CastLabfishing tools · fishing research

Opens a guest profile with one real lake (Lake Harriet, Minneapolis) and a clearly labelled simulated session dataset, so every analysis has something to show.

Accounts are local profiles stored in this browser (password hashed with PBKDF2-SHA-256). There is no CastLab server: nothing you log leaves this device unless you export it. Use Account → Export to move data between devices.

CastLabtools · research · simulation
Today
Go fishing
Fishing map Weather & best times Catch log BaitRadar GearMatch
Science
Research dashboard LureLab STATS CastCheck twin GEO Rod physics MECH Methods & exports
Profile
Account & data

Today

Pick a spot. CastLab shows the best times to fish there, the weather, which fish people have reported nearby, and what to tie on.

Log a catchChoose on the mapSee the shoreline in 3DLog a full trip

Best times to fish

Hour by hour

Pick a spot to see the best times.

Right now

Fish reported near here

from other people, via iNaturalist

What to use

Catch log

Your fish diary: every fish you land, with size, lure and place. See what other people have reported in the same area.

I caught a fish

The catch log is a diary of fish landed. For catch-rate statistics, also log whole trips (including blank ones) in LureLab.

My catches

What others have reported nearby

Research dashboard

One view of everything in your profile: your logged sessions, locations, controlled experiments, field observations and bench tests, and how well the prediction model has done against what actually happened.

Sessions per month

Prediction track record

Open validation

Experiments

Simulation vs measurement

Recent sessions

Fishing map

Your saved fishing locations, logged sessions and GPS field observations. Click the map to add a location; every other section (weather, LureLab, the CastCheck twin) uses these same locations.

Find water

Or click any lake or river on the map: CastLab names it and offers to save it, open its shoreline twin or show the fish reported there.

Add location

LocationSession with catchSession, no catchField observation

Locations

Weather

Hourly conditions from Open-Meteo for any saved location: the past three days and a seven-day forecast. Forecast values are model output, not observations. LureLab stores the hour's values with each session so weather can be analysed against catch outcomes.

Fishing-time index, next 48 hours

How the index is calculated

Air temperature & precipitation

Surface pressure & wind

BaitRadar inventory

Track lures, bait, line and terminal tackle. Items are linked to LureLab sessions, so each item shows how often it has been fished and its observed catch rate with a credible interval, and quantities drop automatically when you log lures lost or bait used.

Best combos for a spot

Add item

List the lures and bait you own. CastLab then tells you which of them suit a spot, tracks how each one performs on your trips, and warns you when you are running low.

More details (optional)

Inventory

GearMatch

Setup recommendations by target species and technique, matched against your BaitRadar inventory and saved rods. The ranges are common angling guidance, not measured data; the physics simulator checks any setup against beam mechanics.

What are you fishing for?

My rods

Choose a species and technique.

LureLab

Statistical analysis of your logged fishing sessions: Bayesian catch-probability and catch-rate estimates with credible intervals, a Bayesian logistic regression on lure, species and conditions, controlled lure-comparison experiments, and forward-in-time validation of every prediction.

Log a session

Log every outing, including blanks. Leaving out unsuccessful sessions biases every estimate upward.

Weather at that hour

Fetch from Open-Meteo (dates within the last 92 days) or type what you measured.

source: manual

Session log

P(at least one fish) by group

Posterior densities

Estimates table

Model specification

Bayesian logistic regression for P(≥1 fish). Choose which predictors enter the model; categorical predictors use the most common level as reference.

Effects (odds ratios, 95% interval)

Predict a trip

Design a controlled comparison

Sample-size planner

Registering freezes the design, the random lure order and the planned sample size before any data are collected.

Experiments

Every session is predicted with a model fitted only to sessions that happened before it, then scored against what actually happened. Nothing from the future leaks into a prediction.

Calibration

Cumulative Brier score

Forward predictions

Logged before the trip; outcome recorded after

CastCheck shoreline twin

A terrain model of a saved location built from public elevation data, with a water level you can raise or lower to see where the shoreline moves, how steep the bank is, and how the simulated shoreline compares with GPS points you record in the field.

Water level

Display

Bank model below the waterline

Elevation models record the water surface, not the lake bed. For drawdowns the bed is extrapolated by continuing the measured bank slope under water, capped at this depth. Lowered-water results are therefore an assumption, labelled in the readout.

2D map

drag to draw a transect
Water at chosen levelReference shorelineSimulated shorelineTransect

3D terrain

drag to orbit, scroll to zoom

Transect profile

Water area vs level

Gage height (USGS parameter 00065) is measured against a local gage datum, not sea level, so only changes in stage are transferred to the twin. A gage on a different water body, or far downstream, may not represent your location.

Daily mean stage vs historical range

Record a field observation

Simulated vs measured

GearMatch physics simulator

Large-deflection beam model of a tapered rod blank loaded by line tension at the tip. Compare up to three rod designs, check a full setup for compatibility, search for designs that balance performance, weight and cost, and calibrate the model against your own bench tests.

Load case

Rod configurations

Deflected shape

colour = bending stress / strength

Tip deflection vs line tension

Bending stress along the blank

Setup

Enter a setup and press Check.

Design search

Every combination in the grid is solved at the design load. Designs below the minimum safety factor are rejected; the rest are scored and the non-dominated (Pareto) set is marked.

Trade-off: mass vs performance (colour = cost index)

Top designs

Static tip-load bench test

  1. Clamp the blank horizontally at the point where the simulated length starts (top of the grip).
  2. Measure the unloaded tip height against a vertical rule.
  3. Hang known masses from the tip-top (weigh them on a kitchen scale) and record the tip height each time.
  4. Enter drop = unloaded height − loaded height. Repeat each mass at least twice.

Predicted vs measured tip drop

Saved bench tests

Methods & exports

Export your datasets, generate a research summary written from your current data, and read how every number in CastLab is calculated. Each graph also has its own PNG and CSV buttons.

Datasets

Research summary

A self-contained HTML report with methodology and results sections, computed from your data at the moment you click.

Verification

The numerical engines are checked against closed-form and published reference values every time the page loads.

Account & data

Your profile, units and the data stored for it on this device.

Profile

Back up and move data

Simulated demo dataset

Generates sessions from a stated logistic model with known coefficients so you can see every analysis working and check that the model recovers the true values. Simulated records are tagged and are never mixed with your observations.

Danger zone

Stored records