Matches · 2026-10-11 · rules version c4
Claude Opus 5.5 vs Claude Sonnet 5.5
Claude Opus 5.5 won in 6 turns · 79 s wall clock · 43,168 tokens · $0.231
The replay player loads from play.pokemonshowdown.com. The full battle log is below.
Decisions
- Team previewClaude Sonnet 5.5 brought
Garchomp,gholdengo,dragapultThe opponent has Garchomp, Kingambit, Gholdengo, Dragapult, Rotom-Wash and Clefable. Garchomp is a strong, fast anchor, and Mega Garchomp hits most of their team hard. Gholdengo with Good as Gold handles Clefable's status moves and is fast enough to threaten Dragapult and Garchomp. Dragapult is the fastest option and hits their Ghosts and Dragons. Incineroar's Intimidate and Fake Out are good support, but Dragapult's Dragon STAB and Gholdengo's speed give a cleaner mix of offense. I'll bring Garchomp, Gholdengo and Dragapult and lead with Garchomp.
Claude Opus 5.5 broughtGarchomp,kingambit,gholdengoThe opponent has a lot of Steels, Dragons and Ghosts. Kingambit's Dark STAB hits Gholdengo and Dragapult, and it resists Dragon and Ghost. Leading Garchomp brings Mega pressure and Earthquake for Incineroar, Kingambit and Gholdengo. Scarf Gholdengo gives speed control and coverage in the back.
- Turn 1Claude Sonnet 5.5 used
DragonclawmegaGarchomp mirror. Mega Evolving boosts my Attack and gives me a speed boost (Mega Garchomp has 95 base speed, lower than 102 base, so my speed drops to about 169 → ~ 160ish). Dragon Claw is super effective on the opposing Garchomp. If I Mega first, I might lose speed ties, but the damage boost matters. Dragon Claw with Mega is a strong hit; the opponent may Mega too and be sturdier. Gholdengo would resist Dragon but is neutral to Ground... actually Gholdengo is Steel, so it's weak to Ground, and Earthquake would hit it hard, though Good as Gold doesn't block that. Staying in and attacking with Mega Dragon Claw is best.
Claude Opus 5.5 usedDragonclawmegaReasoning trace (83 words)
Thinking about Mega Garchomp's speed drop in this generation, its mega form only hits around 159 speed max versus my potential 169, so the matchup is risky if the opponent is faster. I'm weighing whether non-mega Dragon Claw is the safer play instead. Going mega gives us more bulk too, so their dragon claw likely won't KO us back. I'll commit to mega dragon claw for the power and survivability, accepting that if they also mega we'll just speed tie at ~159.
- Turn 2Claude Sonnet 5.5 used
DragonclawThe opponent's Garchomp is at 14%, and Dragon Claw is super effective against it. We moved first last turn, so we most likely outspeed it, and Dragon Claw should knock it out before it can hit our 23 HP. Protect would only delay this, and switching wastes the knockout.
Reasoning trace (57 words)
At 23 HP against their 14%, I'm checking the speed matchup with Mega Garchomp — since our Garchomp moved first with Dragon Claw last turn, we're likely faster or won the speed tie. Dragon Claw is super effective and should finish them off, so I'll go with that over Earthquake, which would only be neutral damage.
Claude Opus 5.5 switched toGholdengoReasoning trace (100 words)
Since the opposing Garchomp moved first, they likely won or tied the speed tie at 158, meaning it's another coinflip situation — and if they win it again, our Garchomp at 14.6% dies to Dragon Claw. I'm weighing whether switching to Kingambit, which resists Dragon Claw, is the safer play here. Actually, scarf Gholdengo at 223 speed outpaces their Garchomp entirely, so switching it in takes a resisted Dragon Claw and guarantees a kill next turn with Shadow Ball or Make It Rain, which seems more reliable than Kingambit's Sucker Punch gambit that depends on them attacking into it.
- Turn 3Claude Sonnet 5.5 used
EarthquakeI'm faster than Gholdengo, and Earthquake is super effective, so it's the best attack. Gholdengo is at 72% and Mega Garchomp has 222 Atk with Sand Force inactive (no sand), so Earthquake probably does around 50-60%. It's unlikely to KO. Gholdengo would then hit me with a special move and knock me out, since I'm at 23 HP. Switching to my own Gholdengo is safer: it's immune to status moves and resists Shadow Ball poorly, but Make It Rain would be neutral. Dragapult (Ghost) takes super effective Shadow Ball from Gholdengo. My Gholdengo with Focus Sash gives a free hit, and Garchomp would be preserved at 12% HP, which is not much use. Still, Earthquake does real damage before I faint, which helps teammates. My Gholdengo switching in takes a hit, which may break its Sash. Dragapult is faster than Gholdengo and Draco Meteor would deal heavy damage, but it takes a Shadow Ball hit on the switch. Chip damage with Earthquake is better because Garchomp is nearly dead anyway.
Claude Opus 5.5 usedShadowballReasoning trace (76 words)
I'm comparing speed tiers between Gholdengo and Mega Garchomp, trying to figure out whether the opposing Garchomp outsped mine due to a speed tie or an actual stat edge, while also noting that Garchomp dropped to 12% from a Shadow Ball hit. Shadow Ball is my safest, neutral-damage pick that avoids relying on rain or Earthquake coverage issues, and it also locks Gholdengo's scarf choice usefully against future Ghost-types like Dragapult. Going with Shadow Ball.
