TL;DR — the July 16 board went 16 of 20 players UP. Every one of our ten high-confidence calls (60%+) upgraded — 10 for 10 — and the best single pick returned +552% in under a week. Buy the whole board, misses and all, and it still returned +37% in six days. Below: the winners, card by card. And the receipt behind the receipts — every prediction the model makes, all 619 of them, graded against SDS's own changelog on one public page.

---

▶ Watch every Diamond Radar short on YouTube — the boards below were called on camera, before the update dropped.

The high-confidence board went 10 for 10

In the week before the July 16 roster update we posted four Diamond Radar boards — 🎲 Flyers, 💸 Best Flips, 📈 Tier Jumps, and 🔒 Locks — each with the model's forecast for specific Live Series cards: the upgrade odds, the projected new overall, the projected price. When the update dropped at 12pm ET, we graded every call, exactly as posted, against SDS's own changelog — matched by exact card ID, no edits, no deletes.

Every single call we made at 60% odds or better upgraded. Ten calls, ten upgrades. Here's the tape.

💸 Best Flips — a perfect board, 5 for 5

Every card upgraded. Four of five beat the projected overall. Griffin was the flip of the week: 71 stubs at post time, 425 after the update — a +499% move on a card that cost less than a pack.

📈 Tier Jumps — the biggest returns of the week

Okert was the single best return of the entire week — 138 stubs to 900. Three cards flagged to cross a rarity line, three that crossed it, every one still profitable.

🎲 Flyers — even the long shots delivered

The Flyers board is where we post the coin-flips — cards we price below even odds, clearly labeled boom-or-bust. Two of them hit anyway, and one made the whole board:

When we posted Yordan at 35%, the comments said flat-out "no way he goes up." We didn't argue — we published a number. He went up two points, past our own projection, and the card gained +12,000 stubs a copy — the biggest raw-stub move of the week, off a call the room laughed at.

🔒 Locks — 2 for 2

The final board, hours before the drop, cleared only two cards — and both upgraded: Jhoan Duran (→85, DIAMOND ✓) and Justin Wrobleski (80→82, +125% at 424 → 952). We said lock, they locked.

Add it up: 16 of 20 players on the week went up, 10 of them at high confidence — a clean 10-for-10 on the calls we told you to trust.

The Duran story — the market graded us before SDS did

This one's the week's anchor:

1. Wednesday morning we flagged Jhoan Duran at 1,449 stubs, 72% to jump Gold→Diamond, with a projected fair price near 3,024.
2. By Thursday 5 AM — before the update even dropped — the market had already bought him to 2,914. The value gap we published had closed to nearly zero overnight.
3. Thursday noon: the update makes him a Diamond, exactly as forecast. Price: ~3,000 — within 1% of the number we published a day early.

The forecast didn't just predict the update. It predicted what the market would pay — 24 hours ahead of it.

The receipts, if you bought the board

Prices captured within hours of the update — an early snapshot, not the ceiling (upgrade repricing keeps drifting for days). One copy each, before tax:

• Steven Okert: +552% (138 → 900)
• Foster Griffin: +499% (71 → 425)
• Bryan Baker: +270% (334 → 1,235)
• Troy Melton: +225% (292 → 950)
• Justin Wrobleski: +125% (424 → 952)
• Jhoan Duran: +107% (1,449 → 3,000)
• Yordan Alvarez: +69% (17,552 → 29,698) — the biggest raw-stub gain of the week

And the honest gut-check: buy every card on the board — one copy each, winners and the few that didn't move — and the whole basket still returned +37% in under a week. The tier-jumpers alone returned +175%. You didn't need to cherry-pick. You needed the board.

---

Why you can trust the streak — the whole model is public

Anybody can screenshot fifteen winners. So here's the part that makes the boards above mean something: every prediction Diamond Radar makes — not the handful we put on YouTube, but all of them, every card, every tier — is graded against SDS's changelog and posted on a public page. No shortlist, no quiet deletes. The receipt behind the receipts.

The whole model, on one public page: every prediction graded against SDS’s changelog — 619 calls, 64.8% precision, all of it public and auditable.

Across the full season so far, the ledger reads:

Two in three, across six hundred–plus calls we can't take back — that's the number the high-confidence boards are drawn from. And where it counts most, the high-confidence bucket is the sharp one: the calls DR stamps at 75%+ are its most reliable signal, which is exactly why the July board's ten high-confidence picks went a perfect ten-for-ten.

We're sharpest where you actually shop, too. By rarity, precision runs 73% on Bronze, 64% on Silver, 61% on Common — the tiers most players are flipping day to day. (Diamonds are thinner and harder to call, and the page says so — we'd rather you know than guess.) The point of putting the whole thing out there, including the windows we'd rather forget, is simple: a forecast you can audit is worth trusting. One you can't is a horoscope.

What's next

The Radar resets tonight and starts pricing the next update window, and the model sharpens with every window it grades — each roster update is a fresh answer key that feeds the next forecast. Every upgrade week from here ends the same way: the board, the receipts, and the full public scoreboard, graded against SDS's own changelog.

See the whole record — live, every window — at /radar/track-record. Then check the live odds, value gaps, and OpScores for every card at diamondops.gg.

---

Diamond Radar is a forecast model — forecasts, not guarantees. Markets move: stub sales, flash sales, and demand swings can shift prices at any time. Nothing here is financial advice.