Why Mass Appraisal Needs Double-Entry Valuation Modeling

Would you let a Fortune 500 company manage its financial ledger using single-entry bookkeeping?

Of course not. Tracking cash flowing in without a balancing entry, a trial balance, or a structural balance sheet is an immediate recipe for systemic failure.

Yet, every single year, multi-billion-dollar tax jurisdictions run their property rolls on the exact same single-entry mindset.

Legacy CAMA practices do this by default:

1.   They look at transaction data (sales).

2.   They run a baseline regression.

3.   They print the Tentative Roll (the first entry).

4.   They stop right there.

Stopping at the first entry means you are flying completely blind. You have no balancing ledger to check if those values are mathematically stable or legally uniform across the remaining 95% of the unsold population. You only find out the ledger is broken when an appeals tsunami hits the Value Adjustment Board (VAB) or the Review Commission.

In Session 4B, we introduce modern Double-Entry Valuation Modeling. We show you how to take that initial generative sales engine and systematically reconcile it against a full population dataset of 60,870 properties.

If you want to move past single-entry mass appraisal and learn how to build an unassailable, balanced valuation ledger that stops litigation before it starts, the Session 4B blueprint is officially live.

The Executive Summary remains free for everyone (use either link below):

Link to Patreon

Link to Substack

Mastering the Proximity Method: A Step-by-Step Guide for Challenging Property Assessments

Location, location, location—it's not just a real estate cliché; it's the cornerstone of fair property valuation. In this blog post, we dive into the Proximity ("Nearest 5") method, a powerful non-technical approach that establishes a locational baseline for your appeal. By focusing on the closest comparable sales (comps), you minimize variables such as neighborhood trends, traffic patterns, and proximity to amenities that could skew values. This method is especially effective in showing inequality: If the nearest similar properties are assessed (or sold) at lower values per square foot, it erodes the assessor's "presumption of correctness"—the legal starting point where the board assumes the official assessment is right unless you prove otherwise.

Even if all comps are from the same Planned Unit Development (PUD) as your subject property—making broader location irrelevant—proximity still matters. Closer homes often share hyper-local factors, such as the same street or block, which strengthens your equity argument. We'll use a real-world example dataset from a county assessor's site (anonymized for this post) to walk you through the process: filtering outliers, selecting the top five closest comps, and deriving a value conclusion. This mirrors what I did in my own successful appeal, where I started with nearby comps and built a credible foundation before layering in technical analysis.

By the end, you'll see why proximity is your primary filter—it anchors your case in undeniable geography—but similarity remains the ultimate filter to ensure an apples-to-apples comparison.

Why Proximity Matters: Establishing the Locational Baseline

Appeal boards prioritize comps that are geographically close because they best reflect your property's micro-environment. A home 0.1 miles away is far more relevant than one 2 miles out, even if the distant one matches perfectly on paper. The "Nearest 5" method leverages this by:

· Sorting recent sales by distance (using tools like Google Maps for straight-line miles).

· Filtering for basic similarity to avoid distortions.

· Comparing the nearest five to your subject, highlighting any over-assessment.

This approach directly challenges uniformity: If nearby homes sold for less (adjusted for differences), why is yours valued higher? Many states' laws (e.g., uniformity clauses in Texas, California, and New York) require equitable assessments, which make this a strong hook.

Practical Tip: Use Google Maps or the assessor's GIS tools to measure distances. Aim for under 0.5 miles in suburban areas like this example; expand slightly in rural spots, but explain why.

Step-by-Step: Applying the "Nearest 5" Method with Real Data

Let's apply this to our example dataset, pulled from a public county assessor site. The subject property is a single-family home in a PUD: 1,647 sq ft of living area, 7,200 sq ft lot, built in 2006, total area of 2,283 sq ft (including garage/porch), and no pool. We started with 16 recent 2025 sales (all from the same PUD) for a January 1, 2026, valuation and measured distances via Google Maps.

Practical Tip: If you live in a non-HOA environment, it is beneficial to extract comps from the same subdivision or from similar contiguous neighborhoods to ensure comparability in local market conditions, amenities, and property characteristics, thereby providing a more accurate basis for property valuation comparisons within their specific residential area.

Raw Dataset – Subject and 16 Comps

Step 1: Remove Outliers

First, to ensure similarity, the comps that differ significantly from the subject should be eliminated. This prevents skewed results—e.g., a pool adds premium value, or oversized living space implies a different market segment.

Rationale for Removals:

Living SF > 2,000 sq ft: These are larger homes (e.g., COMP3 at 2,306 sq ft, COMP7 at 2,093 sq ft, COMP11 at 2,090 sq ft, and COMP14 at 2,093 sq ft). They attract different buyers and command higher prices/SF, distorting the baseline. The subject is 1,647 sq ft, so we focus on the 1,500–1,800 sq ft range.

