MLS statistics

[alebrewer]

That's the great thing about statistics....you can make them say just about whatever you want.

I think the graphs Terrel posted showed the difference in doing a 12 month trend and a 3 month trend. Highlighting the possibility that an appraiser could, indeed, have to apply adjustments that might not line up with an annualized linear trend analysis.

 

[Mark K]

I think the graphs Terrel posted showed the difference in doing a 12 month trend and a 3 month trend. Highlighting the possibility that an appraiser could, indeed, have to apply adjustments that might not line up with an annualized linear trend analysis.

And for your reports, you can cherry pick the data that best fits your argument.

 

[alebrewer]

And for your reports, you can cherry pick the data that best fits your argument.

Of course you can. That is no different than fudging the adjustments using PFA. The fact that statistical analysis can be manipulated does not diminish the validity of statistical analysis.

 

[Mark K]

Of course you can. That is no different than fudging the adjustments using PFA. The fact that statistical analysis can be manipulated does not diminish the validity of statistical analysis.

Absolutely agree. Problem is human beings are developing and using the analysis and humans are pretty far from the objective; they have egos and these egos don't like being questioned. Look at many of the pages in a concurrent thread.

 

[alebrewer]

Absolutely agree. Problem is human beings are developing and using the analysis and humans are pretty far from the objective; they have egos and these egos don't like being questioned. Look at many of the pages in a concurrent thread.

So your argument is to not use statistical analysis because humans can't be trusted to use it meaningfully? I mean - I guess that's one way to deal with it. I'll admit that I tend to be too trusting that folks want to do the right thing, but often just don't understand how. You may be right, but I'll just continue to live in my fantasy world, I guess.

 

[NJ Valuator]

Depends on my subject market and what other indicators show such as spending etc- idk what data you used for - 20% adjustment but since I am not seeing it in my market it is hard for me to comment.

Were all sales identical in all characteristics that impact value? If not, the differences in median prices are not solely due to a change in market conditions.

Identical? No. How often do you get identical comps though?

 

[NJ Valuator]

The tool Jim Amorin put out does similar. It creates a 12 month trend, a 9 month trend, a 6 month trend, and a 3 month trend. Pretty slick actually.

View attachment 97809

how do you acquire the tool from from Jim Amorin?

 

[NJ Valuator]

From Flex last 12 months vs since 1/1/2025 (sales in one town between 1500 and 2500 SF)
View attachment 97805
trend very slightly up
View attachment 97807
Trend decidedly down. Which do you believe?

what program are you using to get these graphs from their respective MLS's?

 

[alebrewer]

how do you acquire the tool from from Jim Amorin?

Bright MLS is on a CoreLogic GUI, which means that you should have a tab called 'Stats'. It will generate graphs similar to the one I posted, but not nearly as 'polished'.

 

[Terrel L. Shields]

what program are you using to get these graphs from their respective MLS's?

MLS allows you to download as a .csv file (comma separated value) which I then load into Excel. I eliminate most of the columns and then run a scatter chart and trendline.