New York Times’ syndicated error about credit scores


From: Greg Fisher []
Sent: Tuesday, July 02, 2013 1:51 PM
To: Tom Troy, reporter, Toledo Blade (Block Communications)
Cc: Tim Grant, reporter, personal finance, housing and banking, Pittsburgh Post-Gazette (Block Communications); Ignazio Messina, reporter, Toledo Blade (Block Communications); John Robinson Block, publisher and editor-in-chief, Pittsburgh Post-Gazette (Block Communications); Allan Block, chairman, Block Communications (via S. Smith)
Subject: RE: credit score, math, Block Communications, election, 0.86 factor II

Please reply.

Greg Fisher
The Credit Scoring Site
Page A2
PO Box 342
Dayton, Ohio  45409-0342

From: Ignazio Messina [Toledo Blade reporter]
Sent: Wednesday, July 03, 2013 10:13 AM
To: Greg Fisher
Subject: Re: credit score, math, Block Communications, election, 0.86 factor II

Sorry for the late reply as I have gotten a larger volume of emails this week than most. As we said, the conversion is not perfect, and provides a rough estimate so people can have an “apples to apples.” The most recent article just put out the scores with out conversion. If you have an insight, please share.


From: Greg Fisher []
Sent: Monday, July 08, 2013 12:48 PM
To: Allan Block, chairman, Block Communications (via K. Franck); Allan Block, chairman, Block Communications (via S. Smith)
Cc: Ignazio Messina, reporter, Toledo Blade (Block Communications); Mary-Beth McLaughlin, reporter, assistant city editor, Toledo Blade (Block Communications); Arthur Sulzberger, Jr., publisher, New York Times
Subject: RE: credit score, math, Block Communications, election, 0.86 factor II, lack of sources

You compare apples, oranges and passenger rail: A FICO score, the VantageScore and the Equifax Credit Score.  But thank you for eventually replying and saving me a 5 a.m. wakeup call.

I have no formula to equate one credit score with another.  And, even though VantageScore recently announced that it would reconfigure its scale to 300 to 850, that does not mean that the curves of a FICO score model with the same endpoints has the same shape, or that the probability for default at 700 is same for both brands of credit scores.

Regarding your conversion formula, an error of 24 percent (131 on a 551-point scale) is, indeed, rough.  Lest I attribute that inaccurate conversion method to you, I asked for your source.  What is the name of your source?  As you can see, in a story from two years ago, your reporter did not provide a name (even after I asked for the source), either—and it’s not like these are matters of national security.

The blithe notion that such a formula can be so simple does not appear to be an original thought.  Providing the name of your source so that I can ask them about the calculation may help another unwitting reporter from being duped.

Your casual translation notwithstanding, the point is moot because you confused one score with another.  You indicate that one candidate for mayor of Toledo had a credit score of 635 “according to Equifax credit-reporting companies, using the FICO model that runs from 300 to 850.”

However, the scale of that candidate’s credit score that you published starts at 280, not 300 (and that fact is not hard to find; it is notated on the next line of the credit report, directly below the score that you cited).  It was the Equifax Credit Score, not a FICO, and that means that the range is 571, as opposed to the range of the FICO, 551.  So, the candidate’s place on the continuum is not 335 points from the lowest score (as you reported), it is 355 points from the lowest.

At this point, I am not as concerned with credit score comprehension, interpretation, bad math logic and an unusual local election tradition as I am with a shadow that I have chased for 5 years.  In 2005, with no substantiation, you reported, “Increasingly, though, such scores are used by landlords, potential employers, and insurance companies to determine someone’s financial health.”

That typical fear-mongering word-series setup (the lions-tigers-and-bears line of this pathetic chapter in journalism), leads many articles about credit scores in order to shock the reader and win attention.  But it isn’t the truth.

As you can see in this email thread, I originally contacted you two years ago—enough time for you to name a source or make a correction.  Subsequently, you even republished a New York Times item from late last year that also makes the inaccurate claim that employers use credit scores.  The downside of syndication is the syndicated error.

Belief of this myth has had serious consequences in other parts of the country, and I want to end that nonsense before it comes to our statehouse again.  Employers do not use credit scores.  I looked into it.

Not many seemed to care about credit scores when I registered the domain 14 years ago.  Similarly, today, few seem to care about corrections on Page A2 of the nation’s newspapers.  That is why I registered—to again address a commonplace problem.  The website at that address deals with the antithesis of the burden of proof, standards of evidence and naming sources: Proving what is not true.  It is disproving the conventional wisdom when affirmative statements are a lot of balderdash.  Credit scores are a convenient vehicle.

The only response I expect from you—and I expect it today—is correction of your errors.  In the case I brought to your attention in 2011, two years is enough time to allow you to do something about it.  Given the demonstration of the fallibility of the New York Times, their three-word name is now only a cliché.  Don’t allow that organization’s error to remain on your website and misinform another person.

Greg Fisher
The Credit Scoring Site
Page A2
PO Box 342
Dayton, Ohio  45409-0342