It really hadn’t occurred to me to go back and evaluate my 2018 race analyses. But when a reader asked, I thought it was an interesting question. I’m currently online at a bar, so a little self-indulgence seems quite appropriate.
With respect to the Senate, I mostly just called a Democratic takeover improbable at best. True enough. I didn’t exactly go out on a limb there.
With respect to the House, however, I did rely on some rather complex custom data analytics. I didn’t so much make predictions as I suggested where money could best be spent in the closing months of the campaign to guarantee that the Democrats took control of the chamber.
- I said there were 11 races where the Democrats would win without much additional help. Democrats won all 11 of those races.
- I picked 20 races where I thought additional money should be concentrated to help the Democratic candidates who had a decent chance to win. Democrats won 14 of those races and another 3 races have yet to be called. To date, Republicans have won only 3 of the races that I picked.
- In Texas, I said there were two seats that the Democrats should flip and, indeed, both flipped. I saw an outside chance of Democrats taking two additional Texas seats and the Democrats took one of them.
Considering only the Democratic wins above, the Democrats would have taken the House. While they won additional races to pad their margin, I think I can claim success. I’ll give my analysis a “B” in the unlikely scenario where the Democrats lose all three of the outstanding races but I think I deserve an “A” if they win them all.
That’s not too bad for my initial attempt at political data analytics and I have the data to tweak the model for the next cycle.
Cheers!