Sick of delegate counts on the proverbial road to 1600 Pennsylvania? Here’s some relief: a state-by-state analysis of the 1600* public Phish performances in the U.S.

The graphic below plots each state as a bubble, at the intersection of the first and last dates Phish played there. The larger the bubble, the more shows Phish has played in that state. (A legend is included in the lower left of the graphic.)

These bubbles appear in several distinct clumps: The large bubbles in the upper left are places Phish started and has performed the most. To the right of that, a group of varying sizes represent states first visited in the band’s early expansion, from 1989 to 1991, and still played recently. Finally, the sparse group of smaller bubbles in the lower middle-to-right represent sporadic stops from 1990-1995 that haven’t been visited in a number of years.

An upward sloping trend, where that third group was higher (i.e. had been played more recently) than the first, would show that Phish was adding new states and shedding old ones. But while they’ve added new states, they aren’t shedding old ones: Though Vermont hasn’t been played in several years, every state played before 1989 has also been played since 2009..

Contrarily, a downward sloping trend, where the first and third groups predominated, would indicate that the band was simply sticking to its origins. Instead, that second group appears: The bulk of the states added from 1989 through 1991 have been played in the most recent year.

The overall trend of the 45 represented states is still slightly downward (indicated by the red dotted line, a simple linear regression of those 45 data points). But take into account all 1600* shows (as the green dashed line does), and the trend is simply forward and steady: Beyond a small number of outliers (such as Oklahoma, not played until 16 years after Nevada, and not shown here at all), Phish has typically expanded their itinerary while continuing to visit most states.

* We're aware of 1823 dates at some point associated with a Phish performance (not counting side projects or guest appearances, which we group as sideshows). Of those, 9 didn't happen (6 cancelled, 3 postponed), and 45 more were either erroneous (e.g. 10/12/88 was actually 12/10/88 set I) or aberrant (e.g. there was no 12/25/88 show, planned or performed). Set aside 16 private events, 13 soundchecks, 14 television appearances, 8 radio shows, 6 sporting events, 3 award events, and 8 miscellaneous events (such as a post-show bluegrass jam in a parking lot), and we’re left with 1701 legit, public, mostly ticketed, "counts for stats purposes" show dates - of which 101 are outside the U.S. (that graph comes later) and 1600 within.

If you liked this blog post, one way you could "like" it is to make a donation to The Mockingbird Foundation, the sponsor of Phish.net. Support music education for children, and you just might change the world.