Entries that look extreme
| Entry (summarised) | What the comment gives | What it does not give |
|---|---|---|
| North Dakota, 2018-19, underbody blades: one set “lasted 3 days” on a truck that uses the underbody blade for 95% of its work; a second set was used 15 hours on and off | Days and hours for two sets; the share of work done with the blade | Hours for the first set, pavement and ice conditions |
| Same file: another comment reports 25 hours of use at 250 to 275 lbs of pressure | Hours and down-pressure | How wear was judged |
| Same file: one set “done in 18 days, 87 hours to be exact” | Both days and hours | Why that is a failure rather than normal life |
| Utah, 2006-07: traditional blades wore out after 21 hours on average; two sets of another blade lasted 124 and 108 hours | Average hours for the control, hours for the trial sets | Route and surface details |
| Illinois, 2009-16: “sets of blades that have lasted 5 years” | Calendar years | Hours of use per year |
What the table shows about outliers
Read the table as a set of different measurements rather than as a ranking.
- “Three days” and “87 hours over 18 days” are not comparable until you know hours per day. A truck that uses the underbody blade for 95% of its work may log many hours in three days.
- “Five years” tells you nothing about hours either, since a truck may plow rarely.
- Pressure, surface and ice type change wear, and comments often omit them.
How to read an outlier
An outlier is usually a data problem before it is a product problem. Work through it in this order.
- Convert to hours of plowing if the entry allows it; if not, mark it as incomplete. Calendar days hide how much the blade actually worked.
- Write down the conditions given and those missing. A three-day life with no hours, surface or speed is not a measurement.
- Check whether the same product has other entries and compare. One entry is an anecdote; several are a pattern.
- Ask the agency for hours and conditions. The Clear Roads contact or your supplier can usually reach the author.
- Do not average outliers with the rest without noting them. One extreme value moves an average much further than it should.
Using outliers well
An extreme low can point to a failure mode worth testing, such as poor fit or too much pressure, and an extreme high can point to a good use case. Treat both as hypotheses for your own trial.
Limits
These are opinions on few units. I make no claim about any product.
Boundary: Entries are opinions on few units with missing conditions; no claim is made about any product, and hours of use are often not given.
Sources and further reading. External references are dated; where a figure is quoted, the link shows where it came from.
Clear Roads Product Experience Feedback 2006-2021: Combined Results, Blades and Plows (Covers state feedback 2006-2021; accessed 2026-10-24) ↗The new blade lasts longer but leaves more cleanup work. How should the trial account for that? ↗Can we compare a front-plow trial with last year’s underbody blade records? ↗Carbide snow plow blades ↗Multiple-Edge Plows: What Three States Found ↗Choosing Which Blades to Trial: A Shortlist Method That Uses Public Data ↗Qualified Products Lists for Plow Blades: Questions to Ask Your State ↗
