PA projections - Watch out: Lots of numbers ahead!

[dbmk]

Hello everyone!

Your (newly) resident Dane here with some numbers.
Like, probably, many of you I have been waiting for the Pennsylvania primary for a while now. Getting jittery.
And yeah, you are probably asking - or will be by the time you get through this; "What makes a Scandinavian guy waste so much time on this? He is not even able to vote nor living here.". Well, I'll tell you: Its just too damn interesting. :)

I have read some very interesting projections and information bits by the brilliant PocketNines on dKos before some of the other primaries, that taught me tons of stuff on the whole primaries/delegates thing and revealed some numbers that the binary polls we normally see are not bringing to light. And I was missing it this time. So I realised that I might as well take it upon myself to give it shot. Might even learn some stuff on the way.

So I present to you:

The Danish Pennsylvania Democratic Primaries Prediction Projection Presentation v1.0

.
Or TDPDPPPP1.0 since more or less obscure abbreviations seem to be the craze around here.

--------------------------------

Pennsylvania

According to the http://padems.com/files/Delegate08Final806.pdf">rules of the PA primary delegate selection, the delegates the candidates are fighting over on Tuesday, April 22. are distributed as follows:

1. 19 congressional districs in which 103 delegates are alotted going in sizes from 3 to 9 based on the popular vote within the district.
   
2. Pledged Party Leaders and Elected Officials (PLEOs) will be splitting 20 delegates based on the popular vote in the state.
   
3. At-Large delegates will be splitting 35 delegates based on the popular vote in the state.
   

For a total of 158 delegates.

Lets first look at the districts. And here I have more or less taken the numbers from http://www.cqpolitics.com/wmspage.cfm?docID=news-000002703375&cpage=1">CQPolitics that have done some predictions, that seem fairly probable. Its probably not good style to poach their district descriptions, and my knowledge of Pennsylvania demographics is, to put it mildly, limited. So descriptions are welcome, if there are some locals willing to give it a shot.

For each district I will list the number of delegates, critical split percentages where delegates changes hands and the prediction.
(Can someone tell me how they handle it if they land precisely on a split like 50% in 5 delegate districts? Who gets the split delegate?).
Then afterwards some comments to chase all those dry numbers down. I don't want to sadden any Hillary supporters giving this a look - but this is where the district allocation system cements that your preferred candidate has little to no chance of making up a significant part of the current difference in the pledged delegate count.

Or you can jump to the bottom where total is listed in nice large font. :)

District #1 South and central Philadelphia; Chester
7 Delegates -- Split percentages | 7,14% | 21,43% | 35,71% | 50,00% | 64,29% | 78,57% | 92,86% |
Prediction: Obama 4 Clinton 3

District #2 West Philadelphia; Chestnut Hill; Cheltenham
9 Delegates -- Split percentages | 5,56% | 16,67% | 27,78% | 38,89% | 50,00% | 61,11% | 72,22% | 83,33% | 94,44% |
Prediction: Obama 7 Clinton 2

District #3 Northwest — Erie
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #4 West — Pittsburgh suburbs
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #5 North central — State College
4 Delegates -- Split percentages | 12,50% | 37,50% | 62,50% | 87,50% |
Prediction: Obama 2 Clinton 2

District #6 Southeast — parts of Berks and Chester counties, Philadelphia suburbs
6 Delegates -- Split percentages | 8,33% | 25,00% | 41,67% | 58,33% | 75,00% | 91,67% |
Prediction: Obama 3 Clinton 3

District #7 Suburban Philadelphia — most of Delaware County
7 Delegates -- Split percentages | 7,14% | 21,43% | 35,71% | 50,00% | 64,29% | 78,57% | 92,86% |
Prediction: Obama 3 Clinton 4

District #8 Northern Philadelphia suburbs — Bucks County
7 Delegates -- Split percentages | 7,14% | 21,43% | 35,71% | 50,00% | 64,29% | 78,57% | 92,86% |
Prediction: Obama 3 Clinton 4

District #9 South central — Altoona
3 Delegates -- Split percentages | 16,67% | 50,00% | 83,33% |
Prediction: Obama 1 Clinton 2