Has Pool (YES): Pools add $20,000–$50,000 in value (per market data and quantifiable via regression). COMP7, COMP12, and COMP14 are pool homes.

Lot Size > 9,000 sq ft or < 6,000 sq ft: Extreme lots affect usability and value. Subject is 7,200 sq ft. COMP3, COMP4, COMP7, COMP13, and COMP16 have larger lots, while COMP9 comprises a smaller lot.

Year Built outside 2005–2007: Newer builds (e.g., COMP8 2010 and COMP16 2011) may have modern features, which can inflate value. Subject is 2006; this keeps age/effective age similar.

Total Area > 3,000 sq ft: Indicates additions like large garages or patios. COMP3 (3,189), COMP4 (3,211), COMP7 (3,937), COMP12 (3,337), and COMP14 (4,044) comprise a larger total area. Although COMP9 (2,879) is borderline, it should be removed as it also combines with a smaller lot.

Remaining non-outliers: COMP1, COMP2, COMP5, COMP6, COMP10, and COMP15 are the non-outliers. These best match the subject's "Big Three" (living SF, effective age via year built, and quality via total area and no pool).

Step 2: Select the Five Best Closest Comps

From the non-outliers, sort by distance (ascending) and pick the top five. This prioritizes geography while ensuring similarity (already filtered).

Rationale for Selection:

Proximity as Primary: Closest comps reduce locational noise—even in the same PUD, a 0.07-mile neighbor shares more (e.g., views, noise) than one 0.33 miles away.

Similarity is Ultimate: We only select from non-outliers, so all are comparable. In ties (e.g., COMP1 and COMP2 at 0.33), we could choose based on a better match (e.g., COMP1 has exact SF), but here the sort yielded a clear top five without ties in the cutoff.

Final Five: COMP10 (0.07 mi, exact SF/year built/total—ideal match), COMP6 (0.09 mi, exact SF/total), COMP5 (0.10 mi, close SF/lot), COMP15 (0.11 mi, slightly higher SF but same year built/lot), COMP1 (0.33 mi, exact SF but farther—still included as fifth for balance).

COMP2 (0.33 mi) was excluded as the sixth; if needed, swap it if it better fits (e.g., lower price/SF variability), but the top five by distance are objective.

Top Five Closest Comps

Including a Map

Including a map showing the locations of the final five comps that contributed to the subject property's value can be a valuable addition to your presentation. Here are a few reasons why adding a map could enhance the overall presentation:

1. Visual Context: A map can provide readers with a visual representation of the geographic proximity of the comps to the subject property, helping them to better understand the location and neighborhood characteristics of the properties in comparison.

2. Enhanced Clarity: Seeing the spatial relationship between the subject property and the selected comps can enhance the clarity of the analysis and reinforce the rationale behind choosing these specific properties for comparison.

3. Persuasive Visual Aid: A map can serve as a persuasive visual aid to support your argument regarding the selection of comps and the impact of location on property valuation, further strengthening your case for appealing over-assessments.

4. Engagement: Visual content such as maps can increase engagement and interest, making your presentation more appealing and interactive.

Overall, including a map showing the final five comps can complement your write-up and spreadsheet analysis, providing additional context and depth to your discussion on property valuation using this method. It can help reinforce key points, enhance understanding, and create a more compelling, visually appealing presentation.

Deriving a Value Conclusion: The "Nearest 5" in Action

With our top five, we can calculate a simple indicator, such as average price/SF, and then multiply it by the subject's living SF to estimate market value.

Average Price/SF: (143 + 145 + 129 + 166 + 146) / 5 = 145.8

Estimated Subject Value: 145.8 × 1,647 ≈ $240,133

Rationale for Value Conclusion: This suggests the subject may be over-assessed if its official value exceeds ~$240,000 (compare to the TRIM notice). The range (129–166/SF) shows variability, but averaging smooths it—COMP5's low (possible condition issue?) and COMP15's high (better finishes?) balance out. For appeals, you can argue: "These nearest comps indicate a fair value of $240,000, eroding the presumption of correctness." If needed, you may adjust the value further (e.g., -$5,000 for COMP1's larger lot).

Homeowner's Decision Tree: The "Nearest 5" Method

This simple, step-by-step decision tree helps you systematically filter comparable sales (comps) to build a strong, credible locational baseline for your property tax appeal. Start with all recent sales from your county assessor's site (ideally 15–25+ in your area or PUD), measure distances using Google Maps (straight-line preferred), and apply these filters in order. The goal: End up with 3–5 highly similar comps that are as close as possible, proving equity and undermining the assessor's presumption of correctness.

