Overwatch Tier list and Hero Meta Report #17: the esl experiment

Click to enlarge!

Opening Thoughts

What’s up guys, CaptainPlanet here to present the Overwatch Hero Tier List and Meta Report #17: The ESL Experiment. For the unacquainted, ESL recently announced a $100,000 prize pool tournament whose group stages just kicked off this week. Despite providing Overwatch's largest purse to date, ESL's Atlantic Showdown was not without controversy. For the past several weeks, the Tournament Meta has revolved around the 1 Hero Limit, 2CP Map Ban Format -- a format proposed by players and implemented and popularized by GosuGamers. Bucking the trend, ESL decided to not limit their Heroes, and to utilize a rotating Map Pool which included 2CP Maps. The motivation behind ESL's decision follows Blizzard’s own announcement and release of Competitive Mode; ESL wishes to provide as similar an experience as possible to the Competitive Ladder you or I could play today. After some extensive Twittering voicing their disgust, Pros realized ESL was not going to budge, and that they would have to start practicing this format to have a shot at the big prize. 

This left this weekend's GosuGamers Tournament Admins in an awkward position. Gosu was the first Tournament Organization to experiment with and eventually popularize the 1 Hero Limit, 2CP Map Ban Format, but continuing to stick to these rules would annoy the Pros at best, and divide the community at worst. Ultimately, Gosu bit the bullet -- becoming ESL and Blizzard's "guinea pig" by removing Hero Limits, opening up the Map Pool, and extending King of the Hill Maps to Best of 5. In doing so they’ve provided the Competitive Community (and more importantly, Blizzard) with hard, objective evidence of what exactly to expect under the "Approved" Ruleset. Let’s just say, it’s not pretty. 


Click to Enlarge!

As always, Raw Data for this and subsequent charts can be found HERE. REMINDER: This Data was collected from the GosuGamers Weekly NA and EU Tourneys, where 0 Hero Limit, Stopwatch Format was used for Payload Maps

The Tiers

S Tier (>=100% Usage Rate*): Tracer, Lucio

A Tier (>80% Usage Rate): Reaper

B Tier (>40% Usage Rate): Winston, Mercy, Reinhardt, Zarya, Pharah, Soldier 76, Roadhog

C Tier (>20% Usage Rate): No one!

D Tier (>5% Usage Rate): Genji, D.Va, Widowmaker, McCree, Torbjorn, Hanzo, Junkrat, Symmetra

F Tier (<5 % Usage Rate): Mei, Bastion, Zenyatta

*What is Usage Rate? It is NOT the percent chance of seeing a particular Hero in a Match, it’s actually a little different than that! Please see this glossary page explaining the term. 

The Tier Ranges I’ve chosen** reflect different states of “Balance” in the Meta. This week, the Meta that is being analyzed is the Meta of 0 Hero Limit, Bo5 King of the Hill Maps, and Stopwatch Scoring for successful attacks on Payload Maps.

The S Tier: The Heroes who are being stacked.  The S Tier is back, with a vengeance. Lucio and Tracer owe their placement in this Tier to their absurd Usage Rates on King of the Hill Maps, where on average you could expect two Tracers on both Teams at all times, and nearly the same amount for Lucio. Lucio's Speed Boost and AOE-based healing are essential on these Maps where short run distances from Spawn and close-quarters fighting are the name of the game. Tracer's ability to contest the point in Overtime and quickly return following an unlucky death makes her absolutely essential as well. 

The A Tier: The “Non-Negotiable” Core Heroes. The A Tier (as well as the rest of the Tiers) are somewhat messed up this week because of the influence of 0 Hero Limit and King of the Hill Maps on the Meta. Due to a lot of factors that I'll be going into more detail later in the Report, King of the Hill Maps and the Heroes used on them account for all four of the top four placing Heroes, as well as five of the top seven places. This week, Reaper just happened to fall into the A Tier due to his high Usage Rate on King of the Hill as a close-range Tank Shredder, destroying the double-Zarya and double-Winston lineups which plagued this Game Mode. 