District #10 Northeast — Central Susquehanna Valley
4 Delegates -- Split percentages | 12,50% | 37,50% | 62,50% | 87,50% |
Prediction: Obama 2 Clinton 2

District #11 -- Northeast — Scranton, Wilke-Barre
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #12 Southwest — Johnstown
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #13 East — Northeast Philadelphia, part of Montgomery County
7 Delegates -- Split percentages | 7,14% | 21,43% | 35,71% | 50,00% | 64,29% | 78,57% | 92,86% |
Prediction: Obama 3 Clinton 4

District #14 Pittsburgh and some close-in suburbs
7 Delegates -- Split percentages | 7,14% | 21,43% | 35,71% | 50,00% | 64,29% | 78,57% | 92,86% |
Prediction: Obama 4 Clinton 3

District #15 East — Allentown, Bethlehem
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #16 Southeast — Lancaster, part of Reading
4 Delegates -- Split percentages | 12,50% | 37,50% | 62,50% | 87,50% |
Prediction: Obama 2 Clinton 2

District #17 East central — Harrisburg, Lebanon, Pottsville
4 Delegates -- Split percentages | 12,50% | 37,50% | 62,50% | 87,50% |
Prediction: Obama 2 Clinton 2

District #18 West — Pittsburgh suburbs, part of Washington and Westmoreland counties
5 Delegates -- Split percentages | 10,00% | 30,00% | 50,00% | 70,00% | 90,00% |
Prediction: Obama 2 Clinton 3

District #19 South central — York, Gettysburg
4 Delegates -- Split percentages | 12,50% | 37,50% | 62,50% | 87,50% |
Prediction: Obama 2 Clinton 2

Comments:
First and foremost lets just acknowledge the fantastic shapes some of the districts have - probably due to being consolidated to 19 districts from previously 21. 12, 6 and 18 are just fantastic.
What will be interesting is perhaps whether Obama can make the cut for the 7-2 split in district 2 and if he can break some of Hillarys 50%+ districts - which as far as I can tell is rather unlikely. The chance of the 4 delegate districts splitting anything other than 2-2 is probably not very high.

Hillary supporters: This is where the game is lost for Senator Clinton. There are quite a few of those in play out there in the remaining contests, where you need landslides to avoid 50-50 splits. Or something close to it. And need it over the entire board. And as far as I can tell Kentucky is the only remaining state where there is something close to that lined up.

Oh. Lets have the total of the district delegates:
Obama 50 Clinton 53

Then we have the 20 PLEO delegates and the 35 at large delegates.
I have chosen to operate from a poll average I guestimated to be Obama 41 Clinton 49, adding a little to make increase the odds of me not underestimating Senator Clintons lead. Splitting the undecided evenly among them.
I would like to hear any comments to these numbers, if you think they need adjustment

That gives us:
PLEO: Obama 9 Clinton 11
At-large: Obama 16 Clinton 19

Giving us a combined total of:

Obama 75 Clinton 83

An 8 delegate gain for Hillary. The percentages in delegates are 47,47% vs. 52,53%. And thats with a 8% spread in the popular votes. If anyone was in doubt about the math here, that should illustrate what percentages we are talking about in terms of popular vote if Senator Clinton is to win 60-70% of the remaining pledged delegates. Which she will probably need to do, to field a reasonable argument to the SDs. All other things being equal.

So. There you have it. Hope you enjoyed it and perhaps even learned/realised something.
I would love it if anyone could add thoughts, corrections, local stories/information.

[FrenchieCat]

1. Will have to bookmark to review tomorrow with a kick!

Thanks!

[Zachstar]

2. I think Obama can do a bit better than this. But this a good modest analysis thanks!!

Clinton only gaining 8 delegates will be a disaster for her campaign. So it is a good honest number!

[dbmk]

5. I have tried to err on the side of Hillary.

The only one that might be a tad optimistic but far from possible is the 7-2 split in district 2.
And all depending on the popular vote the number can quickly shift 2-4 delegates in either direction. But it is probably going to land in that ballpark.

