Fantasy Electionball

Do you think you really know Texas politics? Then try your hand at predicting how many delegates each candidate will get in the Texas Democratic primary a week from today. The Lone Star Project, a Democrat-affiliated research group, has put together an online delegate calculator, in which you can examine Texas' 31 senatorial districts and try to predict how many national convention delegates will come out of each.

Now, why would you need a calculator, and what do the state senate districts have to do with a national election? The 228 Texas delegates will just be divvied up according to the popular vote, right? Nope – leave it to the Democrats to make it all complicated. First off, some of the delegates (67 of them) will be assigned not according to the popular vote, but in the precinct caucuses on March 4 after the polls close. In a feature we’re publishing on Thursday, I’ll explain why it’s very important for you to attend those caucuses. Be sure you read it!

And of course, 35 more are the “superdelegates” you’ve heard so much about – important party dignitaries who can cast their vote however they please (i.e., Austin Congressman Lloyd Doggett, who has already pledged for Obama).

That leaves 126 to be decided by the popular vote. But the Texas Dems couldn’t even make that simple. Rather than dividing up the whole 126 according to a statewide percentage, each senate district gets a certain number of delegates based on Democratic voter turnout in the past two general elections (i.e., Kirk Watson’s Dist. 14 gets eight delegates, while Amarillo Republican Kel Selinger’s Dist. 31 only gets two). Obviously, that will kind of screw districts that have performed poorly in the past but this year see a huge increase in turnout – or, districts that have previously voted Republican but have decided to abandon the sinking Bush ship. But, rules is rules.

So, the delegates in each district will be divided according to the popular vote in that district, and then all the districts are combined to produce the statewide total.

Obviously, this system is a bit of a double-edged sword now that we finally have a hotly contested primary: The delegate allocation easily lets candidates know where all the Democrats are, but at the same time, it makes Democrat levels in different regions self-perpetuating – Obama and Clinton have more incentive to campaign in delegate-rich areas like Austin, Houston, Dallas, and the border and less need to woo potential new voters in the Hill Country or Panhandle, because those votes will count less, even if turnout goes up.

Anyhoo, if you think you know which regions of Texas are pro-Clinton and which lean toward Obama, give it a try. UPDATE: I gave it a try, and damn it's hard to even take a rough shot at it. There are so many variables. I tried plugging in figures that combined the most recent Survey USA poll, mixed with the LSP's figures on racial demographics. But when I came up with a figure that showed Clinton getting 53% of the vote, I knew I had a problem, since Survey USA showed Obama leading Clinton 49%-45%. I realized that LSP's racial figures show percentage of voting age population for blacks and Hispanics, but that's different from both "likely voters" and "percentage of Democratic voters" – the latter meaning that I gave too much weight to whites in the predominantly Republican districts. For what it's worth, my obviously flawed figures showed Clinton with 66 delegates and Obama with 59. Back to the drawing board.