini, and Tviq come to mind. Building on that vein, EU’s Reinhardts tended to have more success on both sides of Eichenwalde – maybe there’s a home-field advantage coming into play for this Germany-based map. One map where NA both picked more often and excelled at was Numbani offense: it looks like many NA teams must have bootcamped specific strategies for attack on this map, while the map was rarely played at all on the EU side.


The rest of the questions are somewhat open-ended, and I have created a series of charts for you all to explore on your own. Let’s check them out!


How often does a Hero Swap lead to a successful offense or defense?

Swap Success Rate: Offense and Defense

This chart requires a little bit of explanation to understand. On the chart, I display Swap Location, Hero Swapped To, Hero Swapped From, and multi-colored, multi-sized dots. As you mouseover the dots, a highlighter appears to better match the heroes swapped to and from. The color of the dots refers to how often the swap led to a successful map finish, again more blue = more successful and more red = less successful. The size of the dots correlates to the number of matches this particular swap occurred, so you can judge the significance of the data on the fly. Finally, if you click on any of the data points, you can see what player(s) contributed to it in the list below. For example, I have clicked on the giant Reaper->Genji dot on Eichenwalde Offense:

Focused Search: Eichenwalde Offense

We can see the three players that made that swap in in the tournament: Surefour, Kyb, and Shadowburn. Of the three, only Shadoburn’s swap lead to a win. I encourage you to click around and see what interesting data you can find!

There are some caveats to this chart, namely that with only a single tournament’s worth of data the color of the dots ends up simply being blue or red – 100% success or 100% failure associated with the swap. As the dataset grows, this chart should become more and more interesting….and significant.


How often does winning the point first on King of the Hill lead to winning the map?

A friend of mine asked me this question a few weeks ago, and I have made another chart to answer this question for the Overwatch Open:

King of the Hill First Point Capture Bias

The X axis of this chart lists the “capture point first” rate, then the Y axis lists how often the team then goes on to win the whole map. The size of the dots indicates how many matches were played, for some semblance of the significance of the results. I found it interesting that FaZe won the point first roughly 85% of the time, then went on to win 85% of those captures. By comparison, EnVyUs only won 55% of points first, but converted on 70% of the points they did capture first – so perhaps they may want to make a stronger effort to secure the point in the future.


What Map Locations were most likely to lead to lineup swaps? What swaps occurred at these locations?

Overwatch Swaps and Locations, What Do They Know? Do They Know Things?? Let’s Find Out!

In case you have not had your fill of swaps analysis, I made another chart! This one ranks the total amount of swaps that occurred during the Overwatch Open, broken down by rough map area in which they occurred. If you click on any of the bars, the window to the right will display each individual hero swap that occurred at that location. I encourage you to explore!


Final thoughts and shoutouts

So many shoutouts this week. Shout out to all the players and all the organizations supporting them who got to compete in the Overwatch Open. Shoutout to the talent who kept us on the edge of our seats throughout the tournament with their casting, hosting, and analyzing. Shoutout to Turner, Eleague, and Faceit for putting on the tournament. Shoutout to ChanmanV and the High Noon Podcast for having me on their respective podcasts this week. Shoutout to caffeine, for being my crutch for recording all of this data. And shoutout to the winners of the tournament, Misfits!!

Until next time,

CaptainPlanet