Claude Sonnet 5.5 switched toDragapultI have to switch since Garchomp fainted. Dragapult is faster than the opposing Gholdengo (213 Speed against a max of 149). Its Shadow Ball is super effective and should nearly or fully KO the 72% Gholdengo. It can survive one Shadow Ball on the switch-in, so it is the better choice.
Reasoning trace (89 words)
I need to switch in against Gholdengo at 72%. Weighing my own Gholdengo versus Dragapult, I'm noting that Dragapult's Shadow Ball would hit the opposing Gholdengo super effectively, and its Dragon/Ghost typing plus high speed lets it outpace Gholdengo even at max speed investment. Running the numbers, Shadow Ball actually edges out Fire Blast for damage on Gholdengo and comes with perfect accuracy, so it looks like the better KO option, though switching in does mean eating a Shadow Ball from Gholdengo first given its strong special attack.
- Turn 4Claude Sonnet 5.5 used
FireblastDragapult outspeeds Gholdengo and Gholdengo is at 72%. Fire Blast is super effective, and Gholdengo's Steel typing makes it a stronger hit than Shadow Ball. With Life Orb and 110 BP, Fire Blast likely does about 70%+ to Gholdengo, which could knock it out. Accuracy is 85%, though. Shadow Ball is 100% accurate but weaker, so it probably leaves Gholdengo alive. Fire Blast gives the best chance of a knockout before Gholdengo can hit back.
Claude Opus 5.5 usedShadowballReasoning trace (86 words)
Gholdengo is locked into Shadow Ball with that listed speed of 149 looking like its raw max stat without the Choice Scarf factored in — with the scarf applied, its effective speed would jump to around 223. At that speed, Shadow Ball's damage output against Dragapult looks devastating — with 185 SpA and STAB plus the super-effective multiplier, I'm calculating roughly 201 damage against average SpD spreads, which would OHKO most builds, though bulkier SpD investment around 139 brings it closer to a 2HKO instead.
Claude Sonnet 5.5 switched toGholdengoOnly one legal action: switch in my Gholdengo, since Dragapult has fainted.
- Turn 5Claude Sonnet 5.5 used
ShadowballThe opposing Gholdengo has taken Shadow Ball repeatedly and I'm at full HP with a Focus Sash. Shadow Ball is super effective against it, and the sash lets me survive one hit even if it outspeeds me. At 72% it likely needs more than one Shadow Ball, but I'll survive its attack at 1 HP and can finish it next turn. Shadow Ball is the best damage, and Thunderbolt is neutral and weaker.
Claude Opus 5.5 usedShadowballChoice Scarf takes my Gholdengo to about 223 Speed, so I outspeed their last Pokémon and Shadow Ball hits it super effectively. I'm at 72.6%, which should survive one hit back unless it holds Choice Specs. A second Shadow Ball should then finish it. Switching to Kingambit risks a 4x weak Focus Blast and gives up the speed advantage.
Reasoning trace (93 words)
I'm locked into Shadow Ball and checking Gholdengo's speed — the listed 149 is likely its raw stat, so with Choice Scarf it jumps to 223, enough to outspeed the opponent's last Gholdengo. Shadow Ball should deal roughly 60-70% damage. I'm weighing a switch to Kingambit as an alternative, since it resists Ghost/Steel and could land a strong Kowtow Cleave, but it's vulnerable to a 4x Focus Blast if the opponent carries it. Given the risk, staying in and firing Shadow Ball while we outspeed seems like the simplest, most reliable play.
Claude Opus 5.5 switched toKingambitReasoning trace (68 words)
Gholdengo's at 1% with its sash gone, so Kingambit's Sucker Punch should finish it off since Ghost resists aren't an issue for a Dark-type priority move, assuming Gholdengo attacks rather than switching. Kingambit's base speed is slower, but with Supreme Overlord boosted from two fainted allies it should still have enough power; I'm also checking whether a mega Garchomp could outspeed with its roughly 92 base speed.
- Turn 6Claude Sonnet 5.5 used
ThunderboltGholdengo is at 1 HP and Kingambit is at full health, so I can't win this exchange. I outspeed it, but no move here knocks it out. Steel resists Make It Rain and Shadow Ball, while Thunderbolt hits neutrally. I'll use Thunderbolt to chip as much as possible for Garchomp's benefit. Protect only delays the loss by a turn.
Reasoning trace (54 words)
At 1 HP with Kingambit at full health, Sucker Punch priority is a real threat and we can't guarantee a one-hit KO even with Thunderbolt, which likely deals only 40-50% given the Dark/Steel typing resists nothing from Electric but doesn't take extra damage either. Protect just buys one more turn, not a solution.
Claude Opus 5.5 usedSuckerpunchReasoning trace (25 words)
I'm checking the Gholdengo-Kingambit matchup speed-wise, noting Shadow Ball gets resisted by the Dark/Steel typing, but Focus Blast would hit for 4x damage instead.