Let’s talk about accuracy and precision for a moment.
To start, shown above is an example of high accuracy but low precision. The marks on the target are close to the central bullseye, but not close to each other.
Advertisement
Accuracy often refers to how close something is to its true or expected value. Let’s say that you aim to drill a 1/4″ hole. If the final result is a 0.2500″ hole, than you have achieved high accuracy.
Precision often refers to how repeatable something is.
Ideally, you want high accuracy and precision.
Let’s say that you aim to drill (5) 1/4″ holes. If your holes are 3x 0.27″, 1x 0.265″, and 1x 0.275″, you have achieved good precision.
There are times when you can get away with prioritizing precision over accuracy. Do your workbench legs need to be exactly 34″ tall, or can they be 34-1/8″?
But other times, dimensions are more critical. If your cabinet opening is exactly 24″, your drawer slides will require the drawers to be a specific width, with just a little leeway. If the opening is less accurate than you anticipated, it will affect the size you need to build your drawers to.
Advertisement
Here, this one’s an example of high precision but low accuracy. The marks are close to each other, but further away from the bullseye.
Which is more desirable? With high precision and low accuracy, each value will be off by a similar amount. With high accuracy and low precision, each value is closer to the true or expected value, but repeatability suffers. Things can sometimes average out.
So here, this is high accuracy and high precision.
This is the goal. But realistically, it takes skill, practice, and sometimes even talent.
Talking about 1/4″ holes again, you might consider two parts of a measurement, the nominal measurement and the tolerance: 0.250″ ±0.001″. Getting close to that 0.250″ size is related to accuracy, and getting multiple measurements within a certain window relates to precision. What this means is that you want your measurements to be within 0.249″ and 0.251″.
At times, that first nominal value can be more flexible. It might not be critical to change a certain dimension, either deliberately or through an unforeseen repeated error, such as not taking into account the exact kerf thickness of a saw blade. But that second value – if you mess up and achieve low precision, your work might look crooked, sloppy, or loose.
Here’s a rectangle where there is a small difference between opposite sides. The target is 400px for the horizontal edges, and 350px for the verticals.
That bottom right hand corner reflects small errors, but compound to create a noticeable gap.
If you don’t notice or correct for the error, and attach everything together, your corners won’t quite be square.
With higher precision, maybe the horizontal components end up being 395 units, and the verticals both 345 units. That might still work.
But if you have a 400 x 350 unit space to fill, you might need higher accuracy and precision.
Let’s say you need two holes, one through the top of a box, one through the bottom. Your holes end up being spot-on in diameter – 0.255″ to friction-fit a 0.250″ shaft. But you’re not accurate enough. The hole in the top was placed 0.1″ to the right of where it was desired, and the hole in the bottom was placed 0.1″ to the left.
IF your shaft fits, there will be a misalignment. If the top and bottom materials are thicker, then your shaft might not fit at all.
It’s not necessary to obsess over accuracy and precision. Well…
But it’s at least a good idea to keep it in mind. It’s hard to know what you can get away with, and which aspects of a project need to be better controlled. Sometimes things become obvious mid-way through a project, or worse – at the end. It’s usually far easier to get things right at the start, than to have to correct things afterwards.
There’s an adage – measure twice, cut once. That’s not always enough.
Unfortunately, there’s no clear cut answer about how to get better at achieving high accuracy and precision. My first 2×4 workbench was an absolute disaster. It was crooked, and wobbly, and I couldn’t for the life of me figure out what I was doing wrong. Well, I built it on the floor, starting with the top, not knowing that the floor wasn’t flat.
I’m sure I’m not the only one who forgot to account for saw blade kerf thickness when making a cut. Cutting a 36″ board in half doesn’t give you two 18″ boards. Especially not if you’re a poor student with no power tools but a drill and Dremel, and who trusted the worker at the home center to cut several boards to accurate sizing at the store.
Achieving high accuracy or high precision, or even better yet – high accuracy and precision – takes a combination of tools, skills, and experience. Even then, nobody is perfect.
firefly
I always thought that precision is easier to achieve than accuracy with less than ideal tooling. For example if I was building a workbench it probably might not matter if the 4 legs is 36″ or 35.9″ as long as they are the same. So if I cut them with the same set up there is a very high probability that they will be the same length. But if it was a mass production workbench the customer might not be so happy if when a replacement leg doesn’t end up being the same. So we can send them four legs at a time or build in small height adjustment at the feet to account for the different. Obviously the later option is more ideal since the customer might not have a level ground to set up the workbench to begin with. So the minor height adjustment serve both purpose. Isn’t that what set design engineer apart from each other?
William
When it comes to precision vs accuracy in a cabinet shop I generally aim for precision over accuracy. To do this I cut things that need to be the same size at the same time. With the way I build my cabinets, any deviation from accuracy will be covered or inconsequential. Stop blocks are key.
fred
Amen
Stop blocks and gang cutting help achieve consistency.
When I was doing most of the built-ins in my house – I learned that my wife preferred inset doors rather than overlay doors. Accurate measuring helps – but then making yourself a reveal gauge to get a consistent gap all around is just as important.