The B Tier: The “Balanced” Heroes. This week, these are the "Everything Else" Heroes. These Heroes are your Payload and 2CP Core Heroes, interspersed with Tanks who were also highly used on King of the Hill Maps like Zarya and Winston. For the most part, Heroes in this Tier were unaffected in their non-King of the Hill Usage, and only Soldier 76 saw significant losses in Overall Usage due to losing spots to stacked Reapers and Tracers. 

The C Tier: The “Meta-Dependent” Heroes. This Tier is loudly absent of Heroes this week. I'm honestly not sure why no Heroes fell into the 20-40% Usage Rate Range even considering King of the Hill and 0 Hero Limit severely warping the Meta. I'm going to chalk this up to just a strange anomaly. 

The D Tier: The “Specialist, or Weak” Heroes. This Tier's description still describes its denizens well. Many of the Heroes in this Tier excel in their particular niche: Symmetra, Junkrat, and Torbjorn on Defense, Genji and D.Va as Offensive Flankers come to mind. Others were the recipients of recent nerfs or have been generally underpowered since Closed Beta: your McCrees, Widowmakers, and Hanzos.

The F Tier: The “We Exist, Barely” Heroes. Oddly enough, Zenyatta had more picks than Mei and Bastion this week! Yay? It seems like regardless of the Format, these three F Tier Heroes are doomed without Blizzardly intervention.

** 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 Picks in Tournament Matches, my data is Objectively determined, and not subjective at all. 

The esl experiment

Hero Stacking is a Problem

If you looked at the Usage Rate Infographic at the top of the page and only just picked up your jaw from the floor, here’s some more interactive data for you to play with. With the removal of the 1 Hero Limit Hero Stacking returned to the Meta en force -- leading to the ridiculous results you see above. Let me just highlight a few:

  • Tracer had 201% Usage Rate on King of the Hill. This means that as an across all King of the Hill Maps played, both Teams had two Tracers on their lineup on average at all times. This means there were at least four Tracers per King of the Hill Map, which is patently ridiculous. 
  • Lucio had 186% Usage Rate on King of the Hill, which basically allows us to conclude roughly the same thing as Tracer. Unacceptable. 
  • Winston had 96% Usage Rate on King of the Hill, but this statistic is actually a little misleading. Whenever Winston was used on King of the Hill Maps, he was primarily used in pairs -- so a near 100% Usage Rate translates to four Winstons on every other Map. Zarya and Reaper starred opposite of Winston, being used in the other half of King of the Hill Maps or as pairs along with the Monkey Menace. 

Starting to see a pattern? 

King of the Hill is a Bigger problem

Click to Enlarge!

Play around with the chart above, and you will notice that King of the Hill Maps are the only Map types where Hero Stacking (i.e. >100% Usage Rate) is prevalent. The design of these Maps unfortunately promotes this tactic: King of the Hill Maps have short runs from Spawn, closed-in areas where most of the fighting happens, and a Meter that counts up whether or not your team is actually on the point. This naturally favors Heroes who have game-changing Ultimates (Zarya, Winston when stacked), fast movement speed (Tracers), or up-close Brawlers (Reaper). 

Here's how most, if not all King of the Hill Rounds proceeded without Hero Limits. First, both teams would start the game as some form of 2/2/2 comp (2 Tracers, 2 Lucios, and 2 of Zaryas or Winstons). Then, one Team would take control of the point initially by capitalizing off a big Tracer pick-off, a Reaper rampage, or a lucky Lucio boop. The other Team would then attempt to either re-take the point once they built their Zarya / Winston Ultimates (or fail to) -- and the clock would count down to Overtime. Whichever Team not currently in control of the objective would gradually morph into a 5 Tracer, 1 Lucio composition in an attempt to stall out the game, abusing the Overtime mechanics with the hopes of landing that one lucky Pulse Bomb that would allow them to flip the control. This abuse of Overtime mechanics led to extremely drawn-out matches, in fact so ridiculously stalled out that Gosu’s Tournament Admins considered pulling the plug on one particularly egregious Reunited stall. 