[Zachstar]

6. Understood! Thanks

[MagickMuffin]

3. Interesting numbers

Of course I HOPE your math is a little off, since I want Obama to win the most delegates and popular vote, but I guess we'll both have to wait and see.

Thanks for doing this dbmk.:hi: from Fort Worth, Texas.

Side note: Hillary didn't WIN that big here. She only won by 3.5% of the popular vote, but Obama WON the MOST DELEGATES

[dbmk]

4. Which in all fairness

..was due to the caucases. ;)

And yeah, I am personally hoping for an Obama upset as well. But even 8 delegates for Hillary will seem like a drop in the ocean, compared to the fact that this primary will burn of a LOT of the remaining delegates in play.

This is her shot. And 8 delegates will not seem like her making it. To her or the media.

[my3boyz]

18. Oh the media will play up her win..even if it is by 1 pt.

They will play the confetti dropping over and over again with her sappy speech. Then the papers will have her picture over them with confetti. Just as they did in Texas even though they knew Obama would probably win the most delegates. They want you to think she has a shot of winning the nomination. My mom called me about a week ago (around first days bittergate) all upset because she said that CNN was saying Clinton was going to win. I don't watch CNN but I'm going to give them the benefit of the doubt. They were probably saying Clinton was going to win PA but my mom thought they meant the primary election. Who knows? It is Clinton News Network so perhaps they were putting spin on her winning the nomination.

[spinbaby]

7. I see you noticed our gerrymandering

I live in the 18th--it's ridiculous. I live outside of Washington, but somehow vote with Greensburg.

[Zachstar]

8. The 18th is beyond ridiculous who the hell did that?

[Califooyah Operative]

9. haha, that looks like some of our California districts.

[dbmk]

14. I searched and found

http://www.mapcenter.org/region/gerrymander.html

It was changed by the 108th congress.

"After the 2000 Census the Republican Party made no secret of their intention to redraw voting districts in PA in such a way that the Democratic Party would lose up to six seats in the U.S. Congress even though the State was losing only two seats due to reapportionment."

[scheming daemons]

16. I live in the 18th.... and you're right... the Repukes drew it to get a guaranteed Repuke seat....

....

[babylonsister]

10. I'm bookmarking this for Tuesday; thank you for all your efforts!

I'm shocked you are so 'into' this! We shall see. Hope to see you around Tuesday. :thumbsup:

[shadowknows69]

11. Good work

Looking this over when I'm awake.

[K Gardner]

12. Bookmark for Tues.. and thank for the hard work ! Good to have another resident Expert on this

math, delegate counting and other "stuff" I don't get. We have a WONDERFUL team here !

[thevoiceofreason]

13. K&R

With a bookmark - thanks for the work

[slinkerwink]

15. kicking

[davidpdx]

17. Very interesting stuff

I love talking politics, but man you are nuts (and I mean that as a compliment). Some of those districts look like crazy jigsaw puzzles.

I'd like to see you do this same thing for Indiana given it could be close.

[dbmk]

20. Now that I have found were to look for the information

I have very intention of doing that for the coming primaries.

And yeah, I might be crazy. I just like numbers. :)
Its especially the details in the effect the congressional districts have on the delegate distribution that interests me - probably because the so called journalists (read: dumbass pundits that are more comfortable talking than thinking) in the media seem to completely ignore those details.

[rug]

19. Ah, but Skagen remains a wild card.

[dbmk]

21. Come again? :)

[rug]

23. Those absentee ballots from Skagen will upset the apple cart.

:P

[RUMMYisFROSTED]

22. Ah, but only a handful of delegates will change hands.

It's over.

Still.

[me b zola]

24. Bookmarked

Thank you for the analysis. :thumbsup:

[TheDoorbellRang]

25. usregimechange has done some great work with state demographics

Here is his/her analysis on Pennsylvania:

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x5088717

Thanks for the analysis. I find it fascinating that so many outside of the States are as interested in this election as we are.

