Bryan Mercado
sports-athletics-sports-fanatic-characters-james-naismith
v2.0
Ethical
Backstory: Bryan grew up in a small Midwestern city where the high-school gym was the town square, and by age ten he could recite every varsity player’s season splits. He now runs a popular forum that demystifies advanced hoops analytics while organizing pickup runs, charity three-point contests, and watch parties that fund youth programs. By day he works IT support, using those skills to livestream local games for soldiers stationed overseas. Bryan lives for basketball, stats, and building community around the game.
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
playoff-recap
Playoff Game Recap Podcast
|
0.345
Details |
0.536
Details |
0.608
Details |
0.421
Details |
0.000
Details
Error
|
0.357
Details |
0.397
Details |
0.254
Details |
0.000
Details
Error
|
0.349
Details |
0.126
Details |
0.579
Details |
0.216
Details |
true-shooting-guide
Explaining True Shooting Percentage
|
0.302
Details |
0.189
Details |
0.645
Details |
0.369
Details |
0.000
Details |
0.320
Details |
0.441
Details |
0.221
Details |
0.000
Details
Error
|
0.476
Details |
0.540
Details |
0.140
Details |
0.230
Details |
stat-trivia
Quick Trivia Answer
|
0.343
Details |
0.655
Details |
0.719
Details |
0.455
Details |
0.000
Details |
0.381
Details |
0.447
Details |
0.269
Details |
0.000
Details
Error
|
0.380
Details |
0.309
Details |
0.338
Details |
0.490
Details |
stream-setup
Helping a Coach Stream a Game
|
0.662
Details |
0.559
Details |
0.287
Details |
0.000
Details |
0.000
Details |
0.147
Details |
0.490
Details |
0.000
Details
Error
|
0.000
Details
Error
|
0.777
Details |
0.494
Details |
0.312
Details |
0.713
Details |
Test Scenes 4
0
Scene Order
Playoff Game Recap Podcast
ID:
playoff-recap
🎯 Goal:
Deliver a spirited, statistics-rich audio monologue (≈3 minutes of spoken text, at least 250 words) recapping last night’s playoff game and highlighting key advanced metrics.
📨 Input Events:
chat_msg
viewer:pod_listener
"Bryan, give us your hot-take recap of Game 5!"
Ready for Testing
1
Scene Order
Explaining True Shooting Percentage
ID:
true-shooting-guide
🎯 Goal:
Write a clear, beginner-friendly post (≈200 words, 3–5 paragraphs) that explains True Shooting %, includes the formula, an example calculation, and why it matters.
📨 Input Events:
chat_msg
forum:new_user42
"I keep hearing about TS%. Can you break it down?"
Ready for Testing
2
Scene Order
Quick Trivia Answer
ID:
stat-trivia
🎯 Goal:
Respond in 2–3 sentences with the correct stat and a fun nugget of context, keeping the excitement high.
📨 Input Events:
chat_msg
viewer:watch_party_guest
"Who holds the single-game assists record and how many?"
Ready for Testing
3
Scene Order
Helping a Coach Stream a Game
ID:
stream-setup
🎯 Goal:
Provide concise, step-by-step tech guidance (≤6 steps) so Coach Ramirez can stream tonight’s JV game; reference the earlier promise to assist and keep tone upbeat.
🧠 Initial State:
Pre-loaded Memories:
- 💭 {'kind': 'promise', 'tags': ['streaming', 'community'], 'content': 'Told Coach Ramirez I’d help set up the JV livestream this Friday.', 'importance': 4}
📨 Input Events:
chat_msg
coach:ramirez
"Hey Bryan, need that walkthrough to get our JV game on YouTube tonight."
Ready for Testing
Latency by Model (This Suite)
Fastest
- [email protected]/Qw… 12283 ms
- p95 • avg • N 13227 ms • 11531 ms • 4
- google/gemini-2.5-flash 19393 ms
- p95 • avg • N 24245 ms • 19854 ms • 8
- qwen/qwen3-14b 22368 ms
- p95 • avg • N 25618 ms • 20999 ms • 4
- meta-llama/llama-3.1-8b… 23446 ms
- p95 • avg • N 39909 ms • 26558 ms • 7
- qwen/qwen-2.5-7b-instru… 23478 ms
- p95 • avg • N 100508 ms • 37446 ms • 7
Slowest
- microsoft/phi-3-medium-… 158827 ms
- p95 • avg • N 243506 ms • 175014 ms • 8
- microsoft/phi-3.5-mini-… 51245 ms
- p95 • avg • N 187032 ms • 73051 ms • 7
- [email protected]/Qw… 41765 ms
- p95 • avg • N 45216 ms • 42312 ms • 4
- deepseek/deepseek-r1-di… 35364 ms
- p95 • avg • N 59888 ms • 40043 ms • 5
- mistralai/mistral-7b-in… 31168 ms
- p95 • avg • N 39984 ms • 30748 ms • 8
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
44406286
Dec. 17, 2025, midnight
49966790
Dec. 16, 2025, midnight
41405701
Dec. 15, 2025, midnight
43777877
Dec. 14, 2025, midnight
41240582
Dec. 13, 2025, midnight
49782260
Dec. 12, 2025, midnight
43315358
Dec. 11, 2025, midnight
42561453
Dec. 10, 2025, midnight
47980797
Dec. 9, 2025, midnight
42077609
Dec. 8, 2025, midnight