The worst part about this phenomenon was that each “Map” played was only one of the 3 to 5 required to win the “Match”. Consider the data below to the right ->

For some reason, Blizzard suddenly decided that King of the Hill Maps were to be Best of 5, instead of Best of 3 when they released Overwatch's Ranked Play Mode. ESL came along and copied this practice, and so Gosu was forced to followed suit. This was...not a good idea by any of the parties involved. My Meta Report’s Usage Rate calculations are based upon “Sides Played” among other things, so each week the Maps have the most Sides Played end up dominating the overall Meta. Payload Maps have a maximum of 5 sides (generally two Rounds of two Sides, with the possibility of one Overtime Side). By contrast, Bo3 King of the Hill Maps had a maximum of 6 sides -- which was generally in line with their Payload counterparts. Unfortunately, by increasing the required Maps to win a King of the Hill Match to Bo5, Blizzard caused this number to jump to 6 minimum, 10 maximum sides to determine a single King of the Hill Match -- and the results are staggering. 

When you also factor in the removal of the 2CP Map Ban -- allowing Pros to dump their Bans on Volskaya/Anubis/Hanamura -- King of the Hill maps ended up being selected for play more often in the Banning process as well. The combination of more King of the Hill Maps being played for more Sides resulted in King of the Hill accounting for 57% of all of this week’s Meta Data. Given King of the Hill’s current position as “Place from which all Evil Emanates”, I may have to completely split my Meta Report in two and begin performing Reports that quarantine King of the Hill Data from the rest of the game.





In fact, I’ve already done so:

Click to Enlarge!

Seems much Healthier, doesn't it?

Overtime is a sneaky problem

It’s always important to keep asking “Why?” to better understand a problem on your way to forming a solution. Here's where we've made it to so far:

Why is the Meta so warped? There’s lot of Hero Stacking going on.

Why are Heroes getting stacked so much? Pros are playing lots of King of the Hill Maps.

Why are Pros playing lots of King of the Hill Maps? King of the Hill Maps are now Best of 5, not Best of 3.

Why are Pros Stacking Heroes on King of the Hill Maps in particular? Certain Heroes excel on these Maps because their strengths are amplified by the Map's Design, but these Heroes are also used to cheese the Overtime mechanic in Overwatch.

The Heroes who get stacked the most -- both in Payload Maps as well as as King of the Hill -- are those that can abuse the Overtime Mechanic in Overwatch. In all of Overwatch's modes, simply touching the point during Overtime resets the 10 second counter entirely. Luckily, on Payload Maps one has to maintain contact with the Payload to continue moving it (or prevent it from moving) to its destination, so this tactic is mitigated in effectiveness. King of the Hill and 2CP Maps offer no such balancing factor. Tapping the point is enough to add another 10 seconds to the counter -- and your Opponents can't do anything to end the game besides keep would-be "Tappers" from touching the point in each 10 second increment. Overrtime in Overwatch seems tailor-made for Tracer and King of the Hill: her incredible movespeed, blinks, full Rewind heal, and a short run distance from Spawn spells "Meta Disaster". Cheesing Overtime and the Hero Stacking that it requires ruins the viewers’, the competitors’, and the Tournament workers’ experiences. Getting back to a little more objective-based opinions, Gosu’s Admins also informed me that this week’s Tourneys lasted roughly four hours longer than previous Tourneys -- which is absolutely unacceptable. 

the meta doctor's recommendations

With all of these “Whys” and their answers in mind, let’s try to think of solutions that Blizzard could implement to nip this growing Tournament Play problem in the bud. 

