Your worst 2 shots at least
Garmin Xero C1 gets the SD wrong
- Thread starter legi0n
- Start date
Your worst 2 shots at least
For the same reason that extreme spread is a poor way to predict and characterize a population of group sizes, it's a poor way on muzzle velocity as well. The SD is more informative than the ES.
How so? Please explain how SD is more informative than ES in relation to "the range of speeds each of your shots"?
ES is exactly that - the range of speeds of the shots. How on earth can some OTHER calculated measure that will ALWAYS be a lower number be a better representation of that?
99.7% of your velocities will be within ±3SD (that is x̄ ±3SD)
95% within ±2SD
68% within ±1SD
this stands true regardless of the extreme spread of the shots cronographed.
Last edited:
The practical reason is marketing. People who use Garmin's will have smaller SDs (the newest fad) than people who use other chronographs. So they are happier, and will use word of mouth to incent others to buy and use.
There is no realistic chance that Garmin applies the wrong formula without a good reason, and I think the above is the reason.
ES can be a very poor representation of your data.
Freshly cleaned barrel with some liquid in it can result in higher pressures for that first shot, seeing a higher velocity (saw that this morning). Following shots cluster about the mean.
One low velocity round probably due to a slightly light charge.
ES will be high (it was!), but the bulk of the shots clustered around the average.
SD will give a much better indicator of the overall shot behaviour, and the ES was an anomaly.
ES, like group size, measures only the two worst shots of the group
SD, like mean radius, factors in every shot of the group and if you fire enough rounds, gives a very repeatable idea of how the ammo/firearm will perform
Mean Radius is under used because it is slow to do by hand (with calipers and some math) and impossible with small groups where the bullet hole touch and overlap. Hence the product "OnTarget TDS". In a five-shot group, group size throws away 60% of the data. In a ten-shot group, 80% of the data is discarded.
SD is easy to calculate by hand, but most / all chronographs do it for you. Except that the Garmin Xero does it incorrectly.
For example, if a long-range shooter is calculating vertical dispersion then the correct SD will be better.
If someone is comparing SDs with a Garmin user and trying to get the same SD from any other chronograph (any optical, or the Labradar LX or Andiscan V3 etc.) they'll have a tough time getting there.
Not the end of the world, just an easily-avoidable speed bump that Garmin decided to build into its product.
If you know your freshly cleaned barrel is going to effect things I would just not count that towards any calculations...
Actually, I would. I would set it aside, and compare it to other freshly cleaned first shots... But I wouldn't consider it with the rest of my data that is coming from a fouled bore.
Now that is an explanation I can get behind. I totally forgot about this, been a while since I took stats and my teacher was dreadful... Of course that is assuming a normal distribution, but that doesn't seem like an unreasonable assumption to make.
Question for you, as you seem to keep pretty meticulous notes and seem to have a fair bit of data at hand: Which tends to be bigger in your data, +-3SD or the ES?
That's what I use. I want to know the worst case scenario, so that when I'm considering drop at long range, I can put in my average, top and bottom and realize that the next round could be anywhere from say 10" low to 15" high (or whatever), but knowing that's the worst case scenario. I load to get my ES to a minimum, the SD is nice to know but less reassuring as I'm a natural pessimist when calculating a firing solution.
In absolute value, sometimes it is the ES other times it is the ±3SD. Mostly it is the ±3SD.
then the ES > ±3SD it means at least one velocity is in the 0.3% that exceeds ±3SD
when ES <= ±3SD it means the ES of the sample is totally within the normal range.
note: the 2 velocities defining the ES are usually not centered about the average (x̄)
edit: a few examples
Last edited:
I had always just assumed that the skew would be 0. So now I'll get some more chronograph data (from my brand-new Labradar LX) and start charting and calculating. Unless you already have enough data to share?
