SD and ES Explained: How Many Shots Do You Need?

Date Posted:9 November 2026 

Reloading guide

SD and ES Explained: How Many Shots Do You Need?

By Xhunter Team · Updated November 2026 · 3 min read

 

Standard deviation (SD) tells you how tightly your velocities cluster around the average. Extreme spread (ES) is simply the gap between your fastest and slowest shot. For a reliable SD, shoot at least 10 rounds; to tell two good loads apart, 20 or more. Five shots will give you a number, but by our calculation it can be out by more than double.

Every radar chronograph in our radar chronograph comparison works these out for you. Knowing what they mean is what turns a screen full of numbers into a decision about your load.

What the two numbers show

Here's a string of 10 shots from a hypothetical .308 load, plotted below. The average is 2,815.2 fps, the SD is 6.2 fps and the ES is 20 fps.

Ten shots: average, SD and ES Average 2,815.2 2,800 2,810 2,820 2,830 ES: 20 fps, slowest to fastest SD band: 6.2 fps either side of average Velocities in fps
Ten shot velocities with the average, SD band and ES marked. Illustrative.

The SD is a kind of typical distance from the average. Garmin's manual puts it plainly: a low SD means your velocities sit close to the average, a high SD means they're spread out. In a normal spread of shots, roughly two-thirds land within one SD of the average, which is why most of the dots sit in or on the edge of the grey band.

ES only looks at two shots, the fastest and slowest, and ignores the eight in between. One odd round, perhaps a slightly light charge or a cold-bore shot, can double your ES while barely moving the SD. That makes ES useful as a warning light and SD the better measure of consistency.

Why five shots isn't enough

Say you're at the reloading bench comparing two powder charges. Load A shows an SD of 9 fps over five shots; load B shows 12. It's tempting to call load A the winner. The chart below shows why you shouldn't.

You measured an SD of 10 fps. How far off could it be? 5 shots 6.5–23.7 10 shots 7.3–16.5 20 shots 7.9–13.7 30 shots 8.3–12.8 0 5 10 15 20 25 Likely true SD, fps (90% confidence) Our calculation from standard statistics.
How much a measured SD of 10 fps can differ from the true figure, by number of shots. Our calculation, 90% confidence.

By our calculation, a five-shot SD of 10 fps is consistent with a true SD anywhere from about 6.5 to 23.7 fps. Both of your loads sit comfortably inside that range, so on five shots each you can't tell them apart. At 10 shots the range tightens to about 7 to 17 fps, and at 20 shots to about 8 to 14 fps. The gains slow after that; 30 shots narrows it only a little further.

Our working rule: use five shots to rule out obviously poor loads, 10 to confirm a load for hunting, and 20 or more when you're choosing between two good ones for long range. The Garmin Xero C2 can resume a saved session, which makes building a 20-shot sample over several cool-down breaks much easier.

What SD should you aim for?

For hunting inside 300 m, we'd be happy with an SD in the low teens. Velocity variation that size has little effect on where the bullet lands at normal hunting distances; wind and your own hold matter far more. For steel past 600 m, a single-digit SD starts to earn its keep, because a slow round drops further the longer it's in the air.

Keep comparisons fair. In Recoil's side-by-side test, the Athlon Velocity Pro read SDs 8 to 15% higher than Garmin units on the same shots, as we covered in our Athlon Velocity Pro review. Compare loads on one chronograph, and treat a mate's SD from a different unit as a ballpark only.

Turning the average into better drops

The average velocity is the number your ballistic app needs. Enter the average from at least 10 shots fired in conditions close to how you'll hunt, not the figure on the ammo box, which comes from a test barrel. If your app's predicted drop still doesn't match real impacts at long range, adjust the velocity in small steps until it does. That process is called truing. The Xero C2 can send its average straight into Applied Ballistics; with the Xero C1 Pro or the Athlon, you enter it yourself.

One habit makes all of this more useful: note the temperature, rifle and lot of powder with each session. ShotView and Athlon Ballistics Lite both let you add notes, and next season you'll be glad you did.

SD and ES FAQ

What is a good SD for a hunting load?

For hunting inside about 300 m, we'd treat a velocity SD in the low teens as perfectly good, since wind and your hold affect the shot far more than small speed differences at that range. Single-digit SDs matter more for long-range target work past 600 m. Measure SD over at least 10 shots before judging a load.

What is extreme spread in reloading?

Extreme spread is the difference between the fastest and slowest velocity in a string of shots. If your slowest round reads 2,805 fps and your fastest 2,825 fps, the ES is 20 fps. It's quick to read but depends on only two shots, so one odd round can inflate it. Standard deviation is the better guide to overall consistency.

Which matters more, SD or ES?

SD matters more for judging a load, because it uses every shot in the string rather than just the fastest and slowest. ES is still useful as a warning: a big ES with a modest SD usually means one outlier shot. Check that shot for a cause, such as a cold bore or a charge-weight error, before you exclude it.

How many shots do I need to measure muzzle velocity?

For a reliable average and SD, fire at least 10 shots. By our calculation, a five-shot SD of 10 fps is consistent with a true SD between about 6.5 and 23.7 fps, while 20 shots narrows it to roughly 7.9 to 13.7 fps. Use five shots to screen loads, 10 to confirm, and 20 or more to compare two good loads.

Why does my SD change from one range session to the next?

Mostly because small samples bounce around. With five or 10 shots, the measured SD can vary widely even if the load hasn't changed. Temperature, barrel fouling, cold-bore shots and different lots of powder or primers can also shift it. Record conditions with each session, and build larger samples before drawing conclusions.

Should I include the cold-bore shot in my SD?

Record it, but tag it. Garmin's ShotView app lets you mark cold-bore and clean-bore shots and exclude them from the average and SD. For a hunting rifle, the cold-bore shot is often the one that matters most, so look at it separately rather than deleting it. Exclude it from load comparisons if it consistently reads differently.

Does SD affect accuracy at 100 m?

Very little. At 100 m, the bullet spends so little time in flight that small velocity differences barely change where it lands; your group size there mostly reflects the rifle, ammunition and shooter. Velocity spread shows up as vertical stringing at longer distances, because slower rounds drop further the longer they're in the air.

What muzzle velocity should I enter in my ballistic app?

Enter the average velocity from at least 10 shots fired from your own rifle, in conditions close to how you'll shoot. Don't use the box figure, which comes from a test barrel. If predicted drops still don't match real impacts at long range, adjust the velocity in small steps until they do, a process called truing.

How is standard deviation calculated for velocity?

Take the average velocity, then find how far each shot is from it, square each difference, add them up, divide by the number of shots minus one, and take the square root. Your chronograph does this automatically. The square root brings the result back into fps, so an SD of 6.2 fps reads directly as a typical deviation from the average.

Does temperature change muzzle velocity?

It can. Many powders produce higher velocities when hot and lower when cold, though how much varies by powder. If you develop a load in summer and hunt in a frosty winter, check your velocity again in cold conditions before relying on your drop chart. Note the temperature with each chronograph session.

Sample-size ranges are our own calculation using standard statistics; example velocities are illustrative.