Regression Analysis

[hastalavista]

Regression has its place.
The quality of the data is key.

I rely on it much of the time for GLA and market-condition adjustments.
Sometimes a much simpler version (putting two different sets of data on the same chart and trending the differences) can be very effective (a picture is worth 1,000 words sometimes.... coming from a guy with many posts of 1,000+ words! ).

Its like any other tool; works very good and is very effective for some problems, not so good/ineffective for others. I'd rather know it and use it when I think it is worthwhile than ignore it or not know it at all.

 

[timd354]

View attachment 32091

Here is a paired sale. It tells me that prices have been stable in this neighborhood between 2015 and 2016. It also tells me C2 vs C1 requires no adjustment when almost new.
Wow, you think that a single paired sale by itself is enough data to conclude that market conditions that neighborhood are stable? I don't think so.
View attachment 32092

Here is another paired sale. It tells me that prices have been stable in this neighborhood between 2014 and 2016. It also tells me C2 vs C1 requires no adjustment when almost new.
See above comment
View attachment 32093

Here is another paired sale using sales in rowhouse condo conversion. It shows me that the middle level units sell for about the same. It also tells me the basement level units sell for $100k to 130k less than middle level units. It also tells me the top floor unit with roof deck sells for about $150k more than the middle level units.
The single basement paired sale may tell you that basement level units sell for $100k to $130k less than the middle level units or it may tell you that the the market has changed in the 8+ months since the middle level units sold, or it may tell you that the basement sale needs more investigation regarding seller motivation. This single paired sale, without additional data and analysis does not provide enough information by itself to adequately support your conclusion
View attachment 32094
Here it is again in the building next door that is new conversion. The units are about 10% larger and is the same developer. It tells a similar story. as the last grouping.
Now you are starting to show a trend...but apparently what you don't realize is that when you do paired sales analysis, you are actually doing a very simple form of regression

I can do paired sales all day and people can understand. Regression, R value T value blah blah blah. BS.

Just because you apparently do not have a sound understanding of regression analysis and cannot explain it in an understandable manner does not mean it is not a useful tool. The fact that you call it BS is all I need to know about your lack of understanding.

 

[timd354]

I'd rather know it and use it when I think it is worthwhile than ignore it or not know it at all.

Apparently, many people are more than willing to wallow in their ignorance. Doing the work to gain a good understanding of regression is hard work for most people, especially those who do not have a strong background in mathematics and statistics. It is easier just to be lazy and dismiss it out of hand as being BS.

 

[timd354]

Regression needs a lot of data. With enough data for regression there is too much noise or differences between the properties in the set. Useless.

This statement shows that you don't understand how and why regression analysis works when done properly. It is precisely because there are differences between the properties in the data set that regression works.

 

[Joe Flacco]

Just because you apparently do not have a sound understanding of regression analysis and cannot explain it in an understandable manner does not mean it is not a useful tool. The fact that you call it BS is all I need to know about your lack of understanding.

I understand the idea of it and that's enough. If you are trying to determine adjustment for attached vs semi-detached is the data going to consider things like if it is semi-detached to alley, semi-detached with a 2 foot gap with the property next door or semi-detached to street? If it has windows on the detached side or not? No. because that kind of data is not even being collected.

 

[Joe Flacco]

You can't beat paired sales for market based adjustments.

 

[timd354]

Appraisers should use old fashioned methods for developing adjustments.

Yeah, keep using just the old fashioned methods for developing adjustments that were contrived more than 50 years ago because nobody possibly could have different and better methods than what was done way back then....lol.

Don't worry about it when you are replaced because the largest intended users of residential appraisals (the mortgage market) sees you and other appraisers who stick solely to the old fashioned methods as dinosaurs

 

[timd354]

You can't beat paired sales for market based adjustments.

Keep telling yourself that. I don't deny that paired sales are a useful tool under certain circumstances, but in many areas and for many appraisal problems, paired sales simply do not exist (especially when the appraiser is forced to use comps that require significant across the board adjustments).

You know what I really love? I love it when the appraiser cannot provide a comp with a particular property attribute because there have been no sales with that particular attribute and the appraiser claims that the resulting across the board adjustment was derived from a paired sale.

 

[Joe Flacco]

Yeah, keep using just the old fashioned methods for developing adjustments that were contrived more than 50 years ago because nobody possibly could have different and better methods than what was done way back then....lol.

Don't worry about it when you are replaced because the largest intended users of residential appraisals (the mortgage market) sees you and other appraisers who stick solely to the old fashioned methods as dinosaurs

50 years ago or not it is the best way still to do it.

 

[Joe Flacco]

Keep telling yourself that. I don't deny that paired sales are a useful tool under certain circumstances, but in many areas and for many appraisal problems, paired sales simply do not exist (especially when the appraiser is forced to use comps that require significant across the board adjustments).

You know what I really love? I love it when the appraiser cannot provide a comp with a particular property attribute because there have been no sales with that particular attribute and the appraiser claims that the resulting across the board adjustment was derived from a paired sale.

Well that is what the problem is with paired sales. Appraisers that say they used paired sales when they didn't. That is the only thing that is wrong with paired sales.