Method
How we source and verify
The source ledger identifies the named document and record date for a quoted measurement.
Every figure published here names the document it came from, the date that document was read, and how often it is rechecked. Where two sources disagree, the disagreement is shown and both documents are named rather than one being quietly adopted. Where a figure is a convention that manufacturers publish and everyone uses but nothing mandates, it is labelled as a convention rather than presented as a rule.
And where a figure would have to be estimated, it is not published at all, which is why model tables need a named manufacturer, test-laboratory or reference source for each stated configuration, and omit unavailable fields instead of filling them with a guessed figure. A specification that is wrong by ten per cent is worse than no specification, because it gets trusted at the moment it matters.
Where figures come from
A quoted measurement needs its source document, the document identifier where available, its stated configuration and the date recorded. An inherited reference entry can still contain a transcription error despite having a named source. The full list is in the source ledger.
Sources are used in this order of authority:
- Federal regulation for anything about legality: size, weight, securement, emissions certification.
- Published standards for dimensional and interface questions: hitch categories, PTO types, coupler and mounting-plate geometry.
- Manufacturer documentation for anything specific to a machine: capacities, intervals, part numbers, ratings.
- Official test reports for measured performance, where an independent laboratory has tested the machine.
Listings, forums and video are used to find out what questions recorded in owner discussions. A dated listing can document an asking price or stated equipment, but cannot establish an independently verified specification. Model tables also cite reference databases such as TractorData and RitchieSpecs where identified; check their configuration against manufacturer documentation.
Standards, conventions and assumptions are labelled differently
Not everything useful is a specification, and presenting a rule of thumb as though it were one is the most common way a reference site becomes quietly unreliable. Three categories are distinguished on every page that uses them:
- Specified. A standard or regulation states it. Hitch pin diameters and the 102-inch width limit are in this category.
- Convention. Widely used, published by manufacturers, but not a standard. The 10–15% tongue weight band is a convention.
- Assumption. A default this site chose so a calculator has somewhere to start. The 0.044 gal/hp-hr fuel burn is an assumption, it is labelled as one, and it is editable.
How the calculators are built and checked
Each calculator is a deterministic function with a version number. Its formula, assumptions and worked examples are on the page, not in a footnote.
Every one carries a set of test cases covering normal values, boundary values, invalid values and safety conditions, with separate test commands for the calculators and page checks. A successful production build alone does not demonstrate that every test ran; the validation report must identify the checks that completed and returned results.
The calculator code and its test cases should be read together when checking validation. Browser behaviour, formula tests and page checks cover different parts of the result; a build success is not a substitute for those individual checks. Out-of-range values are rejected with a message rather than silently adjusted into range, because a calculator that quietly turns an impossible input into a plausible output is worse than one that refuses.
Results carry the version they were produced with. If a formula changes later, an old result still says which version made it.
Information outside the reference scope
- Specifications for a specific machine that we have not confirmed against a document for that machine.
- Repair procedures. Diagnostic reasoning and what a symptom indicates, yes; step-by-step repair instructions, no.
- Unsourced current prices, independent valuations or promises that a listed machine remains available.
- Ratings, reviews or endorsements of products or dealers.
- Any figure presented as verified when the underlying sources disagree. Disagreements are shown as disagreements.
Freshness
Every source has a recheck interval matched to its publication and revision history. Regulations and standards are checked annually or on revision; recall data is checked far more often. A source beyond its stated review interval needs another check against the relevant publication.
Its earlier observation date records what was consulted then and does not establish that the figure remains current.
Where a page depends on something time-sensitive, the date it was reviewed appears on the page.
How the worked-out ratios are calculated
A model page's "What the figures work out to" table is arithmetic run here on the maker's own stated figures for that specific model: it is never a number the maker publishes itself, and never carried in from another model. Each row uses one formula, applied the same way on every model page:
- Engine horsepower per 1,000 lb of weight: stated engine horsepower divided by stated weight, times 1,000.
