How we research

How we research this

Every number here is either a published manufacturer figure with a citation, or computed from one. There is no third category, and this page is the whole process.

Tool storage is described in units that sound precise and are not. Gauge is a weight-derived index with uneven steps, so nobody converts it in their head. Drawer ratings are quoted per pair at full extension, which means half the internet quotes them per drawer and half per slide. Storage volume is published in cubic inches, a unit no human can picture.

The result is a category where a buyer cannot easily compare two products even when both publish their specs. That gap is what this site exists to close, and closing it does not require a test lab. It requires reading the listing, doing the conversion, and publishing the working.

Each product entry carries a Sources list linking to the manufacturer page or retail listing the figures came from. If a spec appears on this site, it appeared on one of those pages first. A spec without a citation would be a spec we remembered, and a remembered spec is a fabricated one.

There is no field in this site's content model for a derived number. Steel thickness in inches, the 80 percent working load of a slide rating, the cubic-foot conversion of a published volume, bench height from a body height - all of them are calculated by shared code from the published input, every time a page renders.

That is not a technical detail, it is what makes the arithmetic trustworthy. A number that cannot be typed in cannot be typed in wrong, and it cannot disagree with the same number on another page.

Steel gauge to thickness

From the Manufacturers' Standard Gauge for sheet steel, which is anchored on 41.82 pounds per square foot per inch of thickness. 18 gauge is 0.0478 in, 19 is 0.0418, 20 is 0.0359. The full table.

Slide rating to working load

Published ratings are per PAIR of slides at full extension. Published design guidance is to operate at no more than 70 to 80 percent of the rating, because of impact loads on closing, uneven weight distribution and fatigue over thousands of cycles. We apply 80 percent - the generous end - and say so wherever the number appears. The full page.

Standing height to bench height

Published ergonomic guidance puts a general work surface at about elbow height, precision assembly 2 to 4 inches above it, and heavy work well below. Elbow height is estimated at 62.5 percent of standing height. That coefficient is the only estimated constant on this site, it is an average across adults rather than a measurement of anyone, and it is labeled as an estimate everywhere it appears. The calculator.

1. 1Does the maker publish the numbers? A product that publishes gauge, slide rating, volume and weight ranks above one that publishes none, all else equal. In a competitive market, makers with good figures tend to print them.
2. 2Are the published numbers internally consistent? A listing claiming a 3,000 lb capacity across drawers rated at 75 lb a pair is telling you something about how carefully it was written.
3. 3Can it actually be bought? A recommendation you cannot act on is a magazine article. Where the best answer is a store exclusive we say so and name a spec-matched alternative.
4. 4Is there a stated reason to skip it? Every roundup on this site carries a What We Would Skip section. A list with no losers is a list nobody edited.

• A price we did not read from live data. Where no live price exists the page says Check price on Amazon rather than showing a remembered number.
• A customer rating or review count. We do not aggregate them and we do not emit review schema for products we have not evaluated.
• A testimonial. This is a new publication and it has no readers to quote yet; inventing one would be fraud.
• A credential the author does not hold, or an implication of tenure. Not once.
• A spec we could not find on a manufacturer or retailer page.

It is worth being explicit about the limits rather than leaving them for a reader to find. Published specifications describe a product as designed and do not describe build consistency, finish quality, how a drawer feels after two years, or what happens when a weld is bad. Those are exactly the things hands-on testing is good at, and we do not have them.

So the honest summary is that this site will tell you which box is specified to do the job and cannot tell you which one was assembled well on the day. Both matter. We publish the half we can actually verify, and we say the other half is missing.

If a number here is wrong, tell us and we will fix it and log the change. The editorial policy has the process. The whole value of published arithmetic is that it is checkable, which only means something if being checked leads to a correction.