Sharp Report
How the Sharp Report is made
Every Sharp Report article is machine-written from our own model data and human-reviewed before it ships. Nothing hidden — here's the machine.
1. Our models generate the numbers
The same prediction models behind our public, automatically-graded track record produce the win probabilities, projected totals, and market comparisons in every article. The models are graded in the open — wins, losses, and pushes — so the numbers an article cites carry a public report card. See our research on honest model win rates for how that grading works.
2. AI writing personas draft the articles
The Sharp Report bylines — Joe Woods, Cliff Wilcher, Erin Lowe, and Tara West — are AI writing personas: pen names our article generator uses so each analytical point of view stays consistent and recognizable. They are not real people, and we don't pretend otherwise. Each persona is a distinct voice pointed at the same live data: one argues from the probabilities, one from player availability and usage, one from how the betting market is moving, and one from verified team news and what it actually changes.
3. A human reviews before publishing
Every article is reviewed and edited by Jordan Hawkes, the founder of Sharper Points and the accountable human behind every byline. That's why each article carries Edited by Jordan Hawkes — if something is wrong, the responsibility is his, not a pen name's.
4. Every number is sourced
Articles link players to the box scores behind their stat lines and cite the model outputs they were written from. When an article makes a call, the follow-up gets graded in public like everything else we publish — see the track record.
Questions about the pipeline? Email support@sharperpoints.com.