d18cea2104
Features: - Multiplayer WebSocket game server (FastAPI) - 8 AI personalities with distinct play styles - 15+ house rule variants - SQLite game logging for AI analysis - Comprehensive test suite (80+ tests) AI improvements: - Fixed Maya bug (taking bad cards, discarding good ones) - Personality traits influence style without overriding competence - Zero blunders detected in 1000+ game simulations Testing infrastructure: - Game rules verification (test_game.py) - AI decision analysis (game_analyzer.py) - Score distribution analysis (score_analysis.py) - House rules testing (test_house_rules.py) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
300 lines
10 KiB
Python
300 lines
10 KiB
Python
"""
|
|
Tests for the GameAnalyzer decision evaluation logic.
|
|
|
|
Verifies that the analyzer correctly identifies:
|
|
- Optimal plays
|
|
- Mistakes
|
|
- Blunders
|
|
|
|
Run with: pytest test_analyzer.py -v
|
|
"""
|
|
|
|
import pytest
|
|
from game_analyzer import (
|
|
DecisionEvaluator, DecisionQuality,
|
|
get_card_value, rank_quality
|
|
)
|
|
|
|
|
|
# =============================================================================
|
|
# Card Value Tests
|
|
# =============================================================================
|
|
|
|
class TestCardValues:
|
|
"""Verify card value lookups."""
|
|
|
|
def test_standard_values(self):
|
|
assert get_card_value('A') == 1
|
|
assert get_card_value('2') == -2
|
|
assert get_card_value('5') == 5
|
|
assert get_card_value('10') == 10
|
|
assert get_card_value('J') == 10
|
|
assert get_card_value('Q') == 10
|
|
assert get_card_value('K') == 0
|
|
assert get_card_value('★') == -2
|
|
|
|
def test_house_rules(self):
|
|
opts = {'lucky_swing': True}
|
|
assert get_card_value('★', opts) == -5
|
|
|
|
opts = {'super_kings': True}
|
|
assert get_card_value('K', opts) == -2
|
|
|
|
opts = {'lucky_sevens': True}
|
|
assert get_card_value('7', opts) == 0
|
|
|
|
opts = {'ten_penny': True}
|
|
assert get_card_value('10', opts) == 1
|
|
|
|
|
|
class TestRankQuality:
|
|
"""Verify card quality classification."""
|
|
|
|
def test_excellent_cards(self):
|
|
assert rank_quality('★') == "excellent"
|
|
assert rank_quality('2') == "excellent"
|
|
|
|
def test_good_cards(self):
|
|
assert rank_quality('K') == "good"
|
|
|
|
def test_decent_cards(self):
|
|
assert rank_quality('A') == "decent"
|
|
|
|
def test_neutral_cards(self):
|
|
assert rank_quality('3') == "neutral"
|
|
assert rank_quality('4') == "neutral"
|
|
assert rank_quality('5') == "neutral"
|
|
|
|
def test_bad_cards(self):
|
|
assert rank_quality('6') == "bad"
|
|
assert rank_quality('7') == "bad"
|
|
|
|
def test_terrible_cards(self):
|
|
assert rank_quality('8') == "terrible"
|
|
assert rank_quality('9') == "terrible"
|
|
assert rank_quality('10') == "terrible"
|
|
assert rank_quality('J') == "terrible"
|
|
assert rank_quality('Q') == "terrible"
|
|
|
|
|
|
# =============================================================================
|
|
# Take Discard Evaluation Tests
|
|
# =============================================================================
|
|
|
|
class TestTakeDiscardEvaluation:
|
|
"""Test evaluation of take discard vs draw deck decisions."""
|
|
|
|
def setup_method(self):
|
|
self.evaluator = DecisionEvaluator()
|
|
# Hand with mix of cards
|
|
self.hand = [
|
|
{'rank': '7', 'face_up': True},
|
|
{'rank': '5', 'face_up': True},
|
|
{'rank': '?', 'face_up': False},
|
|
{'rank': '9', 'face_up': True},
|
|
{'rank': '?', 'face_up': False},
|
|
{'rank': '?', 'face_up': False},
|
|
]
|
|
|
|
def test_taking_joker_is_optimal(self):
|
|
"""Taking a Joker should always be optimal."""
|
|
result = self.evaluator.evaluate_take_discard('★', self.hand, took_discard=True)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_not_taking_joker_is_blunder(self):
|
|
"""Not taking a Joker is a blunder."""
|
|
result = self.evaluator.evaluate_take_discard('★', self.hand, took_discard=False)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
|
|
def test_taking_king_is_optimal(self):
|
|
"""Taking a King should be optimal."""
|
|
result = self.evaluator.evaluate_take_discard('K', self.hand, took_discard=True)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_not_taking_king_is_mistake(self):
|
|
"""Not taking a King is a mistake."""
|
|
result = self.evaluator.evaluate_take_discard('K', self.hand, took_discard=False)
|
|
assert result.quality == DecisionQuality.MISTAKE
|
|
|
|
def test_taking_queen_is_blunder(self):
|
|
"""Taking a Queen (10 points) with decent hand is a blunder."""
|
|
result = self.evaluator.evaluate_take_discard('Q', self.hand, took_discard=True)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
|
|
def test_not_taking_queen_is_optimal(self):
|
|
"""Not taking a Queen is optimal."""
