It's been a while since Jeff Chen and I gave our last update but that doesn't mean we haven't been hard at work. We're not quite ready for our upcoming Big Announcement (although you'll see some hints in this post) but there is some news and I wanted to answer some questions we've been getting.
We're starting to roll out new features
Our Finder Page is becoming an increasingly popular way to look up words or clues. It provides unique features especially useful for constructors so it's no surprise that it now accounts for most of our web traffic. If you search for a word, you see all the clues used for that word (going back to 1942 now!) but if you search for a pattern, that is if you use the ? or * symbols, you see all the words we know that fit, including entries from variety puzzles, acrostics, and a large external dictionary. Pattern searches now sometimes include grayed-out results in the Shortz Era list. These are words that are identified as "bogus" or, more charitably, "thematic." They're unlikely to help you fill a grid. We show them anyway since you may be looking to see if a theme has been used before, or maybe you want to find a puzzle you remember, but we identify them as unlikely to be useful.
Identifying such entries is a manual process, of course, and it's just the beginning of how we are going to be able to help constructors think about their craft. We happen to have a successful and innovative constructor on staff (hi, Jeff!) and together we're going to help you build better crosswords. We still have work to do before the full feature set is ready to go live, but it's coming.
We now have every known NYT crossword going back to the very first one
You've probably already seen that David Steinberg and his army of volunteers known collectively as the Pre-Shortzian Puzzle Project have digitized and edited every known NYT daily crossword, and they're all available on XWord Info. You can view them or even solve them online, even without an XWord Info account. Who cares? It turns out they're fascinating. They reflect the culture of their time. The war-years ones are particularly interesting, although you and your fancy modern sensibilities will likely cringe now and then. Crosswords, like any art form, can be better understood by knowing a little about the history. We put the entire history from one major publisher at your fingertips.
We're getting a better picture of who the most prolific NYT constructors are
Before Will Shortz took over as editor, constructor bylines were usually missing. As pre-Shortz crosswords trickled in, David Steinberg and his team did a lot of work to identify authorship of these older puzzles, and we now have a better list of who the big guns really were. For years, our list was headed by Manny Nosowsky, Elizabeth C. Gorski, and Patrick Berry. Then we decided to count Variety puzzles so Emily Cox and Henry Rathvon shot to the top. Adding the PSPP stats paints a more accurate picture. Eugene T. Maleska himself now sits in third spot with 410 known puzzles. William Lutwiniak (a major talent you can check out online) is next but with 100 fewer, and then it's Jack Luzzatto, Manny, and A. J. Santora. You can see the list here.
These number are far from the final count. There are still many years worth of unidentified puzzles, but the top contributors are likely all on that list.
Ok, enough of that. Time for another Rude Q & A.
Why do I have to pay XWord Info for access when I already pay the NYT for access?
XWord Info is completely independent. We operate under an agreement with the NYT which allows us to provide information about their copyrighted crosswords, but that's all. We run ads and ask for donations to cover the cost of running the site.
What's the difference between contributing at the Casual, Regular, or Angel levels?
Currently there is no difference. People can pay what they think the site is worth. We have several Angel donors and we hugely appreciate each of them. XWord Info wouldn't exist without them.
Currently? You mean there might be differences in the future?
There will be differences in the future. We expect that existing pages will remain open to everyone with an account at any level but some new features will require higher-level accounts.
We tell you what we think your support level is on the top right of our home page. If you paid using PayPal and if you were logged on last time you renewed, we probably guessed right. If you didn't get credit for a high enough donation level or if you're listed as unknown, you'll be able to contact Jeff and we can fix that up manually. Again, it doesn't matter yet, but it will in the future.
I don't want to pay you people anything. What can I get for free?
The site is quite useful even if you don't pay. Full details are available for all puzzles in the past 45 days, and all the pre-Shortz ones. You can get a few samples of any of the other pages without paying. We hope you'll be interested enough to support us but you don't have to if all you want is answers to recent puzzles.
I'm a programmer. What happened to my JSON access, damn it?
Wow, we had no idea how many people were using our JSON data for purposes well beyond what was intended. Let me explain for the non-geeks. We provided a way to get data directly from our private database so that bloggers could add interactive puzzles to their blog posts. This blog has three examples in the left-hand column.
It turns out some people used that access to build up their own database of NYT crossword information. There are two problems with this. One is that we've spent about a million hours of work to edit, massage, and correctly display all these puzzles, including the weird and wacky ones that require manual fixes, so we don't want to just give all that away.
More importantly, we can't give it away even if we wanted to. As mentioned before, XWord Info gets access to copyrighted NYT data through a contractual agreement I signed just before I started the Wordplay blog on nytimes.com back in 2008. (You didn't know I started that blog, did you!) One of the provisions is that we can't make it easy for people without a legal agreement from the Times to extract that information. Fair enough, it's their data.
If you currently use our JSON access to display puzzles on a blog or other website, that still works. If you want to add these puzzles to your own blog, contact us and we can make that happen. But we can't just open up our database to the world. Sorry about that.