Update your Iowa predictions! (KICK ME to see how smart you are)

[Bucky]

I just read the http://www.latimes.com/news/politics/la-na-poll11jan11,1,3764349.story?coll=la-home-headlines">LA Times article Maddezmom posted in another thread. I think this is great news for Dean. So I'm updating my prediction. This is your chance to do so too, with 8 days left until Monday night madness begins. Let's keep this sucker kicked until Monday night, when your balance between objectivity and psychological projection will be put to the test.

:dem: Here's my thinking. While the polls show this:

Dean .... 30% ..... Gephardt ... 23%
Kerry ... 18% ..... Edwards .... 11%

There is this caveat in the Times (the best big city paper in the country, as far as I'm concerned): "The survey also showed likely caucus voters were surprisingly fluid in their commitments to the candidates, even after months of intensive campaigning. About 1 in 11 said they remain undecided in the race. More significantly, nearly 40% who express a preference say they could still switch before the caucuses."

So that's about half the voters are still up for grabs. That leads to the importance of two factors. Along with the Harkin endorsement, Dean has his "Perfect Storm" volunteers cascading into the state. He also is going to get some positive press from winning the non-binding DC "beauty contest" primary on the 13th. These create the ability to give Dean a last minute turbo boost to his plateaued numbers in Iowa. Based on that, I'm going to revise my prediction for the caucuses.

I'm also going to repost the predictions offered in an earlier thread. This is your chance to revise where you put your money and your mouth people.


Dean . Gep. . Kerry. smart aleck . other guesses
35% . . 32% . . 15% . . Bucky (old prediction)
42%. . 28%. . 16%. . Bucky (new prediction)
33% . . 35% . . 14% . . Kathleen04 (old)
40% . . 36% . . 16% . . Kathleen04 (new)
32% . . 34% . . 16% . . Texas Patriot
32% . . 21% . . 22% . . Deesh
35% . . 25% . . 15% . . Daboy
37% . . 34% . . 21% . . Goobergunch (old)
40% . . 38% . . 18% . . Goobergunch (new)
52% . . 28% . . 20% . . H Fishbine
50% . . 34% . . 16% . . Padraig18
30% . . 15% . . 15% . . Paulie (plus 20% for Kucinich)
21% . . 23% . . 19% . . Chimpymustgo
40% . . 36%. . . 8% . . YellowDawgDem
32% . . 28% . . 22% . . Cryingshame
22% . . 24% . . 16% . . Anti-Bush (plus 20% for Edwards)
41% . . 29% . . 15% . . TeDoll78
25% . . 28% . . 19% . . BHunt70
32% . . 28% . . 25% . . SlinkerWink
32% . . 23% . . 19% . . Ozone Man
31% . . 19% . . 21% . . Ghost Consul
47% . . 21% . . 15% . . Askew (plus 17% for Edwards)
42% . . 30% . . 13% . . CynicalSOB1
35% . . 26% . . 18% . . WI Dem
-0- . . 35% . . 25% . . AreWeThereYet (plus 25% for Edwards)
1st . . 2nd . . -0- . . Morgan2 (3rd place for Kucinich) , (not giving numbers)


The average of all these guesses are...

Dean . Gep. . Kerry
35.5% . 28.1% . 17.3%

(edited to fix the times quote)


Send me a DU private message ( ) to become a Teacher for Clark
Teachers for Clark beta site - alpha site coming soon!

[BayCityProgressive]

1. I think

Gephardt will be done after Iowa.

[poskonig]

2. Dean 32, Gephardt 29, Kerry 24, Edwards 7, 8% other

eom

[John_H]

3. Dean 28 Gephardt 26 Kerry 21 Edwards 10

[Insurance_Analyst]

4. Kerry wins Iowa!!!

The tax issue is going to kill Howard Dean! It's over. The foundation is starting to crumble already.

[Bucky]

5. IsnAna, I don't think you get the Dean thing

I'm no Dean fan, but nothing's gonna kill Dean right now. There is a significant part of our party for whom taxes are not a voting issue. Kerry winning Iowa is just silly. He's gonna have to work hard just to keep Edwards from taking 3rd place from him.

Don't get me wrong. He's a great guy. If this were 50 years ago, he'd be a great president. He's the man Adlai Stevenson wanted to be. But Kerry just isn't cutting it in an age of crudeness and forced authenticity. The man, alas, has no Elvis.

