What went wrong?  How did a vibrant profession of experts in property value begin to disappear?  Who’s at fault?  Is it the appraisers?  The Standards?  The “appraisal process”?  The regulators?  Or could it be even the users and clients?

Inertia does not come easy.  Change is even more difficult. Inertia requires defending the old way of doing things.  It requires avoidance, a conscious avoidance or unconscious avoidance – and denial.

The  Problem?

The new replaces the familiar.  Blacksmiths get replaced by auto mechanics.  Then experienced mechanics get replaced by those who had electronic diagnostic training.  Then the problem part gets replaced.

The Solution?

Learn electronic diagnostics – and where to get the replacement part fast.  Oh yea – and check out for any other worn parts which may need to be replaced.  And oh yea – and wash the customer’s car before they come back to pick it up.

The point is:  Help the customer get what they really need.  (Not what they may think they need).  Part of professional behavior is diagnosing the problem. 

So let’s look at what went wrong

I believe we’re so entranced by our own deep competence in the old way of doing things, we forgot to figure out what was needed.  In particular, we may have forgotten that our status as a protected profession is solely “for the public good.”  Yet even with repeated economic melt-downs, we continue to provide what economists call exchange price.  We call it “market value,” and opinionate a market price.  That works – until it doesn’t.  Then everyone points fingers. 

This has been problem #1. 

Exchange price need not equal utility value.  Are we are solving the wrong problem? What went wrong? Things changed.  The problem changed.  The solution changed.  The technology changed. 

Problem #2:

In the face of instant and usually complete market information, we have been taught to discard all but a handful of “comparables.”  Our education teaches us that a comparable is competitive to the subject property.  A property (sale) is competitive if it is similar.  Then it’s explained that a property is similar if it competes with the subject.  Just pick four or five, form an opinion, and support that opinion.

A subjectively selected, partial data set – precludes objective results.

Problem #3:

Statistics became touted as an “advanced” method.  Specifically, the old random sample statistics was what was hyped.  Just as we got complete data – we began to learn perplexing inferential statistics based on random sampling.  Then it got worse.  Random sampling was ignored, but the subjectively selected, partial data set was imagined to somehow be random.  Thus, you could compute clever advanced statistics (such as p-values, chi-squares, hypothesis tests, and confidence intervals) – all proving how really good your data model is.  Gobbledygook.  Advanced gobbledygook.

A clever solution which does not answer the question asked.

Problem #4:

Multiple regression became touted as part of the future.  You could take a bunch of data, throw it into a spreadsheet add-on, get a clever sounding “high” R2 and get an answer.  Somehow this never really worked.  But we were given “advanced” classes, many hours in length – where somehow we could match what takes several years of graduate level econometrics courses to apply effectively.  We were warned:  “You cannot use a regression coefficient as an appraisal adjustment.”  Yet purveyors of appraisal regression software continued to claim such things as “my regression is better than their regression.”

A solution which requires education not canned software.

 Problem #5:

Residential appraisers are forced into a box of box forms.  General (non-residential) appraisers are forced into a box of boxes – called a “spreadsheet.”  Each of these was designed for the way data used to be, and the limitations of the way computers used to be, with limited data, limited power, and limited visualization.

Today we have complete data sets available.  We can objectively identify complete data sets from complete competitive market segments.  We can connect the big and small data to the human brain.  We can create a reproducible work product.  A product which provides an estimate (score) of reliability, similar to the forecast standard deviation used to evaluate AVMs (Automated Valuation Models).

We can connect the value prediction and reliability score to the client needs for collateral assurance, investment potential, and litigation support evidence.

December 8, in San Diego — inaugurates the first known class for valuers in the tool of the future.  It’s open only to graduates of the Valuemetrics.info class Stats, Graphs, and Data Science.  Why?  Because the proper application of the new tool requires knowing when and how to use it.

Modeling knowledge comes before tool application.