Scott Austin on what 405 rounds of .22 Long Rifle taught us at more than 1,500 yards — an interview with the ballistician behind the 4.4-mile world record: 22lr Extreme Long range shooting ballistics
By Shepard Humphries, Nomad Rifleman -based on a video interview with Scott Austin
Disclosure: Some of the equipment links in this article are affiliate links. If you buy through one, Nomad Rifleman may earn a small commission at no extra cost to you. We link only to gear we actually used on this attempt or use ourselves — nobody paid for a mention here, and no manufacturer had any say in what this article reports.
Why we’re publishing a failure
Nine days ago, on July 16, we went out into the high Wyoming desert and tried to break Chris Baxter’s world record for the longest confirmed hit with a .22 Long Rifle. We fired 405 rounds over six hours. We did not hit the target.
That’s the punchline, and I’d rather give it to you up front than bury it.
But here’s the thing I keep coming back to. As we push the limits of extreme long range shooting — or extended extreme long range shooting, which is really what a subsonic rimfire at nearly a mile amounts to — I’ve started thinking of us less as shooters and more as researchers. We had an outdoor laboratory. We ran an experiment. The experiment produced a result. The result was “no,” but a “no” with 405 data points attached to it is worth considerably more to the community than a “yes” with no notes.
So we’re reporting our conclusions. Other people are trying this. Some of them will succeed, probably sooner than we will. If our 22lr Extreme Long range shooting ballistics data shortens their road, good.
I’ve already written a separate article about my own experience on this shoot. I was the forward spotter, which meant I spent six hours folded into a cramped plywood bunker seven yards from the target, listening for impacts, and I whine about it at length elsewhere. That piece is about ears and dirt and the strange discovery that audio spotting is far more accurate than I ever gave it credit for.
This piece is about the brain.
Scott Austin is my founding partner in Nomad Rifleman. He is the ballistician on our projects and he was the shooter’s coach that day. He built the ballistic solution for the 4.4-mile world record shot in September of 2022 — 7,744 yards, connected on the 69th round — and he built the solution for this one. My job on the interview that follows was simply to ask questions and get out of the way. Where his written field report from the day contains a harder number than the one he recalled on the recording, I’ve used the report and said so.
A note on the team, since a laboratory is only as good as its people. Frank D Jennings was our shooter, and he’s a friend who has earned the right to be named. Mike Demetris ran communications and doubled as a second set of ballistic eyes, trying to work out in real time what the bullets were actually doing. Scott coached and ran the numbers. I sat behind a bunker and listened. (My story here)
And one bit of housekeeping before we get into it. I’m turning comments off on this one. I have no interest in being second-guessed by haters who have never spotted a shot past 300 yards. If you are a serious ballistician, if you are interested in one of our ELR shooting experience packages, or if you actually shoot a .22 Long Rifle past 1,000 yards with any consistency, or if you regularly put rounds out past a couple of miles with centerfire — please, feel welcome to be in touch. We will talk your ear off and share almost everything we have. We’re just cutting out the noise this time.
The setting
I’m not going to tell you our secret fishing hole, and I’m not going to give you the exact distance. Further attempts may follow and I’d like them to be ours. What I will say is that the target sat a little more than 1,500 yards away.
The standing record for a .22 Long Rifle, last I checked, is 1,445 yards. That was the number we were trying to beat.
Conditions, for anyone trying to replicate:
| Parameter | Value |
|---|---|
| Date | July 16, 2026 |
| Location | High Wyoming desert, general Boulder – Daniel area |
| Elevation | ~7,200 feet |
| Sky | Overcast |
| Temperature | 65°F at start, 79°F by afternoon |
| Wind | 0–3 mph, left to right, very consistent all day |
| Target | Steel, 24 inches wide × 48 inches tall, on the ground, canted back ~25° |
| Rounds fired | 405 |
| Rounds chronographed | 219 |
| Duration | Six hours, three breaks of about two minutes each |
| Confirmed hits on steel | Zero |
| Closest identified miss | 1–4 inches in front of the target’s right-front corner (video confirmed) |
The ballistic picture
| Parameter | Value |
|---|---|
| Ammunition | Lapua Center-X, 40 grain |
| Average muzzle velocity | 1,104.4 fps |
| Slowest recorded | 1,082 fps |
| Fastest recorded | 1,139 fps |
| Extreme spread | 57 fps |
| Time of flight | 9.694 seconds |
| Total drop | 12,363.21 inches — 1,030.3 feet |
| Remaining velocity at target | 290 fps |
| Remaining energy at target | 7 foot-pounds |
| Spin drift (1:16 RH) | 12.71 MOA left |
| Coriolis (aiming primarily south) | 1.56 MOA left |
| Total elevation | 738 MOA — 700 on the Charlie TARAC, 38 on the scope |
| Maximum hold used | 20 MOA left, inclusive of spin drift and Coriolis |
Two numbers in that table deserve to be read twice.