Placement of hinges, drawer pulls and other hardware is another thing – where a jig (homemade or otherwise) that gives consistent location is better IMO than trying to measure and mark each location with a ruler or tape. Similarly I’m a fan of using things like a story pole or story stick – used them and a spirit level long before I had a rotating laser.
There is also another issue about accuracy when it comes to how we perceive things. If a ceiling or trim is out of level – a picture hung on a wall that is absolutely level may appear to be tilted. Asked to build a deck between a house and a fence – we noticed that the intersecting fence lines were neither square nor parallel with the wall of the house – presenting some choices about how to treat the gaps that would be formed around the deck that would sit in the corner.
firefly
“Stop blocks and gang cutting”. I love fred precision with words.
Alick
Wise words.
To add a small example, a decent sliding combination square is a handy tool as the rule can slide to the required length and dimensions taken, set and transferred to a batch of parts without need to measure each one.
When buying cut lengths of 8020 extrusion, pieces cut in the same batch tend to be closely matched in length as the automated chop saws used by the suppliers have good precision.
KeithinKC
I learned a few years back that using a jig or template for locations of exposed screws or bolts can make results look very professional, even if you are not. It’s not a functional issue if none of your fasteners look like they’re lined up, but the little details are what stand out.
MT_Noob
I prefer to throw my dart against the wall first and then draw a precise and accurate circle around it.
: )
All kidding aside it is amazing how often we conflate/confuse precision and accuracy. It was actually came to my mind as I was reading yesterdays post about post about straight edges and I suspect that is what inspired this one.
Koko The Talking Ape
Re saw kerfs, that reminds me of one of my Dad’s projects. He had to saw the verticals slats for a wood fence about 80 feet long. To create his workflow, he decided to mark each cut using the previous board, forgetting to account for the saw kerf. Then he installed the boards in order.
The result is that over the length of the fence, the boards get progressively shorter. The last one is about a foot shorter than the first.
The odd thing is that he didn’t notice the top of the stack was noticeably shorter than the bottom. But a lot is explained by the fact that he is a mathematician.
ktash
@Koko maybe he was trying to make the yard look longer, by creating a “one-point perspective” effect. A mathematician might try something like this. 😀
Koko The Talking Ape
Yep, and it would have been a good effect too. But sadly, the fence is supported by preexisting posts built of brick. He installed crossbeams top and bottom, and these were the vertical slats to fill in that space. So the slat start out nicely proud of the beams, and end up barely reaching across the distance. But he used cedar and aluminum nails, so it is durable as heck!
Ktash
Ha! Durability, precision, accuracy.
michaelhammer
Accuracy and precision are why I plunk down $120 for a Starrett 12” combo square. There is a difference and it shows in my work.
Stuart
Recently?
My last few Starrett purchasing attempts were highly disappointing. Example: https://toolguyd.com/starrett-tools-tap-wrench-bad-quality/
Alex
I believe that Starrett measuring tools are of high quality, however given that the local hardware store sells drill bits and such of the Starrett brand, i think they might just give their name to some off-brand crap for everything that is outside of their core market (levels, calipers, or any QC tool)
Jonathan
For a second I though TG had segued into a discussion of “tack drivers” (probably a hobby or EDC of many readers here).
I’ve found that demonstrations to my teenage daughters boyfriends to be a most effective communication tool, when sharing accurate and precise grouping of holes in paper from a sub-MOA tack driver, (Minute of Angle = MOA).
Ah the cruel transmogrification of a former Red blooded teenage boy into the father of gorgeous Teenage girls.
Will forgo the clever double entendres that are ready to roll off my fingertips.
I digress…
Sadly, while precision and accuracy can result in in boring cookie cutter houses, products and efficient assembly lines.
The pinnacle of precision and accuracy is a craftsman house or product—Sears offerings 21st century excluded 😉
Joe
Aren’t both expected at all times ? Especially for shop work. Being a framer we strive for both from the ground up…it only helps by the time you reach the roof…..
Dustin L Wright
I personally, as a mechanical drafter think how you define this terms should be flipped. With the target analogy people describe when aiming/shooting a target how accurate the shooter/tool is. “look how dead on accurate they are.” Not – Look how precise they are”
Percision is described as how close to absolute something is.
Just my professional two- cents
Stuart
Respectfully, my analogies draw from accepted scientific definitions and usage practices. The dart board is a common analogy.
Example: https://www.nist.gov/pml/nist-tn-1297-appendix-d-clarification-and-additional-guidance
“D.1.1.1 accuracy of measurement [VIM 3.5] closeness of the agreement between the result of a measurement and the value of the measurand”
“However, ISO 3534-1 [D.2] defines precision to mean “the closeness of agreement between independent test results obtained under stipulated conditions.”
Accuracy is how close something is to a true or target value. Precision is how close measurements or target dimensions are to each other.
So with a target, accuracy is how close a shot is to the center or intended mark, and precision is how close a series of shots are to each other.