I’ve attempted every New York Times Crossword since January 2025 with the goal of a 🌟Gold Star🌟 solve (no hints, same-day solve, no internet) on each puzzle. The only outside help I accept is from my loving wife, Kayla, who has agreed to the Gold Star rules above.
See if you can guess the section titles along the way.
Puzzles Attempted: 562
Gold Star Solves: 461
Gold Rate: 82.03%
Golds per Week: 5.74
Progress at a glance.
Most Consecutive Golds: 72 puzzles in a row, ended by a Wednesday puzzle on 2025-09-17.
I’ve had 61 individual gold runs with an average length of 7.56 puzzles and a median length of 5 puzzles. Pretty indicative of golding Monday-Friday most weeks with a handful of long runs to inflate the mean.
Breakdown of misery:
| Ended By | Count |
|---|---|
| Saturday | 18 |
| Sunday | 17 |
| Friday | 13 |
| Thursday | 6 |
| Wednesday | 6 |
The weekend puzzles are big disruptors.
I track 3 levels of solve outcomes:
● Gold - Same-day solves with no hints
● Blue - Hints/checks used or next day solves
● Grey - Not close to solving
Both blue and grey results are counted as full failures for the purposes of calculating gold solve percentages, but blue solves indicate that a gold was usually a few missed letters away.
Rolling n-Day Gold Solve Percentages:
Each point represents my puzzle result for a given day.
Getting set in my ways.
The NYT does not purposely make crosswords easier or harder during specific times of the year.
The NYT varies the difficulty of crosswords by the day of the week, with Monday being the easiest and Saturday being the hardest. Sunday puzzles are larger (a 21x21 grid compared to 15x15) and are meant to be approximately the difficulty of the Wednesday or Thursday, per the NYT (How to Play). The Friday and Saturday puzzles are the only days that do not feature a theme, which adds difficulty by removing a common thread through 4-5 long answers. The Thursday and Sunday puzzles are the only days that regularly feature rebus squares.
| Day | Puzzles | Golds | GoldPct |
|---|---|---|---|
| Sun | 80 | 43 | 53.75 |
| Mon | 80 | 80 | 100.00 |
| Tue | 80 | 80 | 100.00 |
| Wed | 81 | 75 | 92.59 |
| Thu | 81 | 73 | 90.12 |
| Fri | 80 | 62 | 77.50 |
| Sat | 80 | 48 | 60.00 |
While many squares are blacked out in the final puzzle, the Sunday grid has 441 potential squares which is roughly double the 225 squares in a regular weekday puzzle. A gold solve on Sunday therefore requires roughly double the effort of a normal-sized puzzle. If my solve rate on the Sunday is 53.75% and you assume that this is actually the effort of solving two smaller puzzles in a row, then one estimate for the 15x15 difficulty comparison would be the day that has a gold rate closest to (53.75%)1/2 = 73.31%. For me, that is Friday (77.5%).
On the other hand, a Sunday isn’t really the independent effort of two smaller puzzles since the theme doesn’t “reset” halfway through the puzzle. Not fully understanding the theme leaves a lot of possibilities for missed letters on long answers.
It’s my opinion that the Sunday should be marketed as roughly the difficulty of a Thursday/Friday puzzle (harder clues but with a theme).
Rolling 16-Puzzle Gold Percentages by day of week:
For a particular day of the week, how does my solve rate change over time?
Monday/Tuesday plots omitted for being trivial.
How did my day-level solve percentages come to be?
Overall performance by year.
| Year | Puzzles | GoldPct | BluePct | GreyPct |
|---|---|---|---|---|
| 2025 | 365 | 80.55 | 8.77 | 10.68 |
| 2026 | 197 | 84.77 | 10.15 | 5.08 |
Gold percentage for each day by year:
| Day | 2025 | 2026 |
|---|---|---|
| Sun | 53.8 | 53.6 |
| Mon | 100.0 | 100.0 |
| Tue | 100.0 | 100.0 |
| Wed | 88.7 | 100.0 |
| Thu | 86.5 | 96.6 |
| Fri | 76.9 | 78.6 |
| Sat | 57.7 | 64.3 |
Across: OVERALLSTATS
Down: CALENDAR
Across: STREAKING
Down: RUNNINGAVERAGE
Across: MONTHLY
Down: DAY
Across: DEVELOPMENT
Bonus: CROSSWORDS