feat: add answer confidence differential as scoring factor

When both bots are correct, the one with higher checkAnswer confidence
(exact match 1.0 vs fuzzy match 0.8) gets up to +1.0 bonus points.
This rewards precise answers over approximate ones, adding another
competitive dimension beyond pure speed.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

server/src/engine/scoring.test.ts

@@ -218,6 +218,25 @@ describe('scoreRound', () => {
     expect(typeof result.narration).toBe('string')
     expect(result.narration.length).toBeGreaterThan(10)
   })
+
+  it('answer confidence gives bonus — exact match (1.0) beats fuzzy match (0.8)', () => {
+    // Use answers where one answer is a substring of another
+    // "satoshi" against ["satoshi nakamoto"] scores 0.8 (containment)
+    // "satoshi nakamoto" against ["satoshi nakamoto"] scores 1.0 (exact)
+    const challenge = makeChallenge({ answers: ['satoshi nakamoto'] })
+
+    // Both answer at equal speed, but A has exact match, B has fuzzy
+    const result = scoreRound(
+      challenge, botA, botB,
+      makeResponse('satoshi nakamoto', 500),  // exact → 1.0
+      makeResponse('nakamoto', 500),           // contained → 0.8
+      null, 0, 0,
+    )
+
+    // A should score higher due to confidence differential
+    expect(result.botAScore).toBeGreaterThan(result.botBScore)
+    expect(result.winnerId).toBe('a1')
+  })
 })
 
 describe('calculateElo', () => {