New Web site math challenged

Odds are that Colorado Pols won't be taken very seriously

Aug. 13, 2005

by David Kopel

Peter Blake of the Rocky Mountain Newshas been writing an excellent insider column about Colorado politics for years, and The Denver Post'sDan Haley has recently started a similar one. But where do you go if you want even more Colorado political news and gossip? Colorado readers have two major choices, each with its own virtues and flaws.

For most of the last three decades, The Colorado Statesmanweekly newspaper has been required reading for the state's political professionals. Printed copies of the Statesmanhave always been ubiquitous in the state Capitol and nearby buildings, while newsstand copies are nearly impossible to find anywhere else. The Statesmanis now offering a small amount of its content for free on its Web site,,but almost all the content still requires a subscription.

(Disclosure: my father, Jerry Kopel, has been an award-winning columnist for the paper since 1992.)

This year, the Statesmanis facing serious competition in its niche for the first time ever. The Web site provides daily rumors - some of them true - and commentary for Colorado political junkies. Much of the site's content is simply musing about political stories which have appeared in the Newsor the Post,but the site does have enough original information to make it worthwhile for people who live and breathe politics. And daily information can be much more useful than information delivered every week or two.

However, one part of Colorado Pols that can't be taken seriously is the front-page "line" which the Web site offers on various elections. "Think of it like a betting line," the site claims, "with these being the odds you'd get if you were laying down money today." Not really. In the British bookmaking parlors that take bets on American elections (or in any other casino which hopes to attract rational gamblers), the combined probabilities for every possible event add up to approximately 100 percent. After all, it is 100 percent certain that someone will win the election, so the combined sum of the election probabilities for the various candidates should be close to 100 percent. (Not counting the slight adjustments in the odds necessary to create the "house edge" so that casino makes money no matter who wins.)

Yet consider ColoradoPols' "line" on the 3rd Congressional District race. Incumbent Democrat John Salazar is given 2:1 odds; in other words, Colorado Pols asserts that Salazar has only a 33 percent chance of being re-elected, an amazingly low estimate for an often-moderate incumbent who has encountered no serious problems in his first term.

Salazar's probable opponent, Republican Scott Tipton, is given 8:1 odds, meaning that he has only an 11 percent chance of winning. So according to ColoradoPols, there is only a 44 percent chance that either the incumbent or his likely major-party challenger will win the election. The preposterous implication of the Colorado Pols betting "line" is that there is a 56 percent chance that someone else - not Salazar or Tipton - will win the 3rd District race.

The "line" for the other congressional races is almost as silly. According to ColoradoPols, there is only a 68 percent chance that any of the major candidates will win in the 7th (where incumbent Bob Beauprez is not seeking re-election), and merely a 61 percent chance that a current known candidate will win in the 4th (Marilyn Musgrave's seat).

For a ballot issue, the probability that a measure will either pass or fail is 100 percent, so the probability of it passing plus the probability of it losing should total 100 percent. Yet for Referendums C and D, ColoradoPols claims the odds are 4:1 (20 percent) that the "no" voters will win, whereas the odds are 6:1 (14 percent) that the "yes" voters will win. So the ColoradoPols odds give us, in essence, only a 34 percent chance that C and D will win or lose, and an implicit 66 percent possibility that something else will happen - perhaps the election will be canceled.

I wouldn't take ColoradoPols' election forecasting - or anything else on the Web site involving math - very seriously.

Kudos to Postreligion writer Eric Gorski for an excellent investigative series about the Rev. Harold White, who may have used his position as a priest to molest boys for more than two decades while he was simply transferred from one parish to another after each new accusation of abuse.

Contrary to what appeared in my last column, the Web site is not affiliated with the University of Colorado. Rather, the site is part of a private business that provides specialized sports news, especially about recruiting, for many college teams. On Aug. 4, it was announced that FOXSports will acquire the company.

Dave Kopel is research director at the Independence Institute, an attorney and author of 10 books. He can be reached at .

Share this page:

| More


Kopel RSS feed Click the icon to get RSS/XML updates of this website, and of Dave's blog posts.

Follow Dave on Twitter.

Search Kopel website:

Make a donation to support Dave Kopel's work in defense of constitutional rights and public safety.
Donate Now!

Nothing written here is to be construed as necessarily representing the views of the Independence Institute or as an attempt to influence any election or legislative action. Please send comments to Independence Institute, 727 East 16th Ave., Colorado 80203. Phone 303-279-6536. (email) webmngr @

Copyright 2014