"The Woman in White" February 17, 2023

[BarbaraK]

Who thinks it would be fun to write in whether you had
Any help including zoom calls so we could see how many get it all by themselves. I’m always curious about that. Exposing my inner nerd.
I only know one person who does the meta. We never help each other. And for me any help received would take the thrill away. More nerdishness.

You need to sign up for Matt's contest - https://xwordcontest.com/

When people submit there, they have to indicate whether they did at all alone or worked with a partner, nudger, group, etc. And when Matt reports the results each week, he includes the number of solos as well as the total number of solvers. It is interesting, and you can definitely tell which are the hard ones.

[chart]

I don't know if anyone pointed this out, or noticed, but the central clue of the puzzle was "opposition/ANTI" and the solution was a Democrat first lady when all the clues were Republican first ladies. That was another subtle confirmation for me.

[ricky]

For the people who were wrestling with whether to submit BESS or BESS TRUMAN, you have to understand how the WSJ selects the right answer.

Electronically they have everyone who submits by email and those who submit by the web site in one file. The system randomly selects one entrant. A person looks at the answer to determine if it is correct. In this case, the person would/should and will accept either option as a valid answer. You are not trying to outguess a machine.

And the system calculates the number of each answer so when you see Mike post the percentage of right answers and some of the most common wrong answers, the WSJ does not have some intern sorting through 1000+ entries.

Late to the party on this, but isn't this kind of improper statistically? Shouldn't the winner be randomly selected _from the correct answers_? Or is that just the intuition of a person (me) who's never been great at probability questions?

I can see an advantage of this approach over what I'm thinking would be statistically preferable, though - it means the bad spellers and the typo-prone get a fairer shake because their entries remain in the mix.

[Bob cruise director]

For the people who were wrestling with whether to submit BESS or BESS TRUMAN, you have to understand how the WSJ selects the right answer.

Electronically they have everyone who submits by email and those who submit by the web site in one file. The system randomly selects one entrant. A person looks at the answer to determine if it is correct. In this case, the person would/should and will accept either option as a valid answer. You are not trying to outguess a machine.

And the system calculates the number of each answer so when you see Mike post the percentage of right answers and some of the most common wrong answers, the WSJ does not have some intern sorting through 1000+ entries.

Late to the party on this, but isn't this kind of improper statistically? Shouldn't the winner be randomly selected _from the correct answers_? Or is that just the intuition of a person (me) who's never been great at probability questions?

I can see an advantage of this approach over what I'm thinking would be statistically preferable, though - it means the bad spellers and the typo-prone get a fairer shake because their entries remain in the mix.

Not really. If you take the pool of "right" answers, the odds are no better or worse than if the wrong answers are in there or not. What it does do is take the challenge of having to interpret right and wrong answers by a computer e.g. do you count Bess, Bess Truman or Besstruman as all right or not?

[ricky]

Just wanted to follow up on the reasoning of my wrong answer.

When I looked at presidential administrations, I was looking for a missing member of a series, like when the meta gives you Washington / Jefferson / Madison it points to the answer Adams.

The included First Ladies, from Eisenhower to Bush the second, were the only First Ladies to serve more than one term in that series -- with the exception of Hillary Clinton, who famously wore white on the campaign trail in 2016. So I submitted Hillary Clinton.

The obvious hitch in my reasoning was that "more than one term" is kind of clunky - three of them served two terms, but Pat Nixon served a term and a half. And of course I missed the significance of the numbers until it was too late.

In retrospect, I think this was my brain scrambling to make the subject matter of the puzzle somehow relevant to the year 2023. With all the important female politicians in US history, a puzzle about First Ladies? Where the theme answers are all First Ladies who, to be frank, weren't particularly notable in their own right? And the answer is Bess Truman, who left the White House when my *father* was 10 years old, almost 30 years before I was born, and about whom I know absolutely nothing - which Wikipedia tells me isn't unusual, because she shunned the media and avoided public life?

I'm not saying crosswords should only be about Tik Tok trends, but...