storytime - searchforx

blog/storytime

I'm in a crossword group and we've been solving the New York Times (NYT) and<br>The New Yorker (TNY aka "The Crossword") puzzles daily for almost 2 years now.<br>This is what we saw.

First a brief introduction: The NYT puzzles get harder the next day; MON is<br>the easiest - SAT is the hardest. SUN is a bigger grid with WED-THUR difficulty.

The New Yorker currently runs MTW; MON is the hardest - WED is the easiest.

There are other caveats, but that's the gist.

We didn't always do TNY puzzles. In the beginning, we only did the NYT. We set<br>the pace where folks could self-study MON and TUE. This meant doing it solo or<br>a subset get-together. This would enable our newcomers to have space for<br>introspection and also allowed our seasoned performers to rip on their train to<br>work or whatever. On WED+, it's an Avenger's Assemble event.

This was not a serious decision, but obvious that it was useful. After all, we<br>were just doing this after work hours and usually before bed. It's an effective<br>strategy because newcomers want to get good enough to have fun. Being on a team<br>for SAT but only solving one clue in 45 minutes is not fun. You will clock out<br>and feel melted. So this learning function was the natural, attractive structure<br>for us.

Eventually we grew out of this. Our epicenter was performant enough where we<br>didn't need those days for self-study. An inflection point occurs and we require<br>enough time to justify getting a group of people together.<br>What I mean is that a NYT MON is a The New Yorker is a nice complement to the NYT because MON is their hardest<br>puzzle, easier on TUE, and easiest on WED. This means that if you do both the<br>NYT and TNY on these days, the sum difficulty models a uniform distribution.

Now some other observations:

We all get better, but no one drastically better. The totem pole has kept<br>its shape.

Our average time is strong, but not elite. I cannot be certain if our<br>trajectory is on this curve, but I suspect it will be. Like any growth curve,<br>they can be deceptive. Our curve is earnest. I bet on earnest.

Not everyone has a taste for the mechanics of the crosswords: stuff like<br>punctuation, abbreviation, pluralization, tense, or crosswordese. A simple<br>way to build a serious team would be to make sure every player has a knack<br>for this.

To add to the characteristics of the group, they have simply elected to do this<br>every day, but don't necessarily obsess over getting better. They are ordinary<br>people. We solve and move on. This is usually a bad sign for growth; like how<br>your ELO in chess plateaus. If nothing in your thinking changes, you will never<br>really get better. But seeing my group improve month after month, in a casual<br>setting, warrants reasonable suspicion that there must be something embedded in<br>our microcultures that are actually present everywhere.

Instead of divulging into the "why", I wanted to point out the reasons this is<br>even interesting. Most people would be embarrassed about this topology and start<br>bullshitting their way with questions that don't really matter:

Why don't you just teach the group the mechanics of crosswords?

Why couldn't you just make it so that the members actually study the puzzles?

Wouldn't an obsessed group of people be way faster at learning all of this?

And those are reasonable questions for sure, but going this route forces you to<br>reason against the system, which is not interesting. By the time you can use<br>energy to act against entropy, you can do anything. Instead, position yourself<br>into the hacker's school of thought.

What is innate within the system that the individuals themselves are doing<br>(albeit oblivious) that the overlying system has a clear insight or ruling<br>trajectory over?

One could argue a strong case for consistency. Doing anything every day<br>is going to make you better. But like chess, ripping games daily doesn't make<br>you marginally better. It would be stupid to tell Michael Jordan that he needs<br>to play more basketball. The next marginal improvement is by watching film.

In the case for the crossword group: it's clear that the players effectively<br>reached their p99. The next marginal improvement would be to study crossword<br>mechanics or some form of deeper obsession. But they're not. They don't care.<br>They are normal people; not Scott Wu and his mathcounts gang.

What is interesting is that they are still growing and the signs come out in<br>weird shapes. I remember the daunting days when we spent around 1.5 hrs to solve<br>hard puzzles. But then the unit of work changes from 1.5 hrs to 1 hr. 1 hr to 30<br>minutes. Another PMF signal was that a I think this story is an ode to what I hope people should be looking out for when<br>empathy-hacking with communities.

When a quantum of work (energy) changes, people change.