One thousand thirty feet of drop. The bullet rose and fell through more than a fifth of a mile of vertical travel to cover less than a mile of horizontal. Whatever mental image you have of a rifle shot, discard it. This is artillery.
Seven foot-pounds on arrival. For scale, a .22 Long Rifle at the muzzle carries something like 110 to 140 foot-pounds. A decent air rifle pellet lands with more energy than our bullets did after nine and a half seconds of flight. Hold onto that, because it becomes the whole story of why finding the impacts was the hard part.
Shepard: That wind. I want to dwell on it for a second, because people underestimate what a gift it was.
Scott: After I looked at our dope and whatnot — probably three-ish MPH. At the most. It was the best day we’ve ever had out at this location.
Shepard: We’re probably not going to catch that day again.
Keep that in mind as you read everything below. Every problem Scott is about to describe, we encountered on an overcast day with essentially no wind and remarkably stable conditions. Take our numbers, then add wind.
Part One: Why the last bullet is the only honest witness
I started with a question I’ve wanted Scott to answer on the record for years, because it cuts against how most people assume this ELR shooting thing works.
Shepard: Walk me through why, in extreme long range shooting, we place so much trust in where the last bullet landed — more than what the wind flags are telling us, more than mirage, more than the Kestrel. Why do we trust that last impact point the most?
Scott: It gives you real world data. It gives you exactly — I mean, there’s a couple of guys in the ELR industry that have a couple of awesome quotes. One of those is Ken Oehler, of Oehler Research. He’s an old timer in this. He says it this way: the bullet doesn’t lie, you must listen.
Shepard: I love that. That is good.
Scott: And then a more contemporary to him would be Todd Hodnett out of Texas. And he says it this way. He says that bullet doesn’t get to vote. It doesn’t lie.
The bullet impact itself gives me the best real world data. Then it becomes simply a matter of walking it in — but of course there’s a process for that.
So as much as I would like to say I have all of the numbers, ballistically speaking — it didn’t exactly work out that way. It was a matter of, okay, where was our impact? Where was our heard impact? Where was our known impact? And we were basically walking it into the target.
After the fact, I’m looking up the ballistics of this and going, oh, that’s why this was this way, or that’s why that was that way. And of course, next time I’m going to be a little bit more attentive to the details on that.
I want to pause on this, because it’s the philosophical center of how Scott works and it’s the thing that separates a working field ballistician from a person with a very good app.
A ballistic solver is a model. Models are built from assumptions — muzzle velocity, drag coefficient, atmospheric density, twist rate, sight height, a dozen more. Every one of those assumptions has error bars. Stack ten of them across nine and a half seconds of flight and a thousand feet of vertical travel, and the accumulated uncertainty is measured in tens of yards.
The impact crater has no error bars. It happened. It is the only piece of information in the entire system that is not a prediction.
So the discipline is: use the solver to get into the neighborhood, then throw away your faith in it and let the dirt tell you where you are. Scott’s willingness to say it didn’t exactly work out that way about his own numbers is, to me, the single most credible thing a ballistician can say. The ones who are never wrong are the ones who aren’t measuring.
Part Two: The 57 feet per second problem
Here is where the day was lost, and Scott identified it fast.
Shepard: Do I have my number right — that it was a 57 foot per second velocity spread throughout the day that you recorded?
Scott: Exactly. Exactly.
Shepard: And in precision shooting — even for thousand yard or two thousand yard shooting with centerfire — what’s a good standard? What does a person want to have?
Scott: I will quote David Crandall on this. He shoots 1,000-yard USA Palma. He has two or three gold medals to his name. And in his circles, he says if you have anywhere from 8 feet per second to 12 feet per second variation in your velocities, you’re going to do well in terms of 1,000-yard match competition. That’s what his team has, anywhere between 8 and 12, typically. But that’s with a .308.
Shepard: They shoot 155 grain bullets, right?
Scott: I believe so. You probably know that better than me. But that’s huge — your extreme spreads are huge. And of course we’re dealing there with a centerfire cartridge which is much faster, which is flatter shooting. I think Bryan Litz uses the term danger space. It’s in that danger space longer. So a higher probability of hit — versus what we’re utilizing, which is a .22 Long Rifle at subsonic velocities.
Read that comparison again. A gold-medal Palma shooter considers 8 to 12 fps to be championship-grade consistency at 1,000 yards with a .308.
We had 57 fps, with a bullet going roughly a third as fast, weighing roughly a quarter as much, at a distance half again as far.
Part Three: Longitudinal dispersion, or why this is mortar work
This next answer is the one I’d put in front of anyone who wants to know whether Scott Austin knows what he’s doing. It reframes the entire problem, and I don’t hear anyone else in the sport, except Mr. Litz, talking about it in these terms.
Shepard: At the lowest end of that — the slowest bullet you sent downrange at me, and the fastest one — from your ballistics calculations, how many yards apart would those two impacts be?
Scott: Let me pull up my handy dandy cheat sheet.
