---
title: "Touchdowns are mostly rented, not owned"
slug: touchdown-regression
dek: "A walk-forward test of the fantasy maxim — and a more useful way to apply it."
access: free
email_subject: "The 2.8 touchdowns your projection should take back"
email_preview: "Our first study puts a number on touchdown regression."
published_at: 2026-08-30
data_through: 2024
---

# Touchdowns are mostly rented, not owned

Fantasy analysis says “touchdown regression” so often that it can become a substitute for analysis. We wanted a falsifiable version: **does an opportunity-based expected-touchdown baseline predict next season better than a player's raw touchdown total?**

Across **551 consecutive WR/TE season pairs**, it did. Carrying forward actual touchdowns produced **2.61 TD** of mean absolute error. Carrying forward expected touchdowns cut that to **2.38 TD** — an **8.9% reduction**.

![Forecast error comparison](figures/forecast-error.png)

## The result in one picture

Players in the highest overperformance quintile scored **8.2 touchdowns** in year one and **5.4** the next year: a decline of **2.8**. The biggest underperformers moved the other way, from **3.0** to **4.3**.

![Regression quintiles](figures/regression-quintiles.png)

That pattern is the useful core of regression. It is not “high touchdown players are bad.” It is “the part of touchdown production unexplained by repeatable opportunity should receive less weight.”

## What we modeled

We aggregated nflverse regular-season receiving data for WRs and TEs with at least 40 targets from 2014–2024. For every test season from 2018 onward, the model was trained only on earlier seasons. Five same-season opportunity signals — targets, receptions, yards, air yards, and receiving first downs — estimated a touchdown baseline. Ridge regularization reduced instability from correlated volume measures.

The next-season test compared two deliberately simple forecasts against the same outcome:

1. next season equals this season's actual touchdowns; and
2. next season equals this season's expected touchdowns.

The paired MAE improvement was **0.23 touchdowns**, with a 95% bootstrap interval of **0.09 to 0.38**. In 99.9% of 5,000 resamples, the expected-TD baseline won.

## How to use it

Treat large touchdown-over-expected gaps as a tiebreaker, not a verdict. If two players have similar projected opportunity, prefer the one whose prior production depended less on outlier conversion. Do not mechanically fade elite players with elite volume, and do not turn a historical baseline into a current projection without team, quarterback, health, and market context.

## Where this can break

The test includes only players who returned to at least 40 targets, so it describes volume survivors. Season totals also hide red-zone usage and route context. The model is a benchmark, not a causal estimate of touchdown “skill.” Finally, the public source snapshot ends with 2024; the named examples are historical illustrations, not 2026 draft calls.

## Reproducibility and provenance

Source: nflverse `player_stats.csv`, CC BY 4.0 compilation, retrieved from the project’s GitHub release. The run manifest records the exact SHA-256, code path, seed, model coefficients, walk-forward predictions, bootstrap outputs, reviewer findings, and publication decision. Underlying NFL data remain subject to their owners' terms.