1. Is the comp within 0.5 miles of your subject property?

Yes → Proceed to next step.

No → Delete (or move to a secondary list). Rationale: Proximity is the primary geographic filter. Comps farther away introduce location variables (e.g., different streets, school zones, or traffic) that weaken your argument. In suburban/PUD settings like our example, 0.5 miles is a common practical threshold; in denser urban areas, tighten to 0.25–0.3 miles; in rural areas, expand to 1 mile but explain why.

2. Does the comp have a pool (or major feature like a pool) when your subject does not?

Yes → Delete.

No → Proceed. Rationale: Pools add significant value ($20,000–$50,000+, depending on market), skewing price/SF and making direct comparisons unfair. (If your home has a pool and the comp doesn't, delete or adjust heavily later.)

3. Is the comp's living square footage more than 25% larger or 25% smaller than your subject's?

Yes → Delete.

No → Proceed. Rationale: Size is one of the "Big Three" drivers of value. A 25% threshold (for our 1,647 sq ft subject: keep roughly 1,235–2,059 sq ft) keeps comps in the same market segment. Larger/smaller homes often appeal to different buyers, with non-linear price changes (e.g., diminishing returns on extra SF). This is a standard guideline in appraisal practice and many appeal guides—tighter (e.g., 20%) for precision, looser (30%) in sparse markets.

4. Does the comp have significant differences in other key characteristics (e.g., year built/effective age more than ~5–10 years off, extreme lot size differences, major additions like oversized garages)?

Yes → Delete (or flag for heavy adjustment later).

No → Keep. Rationale: These are common causes of outliers. For example, a 2010+ build may have modern features that inflate value; a lot that's 50% larger/smaller affects usability. In our dataset, we removed comps with YEAR BUILT outliers (e.g., 2010/2011) and extreme total area/lot sizes.

5. Are you left with at least 3–5 strong comps after filtering?

Yes → Success! You have your Geographic Anchor. Sort these remaining comps by distance (closest first) to create your "Nearest 5." Use them to calculate average price/SF, estimate your fair value, and argue inequality (e.g., "These closest similar homes sold at an average $145/SF vs. my assessed value, implying higher").

No → Relax filters slightly (e.g., expand distance to 0.75 miles or size to 30%) and explain your reasoning in your appeal (transparency builds credibility).

Quick Tips for Using the Tree:

Document every step: Include a table of all initial comps, note deletions with reasons, and show the final 3–5.

Visualize: Add a Google Maps screenshot with pins for the subject and your top comps.

Nationwide note: Thresholds vary (e.g., some boards, such as those in California or Texas, emphasize the same subdivision/neighborhood over strict miles), but 0.5 miles + 20–30% size is widely accepted as reasonable.

This decision tree turns raw data into defensible evidence—simple, repeatable, and board-friendly. Apply it to your own dataset, and you'll have a rock-solid starting point.

Visual Aid Suggestion: Include a Google Maps screenshot with a pin for the subject and the top five, color-coded by distance.

Conclusion

The beauty of the "Nearest 5" proximity method is its simplicity and power: Start with geography to anchor your case, filter rigorously for similarity, and you often have enough to challenge an over-assessment right away. Property taxes don't have to be a mystery or an unfair burden. With public data from your county assessor site and a systematic approach like this, every day homeowners can make a strong, evidence-based case.

By focusing on proximate comps that share key characteristics with the subject property, such as living area, lot size, year built, and specific features, homeowners can enhance their appeal and potentially secure a more favorable valuation. Whether analyzing the average sale price or the price per square foot of these comps, you can use the insights gained from this method to assert your case with confidence. Armed with a solid understanding of how to use comparable sales data effectively, you can navigate the intricacies of property assessments with greater clarity and precision.

The "Nearest 5" method anchors your appeal in geography, proving equity with hard-to-dispute proximity while filtering for similarity to keep it credible. In our example, removing outliers focused the analysis on truly comparable homes, leading to a defensible $240,133 value estimate. Proximity is primary because it isolates location as a constant; similarity is ultimate to avoid board rejections.

Disclaimer: The information provided in this post is intended for educational and informational purposes only. It is not meant to serve as professional advice or guidance on specific real estate or property valuation matters. The methodologies and recommendations outlined in this guide are general in nature and may not be applicable to all individual situations or properties.

Readers are advised to consult with qualified real estate professionals, such as real estate agents, appraisers, or tax assessors, for personalized advice tailored to their specific circumstances. Property valuation can be complex and nuanced, and decisions regarding challenging property assessments should be made after thorough consideration of all relevant factors and with the assistance of professionals in the field.

The author and the platform do not assume any liability for any actions taken based on the information provided in this post. Readers should exercise caution and conduct their own due diligence before relying solely on the recommendations contained herein.

