You're working for a supervisor who relies on "the list", which is basically a collection of adjustment factors that are considered acceptable across a variety of properties and which require zero thinking or analysis to apply. It's like autopilot, and it's a valuable tool for those appraisers who are fixated on getting reports out the door in the least amount of time.
As well, "the list" can be pretty functional when most of your assignments involve properties for which there are always a ton or really similar comps. So it's not necessarily an evil thing to rely on the list.
But it would be a good thing to learn how to develop and support your own adjustments at some point in your career.
Opinions on adjustments vary. Some people see an adjustment grid with 20 lines and figure that having numbers in every single one of them is helpful for adjusting their dataset to support the one true value conclusion that can be considered "accurate". Others see adjustments as a (mostly) minor refinement that should be used sparingly, as in, the fewer the better.
You can probably see by the way I just characterized the two that I am among the group that favors the latter viewpoint. In my opinion if you have great comparables you don't need adjustments. You can just rank your subject within an unadjusted dataset and get a reasonable value conclusion from that. Conversely, when you have great comps the factors you use for adjustment will mostly be irrelevant to the conclusion anyway because those adjustments will usually be so small as to not be of effect.
The way I generally develop adjustments is to compare the sales in my dataset against each other to see which adjustments produce the tightest spread, starting with those factors that I think will be of most effect. That comparable dataset usually includes a lot more properties than I end up using as direct comparables in my report.
I usually qualify more sales than I use in my reports - that means that besides the details of the transaction itself I am also getting an idea of quality, condition, site and location attributes as they relate to my subject. Although the statistic geeks like large quantities of data for statistical analysis I don't favor using unqualified data when I'm trying to identify adjustment factors.
Anyways, the way I do it is to array my sales data in reverse chronological order (newest sale or pending first >> most dated sale last). Since I use spreadsheets a lot I've found it's easier to do it on a spreadsheet. If the properties are all in similar locations and have similar site influences then that simplifies the decision of which factor to adjust first, in which case I usually go straight for GLA.
I try out different GLA factors to see which one results in the tightest spread. Sometimes I guess correctly and the first one I start with works better than the rest, but usually I have to try several different iterations. This is where using a spreadsheet can be faster - you can write an equation that references a single cell (which contains your adjustment factor) and copy it into the column or line you're using for your adjustments. You can check on your spread using another set of equations, which on an Excel spreadsheet looks like this:
=max(A28:A38) that cell range refers to the line or column with your adjusted value indicators
=min(a28:A38)
=x-y (x equals the "max" cell above and y = the "min" cell
You're basically telling the spreadsheet to identify which adjusted value out of the range is the highest and which is the lowest and you're looking for the adjustment factors that reduce the spread between the two to the lowest possible margin.
Sometimes GLA is not the primary adjustment factor. Sometimes it's a view amenity or the date of sale or location or physical condition. This is where having an open mind and paying attention during your research pays off. It doesn't pay to start with too many assumptions. If you structure your analysis well you can try a couple different attributes out on a trial/error basis to see which has the most effect. It's often GLA but not always.
Anyways, after you get the first adjustment factor squared away you can test others. If GLA worked best the condition may be the next biggest factor. Or vice-versa; it just depends on the situation.
Sooner or later you're going to reach the point of diminishing returns. Where going further with your adjustment factor doesn't have additional effect or it has an opposing effect on increasing your spread. That's the trick of adjusting, to get to that point and stop.
In many residential assignments you can reach that point of diminishing returns after adjusting for just a few attributes: GLA, condition, date of sale, financing (and seldom in that exact order). Realistically we're supposed to start with financing first, then date of sale; but in my experience it can often work more effectively in a different order.
Anyways, once your spread gets reduced that's the time to stop. You may not be identifying adjustments for 4bd vs 3bd, but then again you wouldn't need to make such an adjustment had you just stuck to 3bd comps in the first place. In this region garage spaces are usually worth adjusting for, but not always. Pools usually aren't worth an adjustment in the older homes but sometimes the data says otherwise.
The main trick is to let the data speak for itself; and having done so, to say that in your report so you don't get some pencil neck "reviewer" hassling you for not adjusting for 1,000 sf of lot area or that 4th bedroom.
In my view, a reasonably adjusted dataset that gets there after only a few adjustments conveys significant support for a value conclusion. If I have to apply a lot of different adjustments that indicates that I started out with a really weak dataset in terms of comparability.
