Say I measure a wall. Disto says 30.1'.
I scan using cubi casa and it comes back 30.8'.
ANSI says to round to the nearest tenth. So if my disto reads 30.12', you would round to 30.1'.
So that is how .8 is not as accurate.
You have feet, tenths and hundreds. You round hundreds to the nearest tenth of a foot. You do not round to the nearest foot. 30.5' is not rounded to 31'. No offense, but if an appraiser pulls a tape on a wall and cannot measure it to an inch....wait that's why pdc use cubicasa. They couldn't measure a home if they tried.
Scan to sketch needs some major improvement. Just not accurate enough for me. The tech guys at alamode are great, but it is not there yet.
Cubicasa will cost me $15 a pop and takes 24 hrs, which is not great. It's somewhat accurate, but define what is accurate enough.
Say I come in low on an appraisal. CUBICASA is off by 70 sqft. That could be $5,600 to a comp. If it's a hybrid what do I do?
That's my point will all of this new tech. Cubicasa, regression, third party data, etc.
USPAP’s Standards Rule 1-4 requires appraisers to collect, verify, and analyze all information necessary for credible results.
Does the pdc know how the black box cubicasa works? Have they ever verified or would they know how to verify or how to know when it might be off?
That's my point. If appraisers start using cubicasa are we going to get the same free pass? It's accurate "enough".
Advisory Opinion 18 (AO-18) in USPAP provides guidance on the use of AVMs, emphasizing that appraisers must understand the model’s functionality and ensure its appropriateness for the assignment.
Do appraisers need to know how cubicasa works? How regression works?
If cubicasa is not accurate and regression analysis is not accurate, can we use it?