MSI Game 3: Bilibili Gaming risk Ezreal into LYON

Viper’s Ezreal is the draft decision that changes the read on this game. Ezreal sits at 40.0% WR over 10G at MSI and only 40.5% globally into Sivir over 126G, so Bilibili Gaming are not taking the safe ADC route; they are betting that ON’s Bard and Xun’s Vi can create enough map chaos for Ezreal to play the mid game on tempo instead of getting dragged into a front-to-back Sivir fight. If that early unlock never comes, LYON (2024 American Team) get a much cleaner scaling path than the overall team strength would suggest.

Compositions

Bilibili Gaming drafted a skirmish-first composition with Bin on Renekton, Xun on Vi, Knight on Ahri, Viper on Ezreal, and ON on Bard. The identity is clear: early prio through solo lanes, fast engage from Vi-Ahri, and Bard roam windows that let BLG snowball side lanes before Sivir and Cassiopeia can stabilize. Renekton and Vi give immediate point-and-click setup, while Ahri adds pick pressure and Bard adds flex tempo through roam and disengage.

LYON (2024 American Team) answered with Dhokla on Zaahen, Inspired on Nocturne, Saint on Cassiopeia, Berserker on Sivir, and Isles on Nautilus. This draft is more conventional in its scaling and teamfight structure. Nocturne-Nautilus can force engage, Cassiopeia punishes dive paths if BLG overcommit, and Sivir gives reliable waveclear plus late-game DPS. Their best windows are the first Nocturne ult cycles and the later 5v5s where Sivir and Cassiopeia can stand and fire.

Key Picks and Stats

Top lane is volatile. Bin’s Renekton owns a 48.6% global WR over 695G, but a 55.6% MSI WR over 9G; against Zaahen he is only 45.1% over 51G, and Bin himself is 0.0% on Renekton at this MSI over 1G with a 1.5 KDA. Dhokla’s Zaahen is steadier at 53.5% globally over 374G, though only 33.3% at MSI over 3G. The direct matchup is close: Zaahen holds 52.9% vs Renekton globally over 51G, while MSI is 50.0% over 2G.

Jungle-mid is where BLG recover the edge. Xun’s Vi has 52.6% global WR over 810G and 87.5% at MSI over 8G, even though Vi is only 41.8% globally into Nocturne over 55G. Knight’s Ahri brings 53.5% global WR over 641G and 80.0% at MSI over 5G; against Cassiopeia she is 52.9% over 17G. Saint’s Cassiopeia is real pressure at 54.6% globally over 350G and 55.6% at MSI over 9G, but the BLG model synergy on Vi+Ahri is 0.5412, one of the strongest pairings in the draft.

Bot lane is the contradiction zone. Viper’s Ezreal has 48.4% global WR over 1049G, 40.0% at MSI over 10G, and just 0.0% for Viper at this event over 1G, even with a 8.0 KDA. Berserker’s Sivir, by contrast, has 53.1% global WR over 478G, 100.0% MSI WR over 1G, and Sivir is 58.7% globally into Ezreal over 126G. Still, ON’s Bard matters: 54.8% global WR over 843G, 66.7% MSI WR over 9G, 60.2% into Nautilus over 83G, though ON is also 0.0% on Bard at this MSI over 1G with a 13.0 KDA.

Draft Edge

The draft edge is narrow in lane stats but wider in execution ceiling, and that favors Bilibili Gaming. LYON (2024 American Team) have the cleaner ADC matchup and a stronger standard teamfight backline, yet BLG’s Vi-Ahri-Bard core gives them more ways to start fights on their terms. Renekton plus Ahri can attack side control, Bard can break map states with roam, and Nocturne cannot freely dive if Cassiopeia is forced to answer picks instead of frontlining space.

For LYON, the win condition is simple: survive the first 15-20 minutes without losing too much on side lanes, then let Berserker’s Sivir and Saint’s Cassiopeia punish BLG’s short range engage. For BLG, the game is about pace: get Knight and Xun on the map early, turn ON loose, and keep Viper on safe two-item timings instead of asking Ezreal to be the primary carry in a static 5v5.

Polymarket Market

Polymarket is far more decisive than the draft model. The Game 3 market sits at 86% for Bilibili Gaming and 14% for LYON (2024 American Team), while the series market now is 99% to 1%. There is no series pre-match number provided here, so the exact move cannot be quantified, but after BLG won G1 12-11 in 32:03 and G2 31-18 in 28:31, the direction is obviously toward Bilibili Gaming.

The gap between Game 3 and Series Now is 13 percentage points, which tells a useful story. The market is overwhelmingly convinced BLG close the series, but it is slightly less absolute on this specific draft because Ezreal into Sivir is the one lane setup that gives LYON a realistic path if they drag the game out. Even so, real-money traders are clearly pricing BLG’s class, momentum, and series state above the red-side scaling advantages.

Prediction

The model opens at 57% for Bilibili Gaming against 43% for LYON (2024 American Team). I would push that slightly upward to 60% for Bilibili Gaming and 40% for LYON (2024 American Team): BLG have the better Vi-Ahri playmaking core, stronger Elo at 0.624 versus 0.376, and they already carry the momentum of two wins in the series. The two factors that could flip it are Viper’s Ezreal failing to convert lane pressure and Saint’s Cassiopeia reaching stable 5v5 positions, but BLG still own more proactive win conditions.