, , 1

Is appraisal based on science?

  • No. It’s an “opinion of value.”
  • Yes. It’s a systematic method to learn about something.

Looks like it’s both!   – – – But wait!

Is every opinion based on science?  No.  Does every method result in an opinion?  No and no.

So now we know how to talk about this.  An opinion may be based on systematic science.  Or not.

The “appraisal” method says form an opinion, and support that opinion in your report.  Support with comps, with adjustments, and by reconciling/explaining why you got different answers.

Does every systematic method result in an opinion?  No.

An AVM (Automated Valuation Model) does not provide an opinion.  Any algorithm only provides a numerical result.

So now we know that all-science (algorithm) or all-judgment (opinion) are probably dumb ways to do things.  But we have yet one other way.  Science (systematic method) merged with opinion (judgment).  Sounds better already.

What would this merged model look like?

The modeler (appraiser) would have to be competent.  This would involve three things:  1) knowledge of the systematic method; 2) knowledge of the market segment; and 3) ability to merge the two competencies – through critical thinking, and the underlying algorithms (the math/statistics).

This merged model (process) is no different from what has always been done.  However, it does optimize the data used, and sharpens the expert appraiser’s judgment.  Today’s merged model simplifies things!  The process is straightforward – the science of data analysis.  The five steps are:

  1. Frame the problem: the question asked, the probable path (abductive reasoning), the probable data set (the assignment data frame), and assumptions, including word definitions.
  2. Collect (download) the data, explore the data, and wrangle (clean and transform) the data. This comprises defining and quantifying the subject, as well as competitive market data.
  3. Reduce the overall Assignment Data Frame (ADF)© to just the relevant Competitive Market Segment (CMS)© and optimize on the principle of the “bias-variance tradeoff.”
  4. Predict to the result, using one or more of the three principle predictor/adjustment algorithms. Traditional software may be used, as well as Ai (artificial intelligence) prompt principles.
  5. Deliver (report) the result(s), to match the aptitude of the intended user, using tables, graphs, maps, and especially explanations of each judgment call.

The above process is the essence of Evidence Based Valuation©.  This is the process technology as promulgated by the Community of Asset Analysts (CAA)©.

We believe the future is a blend of AI, data science, and expert judgment.  These form the core of the Valuemetrics.info curriculum, as in the Stats, Graphs, and Data Science1 14 hour appraiser intro course.