2-2-2 is here and I don’t even know what to think anymore.

Seriously. We began the week with the Los Angeles Gladiators defeating the New York Excelsior, the Vancouver Titans exacting revenge on the Shanghai Dragons, and then ended it with the Houston Outlaws upsetting the very same Gladiators who defeated New York. And, oh yeah, there was Hyo-Jong “Haksal” Kim setting new Dragonblade elimination records against the Florida Mayhem.

OK, maybe that last one wasn’t so unexpected.

What is incredibly apparent is that 2-2-2 has breathed new air into the collective lungs of the cast of damage heroes in Overwatch, and nothing will ever be the same again. If you traveled back in time to Stage 1 of this year—or heck, even Stage 1 of last year—bearing the news of the sights and wonders that you beheld in Stage 4 of 2019, you’d have been thrown into Brig jail.

This is a story of change, of unexpected developments, and of heroes we thought we’d never see again. What follows are the four most interesting statistics from the first week of 2-2-2.

Stat #1: The most common hero composition in Stage 4 features Mei

Imagine grabbing the shoulders of your friend after Jong-Ryeol “Saebyeolbe” Park and NYXL won the Stage 3 Finals last season and telling them that this time next year, Mei would highlight the most meta composition in the Overwatch league. Sometimes, the truth is stranger than fiction.

The most common composition—and one of the most successful from a teamfight win-rate perspective—was the Ana/Mercy/Orisa/Roadhog/Mei/Widowmaker lineup that many have opted to call the “Ice Fishing” or “Ice Pick” comp due to its chilly heroes and pick-based nature. The various elements of this team composition provide both synergy and counters to each other, which is a trait in many meta compositions. Orisa’s Halt combos with Roadhog’s Hook, Ana’s Biotic Grenade, and a Widowmaker taking pot-shots at an immobilized target. On the flip side, however, Mei’s Ice Wall can interrupt enemy Halt-Hook combos or add an extra layer of security to her own team’s eliminations. So far, not every team has gone fishing—but most who have, have found success.

Stat #2: The Washington Justice are leading the league in teamfight win rate

For three stages, the Justice have been disappointing both from a strategic and mechanical standpoint. But now, their time has come. They may have only played one team so far this stage, and that team may have been the Toronto Defiant, but the Justice certainly put on a show that led them to a 63% teamfight win rate, just edging out the Reign’s 62.5%. For all the hype behind Corey “Corey” Nigra’s Widowmaker (which scored first eliminations in 6 out of 26 teamfights, and currently leads the stage), the Justice found most success with Corey on Hanzo and Ethan “Stratus” Yankel on Mei, where they won over 70% of their teamfights. Perhaps the Justice have finally found their groove—and their damage player mind meld—at just the right time.

They’re going to need it to finish the season on a high note… since their new tank line unfortunately is not going to be playing until 2020.

Stat #3: Over a quarter of all Genji play time this stage has been Haksal

This is just the tip of the iceberg. It’s obvious why Vancouver would want to run their most talented damage player on his most famous hero, but let’s take a step back and examine why most teams would not.

First, Genji only has a 40.9% teamfight win rate this season, and only 44.1% in Stage 4. Most teams do not see him as a hero that can get legitimate value outside of his Dragonblade—even with his blade unsheathed, they’re dubious. It’s tough to be Genji these days, with Roadhogs ready to hook, Orisas ready to Halt, and Widowmakers ready to bop you in the head before you can say “Ryūjin no ken wo kurae.”

However, Haksal and the Titans have found a way to make it work. They throw all their damage-boosting resources into Haksal, resulting in the second-fastest average time to Dragonblade—96 seconds, behind only Gui-Un “Decay” Jang’s 92—but with four times the amount of time played. This means Haksal is getting more blades faster, and he’s making the most of them. While the league average teamfight win rate with Genji is 40.9% this season, Vancouver’s is 50.8%—rising to a very spicy 72.7% when Haksal uses Dragonblade.

Stat #4: The most clutch player of the week was Architect

Some time ago I attempted to define what clutching means in Overwatch, which led to the creation of metrics such as Shorthanded Final Blows and Hype Rating. This week, only one player recorded more than one Shorthanded Final Blow while remaining undefeated in all overtime teamfights. That player was Min-Ho “Architect” Park, who saved the Shock against the Dynasty’s last-chance assault on Temple of Anubis:

In this teamfight, Architect recorded five total Final Blows, three of which were shorthanded. Architect, by the way, is no stranger to clutch eliminations. Last season, he notched 3.94 shorthanded final blows/10 mins in overtime teamfight wins, the highest rate of “clutchness” in the league. Thanks to 2-2-2, we get to see more of this monster damage player doing what he does best.

Week 2 of Stage 4 kicks off on Thursday, August 1, when the Charge (11-12) take on the Spitfire (14-9) at 4 p.m. PDT. Watch all 2019 season matches live and on demand on overwatchleague.com, the Overwatch League app, our Twitch channel, MLG.com, and the MLG app.