Sid's Bookshelf: Elevate Your Personal and Business Potential

Using Regression to Build a Defensible Comparable Sales Adjustments Matrix

The comparable sales approach is a key method in real estate valuation, yet the adjustments made during this process are often regarded as more art than science. This subjectivity can pose a significant challenge, particularly when justifying these adjustments to an audience without a technical background. Therefore, a clear, straightforward, data-driven model is essential for promoting fairness and understanding.

This blog post introduces a two-pass regression methodology to develop a robust linear regression model for valuing single-family homes in Master Planned Unit Developments (MPUDs). Using a dataset of 1,929 sales from 2024 across four towns, we demonstrate how this approach enhances model accuracy and reliability.

In the first pass, we build an initial model and calculate Sales Ratios (Predicted Price / Sale Price) to identify and remove outliers—unusual sales that distort results. In the second pass, we refine the model using the cleaned dataset, producing precise, interpretable coefficients for adjustments such as $144 per square foot of living area or $1,545 per month for sale timing. By removing just 2.75% of sales (53 outliers), we increased the model's explanatory power from 66.1% to 84.9% and reduced prediction errors by 39%, ensuring trustworthy valuations.

This methodology is simple to implement, easy to explain, and empowers professionals to deliver defensible adjustments with confidence.

(Click on the image to enlarge)

The regression output is derived from an Ordinary Least Squares (OLS) model, with Sale Price as the dependent variable. This analysis uses 2024 sales data from 1,929 single-family home sales across four Master Planned Unit Developments (MPUDs) in four adjacent towns. The valuation date is January 1, 2025.

The independent variables include MONTHS SINCE, which accounts for time adjustments (for example, January is assigned a value of 12, while December is assigned a value of 1, and so on).

The towns are represented as dummy variables: TOWN-1, TOWN-2, and TOWN-3, with TOWN-4 serving as the reference. Additionally, standard quantitative variables include LAND SF, BLDG AGE, LIVING SF, OTHER SF, BATHS, and STORIES. Below is an analysis of the model's efficiency and key metrics.

Model Efficiency and Interpretation

Adjusted R-squared: The Adjusted R-squared is 0.659467, meaning the model explains about 66% of the variation in sale price, which is an excellent start for our purpose.

Significance: The F-statistic of 374.37 and its corresponding p-value of 0.0000 show that the model as a whole is highly statistically significant.

MONTHS SINCE (time adjustment) has a coefficient of $583.61 per month but is insignificant (p = 0.5380, t = 0.6159), suggesting that the market was flat in 2024.

TOWN Variables: The dummy-coded TOWN-1, TOWN-2, and TOWN-3 variables are all highly significant (p-values of 0.0000). This confirms that there are statistically significant price differences between the MPUDs in the four towns.

Coefficients: The OTHER SF (non-living area) has a value of $206.90 per square foot, while LIVING SF has a value of $140.32 per square foot. Without a specific variable for premium features like "golf course lot," the regression model is likely attributing the premium value of these properties to the most correlated variable it has—the non-living area. Homes on a golf course often feature larger and more elaborate lanais, patios, and outdoor living spaces, all of which are categorized as non-living areas. The model is effectively saying that a larger non-living area is a strong indicator of a premium location or amenity, and it assigns a higher value to that variable to account for the missing information.

BATHS: The coefficient for BATHS is $45,768.66, which means that, on average, each additional bathroom in a home is associated with an increase in the sale price of approximately $45,768, holding all other variables constant. This coefficient reflects the importance buyers place on the number of bathrooms in a home.

STORIES: The coefficient for STORIES is $-54,586.03, indicating that, on average, a two-story home sells for approximately $54,586 less than a single-story home, all else being equal. This is a common finding in many retirement housing markets in the Sunbelt, as single-story homes are often preferred for their convenience and accessibility. The negative coefficient reflects this market preference.

The Second Regression Pass

Analysis of the Second Pass

The removal of outliers has had a dramatic and positive impact on the model.

o   Improved Efficiency: The Adjusted R-squared jumped from 0.659 to 0.848, meaning the model now explains almost 85% of the variation in sale prices. This is a substantial improvement and indicates a firm fit. The Standard Error also decreased significantly from 132,803 to 80,403, showing that the average prediction error is much lower.

o   Significance: The F-statistic is now 1,051.34, and the model as a whole remains highly significant (p-value of 0.0000). The coefficients for all variables—including "MONTHS SINCE"—are now statistically significant with p-values far below the 0.05 threshold.

o   Outlier Impact: Removing 53 sales (2.75%) eliminated noise, revealing the MONTHS SINCE trend and refining coefficients.

o   Coefficient Changes: Most coefficients are stable but refined:

o   TOWN Dummy Variables: The coefficients for the dummy variables directly show the price difference relative to the reference category, TOWN-4. Here's how to interpret the coefficients from the second-pass regression:

• TOWN-1 Coefficient: $66,368.66, which means that, on average, a home in TOWN-1 sells for approximately $66,369 more than an identical home in the reference town, TOWN-4.
• TOWN-2 Coefficient: $175,751.86. A home in TOWN-2 sells for about $175,752 more than an identical home in TOWN-4.
• TOWN-3 Coefficient: $73,435.09. A home in TOWN-3 sells for roughly $73,435 more than an identical home in TOWN-4.