|
|
result = self.evaluator.evaluate_take_discard('Q', self.hand, took_discard=False)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_taking_card_better_than_worst(self):
|
|
"""Taking a card better than worst visible is optimal."""
|
|
# Worst visible is 9
|
|
result = self.evaluator.evaluate_take_discard('3', self.hand, took_discard=True)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_neutral_card_better_than_worst(self):
|
|
"""A card better than worst visible should be taken."""
|
|
# 4 is better than worst visible (9), so taking is correct
|
|
result = self.evaluator.evaluate_take_discard('4', self.hand, took_discard=True)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
# Not taking a 4 when worst is 9 is suboptimal (but not terrible)
|
|
result = self.evaluator.evaluate_take_discard('4', self.hand, took_discard=False)
|
|
assert result.quality == DecisionQuality.QUESTIONABLE
|
|
|
|
|
|
# =============================================================================
|
|
# Swap Evaluation Tests
|
|
# =============================================================================
|
|
|
|
class TestSwapEvaluation:
|
|
"""Test evaluation of swap vs discard decisions."""
|
|
|
|
def setup_method(self):
|
|
self.evaluator = DecisionEvaluator()
|
|
self.hand = [
|
|
{'rank': '7', 'face_up': True},
|
|
{'rank': '5', 'face_up': True},
|
|
{'rank': '?', 'face_up': False},
|
|
{'rank': '9', 'face_up': True},
|
|
{'rank': '?', 'face_up': False},
|
|
{'rank': '?', 'face_up': False},
|
|
]
|
|
|
|
def test_discarding_joker_is_blunder(self):
|
|
"""Discarding a Joker is a severe blunder."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='★',
|
|
hand=self.hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
|
|
def test_discarding_2_is_blunder(self):
|
|
"""Discarding a 2 is a severe blunder."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='2',
|
|
hand=self.hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
|
|
def test_discarding_king_is_mistake(self):
|
|
"""Discarding a King is a mistake."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='K',
|
|
hand=self.hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.MISTAKE
|
|
|
|
def test_discarding_queen_is_optimal(self):
|
|
"""Discarding a Queen is optimal."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='Q',
|
|
hand=self.hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_swap_good_for_bad_is_optimal(self):
|
|
"""Swapping a good card for a bad card is optimal."""
|
|
# Swap King (0) for 9 (9 points)
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='K',
|
|
hand=self.hand,
|
|
swapped=True,
|
|
swap_position=3, # Position of the 9
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
assert result.expected_value > 0
|
|
|
|
def test_swap_bad_for_good_is_mistake(self):
|
|
"""Swapping a bad card for a good card is a mistake."""
|
|
# Swap 9 for 5
|
|
hand_with_known = [
|
|
{'rank': '7', 'face_up': True},
|
|
{'rank': '5', 'face_up': True},
|
|
{'rank': 'K', 'face_up': True}, # Good card
|
|
{'rank': '9', 'face_up': True},
|
|
{'rank': '?', 'face_up': False},
|
|
{'rank': '?', 'face_up': False},
|
|
]
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='9',
|
|
hand=hand_with_known,
|
|
swapped=True,
|
|
swap_position=2, # Position of King
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.MISTAKE
|
|
assert result.expected_value < 0
|
|
|
|
def test_swap_into_facedown_with_good_card(self):
|
|
"""Swapping a good card into face-down position is optimal."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='K',
|
|
hand=self.hand,
|
|
swapped=True,
|
|
swap_position=2, # Face-down position
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.OPTIMAL
|
|
|
|
def test_must_swap_from_discard(self):
|
|
"""Failing to swap when drawing from discard is invalid."""
|
|
result = self.evaluator.evaluate_swap(
|
|
drawn_rank='5',
|
|
hand=self.hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=True
|
|
)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
|
|
|
|
# =============================================================================
|
|
# House Rules Evaluation Tests
|
|
# =============================================================================
|
|
|
|
class TestHouseRulesEvaluation:
|
|
"""Test that house rules affect evaluation correctly."""
|
|
|
|
def test_lucky_swing_joker_more_valuable(self):
|
|
"""With lucky_swing, Joker is worth -5, so discarding is worse."""
|
|
evaluator = DecisionEvaluator({'lucky_swing': True})
|
|
hand = [{'rank': '5', 'face_up': True}] * 6
|
|
|
|
result = evaluator.evaluate_swap(
|
|
drawn_rank='★',
|
|
hand=hand,
|
|
swapped=False,
|
|
swap_position=None,
|
|
was_from_discard=False
|
|
)
|
|
assert result.quality == DecisionQuality.BLUNDER
|
|
# EV loss should be higher with lucky_swing
|
|
assert result.expected_value > 5
|
|
|
|
def test_super_kings_more_valuable(self):
|
|
"""With super_kings, King is -2, so not taking is worse."""
|
|
evaluator = DecisionEvaluator({'super_kings': True})
|
|
hand = [{'rank': '5', 'face_up': True}] * 6
|
|
|
|
result = evaluator.evaluate_take_discard('K', hand, took_discard=False)
|
|
# King is now "excellent" tier
|
|
assert result.quality in (DecisionQuality.MISTAKE, DecisionQuality.BLUNDER)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
pytest.main([__file__, "-v"])
|