feat: surface juror-balanced scores and AI calibration advisory

Adds a shared juror-balancing utility (z-score normalization per juror,
rescaled back onto the raw 1-10 scale) and wires it into:

- Admin reports page: Top-10 project table now shows "Raw Avg" and
  "Balanced" columns side by side, and the summary stats row shows a
  balanced-average tile. Sort defaults to balanced so harsh and lenient
  graders no longer skew the ranking.
- Ranking dashboard: each project row shows a green/amber balanced-score
  chip next to the raw average when the two differ by ≥0.05, making it
  obvious when juror calibration moved a project's effective ranking.

Also adds AI Juror Calibration Advisory — a mutation that takes
anonymized per-juror stats, calls OpenAI, and produces a plain-language
explanation of the cohort's grading patterns plus per-juror severity
(normal / notable / outlier) with a one-sentence narrative. The advisory
describes the statistical balance that already runs; it does not
introduce a new weighting layer. Rendered as a panel in the Juror
Consistency tab when a specific round is selected.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/components/admin/round/ranking-dashboard.tsx

@@ -82,6 +82,7 @@ type SortableProjectRowProps = {
   entry: (RankedProjectEntry & { originalIndex?: number }) | undefined
   projectInfo: ProjectInfo | undefined
   jurorScores: JurorScore[] | undefined
+  balancedScore: number | null
   onSelect: () => void
   isSelected: boolean
   originalRank: number | undefined // from snapshotOrder — always in sync with localOrder
@@ -95,6 +96,7 @@ function SortableProjectRow({
   entry,
   projectInfo,
   jurorScores,
+  balancedScore,
   onSelect,
   isSelected,
   originalRank,
@@ -199,11 +201,25 @@ function SortableProjectRow({
           </span>
         ) : null}
 
-        {/* Average score */}
+        {/* Raw + balanced averages shown side by side */}
         {entry?.avgGlobalScore !== null && entry?.avgGlobalScore !== undefined && jurorScores && jurorScores.length > 1 && (
-          <span className="text-xs font-medium text-muted-foreground" title="Average score">
-            = {entry.avgGlobalScore.toFixed(1)}
-          </span>
+          <div className="flex items-center gap-1.5 text-xs" title="Raw juror average vs. juror-balanced average (z-score normalized per juror, rescaled to 1-10)">
+            <span className="font-medium text-muted-foreground">
+              {entry.avgGlobalScore.toFixed(1)}
+            </span>
+            {balancedScore != null && Math.abs(balancedScore - entry.avgGlobalScore) >= 0.05 && (
+              <span
+                className={cn(
+                  'font-semibold tabular-nums rounded px-1.5 py-0.5 border',
+                  balancedScore > entry.avgGlobalScore
+                    ? 'bg-emerald-50 text-emerald-700 border-emerald-200'
+                    : 'bg-amber-50 text-amber-700 border-amber-200',
+                )}
+              >
+                ⇢ {balancedScore.toFixed(1)}
+              </span>
+            )}
+          </div>
         )}
 
         {/* Advance decision indicator */}
@@ -909,7 +925,8 @@ export function RankingDashboard({ competitionId: _competitionId, roundId }: Ran
                                   currentRank={index + 1}
                                   entry={rankingMap.get(projectId)}
                                   projectInfo={projectInfoMap.get(projectId)}
-                                  jurorScores={evalScores?.[projectId]}
+                                  jurorScores={evalScores?.byProject[projectId]}
+                                  balancedScore={evalScores?.balanced[projectId]?.balancedAverage ?? null}
                                   onSelect={() => setSelectedProjectId(projectId)}
                                   isSelected={selectedProjectId === projectId}
                                   originalRank={hasReorders ? snapshotOrder[projectId] : undefined}