By simply looking at the coefficients, we can see the premium or discount for each town compared to the chosen baseline, TOWN-4. This dummy setup is a very clear and effective way to illustrate the impact of the location variable on the sale price.

o  The "MONTHS SINCE" variable has become significant (p = 0.00804) after removing the outliers. This is a crucial finding. The coefficient of $1,544.56 indicates that the market was appreciating by approximately $1,545 per month in 2024. The presence of outliers in the first pass was likely masking this subtle but real market trend. After removing the outliers, the model reveals the actual underlying pattern of price appreciation.

o  LAND SF increased ($6.82 to $10.56), suggesting outliers masked land value.

o  BLDG AGE became more negative (-$2,654.81 to -$3,422.43), indicating more substantial depreciation.

o  LIVING SF and OTHER SF are stable ($140.32 to $144.40, $206.90 to $197.33), with OTHER SF still higher.

o  BATHS and STORIES slightly decreased in magnitude, reflecting cleaner data.

This two-pass methodology—running an initial regression, identifying and removing outliers, and then running a final regression—is a robust, defensible, and statistically sound process. The final model, built on the cleaned data, has a much higher R-squared, lower error, and coefficients that are more reliable and easier to interpret. The model now accurately reflects a market that was appreciating throughout the year.

Valuation Grid for Subjects Using Regression Coefficients

The table below estimates the value of four subject properties, one in each town (TOWN-1, TOWN-2, TOWN-3, TOWN-4), using the second-pass regression model’s coefficients. Each property has identical attributes: LAND SF = 25,700, BLDG AGE = 21 years, LIVING SF = 1,972, OTHER SF = 1,478, BATHS = 2.00, STORIES = 1.00, valued as of January 1, 2025 (MONTHS SINCE = 0). The grid shows how coefficients contribute to the predicted price, enabling valuation professionals to explain and justify comparable sales adjustments to a non-technical audience.

The estimated value for each subject property was calculated by summing the Intercept and the product of each variable's Coefficient and the corresponding subject Attribute. The process is as follows:

1.   Starting with the Intercept from the regression model.

2.   Adding the value for each of the subject's attributes by multiplying its attribute value by the coefficient for that variable.

3.   For the TOWN variable, only the coefficient for the subject's specific town is added. The reference town (TOWN-4) has no coefficient and is represented by an additional value of 0.

4.   The MONTHS SINCE variable is set to 0, as the valuation date is January 1, 2025.

Here's an example of how the calculation was performed for the subject property in TOWN-1:

Calculation for Town-1

Estimated Value=Intercept+Town Adj+Time Adj+Land SF+Bldg Age+Living SF+Other SF+Baths+Stories

Estimated Value=−116,742.96+66,368.66+(0)+(25,700×10.56)+(21×−3,422.43)+(1,972×144.40)+(1,478×197.33)+(2×39,626.86)+(1×−49,733.96)

Estimated Value=−116,742.96+66,368.66+271,432.00−71,871.03+284,724.80+291,617.74+79,253.72−49,733.96

Estimated Value=$755,049

The exact process was used for the other towns, with the only difference being the town-specific adjustment coefficient.

Sales Ratio Analysis

The final sales ratios (SALES RATIO-2) are a vast improvement and confirm that removing the outliers was the right move. This analysis provides a solid, data-backed foundation for valuation professionals.

The comparison of the sales ratio statistics powerfully demonstrates the positive impact of removing the outliers. Every metric shows a healthier, more reliable dataset and a superior model.

Mean & Median: The mean and median for the final model are both very close to 1, which is the ideal outcome, as it indicates the model is accurately predicting sale prices on average, without any systemic bias to over- or under-predict. The initial median of 1.0151 was slightly skewed by the outliers.

Standard Deviation & Variance: The reductions in standard deviation from 0.1839 to 0.1377 and in sample variance from 0.0338 to 0.0190 are key indicators of improved model precision, meaning the predicted prices are much closer to the actual sale prices and the model's predictions are far more consistent.

