Thats a lot of added complexity to be able to use column names, and to top it all, the use of non standard evaluation make things very complex when one go a little away from basic EDA.
I cannot remember. I emailed him about some results, he said they were interesting, but later I discovered those results were already published. I told him : I will throw them to trash. Very impolite from my part!, that must be 20 years ago.
Long time ago I failed to solve this conjecture. Nice to see it have been solved. I shared some email with S. Hetdetniemi about some special cases. Good old times.
Related: Our first instinct is far too often wrong (finantial times, 10 May 2019, Tim harford, pay walled article).
Another related article:
Heads or Tails: The Impact of a Coin Toss on Major Life Decisions and Subsequent Happiness
Steven D. Levitt https://www.nber.org/papers/w22487
compare base R:
Now hadley solution comment: You can now express that sort of thing fairly elegantly with dplyr:
Thats a lot of added complexity to be able to use column names, and to top it all, the use of non standard evaluation make things very complex when one go a little away from basic EDA.