PM Simulation Advisor

An AI advisor that analyzes your PM Simulation game state and recommends optimal next actions to maximize your win probability.

What It Does

This tool:

• ✅ Analyzes your game log to understand current state
• ✅ Tracks morale, noise, capital, releases, and team composition
• ✅ Recommends specific activities to check each turn
• ✅ Suggests marketing spend amounts
• ✅ Advises on HR decisions (hiring/firing)
• ✅ Follows optimal strategy to maximize: Capital + P&L + Morale + Low Noise

How to Use

Step 1: Run the Advisor

python3 pm_advisor.py

Step 2: Copy Your Log

1. Open your PM Simulation spreadsheet
2. Go to the Log sheet
3. Select ALL data in columns A through D (Timestamp, Turn, Key, Value)
4. Copy it (Cmd+C or Ctrl+C)

Step 3: Paste the Log

1. When the advisor prompts you, paste the log data
2. Press Enter
3. Type END on a new line
4. Press Enter again

Step 4: Follow Recommendations

The advisor will output:

• Current game state (morale, noise, turn, releases)
• Recommended activities to check in the Activities tab
• Marketing spend amount (if applicable)
• HR actions to take
• Reasoning for each recommendation

Step 5: Execute in Game

1. Go to the Activities tab in your spreadsheet
2. Check the boxes for recommended activities
3. Enter marketing spend if indicated
4. Go to HR tab and take hiring actions
5. Select relevant Ideas
6. Prioritize Backlog items
7. Click "Do Turn"

Step 6: Repeat

After each turn:

1. Copy the updated Log
2. Run the advisor again
3. Follow new recommendations

Strategy Overview

The advisor follows a four-phase optimal strategy:

Phase 1: Discovery & Research (Turns 1-4)

• Talk to customers about needs and solutions
• Buy market research
• Conduct surveys
• Generate ideas with low noise

Phase 2: Design & Validation (Turns 5-8)

• Build prototypes
• Test with users
• Conduct design sprints
• Hire initial team (2-3 people)

Phase 3: Build & Release (Turns 9-12)

• Plan iterations
• Execute sprints
• QA testing before releases
• First release by Turn 11-12
• Build team to 4-5 people

Phase 4: Growth & Optimization (Turns 13-18)

• Aggressive marketing campaigns
• SEO investment
• A/B testing
• Continue releasing features
• Maximize revenue

Key Metrics to Maximize

Win Condition: Complete 18 turns with capital > 0

Optimization Goals:

• Capital: Don't run out (game over)
• P&L: Maximize profit
• Morale: Keep > 65 (prevents quitting), aim for > 100
• Noise: Keep low (< 0 is great, < 25 is OK)
• Releases: At least 3-4 ideas released by end

Tips

• Always QA before releasing (reduces bugs and noise)
• Don't fire people after Turn 15 (game prevents it)
• Marketing only works after first release
• Low morale causes team members to quit
• High noise reduces the effectiveness of all outcomes
• Process activities (design sprints, prototypes, QA) reduce noise
• Customer activities (talk about needs/solutions) boost morale

Troubleshooting

"Could not parse log data"

• Make sure you copied ALL columns (A through D) from the Log sheet
• Include the header row

"No log data provided"

• Make sure to paste the log and then type END on a new line

Recommendations seem off

• The advisor analyzes based on the log
• Make sure the log is from your current game (not an old one)

Example Output

╔══════════════════════════════════════════════════════════════╗
║                     CURRENT GAME STATE                       ║
╚══════════════════════════════════════════════════════════════╝

Turn: 5/18
Morale: 87.0 ✓
Noise: -4.0 (Factor: 1.04) ✓
Headcount: 3
Released: No
Released Ideas: 0

╔══════════════════════════════════════════════════════════════╗
║                   RECOMMENDED NEXT ACTIONS                   ║
╚══════════════════════════════════════════════════════════════╝

PHASE: DESIGN & VALIDATION

ACTIVITIES TO CHECK:
  [1] Build prototype
  [2] Plan iteration
  [3] Conduct sprint

HR ACTIONS:
  • Hire another engineer or designer for capacity

REASONING:
  • Build prototype to test hypotheses with users
  • This reduces noise and generates feedback

---

Good luck achieving the highest possible win! 🚀