Sombra Moves to Support and Roadhog Loses the One-Shot Combo: Season 5 Rewrites the Overwatch 2 Board

Three data points sit next to each other in Overwatch 2's Season 5 update. One: the number of one-shot kill combos left in Roadhog's kit. Three: the number of heroes that have ever changed role in the game's history, with Sombra the newest name on that list. Zero: the number of escape routes for teams that spent an entire preparation cycle building around Sombra as a damage hero. In the language of professional play, a perma-banned hero does not exist on the map. It exists on the ban sheet. Sombra lived that way for multiple seasons as a damage pick: no wins, no losses, no data, no trace beyond a line of text on the ban board. From Season 5, she returns to the match in a different role entirely: support. Overwatch 2 launched in October 2026, cutting team size from six to five and splitting the roster into three roles: damage, support, and tank. Inside that structure, a role change is not a power adjustment. It plugs one strategic slot and opens another, dragging composition, fight tempo and ultimate economy along with it. This patch reworks two heroes at once, and neither fix is a numbers cut. Sombra keeps most of her ability skeleton, but the effects are rewritten for a support function. The official phrasing is similar behaviours, new effects. To me that is the most uncomfortable kind of change for a professional: familiar hands, unfamiliar head. Mechanical memory of the hero survives; tactical memory is wiped. Roadhog loses the thing that defined him for years, the hook-and-one-shot combo. A note on method before the detail. I grew up on European football models, where xG and PPDA are used to hear what players do not say out loud. PPDA was never about predicting Croatia; it was about hearing the intention Modric never put into words. But a hero shooter is not football, and I have learned enough not to force a pressing-intensity metric onto a game that runs on cooldown budgets and ultimate timing. EMP stays on the support role. In Overwatch, EMP is one of the heaviest ultimates in the game: it disables abilities and barriers across most of the enemy formation for a short window. Keeping it while moving Sombra to support creates a structural paradox, a role designed to keep people alive carrying the strongest fight-opening tool in the game. If EMP casts per ten minutes rise while eliminations per ten minutes fall, the hero has changed value axis rather than changed labels. If Sombra's ban rate as a support stays as high as it was as damage, the redesign has moved the problem rather than solved it. Based on my own experience watching matches, I check three data layers before trusting a new hero: pick rate, ban rate, and win rate split by match length. This patch supplies none of them, so everything below is a conditional hypothesis. Roadhog changes the unit of measurement. Scrap Gun now fires in two bursts. He gains an alt-fire called Trash Compactor, which absorbs enemy projectiles and returns them as an explosive ball, mechanically close to D.Va's Defense Matrix. The Scrap Hook Minor Perk now reloads four ammo instead of two. Read the mechanism and you see a switch in measurement units: a tank that kills quickly creates an immediate numbers advantage; a tank that absorbs damage creates a time advantage. Those need different stat sets, and the old set will make Roadhog look worse than he is. The three metrics I will watch instead are average survival time per fight, damage absorbed per ten minutes, and the share of fights opened by a hook that produces no immediate kill. The Perk system adds a preparation dimension. The list includes CTRL ALT ESC, Life Hack, Data Packet, Cybersecurity, Piggy Piggy, Toxic Exhaust and Scrap Hook Minor. Teams must now optimise hero selection and Perk configuration against each opponent. In BO3 or BO5, that adds a between-game dimension: a team can lose game one on one Perk build and win game two on another with the same hero. For teams with thin analytics staff, that is a real cost; for teams with dedicated analysts, it is an advantage that repeats all season. Numbers do not lie; only readings do. A Perk table is not an advantage unless a team has a testing process. And a team without one can still win by reading the opponent's mistakes faster. From a market angle: the transfer market is where emotion gets priced, and I only stand outside that room. But I stand close enough to see an asset class being repriced, the Sombra damage specialists. Value tied to a role that no longer exists falls in the short term, regardless of individual skill. The contrarian angle. Two heroes reworked at once means teams must learn two new variables inside one preparation window. When the meta is unstable, the advantage of strong teams compresses, because past experience loses part of its reference value. My estimate: the probability of an upset result in the first stage of the season rises to roughly 55 to 65 percent, provided neither design is hotfixed within two weeks. Second contrarian point: the community will call Roadhog nerfed. Eliminations will fall, and the familiar stat sheet will confirm the feeling. But a tank that absorbs eight hundred damage and survives four extra seconds can generate more value than one kill in a contested objective fight. The easiest stat to read is rarely the right one. Third, the methodological warning: correlation is not causation. If teams running Sombra support win more in week one, the cause may be opponents who do not yet know how to answer her, not a hero that is too strong. Separating those requires a lagged variable, week one against week three, once teams have film to study. Data is where I hide, and also where I learn to distrust every claim. This patch does not give me enough to say anything certain about either hero's true strength. It only tells me exactly what to measure. Signals to track: Sombra support pick rate at professional level; her ban rate, because if EMP still forces bans, the redesign has not produced the diversity it promised; Roadhog's average survival time in fights; the frequency of hooks that produce no kill; and the upset rate across the first two weeks. If those signals confirm, we are watching one of Overwatch 2's largest value-axis shifts, where two heroes trade an old identity for a different seat in the composition structure. If they do not, we watched a loud patch. And if a team can win simply by understanding the patch a few days faster than its rivals, is the most expensive position on this season's transfer market still a player?