add pattern analysis feature with web dashboard and CLI

New /patterns page with 9 analyses: chair win bias, bet rank
correlations, hand type distributions, pot size buckets, streaks,
hourly patterns, and recent-vs-overall comparison. Also adds a
standalone analyze.py CLI script for terminal output.

analyze.py

@@ -0,0 +1,166 @@
+#!/usr/bin/env python3
+"""
+Standalone CLI script for Teen Patti pattern analysis.
+
+Usage:
+    python analyze.py --host localhost --port 8123
+"""
+
+import argparse
+import sys
+
+
+def fmt_pct(n, total):
+    return f"{n/total*100:.1f}%" if total else "0.0%"
+
+
+def print_table(headers, rows, col_widths=None):
+    """Print a simple formatted table."""
+    if not col_widths:
+        col_widths = [max(len(str(h)), *(len(str(r[i])) for r in rows))
+                      for i, h in enumerate(headers)]
+    # Header
+    hdr = "  ".join(str(h).ljust(w) for h, w in zip(headers, col_widths))
+    print(hdr)
+    print("-" * len(hdr))
+    for row in rows:
+        print("  ".join(str(c).ljust(w) for c, w in zip(row, col_widths)))
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Teen Patti Pattern Analysis CLI")
+    parser.add_argument("--host", default="localhost", help="ClickHouse host")
+    parser.add_argument("--port", type=int, default=8123, help="ClickHouse HTTP port")
+    args = parser.parse_args()
+
+    # Set config before importing db
+    from app import config
+    config.CLICKHOUSE_HOST = args.host
+    config.CLICKHOUSE_PORT = args.port
+
+    from app import db
+
+    print(f"Connecting to ClickHouse at {args.host}:{args.port}...")
+    try:
+        data = db.get_pattern_analysis()
+    except Exception as e:
+        print(f"Error: {e}", file=sys.stderr)
+        sys.exit(1)
+
+    total = data["chair_bias"]["total_games"]
+    print(f"\n{'='*60}")
+    print(f"  TEEN PATTI PATTERN ANALYSIS  ({total:,} games)")
+    print(f"{'='*60}\n")
+
+    # 1. Chair Win Bias
+    print("1. CHAIR WIN BIAS (expected 33.3%)")
+    cb = data["chair_bias"]
+    rows = []
+    for ch in ("A", "B", "C"):
+        d = cb[ch]
+        diff = d["pct"] - 33.3
+        sign = "+" if diff >= 0 else ""
+        rows.append([ch, f"{d['wins']:,}", f"{d['pct']:.1f}%", f"{sign}{diff:.1f}%"])
+    print_table(["Chair", "Wins", "Win %", "vs Expected"], rows)
+
+    # 2. Bet Rank Analysis
+    print("\n2. BET RANK ANALYSIS")
+    br = data["bet_rank"]
+    br_total = br["high"] + br["mid"] + br["low"]
+    rows = []
+    for rank in ("high", "mid", "low"):
+        rows.append([rank.capitalize(), f"{br[rank]:,}", fmt_pct(br[rank], br_total)])
+    print_table(["Rank", "Wins", "Win %"], rows)
+
+    # 3. Per-Chair Bet Rank
+    print("\n3. PER-CHAIR: HIGHEST BET WIN RATE")
+    print("   When chair X has the highest bet, how often does X win?")
+    pcr = data["per_chair_rank"]
+    rows = []
+    for ch in ("A", "B", "C"):
+        d = pcr.get(ch, {})
+        rows.append([ch, f"{d.get('has_highest', 0):,}",
+                      f"{d.get('wins', 0):,}", f"{d.get('win_pct', 0):.1f}%"])
+    print_table(["Chair", "Times Highest", "Wins", "Win %"], rows)
+
+    # 4. Hand Type Distribution by Chair
+    print("\n4. HAND TYPE DISTRIBUTION BY CHAIR")
+    htbc = data["hand_types_by_chair"]
+    hand_order = ["Trail", "Straight Flush", "Straight", "Flush", "Pair", "High Card"]
+    rows = []
+    for ht in hand_order:
+        a = htbc["A"].get(ht, 0)
+        b = htbc["B"].get(ht, 0)
+        c = htbc["C"].get(ht, 0)
+        if a + b + c == 0:
+            continue
+        rows.append([ht, f"{a:,}", f"{b:,}", f"{c:,}"])
+    print_table(["Hand Type", "Chair A", "Chair B", "Chair C"], rows)
+
+    # 5. Hand Type Win Rates
+    print("\n5. HAND TYPE WIN RATES")
+    htw = data["hand_type_wins"]
+    htw_total = sum(htw.values())
+    rows = []
+    for ht in hand_order:
+        v = htw.get(ht, 0)
+        if v == 0:
+            continue
+        rows.append([ht, f"{v:,}", fmt_pct(v, htw_total)])
+    print_table(["Hand Type", "Wins", "Win %"], rows)
+
+    # 6. Pot Size Buckets
+    print("\n6. WIN RATES BY POT SIZE")
+    pb = data["pot_buckets"]
+    ranges = pb.get("_ranges", {})
+    rows = []
+    for bucket in ("small", "medium", "large", "whale"):
+        d = pb.get(bucket)
+        if not d:
+            continue
+        t = d["total"] or 1
+        rows.append([
+            bucket.capitalize(), ranges.get(bucket, ""),
+            f"{d['total']:,}",
+            f"{d['A']/t*100:.1f}%", f"{d['B']/t*100:.1f}%", f"{d['C']/t*100:.1f}%",
+        ])
+    print_table(["Bucket", "Range", "Games", "A %", "B %", "C %"], rows)
+
+    # 7. Streaks
+    print("\n7. STREAK ANALYSIS")
+    streaks = data["streaks"]
+    rows = []
+    for ch in ("A", "B", "C"):
+        s = streaks[ch]
+        rows.append([ch, str(s["max_streak"]), str(s["current_streak"])])
+    print_table(["Chair", "Max Streak", "Current Streak"], rows)
+
+    # 8. Hourly Patterns
+    print("\n8. HOURLY PATTERNS (win % by hour)")
+    hourly = data["hourly"]
+    hours = sorted(hourly.keys(), key=lambda h: int(h))
+    rows = []
+    for h in hours:
+        d = hourly[h]
+        t = d["total"] or 1
+        rows.append([
+            f"{int(h):02d}:00", str(d["total"]),
+            f"{d['A']/t*100:.1f}%", f"{d['B']/t*100:.1f}%", f"{d['C']/t*100:.1f}%",
+        ])
+    print_table(["Hour", "Games", "A %", "B %", "C %"], rows)
+
+    # 9. Recent vs Overall
+    print("\n9. RECENT (LAST 100) vs ALL-TIME")
+    rva = data["recent_vs_all"]
+    for label, section in [("All-Time", rva["all"]), ("Last 100", rva["recent"])]:
+        t = section["total"] or 1
+        d = section["dist"]
+        parts = " | ".join(f"{ch}: {d[ch]:>4} ({d[ch]/t*100:.1f}%)" for ch in ("A", "B", "C"))
+        print(f"  {label:>10}  [{t:>5} games]  {parts}")
+
+    print(f"\n{'='*60}")
+    print("  Done.")
+
+
+if __name__ == "__main__":
+    main()