  1. Fundamentally change King of the Hill Map Design: If Blizzard insists on sticking to their 0 Hero Limit design philosophy,  King of the Hill Maps will need their relevant areas (Point, Hallways) increased in size and/or openness. This would place less emphasis on Heroes who can control small areas of space incredibly effectively, and open the door for more effective midrange hitscan heroes like Soldier 76 and McCree. 
  2. Overtime needs an OverHaul: It just doesn't make sense that you have to physically be on the point to prevent your Opponent from building Meter, but you can prevent them from winning the game simply by tapping the point. I’ve seen a lot of great suggestions for how Blizzard could change Overtime without taking away from its excitement factor. You could make contact with the point slowly “builds” time rather than resetting it. You could limit the amount of respawns each player gets to a certain set amount before staying permanently dead. You could even remove Overtime entirely!
  3. Add more Heroes whose cross-Hero Synergies are stronger than same-Hero Synergies: This is a bit more vague, but hear me out. I have always been an advocate for more Hero releases to Overwatch, and I believe that once the Hero pool is sufficiently large all of these issues plaguing Ranked Play will be a funny blip in the past. Additional Heroes who work better supporting other Heroes than in a stack (think Pharah-Mercy) would help crowd out the stacked Heroes simply by being better options for the same lineup slots. 

"But Captain," you say, "What if Blizzard doesn't want to or can't do any of these things yet? What can we do??"

You're absolutely right -- we can't expect Blizzard to implement sweeping changes just because of one weekend of Tournament play. This is where Tournament organizations have to (and already have in the past) take things into their own hands. Best of 5 King of the Hill needs to be abolished for Tournament play at a minimum -- these Maps last far too long and have far too much influence on the Overall Competitive Scene as they currently stand. 2CP Maps may also need a look as well, as 2CP Maps now also suffer from prolonged matches at the Pro level. Lifting the Ban on 2CP Maps resulted in only two Matches played (both on Volskaya Industries), but each of these Matches accounted for 6 and 5 matches respectively due to the new Time Bank system implemented by Blizzard. Luckily, Pros already avoid 2CP Maps like the plague, but ESL intends to feature different Maps in their Map Pool each week so 2CP matchups may be unavoidable. Unfortunately, there’s no easy solution to this problem: the best practice is simply to avoid 2CP at all costs. 


Final Verdict

There are glaring flaws in ESL’s current ruleset, and now we have data to back up what the Pros have been upset about since their announcement. If they change nothing else, Best of 5 King of the Hill needs to be changed to Best of 3. On top of that, re-introducing either 1 Hero Limit or experimenting with formats like the 2-1-1-1-1 Lineup proposed by some Pro players can be another way to to help mitigate the severe warping of the Meta. Banning the now long-lasting 2CP Maps may also prove necessary if Pros begin to ban King of the Hill out of sheer disgust. 

Of course, I do not actually expect ESL to implement any of these changes without some indication from the big guys: Blizzard themselves. Blizzard has all of the tools to make the fundamental changes needed to address the true offenders: Overtime, Map Design, and a Roster of Heroes which is looking smaller and smaller by the day. ESL’s Tourney qualifying rounds run for the entire month of July and most of August -- ending at Gamescom for a LAN Finals.

Here's where things get interesting. As stated many times by Jeff Kaplan himself, Blizzard intends to keep working on Competitive Mode on a Season-by-season basis. This year’s inaugural Summer Season ends just two days before ESL’s Finals, so it’s not completely out of the question that news of sweeping changes coming to the Fall Season’s Competitive offering could be announced by then. The best we, the Overwatch Community can do at present is to keep providing Jeff and his crew with as much good evidence and feedback and hope for a well-Designed solution come August.  

Until next time,


Icarus' corner

This week, as always, I partnered with Icarus to compile the data necessary for both of our Meta Reports. If you'd prefer your Meta Reports in audio/video form, head over to his YouTube channel, or check back here around every Wednesday of the week! Icarus finished early this week, check him out below!

Like this content? Tweet  and Share it with your followers!