Why I Built USPoliticalRank

A puzzle has one right answer

I make puzzles for a living. Guinness World Records lists me as the world's most syndicated puzzle compiler (Guinness World Records, Most Syndicated Puzzle Compiler). I edited the Universal Crossword for more than 21 years and over 8,100 puzzles (PR.com, 2018). I was senior crossword editor at USA Today for 13 years. I wrote all 225 episodes of Merv Griffin's Crosswords (timothyeparker.com). I say all that for one reason. It is my credential for exactly one claim: I know what a fair test looks like.

A crossword does not care who you voted for. Seventeen across is either right or wrong. The grid is the same for the doctor in Boston and the trucker in Tulsa. Nobody gets an easier clue because the editor likes their politics. That is the entire appeal. A puzzle is a contest where the rules are printed on the page and applied to everyone.

Then I would put the puzzle page down and read the news pages. Different sport entirely. Same politician, same week, same actions. One outlet scores it a triumph. Another scores it a scandal. The reader is not being given a test. The reader is being given a verdict and asked to sign it.

The scoreboard should not care who is playing

So I built the thing I wanted to read. USPoliticalRank ranks presidents, governors, senators, and states on published metrics from named sources. Bureau of Labor Statistics. Census. U.S. Sentencing Commission. GovTrack. Official election returns. The methodology runs at the top of every ranking, before the results, so you can check the ruler before you look at the measurements. Both parties get the same ruler. Every claim gets a citation. If a fact cannot be verified, it does not run.

This is not a new idea. It is the oldest idea in sports. Nobody asks the scoreboard at a football game to consider the quarterback's intentions. Points are points. Trust in political media sits at 31 percent, tied for the lowest Gallup has ever measured (Gallup, 2024). You do not fix that number with better opinions. You fix it by removing opinion from the parts of the page where opinion does not belong.

Data, not narrative

Here is my creed, and it fits in four sentences. Only actions matter. Speeches are not deeds. Numbers from named sources beat adjectives from anonymous ones. When the data embarrasses my expectations, the data wins.

That last sentence is the hard one. Everyone loves data until it disagrees with them. I have published rankings on my other project, Black Intervention, that made people on every side angry, because the documented record refused to match anybody's preferred story (blackintervention.com). Good. A method that never surprises you is not a method. It is a costume your opinions wear.

I am 30 years into a career built on a simple contract with solvers: the puzzle is fair, the answer is checkable, and the editor does not cheat. USPoliticalRank is the same contract, signed with readers instead of solvers. The grid is the same for everyone. Check my answers. That is what they are there for.