What did the desk's first two weeks look like?

I started Steezanomics with one goal: automate the system that made me about +$74k betting on sportsbook promos, and run it in public where nobody has to take my word for anything. June 24 was day one of the public desk. This is what the first two weeks actually looked like, straight from the settlement pipeline, with nothing cleaned up.

This post is a frozen snapshot as of July 6, 2026. The live ledger keeps moving every morning, which is exactly why this is dated: you can hold the two against each other any time you want.

The headline numbers

Since June 24 the desk has graded 179 plays: 53 wins, 124 losses, 2 pushes. Total profit +$2,699.04 on $9,942 staked, a +27.1% ROI at an average stake of $56.

A 53-124 record reads like a disaster. Here’s the missing context: these are plus-money plays, averaging +434 odds. A single bet at +434 breaks even at 18.7%, but a real portfolio’s bar has to be computed play by play, because every slip carries different odds. Do that across all 177 decided plays and the average play needed to win 22.3% of the time, or 25.2% once you weight by stake. The desk hit 29.9%.

That gap is where all the profit comes from, and it’s why win rate alone tells you nothing about whether a bettor is making money.

Every slip type is green so far

Two things worth noticing, and one warning.

First, parlays, the bet type famous for losing money, are at +23.6% ROI here while hitting only 18.8% of the time. That’s the boost math working: these aren’t parlays at standard prices, they’re parlays where a promo pushed the payout past fair value.

Second, SGPs lead at +44.7% ROI on a 44.4% hit rate, carried by boosted prices on short slips.

The warning: these are two-week samples sliced four ways, so the per-type numbers are noisy and will move. What matters is the direction of the logic, not the third decimal.

The variance, shown instead of hidden

Twelve betting days so far: 9 green, 3 red. Best day +$897.63 (June 26). Worst day -$497.00 (June 25). Those two days were back to back, which is what betting at plus-money averages feels like in real life.

June closed at +40.6% ROI. July so far is at +4.8%. That spread isn’t a strategy change, it’s what small samples do, and it’s exactly why the desk publishes everything instead of screenshotting the good weeks.

How to read all of this

The method behind these plays is boring on purpose: fair value from multi-book market consensus, a two-sided devig, and an EV bar a promo has to clear before anything posts. No predictions, no gut calls. The about page explains it in plain terms.

Two weeks proves very little on its own. The reason this post exists is that it’s dated, it’s checkable against a ledger that updates every morning, and the next snapshot will be compared to it in public, whichever direction the numbers go.