Walleye — Research & Methodology
Last updated: 2026-03-18
This page explains the live Walleye model in plain English. We show the drivers, the open-access sources behind them, what is research-backed, and what is calibrated. We do not publish the private numeric recipe.
How it works
PrimeBite scores Walleye Major windows only for the live Walleye model. A Major window is when the moon is overhead or underfoot. Each Major window gets a 1.0-5.0 rating, and the day rating is the average of the two Major-window ratings.
The live Walleye model combines four things: moon position, moon phase, one unified low-light driver, and a modest seasonal adjustment. That light driver combines sunrise/sunset timing with real solar radiation when the weather feed has it. If real solar radiation is not available, the solar part of the light driver goes neutral instead of guessing.
A high rating means conditions look better relative to the Walleye benchmark set. It does not guarantee a catch.
Drivers we use
Research-backed means the source supports the direction of the effect. Calibrated means PrimeBite chose the exact curve, cap, or weighting to keep ratings stable and balanced.
1. Moon position (Majors only)
Research-backed: trips were more successful when the moon was overhead or underfoot.
Calibrated: the exact way that effect is turned into a 1-5 score.
Source: Shaw et al. 2021 — PLOS ONE
“Anglers were observed to be about 10–12% more likely to be successful when the moon was overhead or underfoot…”
(PDF p.13)
2. Moon phase
Research-backed: gibbous phases were best in the study and quarter phases were worst.
Calibrated: the exact curve shape and weight.
Source: Shaw et al. 2021 — PDF
“The odds of a successful trip were highest during the gibbous phases… and lowest during the quarter phases…”
(PDF p.13)
3. Unified light driver
Research-backed: Walleye success improves in lower light, especially around dawn, dusk, and lower solar radiation.
Calibrated: the way PrimeBite blends sunrise/sunset timing with solar radiation into one light score.
Implementation note: the live model uses real Open-Meteo solar-radiation data when available instead of a cloud-cover stand-in.
Sources:
“…The predicted odds of a successful trip was highest at dusk … followed by dawn … and then day …”
(Shaw 2021 PDF p.8)
“Mean daily solar radiation had a negative effect on walleye trip success…”
(Shaw 2021 PDF p.11)
“Walleye … trip success and CPUE … generally increased as light conditions decreased…”
(Shaw 2021 PDF p.25)
“Observations 30 mins before and after sunrise were classified as dawn and observations 30 mins before and after sunset were classified as dusk.”
(Vasquez 2024 printed p.30)
4. Seasonality
Research-backed: open-access sources support a spring-heavy Walleye pattern, plus weaker seasonal movement across the rest of the year.
Calibrated: the exact month-by-month curve and cap.
Sources:
“During spawning activities, individuals congregate in the littoral zone of the lake and fish may be more active than at other times of the year… making them more vulnerable to anglers.”
(Shaw 2021 PDF p.26)
“The average harvest rate during the angling season differed between day types and peaked in May on weekdays but also December on weekends.”
(Deroba 2007 printed p.717)
“For many freshwater fishes and common sampling gears, CPUE, size and age structure, and condition are highest in the spring and fall…”
(Pope & Willis 1996 printed p.57)
What we do not use
- Minor windows: not used for live Walleye scoring.
- Overlap as a separate driver: removed from Walleye v2.
- Cloud cover as a solar proxy: removed now that real solar-radiation data is wired in.
- Wind speed, precipitation, and pressure: not used in the live Walleye score.
Key definitions and checks
Moon phase position vs moon illumination
The phase model uses phase position, not just percent illuminated. That matters because first quarter and last quarter can look similar in brightness but are different phases.
Source: SunCalc docs
“Moon phase value should be interpreted like this: … 0 | New Moon … 0.25 | First Quarter … 0.5 | Full Moon … 0.75 | Last Quarter …”
What PrimeBite is and is not claiming
- We are claiming these drivers have evidence behind them.
- We are not claiming any rating guarantees a catch.
- We are not publishing the private numeric tuning used to turn those drivers into a stable 1-5 score.