Computational Complexity

Wednesday, September 03, 2008

 
I would bet on INTRADE that INTRADE will do badly picking VP Nominations

Posted by GASARCH

(Lance and I independently made a post on VP selection. This post is not related to his, nor is his related to mine. Dave Barrington helped me with some of the history in this post, as did some folks in their 80's who assure me that Nixon was a surprise VP pick.)

The day before McCain picked Palin a pundit said the following: I don't know who it will be but INTRADE has had a big spike for Romney. INTRADE is always right, hence I predict that some insiders know that its Romney and that is who it will be

Well, INTRADE is not always right, even before the Palin Pick. And would an insider be guilty of insider trading? I predict that in the future VP will be one of those things INTRADE does badly on. Why? Because it is idiosyncratic. Picking the Prez Nominees is like picking the Oscar: small number of possibilities, and one has a sense of things. Picking the VP is like picking what movie Lance Fortnow favorite movie of 2008: too many possibilities, too ill defined (what if he saw a movie made in 1998 in the years 2008, does that count?) and too dependent on his mood.

In the past there was no INTRADE, but there was a short list of VPs. Below is a list of VP candidates (not including incumbents VPs) and whether I think they would have done well on INTRADE. My speculation is based mostly on if they would have been on the short list. I also include the Prez Candidate, the party, and WON/LOST. BADLY means would do badly on INTRADE. GOODLY means would do goodly on INTRADE. OKAY is inbetween.

1. 2008: McCain picks Palin. Rep. BADLY.
2. 2008: Obama picks Biden. Dem. GOODLY.
3. 2004: Kerry picks Edwards. Dem. GOODLY. LOST
4. 2000: Gore picks Lieberman. Dem. OKAY. LOST
5. 2000: Bush picks Cheney. Rep. BADLY. (Cheney was head of VP selection committee, so really Cheney picked Cheney.) WON.
6. 1996: Dole picks Kemp. Rep. BADLY. LOST
7. 1992: Clinton picks Gore. Dem. OKAY. WON
8. 1988: Bush picks Quayle. Rep. BADLY. WON
9. 1988: Dukakis picks Bentson. Dem. OKAY. LOST
10. 1984: Mondale picks Ferraro. Dem. BADLY. LOST
11. 1980: Regean picks Bush. Rep. GOODLY. WON
12. 1976: Ford picks Dole. Rep. OKAY. LOST
13. 1976: Carter picks Mondale. Dem. OKAY. WON
14. 1972: McGovern picks Eagleton/Shriver. Dem. BADLY. LOST
15. 1968: Nixon picks Agnew. Rep. BADLY. WON
16. 1968: Humphrey picks Muskie. Dem. GOODLY. LOST
17. 1964: Goldwater picks Miller. Rep. BADLY. (Miller was in House not senate, so a surprise.) LOST
18. 1960: Kennedy picks Johnson. Dem. GOODLY. WON
19. 1956: Stevnson picks Kefauver. Dem. GOODLY. LOST. (Was serious contender for nomination.)
20. 1952: Stevenson picks Sparkman. Dem. BADLY. Speculation- (Was not a contender for nomination.) LOST
21. 1952: Eisenhower picks Nixon. BADLY. He was a 39 years old unknown at the time and a surprise. WON

1. 10 BADLY, 8 GOODLY, 5 OKAY. INTRADE usually does much better than this.
2. Dems: 6 GOODLY, 3 OKAY, 3 BADLY.
3. Reps: 1 GOODLY, 1 OKAY, 6 BADLY.
4. WINNERS: 2 GOODLY, 2 OKAY, 4 BADLY.
5. LOSERS: 3 GOODLY, 3 OKAY, 5 BADLY.

The winners/losers things may be unfair since these were all non-incumbents; however, some of the incumbents lost (Bush-Quayle and Carter-Mondale) so I leave it be. The stat I find most interesting is that INTRADE doesn't do that well here. Or wouldn't have. I may return to this study 10 years from now when I have real INTRADE data.

10:26 AM # 2 comments

1.  Anonymous says:  
   
   Um, "WELL"?
   

   8:18 PM, September 03, 2008

2.  Dave says:  
   
   As far as I know there is no insider trading prohibition on intrade. In general, when the goal is good predictions, insiders should be encouraged to trade.
   

   5:09 PM, September 04, 2008

Comment Feeds: This Post All