Calibration Challenge — submit 10 posts (5 winners, 5 flops) and we'll score them blind

[CaptainFredric]

ContentForge is built on a premise: deterministic heuristics can predict content performance better than gut feel. Before we claim that publicly on Product Hunt, we want to prove it.

Here's the challenge:

Submit 10 historical posts — 5 that performed well, 5 that flopped — without telling us which is which. We'll run them through the scoring engine and rank them. Then you tell us if we got it right.

---

How to participate

Copy this template into a comment:

Platform: [Twitter/LinkedIn/Instagram/TikTok/etc.]

Post 1: [paste text]
Post 2: [paste text]
Post 3: [paste text]
Post 4: [paste text]
Post 5: [paste text]
Post 6: [paste text]
Post 7: [paste text]
Post 8: [paste text]
Post 9: [paste text]
Post 10: [paste text]

Actual result: [e.g. "posts 2, 5, 7, 8, 10 were the top performers"]

DMs welcome if you'd rather keep your content private — results reported without attribution.

---

What you get

• Complete score breakdown for all 10 posts: score, grade, quality_gate (PASSED/REVIEW/FAILED), and itemized deductions
• If we rank your top 5 correctly: 500 free API credits, no expiry
• If we get it wrong: you've just identified a miscalibrated signal — that's a direct improvement to the engine for everyone

---

Why public calibration

The weights are based on platform best-practice documentation. Real engagement data is the ground truth. We'd rather find the gaps now than after PH launch.

Every submission goes into /docs/validation.md. Results are posted openly.

---

Running until we hit 20 submissions across ≥3 platforms. Questions? Drop them below.

[CaptainFredric]

Submit Here!