Stupid Excel Tricks: Tsunami Tuesday Delegate Count

Inspired by this incredibly detailed diary by crazymoloch, I broke out polling data from pollster.com, Real Clear Politics, and OurCampaigns to try some Excel tricks of my own to determine the delegate count next Tuesday.

My methodology is pretty simple.  I take the most recent poll for each state and normalize it to 100% for Clinton, Obama, and Edwards.  Then I adjust it to the present date by the amount a 5-poll average of the national polls has changed between the date of the state poll and now.  The last step is to impose a 15% viability on Edwards by scaling his delegate share from 0 at 7.5% to full share at 22.5%.  Although many states don't have explicit viability thresholds, my understanding is that they apply anyway for district based allocation of delegates.

So this is what I come up with for each state's delegates:

State Clinton Obama Edwards
AL 27 18 7
AK 7 5 1
AZ 26 22 8
AR 24 9 2
CA 205 165 0
CO 22 24 9
CT 23 24 1
DE 9 6 0
GA 37 44 6
ID 7 9 2
IL 40 95 17
KS 12 11 9
MA 48 42 3
MN 39 25 8
MO 32 18 21
NJ 63 42 2
NM 12 8 6
NY 131 96 5
ND 7 5 1
OK 18 8 11
TN 34 26 8
UT 11 11 1
TOTAL 833 712 127

By this methodology, Obama only will win 5 states (CO, CT, GA, ID, IL) but still do respectably with the delegates.  The lack of recent polls in some states probably understates Obama support.  Marc Ambinder combines polls and information from campaigns and indicates that Obama should also be favored in AL, KS, MN, and ND.  This projection will change as we get closer to Tuesday, find out where Edwards voters go and as more polls come in, but for now it's Clinton getting 50% of the delegates on Tuesday.

•  Interesting count, but I think it's off

  in a few cases.

  by AUBoy2007 on Wed Jan 30, 2008 at 07:57:00 PM PDT

  •  Yes...

    Besides the states that Arbinder says are leaning Obama which my spreadsheet doesn't have that way (AL, KS, MN, ND), I expect Arkansas and Delaware won't turn out as lopsided as they appear right now.  There isn't any polling data out of AK or ND, so I just mirrored the latest national poll for them.

    All in all, it is a pleasant surprise for me as an Obama supporter.  For Clinton to win by 120 delegates wouldn't be so bad considering the following states are better for Obama, and I expect things to tighten before Tuesday.

    by CA Pol Junkie on Wed Jan 30, 2008 at 08:07:58 PM PDT

  •  rec for Your sig

    i am glad that You are not basing it on DK surrogates,  i am frankly embarrassed sometimes by the actions of my candidates supporters and i can't help but think that the other side should be embarrassed as well ...

    by josephk on Wed Jan 30, 2008 at 08:30:08 PM PDT

•  Fascinating. If you tack on the current totals...

  ...and the numbers come up the way you predict, they'll end up with (including declared superdelegates):

  Clinton: 833 + 232 = 1,065
  Obama: 712 + 158 = 870
  Edwards: 127 + 62 = 189

  ...which means that Obama+Edwards = 1,059, or essentially equal to Hillary's tally.

  ...except, of course, that Edwards dropping out of the race means that most of his superdelegates will wander off elsewhere, and no one knows what his plans are for his pledged delegates (hang onto them? ask them to go to Obama or Clinton?)

  by Brainwrap on Wed Jan 30, 2008 at 07:59:04 PM PDT

•  Don't Forget Early Voting

  Obama may do great on the day of, but the early voting will favor the poll numbers from earlier. So, the errors should cancel out somewhat.

  In other words, you probably won't be as far off as you think, especially overall.

  BG

  by BlackGriffen on Wed Jan 30, 2008 at 07:59:35 PM PDT

  •  pollsters try to take that into account

    IIRC, the latest Survey USA poll of CA had 23% of their sample already voted (strongly for Clinton).  California probably has a higher percentage of early voters than most states.

    by CA Pol Junkie on Wed Jan 30, 2008 at 08:09:29 PM PDT

•  He's going to win

  way more states than that!  But even with skewed #s it's lookin' good!

  by Hope08 on Wed Jan 30, 2008 at 08:08:36 PM PDT

  •  fingers crossed

    The Rasmussen poll from Massachusetts showed a much closer race (now Clinton +6) than before.  Of course, the Kennedys might have a little influence there...

    by CA Pol Junkie on Wed Jan 30, 2008 at 08:10:52 PM PDT

•  Feb. 5 is so last week!

  Ok, maybe not.

  But one thing is increasingly clear: neither HRC or Obama is likely to "win" super tuesday when it comes to the delegates.

  However, what matters most is what the press says.  and if HRC take 17 of 22 states, that will be the story.  Regardless of how close the delegate count will be (and it will be within 100-200 even if HRC has a good day).

  What's more interesting is how long the slog is going to be if Obama survives next week.  What is striking about this list of the remaining states is that it shows that most of HRC's natural fertile ground will have already cast ballots:

  Feb. 9: I can easily see this as a "wash" delegate and win wise.

  Louisiana (56)
  Nebraska (24)
  Washington (78)
  Virgin Islands (3)

  Feb. 10: probably leans Obama

  Maine (24)

  Feb. 12: leans HRC

  DC (12)
  Maryland (70)
  Virginia (83)

  Feb. 19: Obama will take them both

  Hawaii (20)
  Wisconsin (74)

  Which takes us all the way to March 4 and Texas and Ohio (over 300 delegates total).

  Normally, I would say this is HRC country...but we're going to have two weeks of feverish campaigning and with Columbus and Austin I think Obama could take them both.

  And if these aren't enough, the only states left in March are Wyoming and Mississippi (March 8 and 11).

  After that it's Pennsylvania (and the nomination?) on April 22.

  I'm telling you, if it's the draw I think it's going to be next week, we're going all the way to Spring.

  by jonnykilbane on Wed Jan 30, 2008 at 08:29:54 PM PDT

  •  If 2/5 doesn't sew things up

    then Feb 9 probably means Obama in Louisiana and Nebraska, and Clinton in Washington. Maine I'd say is Clinton (Feb 10). Feb 12: DC for Obama, Virginia probably for Clinton, Maryland leaning Clinton. Feb 19: Agreed, Obama wins both. And as for Texas and Ohio, if it comes to that you can bet both campaigns will go all out. But chances are it won't come to that, IMHO.

    by Nathaniel Ament Stone on Wed Jan 30, 2008 at 08:47:15 PM PDT

•  Other variables:

  An interesting shot... but I'd point out the other variables, the biggest being that the Obama campaign has been playing all along to maximize the delegate count -- which is how they won the delegate counts out of Nevada despite losing the popular.

  They've put a lot of resources into organizing the caucus states, and the organization and planning is going to magnify any successes.

  They're also pretty tight, I suspect, on the CD races -- they're narrow-casting for particular vote targets in particular congressional district races, which will likely shave delegates off of Hillary's wins.

  by Carrington Ward on Fri Feb 01, 2008 at 05:34:00 PM PDT