What AI Is Telling Us? There Is Still No Moneyball in CRE.

August 26, 2026

Six years ago, I wrote an article arguing that there is no Moneyball in commercial real estate.

The premise was simple. Baseball lends itself beautifully to statistical analysis because it produces enormous quantities of standardized data. Commercial real estate does not. Every property is unique, markets are hyperlocal, transactions are relatively infrequent, and many of the variables that matter most are difficult to quantify.

Six years and one AI revolution later, I think the argument still holds. But the implications have changed.

AI gives us extraordinary new tools to gather, organize, summarize and analyze information. Tasks that once consumed hours can now be completed in minutes. Extrapolating this trend, it’s obvious that, over time, the rote work involved in commercial real estate analysis will disappear.

But faster analysis does not necessarily produce better answers.

Put good AI on top of bad data and you simply get to the wrong answer faster.

That makes the human task of data curation more important, not less.

Consider something as basic as comparable sales. An AI model can ingest hundreds of transactions and identify patterns almost instantaneously. But should all those transactions be considered? Was one sale distressed? Did another involve an unusual buyer motivation? Is a nearby property actually competitive, or does an invisible boundary separate two very different submarkets? Does a reported cap rate reflect stabilized income, trailing income, or a buyer’s expectations about future performance?

The difficult part is no longer processing the information. It is deciding what information matters.

Real estate is particularly resistant to pure algorithmic analysis because buildings exist in the physical world and transactions occur between human beings. AI does not walk a property, experience a neighborhood, sit across the table from a buyer, or understand why two apparently similar assets may trade very differently. It can identify correlations across enormous datasets, but correlation without context can create false confidence.

And AI introduces another problem: provenance. Large language models can synthesize astonishing amounts of information, but their answers are only as reliable as the information underneath them. A polished response assembled from stale records, inconsistent definitions, unverified sources, or internet commentary is still a bad answer.

For CRE firms, the competitive advantage is therefore shifting.

The old advantage was access to information. The emerging advantage is disciplined curation and contextual judgment: maintaining clean data, understanding its provenance, identifying anomalies, rejecting false comparables, and recognizing when listings or unusual circumstances surrounding a transaction distort the market evidence.

AI will increasingly handle the mechanics. Humans will be responsible for deciding what deserves to go into the machine and whether what comes out makes sense.

There still isn’t a Moneyball for commercial real estate.

But there is an enormous opportunity to combine AI’s ability to process information with something much harder to automate: the judgment to know which information matters.

📬 Let’s Talk

Capright specializes in institutional valuation and advisory services throughout the United States. As a truly independent third-party, Capright has emerged as the leading provider mark-to-market valuations for many of the largest privately-owned commercial real estate platforms. In addition to valuation services, Capright offers NAV calculation, daily pricing, valuation process implementation, debt mark-to-market, option and JV interest valuation, compliance review, audit assistance, litigation support, and managed services. In partnership with several of the largest institutional investment managers and data platforms, Capright has pioneered industry-leading analytical tools to set higher levels of accuracy and credibility for assets requiring higher-frequency valuation compliance.

For more information or if you would like to discuss this article, please reach out to: