b140d4d82a
- 313 new markdown files created - 30 relationships embedded - 313 entries indexed - State initialized with usage data
2.0 KiB
2.0 KiB
| id | type | title | tags | importance | confidence | created | updated | relations | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2937a962-4a6e-44fb-b7c7-7d472df7fc0e | workflow | Paper Dynasty: 2005 Live Series card update workflow (mid-season) |
|
0.7 | 0.8 | 2026-02-02T01:39:26.024355+00:00 | 2026-02-02T01:39:26.024355+00:00 |
|
Task: Update 2005 Live cardset through end of July (mid-season update)
Configuration (retrosheet_data.py):
- START_DATE: 20050403 (Opening Day)
- END_DATE: 20050731 (end of July)
- SEASON_PCT: 81/162 (0.5 for half season)
- CARDSET_ID: 27 (2005 Live)
- MIN_PA_VL/VR: 20/40 (live series thresholds)
Workflow Steps:
-
Generate Arm Ratings:
python generate_arm_ratings_csv.py --year 2005 --events data-input/retrosheet/retrosheets_events_2005.csv- Creates data-output/retrosheet_arm_ratings_2005.csv
- Used for accurate OF arm ratings from play-by-play data
-
Run Card Creation:
python retrosheet_data.py- Result: 383 batters, 373 pitchers posted
-
Verify Positions:
./scripts/check_positions.sh 27- Validates DH counts (<5 normal), outfield positions present
-
Update Scouting:
pd-cards scouting all && pd-cards scouting upload- CRITICAL: Always run for ALL cardsets (no --cardset-id filter)
- Scouting is unified view across all players
-
Upload to S3:
pd-cards upload s3 --cardset '2005 Live'- Fetches card images from API, uploads to S3
- Updates player.image URLs with cache-busting timestamps
- Result: 756/756 cards uploaded (16 minutes)
Results:
- Cards: 756 total (383 batters + 373 pitchers)
- Position validation: ✓ 4 DH, 133 OF with proper positions
- S3 URLs: https://paper-dynasty.s3.us-east-1.amazonaws.com/cards/cardset-027/player-{id}/battingcard.png?d={date}
Key Insight: After PostgreSQL migration, needed to fix missing templates before S3 upload would work.