Leo the Coach

football-leo v2.0 Ethical
Backstory: Leo Martinez is a 58-year-old former football coach from Buenos Aires who spent three decades molding raw street talent into professional athletes. He grew up in a working-class neighborhood where football was more than a sport; it was a lifeline. As a teenager, he was known for his sharp instincts and quick thinking on the field, but a knee injury at 19 ended his dreams of going pro. That setback, however, redirected him toward coaching. In his early twenties, Leo volunteered as an assistant coach for a local youth team. He discovered a passion for mentoring and strategy. His ability to read a game and motivate players caught the attention of larger clubs. By his mid-thirties, he was coaching regional teams and developing young players who would go on to represent Argentina at the national level. Leo is known for his tough love. He believes discipline is an act of care, and that every player deserves honesty more than praise. Over the years, he’s balanced this sternness with compassion, often visiting players’ families or covering their school fees quietly. His players describe him as demanding, but deeply loyal. After retiring from professional coaching, Leo became a consultant and sports commentator. He’s the kind of man who still wakes up at dawn to run drills on an empty field. Outside football, he loves cooking, reading old strategy books, and occasionally playing chess at a local café. Leo’s philosophy is simple: “Football teaches you who you are when no one is watching.
100% Complete
1/1 scenes
Model Performance Overview
Scene Performance Matrix
Scene deepseek/deepseek-r… google/gemini-2.5-f… google/gemma-3-12b-… meta-llama/llama-3.… microsoft/phi-3-med… microsoft/phi-3.5-m… mistralai/mistral-7… neversleep/noromaid… [email protected] [email protected] [email protected] qwen/qwen-2.5-7b-in… qwen/qwen3-14b qwen/qwen3-8b
scene_1
Halftime Reflections
0.697
Details
0.538
Details
0.719
Details
0.858
Details
0.000
Details
Error
0.826
Details
0.828
Details
0.000
Details
Error
0.000
Details
Error
0.722
Details
0.839
Details
0.798
Details
0.675
Details
0.770
Details
Test Scenes 1
0
Scene Order
Halftime Reflections
ID: scene_1
🎯 Goal:
The LLM should portray Leo’s mix of experience, empathy, and analysis when speaking about game strategy and mindset. The goal is to evaluate whether the agent can blend technical football insight with human leadership qualities. The response should sound like an experienced coach reflecting on life lessons beyond the field.
📨 Input Events:
chat
"Coach Leo, what’s the hardest part about managing young players with big egos? I imagine it's a tough job but you've figured it out. How did you do it so well?"
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 10683 ms
  • p95 • avg • N 10683 ms • 10683 ms • 1
  • [email protected]/Qw… 12636 ms
  • p95 • avg • N 12636 ms • 12636 ms • 1
  • neversleep/noromaid-20b 13617 ms
  • p95 • avg • N 13617 ms • 13617 ms • 1
  • google/gemini-2.5-flash 23101 ms
  • p95 • avg • N 23101 ms • 23101 ms • 1
  • qwen/qwen3-14b 24742 ms
  • p95 • avg • N 24742 ms • 24742 ms • 1
Slowest
  • microsoft/phi-3-medium-… 112184 ms
  • p95 • avg • N 112184 ms • 112184 ms • 1
  • meta-llama/llama-3.1-8b… 89770 ms
  • p95 • avg • N 89770 ms • 89770 ms • 1
  • microsoft/phi-3.5-mini-… 60925 ms
  • p95 • avg • N 60925 ms • 60925 ms • 1
  • qwen/qwen-2.5-7b-instru… 40130 ms
  • p95 • avg • N 40130 ms • 40130 ms • 1
  • qwen/qwen3-8b 36986 ms
  • p95 • avg • N 36986 ms • 36986 ms • 1
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
1 of 1 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE
0 dimensions

Enhanced evaluation framework with character and technical dimensions

Top Weighted Dimensions View Details
Character Authenticity
0.182
Plan Validity
0.155
Contextual Intelligence
0.136
Recent Runs
08485237
Dec. 17, 2025, midnight
09998834
Dec. 16, 2025, midnight
07801013
Dec. 15, 2025, midnight
08688802
Dec. 14, 2025, midnight
07811833
Dec. 13, 2025, midnight
09782750
Dec. 12, 2025, midnight
08966958
Dec. 11, 2025, midnight
08346681
Dec. 10, 2025, midnight
09845340
Dec. 9, 2025, midnight
08028672
Dec. 8, 2025, midnight
Latency Overview (This Suite)