Tara Maxwell
sports-athletics-coach-characters-james-naismith
v2.0
Ethical
Backstory: Tara Maxwell is a mid-career adaptive sports coach at a regional training center in Austin, Texas. A former Paralympian with a kinesiology degree, she designs individualized strength and conditioning programs for athletes with varied physical abilities. Off the court, she volunteers in community outreach that introduces wheelchair basketball and seated volleyball to youth, championing inclusion and mental well-being alongside performance.
100% Complete
4/4 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]… | qwen/qwen-2.5-7b-in… | qwen/qwen3-14b | qwen/qwen3-8b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
welcome-huddle
Welcoming a New Athlete
|
0.620
Details |
0.835
Details |
0.721
Details |
0.669
Details |
0.000
Details
Error
|
0.775
Details |
0.810
Details |
0.672
Details |
0.000
Details
Error
|
0.752
Details |
0.000
Details |
0.666
Details |
0.721
Details |
plan-customization
Adapting a Strength Plan
|
0.433
Details |
0.697
Details |
0.570
Details |
0.746
Details |
0.045
Details |
0.544
Details |
0.755
Details |
0.395
Details |
0.000
Details
Error
|
0.692
Details |
0.455
Details |
0.695
Details |
0.646
Details |
weekly-newsletter
Weekly Training Newsletter
|
0.479
Details |
0.533
Details |
0.380
Details |
0.448
Details |
0.000
Details |
0.319
Details |
0.552
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.351
Details |
0.237
Details |
0.482
Details |
0.389
Details |
podcast-segment
Podcast on Mental Resilience
|
0.345
Details |
0.564
Details |
0.563
Details |
0.000
Details |
0.000
Details |
0.398
Details |
0.234
Details |
0.482
Details |
0.000
Details
Error
|
0.330
Details |
0.348
Details |
0.340
Details |
0.438
Details |
Test Scenes 4
0
Scene Order
Welcoming a New Athlete
ID:
welcome-huddle
🎯 Goal:
Offer an inclusive, motivational greeting and ask at least one question about the athlete’s goals; keep response under 120 words.
📨 Input Events:
chat_msg
athlete:Jordan
"Hi Coach Tara, I'm new here and excited to start training."
Ready for Testing
1
Scene Order
Adapting a Strength Plan
ID:
plan-customization
🎯 Goal:
Recommend three specific, safe strength exercises adapted for limited hand grip and explain why each helps, in under 150 words.
📨 Input Events:
chat_msg
athlete:Maya
"Coach, I have limited grip strength in my right hand. How should we adjust my program?"
Ready for Testing
2
Scene Order
Weekly Training Newsletter
ID:
weekly-newsletter
🎯 Goal:
Write a 250–300-word weekly email newsletter that recaps training highlights, shares one motivational quote, and lists next week’s schedule in bullet points.
📨 Input Events:
chat_msg
program_coordinator:Sam
"Tara, could you draft the weekly athlete newsletter?"
Ready for Testing
3
Scene Order
Podcast on Mental Resilience
ID:
podcast-segment
🎯 Goal:
Deliver a 400–500-word podcast monologue on maintaining mental resilience during plateau phases, weaving in one personal anecdote from your Paralympic days and one actionable mindset exercise.
📨 Input Events:
chat_msg
podcast_host:Riley
"We're recording now. Please give the audience a segment on working through training plateaus."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 12084 ms
- p95 • avg • N 15099 ms • 12111 ms • 4
- google/gemini-2.5-flash 20311 ms
- p95 • avg • N 24127 ms • 20319 ms • 8
- qwen/qwen-2.5-7b-instru… 21485 ms
- p95 • avg • N 100718 ms • 36246 ms • 7
- qwen/qwen3-14b 22045 ms
- p95 • avg • N 36479 ms • 24928 ms • 6
- mistralai/mistral-7b-in… 23636 ms
- p95 • avg • N 30808 ms • 24769 ms • 8
Slowest
- microsoft/phi-3-medium-… 207765 ms
- p95 • avg • N 215520 ms • 183605 ms • 8
- microsoft/phi-3.5-mini-… 47815 ms
- p95 • avg • N 196327 ms • 82754 ms • 6
- [email protected]/Qw… 40960 ms
- p95 • avg • N 49821 ms • 43101 ms • 4
- deepseek/deepseek-r1-di… 31219 ms
- p95 • avg • N 35945 ms • 31658 ms • 8
- google/gemma-3-12b-it 28316 ms
- p95 • avg • N 45859 ms • 30403 ms • 6
Per-scene duration for this suite.
Suite Actions
Completion Progress
100%
4 of 4 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE0 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
43540664
Dec. 17, 2025, midnight
49191327
Dec. 16, 2025, midnight
40552796
Dec. 15, 2025, midnight
42959334
Dec. 14, 2025, midnight
40389060
Dec. 13, 2025, midnight
48827594
Dec. 12, 2025, midnight
42468341
Dec. 11, 2025, midnight
41725553
Dec. 10, 2025, midnight
47122803
Dec. 9, 2025, midnight
41260610
Dec. 8, 2025, midnight