[Padraig18]

6. I will stand with my prediction.

Dean 50%

Gephart 34%

Kerry 16%

[Bucky]

9. I admit I'm more timidly nudging my assessment. But Dean still has Neg's

I'm not sure how Dean can break 50% if Gep is still getting 1/3 of the votes. This prediction totals up to 100%. You're saying that no one at all will caucus for Kucinich or for Undeclared. Edwards currently has one in nine Democrats in the state supporting him. Are you saying he's going to lose virtually all of his support in the coming week?

[Padraig18]

11. No one gets any delegates if they don't break 15%.

I'm saying those 3 will get the only delegates, and what % of the delegates they will get. :)

[Bucky]

13. Ah! Everyone else is predicting the actual vote. What's your take on that?

To keep the comparisons valid, we should all be predicting on the same question. The question is precinct level turn out, not final delegate count.

[Padraig18]

14. That is precinct level.

You don't get ANY delegates if you don't get 15%. There will be no 'count' of any kind for anyone who doesn't get 15%, unless I'm mistaken. :)

[Bucky]

15. are there two points of cut off?

If the 15% cut off is at the precinct level, then you're predicting that there won't be a single precinct in all 99 counties in which any candidate gets 15% other than those three. Unless there's a second 15% threshold cut off once you get to the state level, then I don't see how no one besides these three can get delegates.

We have the same 15% threshold in Texas, but the cut off point is not at the state-wide level. This allows for a regional appeal like with suburban, small town, or inner city precincts.

[iowapeacechief]

21. One delegate won at precinct level...

...means a ranking at the state level; thus there will be a full range of percentages. That's why it matters if my guy (Kucinich) is expected to place dead last but manages to come in sixth. That will constitute his Iowa Surprise.

[iowapeacechief]

19. 15 percent or nothing

Delegate count is all that matters.

Any effort that does not yield at least one delegate counts ZERO.

Fifteen percent is the barest minimum needed to win one delegate. It may take lots more at many caucuses, e.g., if your precinct gets to choose only one delegate, you need 50-plus percent.

[ISUGRADIA]

23. An Example from a Caucus

There is no reporting of candidate votes, only delegates. And that vote is based on the 15% at least at a precinct. So a precinct that has this breakdown for example after all the realigning is done for those candidates who don't have 15%:

Caucus X in Y County 200 people turn out.

13 delegates, 15% needed for a delegate, 15% = 30

First Breakout

Dean 60

Kerry 39

Gephardt 37

Edwards 31

Kucinich 16

Clark 4

Lieberman 3

Sharpton 1

Undecided 9

All those in groups 30 or less have to reallign. A few go to viable candidates, the rest stay in a undecided group.<P>
2nd Breakout, with delegates.

Dean 60 4 delegates

Kerry 40 3 delegates

Gephardt 37 2 delegates

Edwards 31 2 delegates

Undecided 30 2 delegates

The break down of the 200 voters is not reported by the party as results. The only numbers the state party will report will be delegate numbers. So in this caucus 0 kucinich votes are reported even though 16 people came for him.

This is even more complex as those individual caucus delegate counts are apportioned among the delegates each county has been assigned the state party based on their perfomance for Tom Vilsack in 2002.

So county x with 25,000 people may have 35 state delegates while county y with 35,000 people may have 32 state delegates because it was a poor county for Vilsack.

Yeah, I know I may have lost some people along the way; the caucuses and the way results are reported are hard to explain. This said, there will not be a breakdown by the state party on caucus night of the 100,000 or 150,000 attendees by candidate, only a % of the 3000 state delgates elected. This is how it was done in 2000.
So statewide we will see reports of candidates of less than 15%.

Check out www.iowademocrats.org for details on the caucus process, caucus math and vote reporting.

[HFishbine]

29. That's a great example

Very clear. Thank you.

[AP]

7. Dean and Gep TIE! Kerry right behind them , and Edwards in 4th

with 15%.

[tedoll78]

8. Okay,

here's my revision:

Dean 41%
Gephardt 22%
Kerry 22%

If Dean is around 30% in the polls and many can change their minds, I'm betting that the sight of Dean with Harkin and Gore next week weight heavily on voters' minds.

[Padraig18]

10. You read it correctly.

Gore + Harkin with Dean = powerful message.


Iowa will be a blowout for Dean, IMO. :)

[Progressive420]

12. here we go

Dean 32% Gep 28% Kerry 20% and Edwards 20$

[deminflorida]

16. Dean will lose his front-runner status next week to Clark nationally

I wonder how that will effect Dean in Iowa?

[WI_DEM]

17. Just like Clark lost it to Dean?

It is worth remembering that when Wes Clark entered the race in September with tremendous publicity because of his "Draft" efforts and status as a TV commentator that he skyrocketed to the lead in many national polls and actually stayed there for a month or so, and then Dean started to climb.