- PTO power as a share of engine power: stated PTO horsepower divided by stated engine horsepower.
- Hitch lift as a share of machine weight: stated three-point lift capacity divided by stated weight.
- Pump flow per engine horsepower: stated pump flow in gallons per minute divided by stated engine horsepower.
- Hydraulic horsepower from flow and pressure: listed flow (gpm) times listed pressure (psi), divided by 1,714. Some makers print their own figure for a given model, shown alongside where it differs.
- Rated operating capacity as a share of tipping load: stated rated operating capacity divided by stated tipping load. The wheel loader and Mini excavators: excavator class pages explain why this ratio sits near half on a wheeled machine and nearer a third on a tracked one.
- Fuel tank per engine horsepower: stated fuel capacity divided by stated engine horsepower. It sizes the tank against the engine; it is not a fuel-burn rate.
- Rated payload as a share of empty weight, and per cubic yard of heaped body: stated rated payload divided by stated empty weight, and divided by stated heaped capacity.
A missing figure is not a zero and does not establish that an option never existed. Tables omit unavailable measurements rather than estimate them. For a mixed-version comparison, a populated value is attached to the named configuration; consult its manual for an omitted field.
Corrections
Errors get fixed and the page's review date changes. Where an error was consequential, a figure someone could have acted on, the correction is noted on the page rather than quietly overwritten. Corrections: Reporting an error takes a moment and is genuinely welcome; the most valuable reports come from people holding the manual we got wrong.
| Kind | Example | How much weight it deserves |
|---|---|---|
| Specified | The 102-inch federal width limit; the federal tiedown rules; hitch pin diameters under ASABE S217 | It is the rule or the standard applied to your numbers. It does not depend on our judgement. |
| Convention | The 10–15% tongue weight band; the 35% rated-capacity fraction for tracked loaders | These are conventions in particular documents or configurations, not universal settings or legal clearance. Use the actual trailer instructions or loader capacity definition; do not transfer the percentage to another machine. |
| Assumption | The default fuel burn in the cost calculator; the 80% field efficiency in the mowing calculator | A starting point we chose so a tool has somewhere to begin. Replace it with your own measurement. |
| Maker-stated | A model's engine power, weight, PTO, hydraulic flow and lift capacity as printed on the manufacturer's own product page | Taken from the maker's page and dated on each model page. An unavailable figure is omitted rather than estimated; the exact version and source remain important. |
Straight answers
How we source and verify: Questions and answers
Where do the figures on this site come from?
Primary sources: federal regulations read on the eCFR, published standards from ASABE, SAE and ISO cited by designation, and independent test data such as the Nebraska Tractor Test Laboratory reports. The document list records source names, observation dates and stated recheck intervals. Those fields help distinguish an earlier reference edition from the current manufacturer publication.
Where a standard is sold rather than published freely, the designation is the citation and no link is given.
Where do the model-level specifications come from?
From the named manufacturer document or the identified model-reference listing, with configuration and source distinctions; the machine models section lists what is covered. Publishing them accurately takes a data-acquisition programme, and an approximate specification is worse than none. Tyre choice, loader model, ballast and factory options all move lift capacity, operating weight and transport height for the same model name.
A figure that is wrong by ten per cent gets trusted at the point of purchase and discovered on a trailer ramp. Class-level reference plus the machine's own documentation keeps the general rule separate from the machine configuration.
What happens when two sources disagree?
Both are named and the disagreement is shown, rather than one being quietly adopted. Where the conflict is between a regulation and a secondary source, the regulation governs. Where it is between a standard and a machine, the machine governs, because a standard describes what a compliant part should measure rather than what a particular manufacturer built.
Where the evidence remains unresolved, the page says so.
How often is the information rechecked?
At an interval chosen for the type of source. Federal regulations are rechecked annually, because they change on a legislative timescale. Published standards are rechecked on revision, because that is when they change.
Test data is rechecked quarterly. The ledger records a review interval for each source. An elapsed interval calls for another source check; the date alone does not prove that a fresh document review has taken place.