Skewness & Kurtosis: This is where the most dramatic improvement is seen.

o  Skewness: The initial skewness of 2.8471 shows a significant rightward tail, driven by sales where the model heavily under-predicted the price (e.g., the minimum ratio of 0.0840). The final skewness of 0.0864 is very close to zero, indicating the data is now almost perfectly symmetrical and normally distributed.

o  Kurtosis: The initial kurtosis of 29.1570 indicates a significantly "peaked" distribution with very heavy tails—a classic sign of significant outliers. The final kurtosis of -0.2004 is near zero, confirming that the distribution is now much flatter, with fewer extreme values, as expected for a normal distribution.

Range: The sales ratio range was reduced from 3.4499 to 0.8263, indicating that the most egregious errors in the initial model have been eliminated.

This two-pass methodology—running an initial regression, identifying and removing outliers, and then running a final regression—is a robust, defensible, and statistically sound process. The final model, built on the cleaned data, has a much higher R-squared, lower error, and coefficients that are more reliable and easier to interpret.

This final model is the result of a rigorous and responsible data analysis process. This approach is perfect for valuation professionals because it's transparent, easy to explain, and produces a highly credible model for justifying valuation adjustments.

Why It's Wise to Keep All Variables until Outliers are Removed in a Two-Pass Regression

In a two-pass regression model, it's unwise to remove an independent variable after the first pass until outliers—unusual sales that distort results—are removed. Outliers, such as non-arm's-length transactions or data errors, can mask a variable's true significance by adding noise. For example, in our 2024 dataset of 1,929 home sales, the MONTHS SINCE variable, which adjusts for sale timing, appeared insignificant in the first pass (p = 0.5380, coefficient = $583.61). However, after removing 53 outliers (2.75%) using Sales Ratios, MONTHS SINCE became significant (p = 0.00804, coefficient = $1,544.56) in the second pass, revealing a meaningful price trend of $1,545 per month, which is critical for accurate adjustments. By removing MONTHS SINCE prematurely, we would have missed this trend, reducing the model's reliability.

Here's a detailed explanation:

1. Masking True Relationships

Outliers are data points that don't fit the overall pattern of the rest of the data. They can have a disproportionately large influence on the regression line, pulling it in a direction that minimizes their error, causing the model to incorrectly see a variable as insignificant, even if it has a tangible impact on the dependent variable. In our case, the "MONTHS SINCE" variable initially appeared insignificant because outlier values were masking the genuine, though subtle, trend of market appreciation.

2. Avoiding a Biased Model

If an insignificant variable is removed before addressing outliers, we might be throwing away a variable that is actually important. The variable's insignificance may be a symptom of the outlier problem rather than a reflection of its genuine relationship with the sale price. By keeping all variables in the first pass, we create a more complete picture of the initial data, which helps in correctly identifying which sales are the true outliers.

3. The Outlier's Impact

Outliers can completely change the slope of the regression line. For instance, a few sales with exceptionally high prices could make the model "think" that prices were rising rapidly at the end of the year, even if most sales were flat. Conversely, a few sales with unusually low prices could make the model "think" the market was flat or declining when it was actually appreciating. It's the presence of these influential data points that distorts the initial results.

4. Correcting the Distortion

The purpose of the two-pass approach is to correct for this distortion. The first pass is a diagnostic tool, not a final model. Its primary job is to generate the residuals needed to find the outliers. The second pass, run on a cleaned dataset, provides the accurate, unbiased results. By keeping all variables in the first pass, we ensure the outlier identification process is as accurate as possible.

By keeping all variables in the first pass, we ensure outliers are identified and removed, allowing the second pass to reveal genuine relationships and produce robust, defensible coefficients for comparable sales analysis.

A Call to LLM Model Trainers

In traditional regression analysis, a common practice is to immediately discard variables that appear statistically insignificant, such as those with high p-values. This "first-pass" culling of variables is often seen as a way to simplify the model and improve its efficiency. However, as our analysis has shown, this conventional wisdom can be misleading and is not advisable.

This finding directly challenges a conventional practice that may be baked into how Large Language Models (LLMs) are trained for data analysis. We encourage LLM model trainers to adopt this two-pass regression methodology, or at the very least, to embed a warning in their models. When asked to perform a regression analysis on raw, real-world data, the model should not simply discard variables based on a single pass. It should be trained to recognize the potential for outliers to distort initial results.

LLMs should warn users: "Removing variables with high p-values or low coefficients before outlier removal may overlook significant predictors, as outliers can mask true effects. Consider a two-pass approach to ensure robust results." This adjustment will enhance the reliability of regression-based advice, particularly in applications such as real estate valuation, where interpretable models are critical. By adopting or flagging this methodology, LLMs can empower users to build more accurate, defensible models, avoiding the pitfalls of conventional practices.