[WI_DEM]

18. my guesses

Dean: 35%
Gep: 23%
Kerry: 17%
Edwards 14%
remaining 11% divided among remaining candidates.

[HFishbine]

20. I have the unique distinction

Of having the highest percentage for Dean. After the Harkin endorsement, I choose to leave my prdictions as they stand.

Dean: 52
Gephardt: 28
Kerry: 20

Thanks for keeping track of this for us, Bucky.

[Bucky]

22. How these predictions will be scored (to see if you're a smarty or not)

On Monday night I'm gonna follow the returns. Then I'm going to measure the returns against each predictor's numbers. I'm only accepting full numbers, but he returns will be measured in 10ths of a percent. That way there is little chance of there being a tie.

Example: Let's say the returns are these:
Candidate X -- 45.2%
Candidate Y -- 20.3%
Candidate Z -- 16.8%

If I predicted the returns for the top three (as per the original thread) as 50-25-15, then I'd get 4.8 points for X, 4.7 for Y, and 1.8 for Z. That's a total score of 11.3.

But then say you predicted 40-20-22. You'd get 5.2 for X, 0.3 for Y, and 5.2 for Z. Even though you got the wrong order of candidates (you thought Z would beat Y, but Z is a punk loser), your total is 10.7, so you just barely beat me.

Now almost everyone picked the same top three finishers. So you might be wondering what will happen if Kucinich or Edwards suddenly surges at the last minute and takes 3rd or even 2nd place. If that happens, well, then there was no point in us doing these stupid predictions.

[HFishbine]

24. May I make a suggestion

I think there should be reward for getting the order correct. I humbly propose this:

20 points for a correct 1st place pick who finishes 1st.
15 points for a correct 2nd place pick who finishes 2nd or better
10 points for a correct 3rd place pick who finishes 3rd or better.

(A trifecta earns 45 points.)

Then, from this number, subtract the "point difference" between predicted and actual results. A higher score is better.

To use your example above. The first guy got all three candidates in order. He's awarded 45 points. Subtract the point difference, and he ends up with 33.7

The second guy gets 20 points for his correct 1st place pick and ten points for his 3rd place pick who finished 2nd. (Zero points for a third place finish of a second place pick.) He's awarded 30 points. Subtract his point difference, and he ends up with 19.3.

What do you think?

[Bucky]

27. You system is more fair. My system is easier to do on a spreadsheet

So guess who wins, bubie!

[HFishbine]

28. Uh, oh

Looks like we're in for a shared national title. I'll offer my scoring system as an alternative.

Who deserves first place in this scenario:

ACTUAL RETURNS:

X: 34%
Y: 32%
Z: 28%

PLAYER 1:

X: 39%
Y: 28%
Z: 24%

(Bucky score: 13. Fishbine score: 32)

PLAYER 2:

X: 40%
Y: 28%
Z: 24%

(Bucky score: 14. Fishbine score: 31)

PLAYER 3:

Y: 34%
Z: 32%
X: 30%

(Bucky score: 12. Fishbine score: -2)

Under the Bucky system, Player 3 wins even though he didn't get anybody in correct order and even though he has the first place finisher in third. He beats the other two players who got the order correct.

Under the Fishbine system, Player one, who got the candidates in order and was only one point less accurate in percentage predictions than player 3, wins it. Player 2, who also got the candidates in correct order and was only 2 points less accurate in percentage predictions than player three, comes in second. Player three who was very close in his percentage predictions, but failed to get annybody in their proper order of finish, comes in third.

[Bucky]

30. Jeepers, Fishbine, don't you have a job? :o)

I would argue that the horserace aspect of who came in first and who came in second is a distortion of the value of the actual votes cast. Under your system the person who picks the 7th and 8th place finishers to come in 2nd and 3rd gets just as many starter points as the player who only gets the 2nd and 3rd place finisher in the wrong order.

If the number 2 slot is really competitive, then getting the order reversed for two candidates isn't that big an error. I'm interested in total volume of support for each candidate, not the relative polling positions between the candidates' final tallies.

If you predict the percentage of votes earned within only a tiny difference between prediction and outcome for one candidate you should be penalized based only on how far off the mark you are. You shouldn't get penalized a second time for a larger miss in guessing the outcome for another candidate. Under the Fishbine system, if I'm way off on all my predictions but I'm still in the right order, I'm still getting a pretty decent score.

Like say I guess 60-15-10, but it's really a 35-30-25 outcome. I got the right order and I've beaten someone who guessed 30-32-31. Yet who is the better prognisticator?