in the example above, I added the min/max of the sample and their respective diffs from the average
perhaps in a larger sample they would be closer to symmetrical
| SD= | 14.31 | 20.07 | 23.68 | 18.46 | 21.04 | 14.56 | ||||||
| AVG= | 2605.46 | 2591.74 | 2609.94 | 2669.71 | 2678.94 | 2661.78 | ||||||
| ES= | 37.00 | diff from AVG | 53.10 | diff from AVG | 90.20 | diff from AVG | 56.30 | diff from AVG | 62.90 | diff from AVG | 43.50 | diff from AVG |
| min= | 2590.80 | 14.66 | 2561.40 | 30.34 | 2568.10 | 41.84 | 2635.30 | 34.41 | 2656.10 | 22.84 | 2639.20 | 22.58 |
| max= | 2627.80 | 22.34 | 2614.50 | 22.76 | 2658.30 | 48.36 | 2691.60 | 21.89 | 2719.00 | 40.06 | 2682.70 | 20.92 |
| 1 | 2596.7 | 2566.8 | 2621.8 | 2644.0 | 2691.1 | 2675.2 | ||||||
| 2 | 2605.7 | 2614.5 | 2610.8 | 2684.5 | 2707.2 | 2682.7 | ||||||
| 3 | 2590.8 | 2602.3 | 2599.6 | 2682.2 | 2681.0 | 2647.7 | ||||||
| 4 | 2611.7 | 2606.7 | 2626.0 | 2679.7 | 2664.7 | 2666.7 | ||||||
| 5 | 2592.6 | 2563.7 | 2568.1 | 2656.3 | 2719.0 | 2643.4 | ||||||
| 6 | 2596.2 | 2561.4 | 2610.6 | 2635.3 | 2656.7 | 2675.4 | ||||||
| 7 | 2615.0 | 2590.5 | 2611.4 | 2678.1 | 2674.3 | 2664.4 | ||||||
| 8 | 2627.8 | 2604.7 | 2590.0 | 2691.6 | 2667.0 | 2657.9 | ||||||
| 9 | 2627.0 | 2603.4 | 2602.8 | 2673.5 | 2672.3 | 2665.2 | ||||||
| 10 | 2591.1 | 2603.4 | 2658.3 | 2671.9 | 2656.1 | 2639.2 |
but feel free to show us your LX and the wonderful data it produces.
If I wasn't gifted the Garmin I would have likely bought an LX come Christmas.
The trick is getting enough data to look statistically convincing without just going to a range and burning up a barrel for this purpose. This weekend I'm doing a Precision match and will have the LX beside throughout, should be enough.
re: getting an LX, Labradar has announced that they will do BC with a firmware update. As it stands, counter to what I'd been told, the data tracks the LX captures can be used to do BC with Excel and some formulae. Since BC is the reason I'm keeping a Labradar V1, after some testing it will go up on the EE.
You can't calculate ballistic coefficient from any of our consumer radars. Only the $50k huge ones that Hornady and the like use can do it. The radar has to be capable of tracking the bullet out to 1000+ yards to get any useable data for BC calculation.
Yes you can, and I've done it to good effect. The easy way, if you have a Labradar V1 : https://bc.geladen.ch/labrabaco/labrabaco.html
This is how I determined the G7 BC for a DTAC 6mm 115gr. Then I zero'd at 100, and later hit a 1MOA target first round at 617.
Labradar LX will have built-in BC at some near-future point. I look forward to trying it out.
I disagree for most purposes, having done it with short-range radar returns as above.
I do agree that ELR shooting uses adjustments from DOPE most of the time.
Having said that, one can calculate these ranges (e.g., 2 miles) rather than relying on DOPE with a CDM for the specific bullet you're using and a ballistics program that can work with it. Like what GeoBallistics does with the CDMs supplied by Barnes, among others.
But even here, the G7 is a very close match to CDM. https://bergerbullets.com/nobsbc/custom-drag-models-for-extreme-long-range/ , so I'm comfortable with G7s for most things most of the time.
