Nobody can accurately predict appreciation, particularly several years out. 

And predictable and consistent in the amount and timing of late payments (so at least it's a known.)

Mechanicals - The core set of AC, Electrical, and plumbing are pretty easy to predict. 

(I'm kind of copying @joederobertis' other post asking for predictions about beachcities poised to boom.)Since so many neighborhoods in LA have been gentrifying recently into "cool yuppy/hipster" areas.  

For example, someone could present to me an investment with very attractive pro forma returns but, if the predictions include rent increases beyond what evidence supports, I would question whether this person takes their responsibility seriously.

And with timber, just like Self-Storage, you can conduct a feasibility study to fairly accurately predict demand, and model a business based upon historical market data to derive projections for the project.  

Fico scores and what goes into them and what makes them move up or down is well studied and predictable at this point.

I'd be crazy to think that new models wouldn't arise, but I certainly don't know what those models are (just like I couldn't predict new methods of communication (the Internet), new methods of shared transport (Uber) or new methods of short-stay housing (AirBnB)). 

Until newbies learn to make a "more astute analysis", it's probably best for them to stick to less risky / more predictable options.

Obviously I cant predict everything but the savings and e-fund have helped me gain wiggle room in my monthly budget, enough that I'm ready to take this plunge.