Another analysis that I find intriguing is the Obama campaign spreadsheet that was "accidentally" leaked after February 5th. It has only been incorrect in one of its predictions -- the spreadsheet showed Obama losing Maine when in actuality he ended up winning it. The spreadsheet's prediction for Pennsylvania is a Clinton win 52-47%. If anything, the spreadsheet underestimated how much he would win by in many of the predictions. Here is a copy of the spreadsheet.

[dbmk]

48. Yeah, too bad its been archived.

Would have loved to kick that back into life.

Could def use his/her insight into the districts.

[City Lights]

26. K&R, and B for Tuesday.

Great stuff, dbmk! Thanks for taking the time to put it together.

[grantcart]

27. Tat for Sitz

Did I get it right?

Thanks for everything.

[dbmk]

39. lol

:D

Took me a few brain revolutions to get what you were going for. And I think you failed. Miserably. :)
Unless you really wanted to say thanks for the great time we had last time we met. Which I would probably remember.

But definately points for trying. :)

[truthpusher]

28. i have been at du for a many years... i have not contributed anything even remotely close to this...

...thanks for making me feel like a total ass.

You contribution to help educate us that are to caught up in lapel pins and middle fingers, is humbling.

(love the fantastic shapes part. I am on my blackberry right now, so the fantasticness of the district shapes is hard to distinguish...can't wait to get to the laptop to get a closer look!)

[JimGinPA]

29. Great Work!

I bookmarked this too.

K&R

[phrigndumass]

30. Big K and R

Excellent!

I'm projecting voter turnout of about 2,100,000. An 8-point spread (54-46) would result in:
Senator Clinton - 1,134,000 (+168,000)
Senator Obama - 966,000

+168,000 for Clinton in Pennsylvania, and Obama will gain back about 163,000 in North Carolina (see link, further down in the replies):
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x5576204#5578398

:thumbsup:

[livetohike]

31. Thank you for all this work!

:hi: Let's keep it kicked until after the primary.

[phrigndumass]

32. Kick

More people should see this

[phrigndumass]

38. Double-kick

:kick:

[phrigndumass]

52. Hat trick kick

:kick:

[dbmk]

53. Thanks :)

[Kukesa]

63. Kickety-kick kick. n/t

[cliffordu]

33. Good work! K&R

:kick:

[grantcart]

34. demconwatch has added a special page for PA primary

http://demconwatch.blogspot.com/2008/04/pennsylvania-primary-results.html

[malik flavors]

35. So, I take it we're all hoping for about an 8 point difference? I'd be really happy with that.

[otohara]

36. Time to Change The Rules

AGAIN!

[crankychatter]

37. YOU, are a damned HOOT... are all Danes this much fun?

seriously

[AtomicKitten]

40. That's an impressive analysis.

Pennsylvania is going to be fascinating. Throw EVMs in the mix and very motivated opposition, and who the hell knows what can happen. Oy.

K&R

[denem]

41. Funny - 52% HRC, 47% BHO was the Obama campaign's

working numbers after Super Tuesday. Their Texas and Ohio prediction were spot on. Only WI was way off - predicted narrow Hillary win. (as per reported NBC after TX/OH primaries).

[dbmk]

44. Yeah, I purposefully gave HRC a little more than expected.

I am going to play around a little today with the popular vote delegates, to so where it breaks delegatewise.

[dbmk]

47. Funny thing is

When I punch in the Obama campaigns numbers, 47-52 in terms of popular vote, they give me a 4 delegate win for Hillary.

But the Obama campaign themselves predicted that to give the same numbers I arrived at first time around, in terms of delegates.

So they must have estimated some of the districts more cautiously.

[stillcool]

42. well....I read it all the way through...

and I will just say thank you, and leave it at that. I'm sure there are some who will really appreciate your work, but I am math challenged, so my eyes are now crossed... but I get the 8+

[ORDem]

43. If you're right, then Obama will need only 136 more PDs to reach a majority of PDs.

That's the tipping point when the SDs will be standing in line to endorse him!

:dem:

[barack the house]

45. Abbreviations are hardly a craze. Just how a lot of terms are referenced in America conveys ...

information more quickly especially wiht some very longed named states and they are in list form.

[quantass]

46. Another Fellow non-American who ALSO finds all this "too damn interesting"!

:hi: Fantastic work on the math and predictions!

[dbmk]

49. Been looking at the state wide delegates

So.. we have our two sets of statewide delegates:

1. Pledged Party Leaders and Elected Officials (PLEOs)
   
2. At-Large Delegates
   

Pledged Party Leaders and Elected Officials (PLEOs):
20 delegates
The delegates breaks at these percentages
9 42,500%
10 47,500%
11 52,500%
12 57,500%

So far it seems unlikely that we will see a 8-12 split , either way. It is possible - there has been a poll with that spread lately - but the rest of polls doesn't really support it. It would mean a 15 point+ win for Hillary.

Atm it seems like the polls are hovering right around the 4-5-6 point win for Hillary. If Obama keeps it under a 5 point loss, this category gets tied delegatewise.

And its probably also very unlikely that Obama flips it to an 11-9 his way.

So this should be a tie or a 2 delegate win for Hillary. Atm I am inclined to project this to be a tie. (Note: in the OP its a 2 delegate win for Hillary. )

At-Large Delegates
35 delegates
The delegates breaks at these percentages
15 41,429%
16 44,286%
17 47,143%
18 50,000%
19 52,857%
20 55,714%
21 58,571%

I don't think Obama is going to beat Hillary on the popular vote. Close perhaps, but there is nothing to indicate he will pass her atm.

So the question is if he can break the 47,143% to get the 17-18 result and keep the loss at 1 delegate in this category.
The lastest Rasmussen poll:
http://www.rasmussenreports.com/public_content/politics/election_20082/2008_presidential_election/pennsylvania/pennsylvania_democratic_presidential_primary
suggests this to within the real of possibility.

But so is a 19-16 win for Hillary. The 20-15 needs a 10 point win for Hillary - and I dont see that coming atm.
I am inclined to go for the 1 delegate projection atm.

From what I have read over the last couple of days, there is a possibility that the large amounts of colleges in Pennsylvania could be underrepresented in the polls. Which, if true, could mean the polls are a bit skewed towards Clinton.

But overall I would be VERY surprised if these two categories would yield more than 5 delegates for Hillary. And expect less.

[davidpdx]

50. FYI

I think you updated these numbers (or it looks like you did). If you do, please put something like this at the top:

*Updated 4/20/08 8pm EST*

It will help people be able to keep track of changes if they are watching your thread.

[dbmk]

51. I havent. And I can't.

There is a limited time where you can edit a post after posting it. And we are past that.

Would love to be able to though. :)

[mckeown1128]

54. Obama has a good shot actually of winning the odd delegate in #15.

Check it out... It looks like good news.

http://www.dailykos.com/story/2008/4/19/16825/4398/287/499078

Let me know what you think???

[dbmk]

57. Hmm..

.. I am not so sure that the numbers support Obama breaking 50%. Could be close, but so far I would have to project no cigar. District 8 is, as far as I can read, also possibly in play.

But thanks for the link!

[mckeown1128]

58. I don't know if you have seen these Delegate predictions...

they bothshow CD 8 breaking for Obama.

http://www.dailykos.com/story/2008/4/19/141034/315/327/499022

http://ruralvotes.com/thefield/?p=888

They both show 15 breaking for Clinton.

[dbmk]

59. I hadn't. Thanks.

Will have to look those over tomorrow. 1:45 AM here, so I will need some sleep first I think. :)

[ihelpu2see]

55. Nicely done. I think Obama would be happy with that and Hillary would be

just that much closer to having to step out... imo

[msallied]

56. Fantastic analysis!

And one I could be happy with if we can't pull out an all-out win for Obama. And thank you for doing this. It's fun knowing that this election (the Democratic one) is capturing the attention of the international audience.

[npincus]

60. Impressive, Danish friend!

:hi:

[dbmk]

61. If anyone see anything resembling district predictions

..please keep me in mind and drop me a link. :)

Will be working on an analysis of the ones I can find today.

[dbmk]

62. For those that had this thread bm'ed

v2.0 can be found here
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=132x5608247