fix(anomaly): exclude non-directional kinds from grading and backtest

Review follow-up (HIGH): the three detectors fed the same evaluator/backtest, but
SuspensionFreeze is non-directional (favourite unchanged) — grading it as "favourite
won" polluted the hit-rate with the base favourite-win rate, and its high frozen-ness
score always cleared the backtest threshold.

- Add AnomalyKind.IsDirectional() (flip + steam = true, freeze = false).
- AnomalyOutcomeEvaluator returns Unresolved for non-directional kinds (favourites
  still surfaced for display) so they don't distort calibration.
- RunBacktestUseCase skips non-directional anomalies when building candidates.
- Tests for the classification, the evaluator path, and the backtest skip.

tests/Marathon.Application.Tests/UseCases/RunBacktestUseCaseTests.cs

@@ -85,6 +85,25 @@ public sealed class RunBacktestUseCaseTests
         result.BetsPlaced.Should().BeGreaterThan(0, "the in-range graded anomaly produces a bet");
     }
 
+    [Fact]
+    public async Task Should_SkipNonDirectionalAnomalies_When_BuildingCandidates()
+    {
+        // A SuspensionFreeze anomaly (favourite unchanged) must not be staked.
+        var id = new EventId("88888888");
+        var freeze = new Anomaly(
+            Guid.NewGuid(), id, BaseTime, AnomalyKind.SuspensionFreeze, 0.9m, FlipEvidence);
+
+        _anomalies.ListAsync(Arg.Any<CancellationToken>())
+            .Returns(new[] { freeze }.ToList().AsReadOnly());
+        _events.GetAsync(id, Arg.Any<CancellationToken>()).Returns(MakeEvent(id));
+        _results.GetAsync(id, Arg.Any<CancellationToken>())
+            .Returns(new EventResult(id, 0, 2, Side.Side2, DateTimeOffset.UtcNow));
+
+        var result = await CreateSut().ExecuteAsync(DefaultStrategy(), CancellationToken.None);
+
+        result.BetsPlaced.Should().Be(0, "SuspensionFreeze is non-directional and must not be staked");
+    }
+
     [Fact]
     public async Task Should_ReturnEmptyResult_When_NoAnomaliesExist()
     {