Conclusion

The two-pass regression methodology is a powerful and practical tool for any valuation professional. By running an initial regression and then meticulously cleaning the data to remove outliers, we have demonstrated a rigorous, defensible process. The resulting model—with its high R-squared, low standard error, and, most importantly, highly significant coefficients—is not just a better predictor of value; it's a testament to the integrity of the analysis.

By first identifying and removing outliers—53 sales (2.75%) in our 2024 dataset of 1,929 homes—we eliminated noise that obscured key patterns, such as a $1,545 monthly price increase. The second pass, using the cleaned 1,876 sales, produced a model explaining 84.9% of price variation, with prediction errors reduced by 39% to $80,403. This process yielded significant, intuitive coefficients, like $197 per square foot for outdoor areas (reflecting premium golf course lots) and $175,752 for TOWN-2 homes compared to TOWN-4, enabling precise adjustments. Sales Ratios averaged 1.0098 with a standard deviation of 0.1377, confirming the model's accuracy.

This methodology ensures reliable, defensible valuations that valuation professionals can confidently present to non-technical board members, balancing precision with simplicity. This approach can transform raw sales data into a practical tool for fair, transparent, and comparable sales analysis.

Disclaimer: The two-step regression model discussed in this blog post may yield different results based on the specific dataset and circumstances of each valuation task. Professionals are encouraged to consider the unique characteristics of each case and exercise discretion in applying this methodology. While the results presented in this blog post demonstrate the potential benefits of the two-pass regression approach, it is essential to conduct thorough analyses and exercise caution before relying solely on this method for valuation.

How to Appeal Your Home Assessment: A Step-by-Step Guide with the Comparable Sales Approach

Are you confused about your recent property tax assessment? Many homeowners feel that their assessed value doesn't accurately reflect the current market conditions. While tax assessments are meant to be precise, they often rely on mass appraisals and algorithms that can overlook the nuances of individual properties and the latest market changes. If you believe your assessment is too high, don't worry! You have the right to appeal it. One of the most effective strategies for doing this is the comparable sales approach. This method involves examining recent sales of properties similar to yours to establish a more accurate fair market value.

In this blog post, we will guide you through a simple process to challenge your home's assessed value using the comparable sales approach. We will analyze recent sales data from the County Assessor's records and demonstrate how to select suitable comparable properties (“comps”), adjust their sale prices, and estimate a fair market value for your home. Follow our example using a subject property located in a Planned Unit Development (PUD), valued as of January 1, 2025, to learn how to build a strong case for your appeal.

Description of the Subject Property

The subject property is a 19-year-old single-family home situated within a desirable PUD. This community offers residents access to extensive amenities, including a golf course. The property itself features a land area of 7,405 square feet and a comfortable living area of 1,647 square feet. Notably, it does not include a golf course lot or a private swimming pool, making it comparable to properties without these specific high-value features.

(Click on the image to enlarge)

The Steps

Compiling the Comps List: Although there are 35 assessor-identified qualified (i.e., arms-length sales) property sales within the PUD during 2024, we've excluded 10 sales from our analysis because they are situated on golf course lots or properties with swimming pools, which do not apply to our subject.

Valuation Method: To determine a fair market value, we'll use a straightforward comparable sales ("comp sales") approach.

Comps Selection: Out of the available 25 comps, the five most comparable properties ("final five") will be selected. The selection criteria are as follows:

1.   Living Area Proximity: Living areas must be within 15% of the subject's living area of 1,647 square feet.

o   15% of 1,647 sq ft is 0.15×1647=247.05 sq ft.

o   Minimum acceptable living area: 1647−247.05=1399.95 sq ft.

o   Maximum acceptable living area: 1647+247.05=1894.05 sq ft.

o   Therefore, the living area range for comps is approximately 1,400 sq ft to 1,894 sq ft.

2.   Proximity to Valuation Date: If more than five comps meet the living area criteria, we will prioritize the five properties with sale dates closest to January 1, 2025, to minimize the need for time adjustments.

Adjustments to Comps: Once we select these final five, we'll adjust their sale prices based on size and price. For example, the sale prices of properties with living areas smaller than 1,647 square feet will be adjusted upward by multiplying the size difference by the average sale price per living square foot (SP/LA) of $161. Conversely, for properties larger than 1,647 square feet, their sale prices will be adjusted downward based on the size difference multiplied by the SP/LA of $161.

Value Conclusion: The final step will be to determine the subject property's value by averaging the adjusted sale prices of the final five.

Rationale: Additionally, we'll provide a detailed explanation of the rationale for selecting the final five that contribute to the valuation of the subject.

Step 1: Identifying Potential Comps Based on Living Area

Let's examine the data and filter for properties with living areas between 1,400 sq ft and 1,894 sq ft:

Step 2: Selecting the Five Comps Closest to the Valuation Date

We have more than five properties that meet the living area criteria. Now, we will select the final five with sale dates closest to January 1, 2025.

