NYT Crossword Stats

Fetch your NYT Crossword Puzzle solve stats and export them as CSV.

The NYT app shows minimal information around your streaks and average solve times. This gets you the raw data so you can do your own analysis.

What's a GitHub?

Not well-versed in Python or the command-line? Consider signing up for XW Stats. It's a free site that automatically fetches your solves and allows you to download them as a CSV.

Requirements

1. Python
2. A NYT Crossword subscription

Installation

1. Clone this repository
2. pip install -r requirements.txt

Usage

Fetch all solve stats since January 1, 2019. Use your NYT email and passwords as arguments here

python fetch_puzzle_stats.py -u [email protected] -p yourpass -s 2019-01-01

Options

Follow the command python fetch_puzzle_stats.py with these options if you want to:

• -u or --username The email address for your NYT account
• -p or --password The password for your NYT account
• -s or --start_date The date to start fetching stats from (default: 30 days ago)
• -e or --end_date The date to stop fetching stats from (default: today)
• -o or --output The name of the CSV file to output (default: data.csv)
• -t or --type The type of puzzle to fetch. One of "daily", "mini", or "bonus" (default: Daily)

Login Issues

If you are experiencing 403 errors and you are sure your username and password are correct, you may need to provide your login cookie manually. It's a two-step process:

1. Follow the instruction steps here to get your cookie.
2. Rename the file .env.template to .env and paste your cookie into it like this:

NYT_COOKIE=1ELEHES...yDo40

• If your e-mail or password have some unusual characters, be sure to escape them properly, or wrap the password in quotes (' or ").

The resulting CSV file (data.csv by default, override with -o flag) has your solve stats.

Data Format

Example CSV:

(my real stats...don't judge me you pros out there...)

author,editor,format_type,print_date,day_of_week_name,day_of_week_integer,publish_type,puzzle_id,title,version,percent_filled,solved,star,solving_seconds
Will Nediger,Will Shortz,Normal,2023-12-23,Saturday,6,Daily,21560,,0,100,True,Gold,675
Drew Schmenner,Will Shortz,Normal,2023-12-24,Sunday,0,Daily,21568,Wrap Stars,0,100,True,Gold,1449
Amie Walker,Will Shortz,Normal,2023-12-25,Monday,1,Daily,21567,,0,100,True,Gold,243

author	editor	format_type	print_date	day_of_week_name	day_of_week_integer	publish_type	puzzle_id	title	version	percent_filled	solved	star	solving_seconds
Will Nediger	Will Shortz	Normal	2023-12-23	Saturday	6	Daily	21560		0	100	True	Gold	675
Drew Schmenner	Will Shortz	Normal	2023-12-24	Sunday	0	Daily	21568	Wrap Stars	0	100	True	Gold	1449
Amie Walker	Will Shortz	Normal	2023-12-25	Monday	1	Daily	21567		0	100	True	Gold	243

Fields in CSV:

• author - The author(s) of the puzzle
• editor - Almost always Will Shortz
• format_type - Almost always "Normal"
• print_date - Date the puzzle was published
• day_of_week_name - Pretty name of the puzzle's day
• day_of_week_integer - Integer value, Sunday = 0, Monday = 1 ...
• publish_type - Normal, Mini, Bonus
• puzzle_id - NYT's reference ID for the puzzle
• title - The puzzle's title (for Sundays)
• version - NYT's version of the puzzle
• solved - Whether the puzzle was solved
• star - Gold if solved before midnight Pacific of its date (value available starting 2018-12-12 )
• solving_seconds - how long it took to solve the puzzle

Example

For an example of how to use this data, I plotted my solve times over the last ~8 months, grouped by day: