What a week of ChessIntel looks like
One kid, one pattern, eleven days. Every quote below is copied verbatim from his actual reports. The kid is Daniel, the founder's son and the first user; the games and the engine numbers are real.
Where it started — a loss
The arc begins with a tournament game Daniel lost, weeks earlier. No blunder to point at — which is exactly why a single-game engine review would have had nothing to say. His report did:
"From move 21 to 37 your queen wandered a7 → b6 → b7 → c6 → b6 with no target. No blunder — you just drifted, and a roughly equal position slowly became a clearly worse one. This is the exact thing online never tests, because online games don't stay quiet."
That became the tracked pattern — quiet positions, no plan — and a drill was assigned: practice games from his own positions, naming a plan before each move.
Day 1 — the drill isn't working yet
This is the part a brochure would skip. The report on his first practice games:
"The drill's purpose — learning to play with a plan in quiet-ish positions — isn't landing yet…"
Day 2 — movement
"The threat-naming shows real progress — he's thinking forward now… The pattern is improving."
Day 3 — it shows up in a real game
Same evening as the drills, a rated game against a 1772 opponent:
"The quiet-position planning we've been working on showed up when it mattered… The pattern is fading. Keep drilling."
And the next rated game after that:
"Clean, plan-driven chess from start to finish… Pattern 1 (quiet positions / no plan) did not fire. One game isn't graduation, but this is the second straight positive data point."
Day 6 — a failure, and the memory earns its keep
Progress isn't a straight line. Against an 1845 player, Daniel built a completely winning endgame — and didn't win it. The report doesn't soften it:
"You built a winning position out of a sharp Petrov and reached a completely won king-and-pawn endgame — then let it slip to a draw on the final moves… One move away from the full point."
A single-game engine review would call the draw a decent result and move on. The report, because it remembers, saw what the slip actually was:
"The quiet-position planning pattern we've been tracking didn't fire in the middlegame (good news), but it showed up in a new form: converting a technical endgame when there's no tactic, just technique."
The old pattern fading, and its next form caught in the same breath — that became the new drill.
Day 8 — over the board
A tournament game, digitized from the paper scoresheet:
"Daniel won a sharp Scotch Game, converting a back-rank mate… The middlegame planning pattern continues to hold."
Day 11 — the new normal
"You won a solid game against an 1836 opponent, and the middlegame planning pattern stayed quiet — no drift, no passive retreats."
What to take from this
Not that every kid fixes a habit in eleven days — arcs differ, and this pattern had already been worked on before this stretch. The point is the shape: a pattern named from real games, a drill from the kid's own positions, honest reports when it isn't working, evidence when it is, and a memory that notices the next thing before it costs points. A single-game engine review can't do any of that — it forgets each game the moment it ends.
Full disclosure: Daniel is the founder's son — he's the kid the product was built for. His reports are quoted with his (and his dad's) permission, opponents appear by rating only, and none of the quotes were edited beyond trimming.
No card needed. See how it works or read a full sample report.