An 8×8 board, top-left corner. The teal region started as a 2×3 block covering rows 1 and 2, columns 1 through 3. A queen you placed over on the right side has since wiped out all of row 1, so teal is down to three cells in row 2. Your finger is already hovering over row 2, column 1, because the corner feels right.
Stop. That placement is the whole problem, and it has four cousins.
You already know the rules you're breaking. One queen per row, one per column, one per color region, no two touching. Nobody gets stuck on Queens because rule four slipped their mind. You get stuck because under mild time pressure you skip checks you know how to run. The gap is in discipline rather than knowledge, which is good news: discipline is fixable, and each fix below is one habit.
Mistake one: placing a queen when you've only proved a line
Back to that corner. What you actually proved is that teal's queen lives somewhere in row 2, columns 1 through 3. You did not prove which cell. But the proof you do have is worth plenty: if teal's queen is in row 2, then row 2 is teal's row, and every cell in it from column 4 to column 8 is dead. Cross them off. Five eliminations, and one of them usually strangles a neighboring region into its own forced move.
The mistake is converting "this region got small" into "place a queen" when the position only supports "claim a line." A region squeezed into one row or column hands you the line for free. The exact cell comes later, once those eliminations do their work elsewhere.
The check you skipped takes three seconds: do all of this region's remaining cells share a row or column? If yes, mark the rest of that line before you even think about placing. This line-claiming move is the backbone of our guide to solving Queens without guessing.
Mistake two: doing eliminations in your head
Medium board, five minutes in. You reason it out cleanly: the orange region can't take row 5, because if it did, the yellow region beside it would lose its last cell. Nice deduction. You don't mark it. Three moves later you're seriously weighing a queen at row 5, column 7 — inside orange — and the earlier logic is just gone.
Working memory holds maybe three of these before one leaks. Hard boards need a dozen. The fix isn't concentration, it's marks: the moment you deduce a cell is impossible, tap it once and X it out.
Our boards have an auto-X setting that crosses off every square your placed queens rule out. Use it, but understand what it does: auto-X marks the consequences of queens already on the board. Eliminations you derive by pure reasoning, before any queen exists, are invisible to it. Those are exactly the ones to write down, because those are the ones you'll forget.
If you play Star Battle too, you've met this failure under a different name. It's one of the classic Star Battle mistakes, where two stars per row makes unmarked reasoning collapse even faster.
Mistake three: only scanning in one direction
Everyone scans regions. You look at each color, ask where its queen can still go, and act on the small ones. Almost nobody runs the reverse scan, and the reverse scan is where hard boards break open.
Concrete version, from a 9×9 hard board: after a round of eliminations, look down column 6. The only open cells left in it are rows 5, 6, and 7, and all three are purple. So column 6's queen, whoever she turns out to be, is purple's queen — which means purple's cells outside column 6 (say it also owned two cells over in column 7) are dead, even though you have no idea which of the three candidates wins.
Region-to-line and line-to-region are mirror images. Our step-by-step solutions treat them as two separate moves because solvers miss the second one constantly. The habit: every time your region scan stalls, walk the rows and columns once each and ask whether that line's survivors are all one color. It fires more often than you'd expect, and precisely on the boards where you're stuck.
Mistake four: waiting for certainty a pair already gives you
Green is down to two cells sitting side by side: row 4, columns 2 and 3. You don't know which one gets the queen, so you shrug and look elsewhere. That shrug throws away four guaranteed eliminations.
Run both futures. If the queen lands on (4,2), the no-touching rule kills the cells around it, including (3,2), (3,3), (5,2), and (5,3). If she lands on (4,3) instead, it kills (3,2), (3,3), (5,2), and (5,3) again. Every cell directly above and below a side-by-side pair dies no matter which cell wins, because each one touches both candidates. Cross off all four now. (Row 4 outside the pair is also gone, but that's mistake one's line-claim doing its usual job.)
The general habit is either/or reasoning: when a region is down to two cells, ask what both futures agree on, and bank that. Certainty about the queen can wait; certainty about her neighbors is already here.
Mistake five: using Check as a coin flip
Our boards have a Check button. It tells you whether the queens you've placed so far are right, and nothing more. It also keeps a running "checks used" count under the board, and that number doesn't lie. If you're pressing Check to find out whether a placement was correct, you didn't deduce that placement. You guessed, and you're asking the button to do the logic for you.
The obvious counterargument: you finished the board, who cares how. Fair, once. And LinkedIn's one-a-day format makes guessing feel nearly free — low stakes, undo a tap away, a clean slate tomorrow. That cheapness is the trap. A guessed solve skips the deduction the board was built to teach, which is why the hard boards stay hard month after month.
Two replacements. First, treat the urge to press Check as an alarm. It means there's a deduction on the board you haven't found, so run the scans from mistakes one and three before touching the button. Second, do post-mortems. Every daily board we publish has a step-by-step solution showing the pure-logic line from empty grid to final queen. When you finish a board you're not proud of, read the day's solve and find the exact step you missed. It's usually one of the five in this article, and seeing which one is how the leak actually gets fixed.
Make it a routine, not a resolution
None of this needs to be memorized under pressure. It needs reps. Today's Easy board takes a few minutes and is exactly the right size for drilling one habit at a time: claim lines before placing, mark every deduction, run the reverse scan once, milk your pairs, and leave Check alone. Do that on the Easy, then carry it into the Medium and Hard.
A week of that and "stuck" starts meaning something different. Not "I don't know what to do" — "I haven't run my checklist yet." The second one always has an exit.