, 0

The Stats, Graphs, and Data Science classes are appraisal science.

Why should an appraiser care about appraisal science?  Is there an art part?  Or is it a hybrid?

The appraiser’s future is appraisal science, and sharpened judgment with AI assistance. TYN2K The Webinar with George Dell Did you miss our live webinar? Buy Now. AI for Appraisal with Chandra Mast and Justin Gohn

Appraisal science:

  • Appraisal is an opinion of value, or commonly, “the report.”
  • Science is a “systematic study” through observation and evidence-based reasoning.
  • Art reflects human creativity through ideas, emotions, and beauty, with skill and imagination.

If we are to believe that appraisal is both an art and science, then it is worthwhile to look at these two worlds, to see how they compete!  In coming issues, we will explore the use of AI in valuation.

The Stats, Graphs, and Data Science classes focus on evidence, reasoning, and interaction with your expert judgment.  Thus, you exploit the power of modern computation through evidence-based valuation.  Please note that the word “statistics” is not part of the title.  Why?

  • Stats: Appraisers do not need the mystical random-sample inferential statistics!   However, valuation analysts do need to be comfortable with the use of summary numbers from the competitive comparable data set – descriptive measures of groups of data – like the mean, median, range, quartiles, and a couple of similarity measures.  That’s it!
  • Graphs: Visuals provide the way for your expert brain to understand markets as a whole. “Comps” are a way to do that, and work well where the market is sparse.  Today, graphs are created in seconds.  Brain-understanding is the art.  The art!
  • Art: A vast field expressing ideas, emotions, and beauty.  (Per Google©.)

The science of data – “data science” – is the foundation of EBV (Evidence Based Valuation).  Here is the clincher: “statistics” and data science are not the same!  Statistics is theoretical.  Inferential statistics assumes you are representing a “population” with a random sample.  “Appraisers don’t do no random samples!”

Data science differs from statistics in two major ways:

  • It explicitly recognizes the need for the “field-related expert – YOU.
  • It works with complete (or substantially complete) data – the whole relevant “population.”

It is easier, faster, and more accurate than the legacy “pick comps, then adjust” habit.

Instead, the analyst does this:

  • Establish the scope of work, subject features, and download the entire potentially useful data.
  • Reduce (“wrangle”) the data, including CMS (Competitive Market Segment) parameters.
  • Identify/weight the influential predictors (the “elements of comparison”).
  • Select relevant model(s) and apply the appropriate algorithm.

To sum up:

EBV© (Evidence Based Valuation):

  • Is factually/logically more exact and accurate.
  • Requires professional judgment beyond just “experience.”
  • Is Complete and optimal to start, so does not need to be “defensible”.
  • Is more fun!
  • AI needs EBV, and makes it easier.

At George Dell’s Valuemetrics Appraiser Education, we analyze markets, not compare comps. Our next class is coming soon. For more information, visit, Valuemetrics.Info.