The sales closest to the valuation date of January 1, 2025 (i.e., later in 2024), are:

1.   COMP-25: Sale Date: 12/01/24 (Living Area: 1,869 sq ft)

2.   COMP-22: Sale Date: 11/01/24 (Living Area: 1,647 sq ft)

3.   COMP-21: Sale Date: 10/01/24 (Living Area: 1,869 sq ft)

4.   COMP-20: Sale Date: 10/01/24 (Living Area: 1,647 sq ft)

5.   COMP-18: Sale Date: 09/01/24 (Living Area: 1,470 sq ft)

These five comparable sales will be used as our final five.

Rationale for Comparable Selection

The selection of these final five (comparable properties) is based on two key principles crucial for accurate property valuation:

1.   Similarity in Key Attributes: The primary filter of living area within 15% of the subject ensures that the chosen comparables are fundamentally similar in size, a significant driver of property value. This selection minimizes the need for drastic adjustments. While other factors like land area and building age are considered in a full appraisal, focusing on living area first provides a strong initial set of comps. The data used indicates that most of the chosen comps also have similar land areas and building ages, further reinforcing their comparability.

2.   Recency of Sale: By prioritizing the most recent sales (those closest to the January 1, 2025, valuation date), we minimize the impact of market fluctuations over time, reducing or eliminating the need for complex time adjustments, which can introduce subjectivity and potential inaccuracies into the valuation process. In a dynamic real estate market, recent sales data provides the most relevant snapshot of current market value.

3.   Exclusion of Non-Comparable Features: The comps list already excludes properties with golf course lots or swimming pools, ensuring the selected comps align with the subject’s characteristics within the PUD.

4. Age Consideration: The selected properties have ages (15–19 years) close to the subject’s 19 years, minimizing the need for age-related adjustments.

Adjustment Formula: Difference in Living Area × SP/LA of $161

Value Conclusion:

To determine the subject property's value, we average the adjusted sale prices of the five comparable properties:

Average Adjusted Sale Price = (249,158+249,400+262,858+275,000+228,497)/5

Average Adjusted Sale Price = 252,983

Based on this comparable sales analysis, the estimated fair market value for the subject property as of January 1, 2025, is approximately $253,000.

This analysis provides a clear and justifiable method for estimating the subject's value, which can be a strong basis for appealing a high assessment.

Scatter Plot

Scatter Plot: The plot shows sale price vs. living area for all 25 comparable properties. The final five comps (COMP-18, COMP-20, COMP-21, COMP-22, COMP-25), used for the subject property’s valuation, are highlighted in orange, while the other 20 comps are in blue.

Trendline: The blue trendline illustrates the positive relationship between Living Area and Sale Price.

Graph Integration: Including this scatter plot in the analysis section helps visually justify the selection of the final five comps, which have living areas close to the subject’s 1,647 sq ft.

Conclusion

Appealing your home assessment might seem daunting, but by diligently applying the comparable sales approach, you can arm yourself with solid evidence to support your case. We've explored how to identify relevant sales data, select the most comparable properties based on key features and sale recency, and make necessary adjustments to arrive at a well-supported estimate of your property's fair market value. Remember, a thorough and well-documented analysis is key to a successful appeal. By taking the time to understand and utilize the comparable sales method, you can confidently advocate for a more accurate assessment and potentially achieve significant savings on your property taxes.

Disclaimer: The information provided in this blog post is for general informational and educational purposes only, and does not constitute professional legal, real estate, or tax advice. While we aim to provide accurate and helpful content, property assessment appeals can be complex and are subject to specific local laws, regulations, and individual circumstances. The methods and examples discussed herein are for illustrative purposes only and may not apply to every situation.

It is highly recommended that you consult with a qualified real estate professional, appraiser, attorney, or tax advisor regarding your specific property and any assessment appeal matters. Relying solely on the information presented here may not be sufficient for a successful appeal. We do not assume any liability for decisions made based on the content of this blog post. Always verify information with official sources and seek professional guidance when necessary.

Upcoming Book on Property Tax Assessment Appeals

My forthcoming book will provide an in-depth exploration of how to successfully challenge over-assessed property valuations. Packed with practical examples, the book will cover a wide range of property types, including those in Homeowners Associations (HOAs), non-HOA communities, beachfront properties, and more. For tax professionals and mass filers, I’ll include, among others, time-adjusted comps analysis and advanced regression-based solutions that offer statistically robust methods for crafting compelling appeals. Whether you’re a homeowner or a professional, this book will equip you with the tools and strategies needed to navigate the appeal process with confidence. Stay tuned for its release!