[HFishbine]

31. You Aren't Quite Getting It

"Under your system the person who picks the 7th and 8th place finishers to come in 2nd and 3rd gets just as many starter points as the player who only gets the 2nd and 3rd place finisher in the wrong order. "

That's incorrect. Under my system, the person who picks the 7th and 8th place finishers to come in 2nd and 3rd gets 20 starter points (assuming he got 1st place correct). The person who gets the 2nd and 3rd place finisher in the wrong order gets 30 starter points (assuming he got 1st place correct) and, when you subtract Bucky points, the margin of difference between the two widens even more.

"Like say I guess 60-15-10, but it's really a 35-30-25 outcome. I got the right order and I've beaten someone who guessed 30-32-31. Yet who is the better prognisticator?"

Actually, you don't win in that scenario. You would have -10 under the Fishbine system and your opponent would have -3. The better prognosticator does indeed win.

Your system completely disregards any reward for order of finish. The Fishbine system rewards accuracy AND the order of finish. Yours is like predicting the total points scored in a basketball game, mine is like predicting the total points scored and picking the winner. Which is the better test of predictive ability?

[Bucky]

32. We're not gonna argue about the designated hitter rule next, are we?

Fine. There's gonna be two titles then. Gross Points accuracy (Rea system) and Adjusted Placement accuracy (Fishbine system). The two winners will then square off for New Hampshire predictions to determine who's the national champeen. We can call it the "Final Stand in Yankeeland" like Don King and sell tickets. Or maybe call it the "Guesser in Manchester". Or "Prognosticate for the Granite State".

PM me if you want me to send you the spreadsheet file. I really can't figure your system out.

[HFishbine]

34. Of course

It's not out of the question that someone could win both. This is going to be fun.

[ISUGRADIA]

25. My Predictions: Dean 32, Gephardt 23, Kerry 22, Edwards 10...

Dean 32%

Gephardt 23%

Kerry 22%

Edwards 10%

Undecided 7%

Kucinich 4%

Clark 1%

Lieberman, Sharpton, Moseley-Braun 1% combined

[slinkerwink]

26. okay, with the Harkin endorsement and Gore stumping for Dean

this weekend, I think it'll be more like 38% for Dean, 26% for Geppy, and 25% for Kerry.

[slinkerwink]

47. that's my story and I'm sticking to it!

[Paulie]

33. No change in mine

I think we'll see some happy Kucitizens. :)

[slinkerwink]

35. kick

[goobergunch]

36. You miscopied "Goobergunch (new)", which I still stand by

Dean: 40%
Gephardt: 30%
Kerry: 18%

http://www.democraticunderground.com/discuss/duboard.php?az=show_topic&forum=132&topic_id=77987#78505

That remains my prediction.

[Bucky]

37. D'oh! Sorry...

It's now corrected.

By the by, I've always wanted to thank you for posting the C-Span links in your sig file. I never knew about that service till I saw your messages. I don't have a TV but every now and then CSpan has something I wanna watch. Now thanks to you I don't have to go bother my ex-GF just to see Booknotes or Road to the White House.

[hellhathnofury]

38. My revision:

Dean 39% Gep 32% Kerry 14%

[NewYorkerfromMass]

39. I predict a perfect muddle

Dean 25% Kerry 24% Gephardt 22% and Edwards 20%

[TexasPatriot]

40. i didnt have a first prediciton...

but its early still...

Gephardt 36
Dean 34
Kerry 15
Edwards 15

[Bucky]

44. Well, actually you did. (plus a kick)

This thread moved to page 18 in just two days. My but we're busy round here. Tex-Pat, you previously predicted Gep-34, Dean-32, Kerry-16. That clearly leaves room for Edwards to nuzzle in there, as you seem to be predicting. OK, changes noted.

OK, consider this dog :dem:'ed

[windansea]

41. nailbiter!!!!!!!!!!!!!!!

Gep 32%
Dean 28%
Kerry 21%

[askew]

42. New picks

Well, with the news of the Harkin endorsement and the Edwards endorsement, I am going to go out on a limb here with my new picks:

50% Dean
20% Edwards
18% Gephardt

Everyone else less than 15%.

[anti-bush]

43. Keeping Same Picks - G-24 D - 22 E- 20 K-16

And I am still cautiously optimistic that Edwards will do even better than second.

[MrMoJoeRisin]

45. I think Joe is gonna surprise a lot of people with a strong showing in IA

he was very impressive during the recent debates.

[Bucky]

46. Nice try

But that is like so Tuesday... :eyes: