LYON (2024 American Team) 3-0 G2 Esports — MSI 2026 Results & Stats

TL;DR: LYON (2024 American Team) swept G2 Esports 0-3 at MSI 2026, turning a supposed mismatch into a statement series. They won because their drafts became more coherent as the BO5 went on, their teamfighting stayed cleaner under pressure, and their carries repeatedly converted early edges into complete map control.

LYON (2024 American Team) did not just beat G2 Esports; they shut the door on them without giving back a single game. In a series that began with G2 as the clear public favorite, LYON made the result feel more convincing with each map, transforming MSI 2026 from a routine favorite’s march into a sweeping 0-3 upset with real international weight.

Key Takeaways

• Berserker was the series MVP, and the closing image says why: after helping build bot-side control in Game 1, he exploded in Game 3 on Caitlyn at 11/0/8 with +1338 GoldDiff@15, while LYON finished that map with a brutal 23-5 kill score and 11 towers.
• The most decisive series moment came in Game 2, when LYON turned a slim live draft-model edge of 50% into an 18-7 win; once Inspired on Lee Sin reached 6/2/9 and Saint on Viktor hit 6/0/5, the BO5 stopped feeling like a one-game surprise and started feeling like a takeover.
• The final 0-3 score was not a fluke hidden inside close endings: LYON won despite pre-match series odds of only 22%, then kept validating the upset with cleaner objective play, taking Game 1 in 35:42, Game 2 in 30:38, and Game 3 in 38:47.

Before the Series

Coming in, the draft looked like the clearest battleground. The pre-match read centered on Orianna at 72.4% presence and 56.9% ban rate, plus Vi at 70.7% presence and 72.7% win rate, with the idea that G2 might need to deny control in mid before the map got reactive around Caps. The nightly pre-draft analysis also flagged Orianna, Jarvan IV, Jayce, and Bard as champions likely to matter if they appeared.

Across the full arc, that prediction landed unevenly but meaningfully. Orianna showed up in Game 1 and gave Caps a lane edge at +170 GoldDiff@15, yet it did not become a winning answer. Jarvan IV appeared in Game 1 as well, but SkewMond’s 1/2/6 line captured the problem: present, useful, not decisive. Bard, on the other hand, absolutely delivered when the series was on the line, with Isles posting 3/4/16 in Game 3 and giving LYON the roam and engage timing the pre-draft model promised. Jayce was one of the highlighted danger picks from the meta conversation, but in this series it never became the defining lever those other champions did.

Game 1 — Setting the Tone

Game 1 was the first crack in the expected script. The live draft model gave G2 Esports 53%, and that is the one draft edge in this BO5 that did not translate to a win. You could hear why the moment the game got messy. G2 had small lane advantages, especially Caps on Orianna and BrokenBlade’s Anivia, but LYON kept refusing the clean, front-to-back rhythm G2 wanted.

Instead, the voice of the game became Saint on Sylas, finishing 6/1/6 and turning every uncertain fight into a louder LYON moment. Bot lane mattered too: Berserker and Isles built control through +595 and +54 GoldDiff@15, and that translated into 4 dragons, 1 barons, and 10 towers. It was not only an upset at 28% pre-game odds; it was a warning that LYON’s read on tempo was sharper than the market believed.

Game 2 — The Pivot

Every sweep has a hinge, and this was it. Down 0-1, G2 still had room to reset the series, but Game 2 became the map where LYON’s superiority stopped looking temporary. This time the live draft model leaned LYON at 50%, and unlike Game 1, the predicted draft edge arrived exactly on schedule.

Top side opened first, with Dhokla’s Gnar building +1791 GoldDiff@15 over BrokenBlade’s Yasuo. Then the jungle and mid pairing took over the soundscape of the game. Inspired on Lee Sin finished 6/2/9 with +1039 GoldDiff@15, while Saint on Viktor ended 6/0/5 with +1029 at 15. Prediction item 2 had to be judged across the series, and Game 2 was the clearest confirmation: the model’s edge for LYON translated directly into result, pressure, and control.

Game 3 — The Climax

At 0-2, G2 needed the series to slow down. LYON made sure it never did. The live draft model favored LYON again at 55%, and this was the sharpest example of draft edge turning into a complete on-stage avalanche. If Game 2 was the pivot, Game 3 was the stamp.

The bot lane broke the map open. Berserker's Caitlyn reached 11/0/8 and a crushing +1338 GoldDiff@15 over Hans Sama's Yunara, while Isles on Bard roamed exactly as predicted, ending 3/4/16 and helping LYON secure 4 dragons. That explicitly resolves prediction item 1: of the highlighted picks, Bard delivered hardest when it mattered most, Orianna and Jarvan IV appeared but failed to swing the series toward G2, and Jayce never became a decisive factor in practice. By the end, the 23-5 kill score told the truth: this sweep was earned lane by lane, fight by fight.

Aftermath

What makes this 0-3 memorable is not only that LYON (2024 American Team) beat G2 Esports at MSI 2026. It is that the series became more convincing for LYON as it went on. Game 1 challenged expectation, Game 2 reordered the balance of the BO5, and Game 3 removed all doubt. Saint, Berserker, Inspired, Isles, and Dhokla all left fingerprints on the result, but the larger lesson was collective: LYON drafted with clarity, fought with discipline, and punished every opening before G2 could snowball.

Polymarket Trajectory

The market began with G2 Esports as a heavy 78% series favorite, which made sense on reputation but proved slow to absorb what the games were actually saying. Even after LYON took Game 1, G2 still held a 62% series edge, a sign that the public continued to trust the brand and the expected ceiling more than the evidence on the Rift. The real correction only arrived after Game 2, when the series price swung hard and finally treated LYON like the team in control. In retrospect, the warning sign was visible earlier: the live draft model was already more skeptical of G2 than the broader market, and once LYON’s mid-jungle and bot-side execution kept outperforming those public odds, the upset stopped being a surprise and became the accurate read.

Series Stats

FAQ

*Odds via Polymarket, 2026-07-10 11:26 UTC.*