So this is called longitudinal dispersion versus vertical dispersion. Normally when you’re shooting with a supersonic bullet at normal ELR distances, you’re dealing with that variation vertically. We’re looking at it from basically the sky down, and from a longitudinal perspective — almost like somebody sending a mortar.
So the difference in those from the target, roughly, with our velocities: at 1,082 fps for our lowest, we were roughly 55 yards short. With our highest velocity, 1,139 feet per second, we were 103 yards long.
Stop and hold that.
A 57 foot-per-second spread in muzzle velocity produced roughly 158 yards of impact scatter along the line of fire. From 55 yards short to 103 yards long. That is nearly a tenth of a mile of dispersion, generated entirely by ammunition inconsistency, on a nearly windless day.
Work the arithmetic and it gets worse. One hundred fifty-eight yards divided by 57 fps is about 2.8 yards of impact shift for every single foot per second of muzzle velocity variation. Call it eight and a third feet of ground, per one fps.
Now consider the target. Twenty-four inches wide, forty-eight inches tall, lying on the ground and canted back about twenty-five degrees. Even being generous about the landing window a descending bullet has to work with, you are trying to place a projectile into something on the order of four feet of longitudinal real estate.
Which means — and this is my arithmetic from Scott’s figures, not a claim he made — the ammunition tolerance actually required to reliably strike that target is somewhere in the neighborhood of one foot per second of total spread.
Not 8. Not 12. Roughly one. Yes, 1.
That number does not exist in rimfire. It barely exists anywhere.
But there’s a second number, and it’s Scott’s, and it’s the one that actually matters. Asked in his field notes what has to be true before a shot like this can be made on demand rather than by probability, his answer is: perfect environmentals, a perfect shooter, and ammunition with extreme spreads inside 3 to 4 fps.
Those two figures aren’t in conflict once you see what each one is measuring. Roughly one fps is what it takes to make the hit a certainty, and nothing on the market will get you there. Three and a half fps puts about thirty feet of longitudinal scatter on the ground — which, against a four-foot window, is better than a one-in-eight chance on every in-node round you fire. At that rate you don’t need luck. You need an afternoon.
That’s the whole game, and it’s the difference between a team that fails and learns something and a team that fails and just buys more ammunition. Nobody is going to engineer certainty out of a rimfire cartridge. Scott’s 3-to-4 threshold is where the attempt stops being a lottery ticket and becomes a reasonable way to spend a day and a case of ammunition. We were roughly sixteen times outside it.
Scott: That’s the reason you weren’t seeing a higher percentage of impacts. You weren’t seeing those — it was more hearing that. Sometimes you would hear it and it was well beyond. And we would look at the chronograph and go, it could be that he’s hearing it but not sensing forward and backwards quite right.
And actually, let me take that back. If one hit behind you, you may actually think it was in front of you, based on the sound deflecting off of your target. That was our theorizing. Because if we looked at the velocity, the velocity didn’t match with that. And of course our best information was the true visuals.
Shepard: Right. Though I’ll say — having now had a lot more experience forward spotting, I’m placing a way higher value on audio than I used to. I’ve got 405 shots of experience in six hours now. I’d done it before, but at two or three mile shots where I was a few hundred yards back, or even at the 4.4 mile shot where I did some of the forward spotting while we were prepping. This was where I got most of my education, and I was really surprised at how accurate audio was. But more about that in my other article.
That exchange is worth noticing as a piece of method. Scott had two independent data streams — my audio call from the bunker, and the chronograph reading at the firing line — and when they disagreed, he didn’t discard either one. He asked what physical mechanism could produce the disagreement, and proposed a specific, testable explanation: sound reflecting off the steel face inverting my sense of front-versus-behind.
That’s a hypothesis, not an excuse. It can be checked next time with a microphone array or a different bunker geometry. Scott generates hypotheses about his own instruments. Field testers who don’t do that end up with beautiful data and wrong conclusions.
Part Four: What to do about a velocity you can’t control
Shepard: So we’re looking at roughly 158 yards of difference from a 57 foot per second muzzle velocity spread. Which makes me think about something you and Frank have been discussing — next time we do this, or if someone else does it, take your ten fastest and your ten slowest and ignore those results completely. Only record or care about that middle band. Is that what you two have been talking about? I don’t understand this part as well as you do.
Scott: You know, I think from a practical standpoint — if you send one and you’re more than 10 feet per second high or low, you might as well just plan on working on sending your next one. Because we’re talking a nine-and-a-half second flight time. And if you’re shooting in what we call a season of wind, you might as well maximize your time. As soon as that one’s sent and your velocity is out of the velocity node that you want, just send another one.
Again, we’re dealing with an almost 15 degree difference from the morning to the afternoon. Next time we do this, I’m going to track that better and actually see, at each temperature setting, how it changes that specific velocity.
Three ideas in there, and they’re all good.
The discard band. Every round still has to be fired to learn what it did — you cannot pre-select a good round out of a box. But once the chronograph reports, you know immediately whether that shot’s data is worth anything. If it left the muzzle at 1,082 or 1,139, it was never going to hit, and its impact tells you nothing about your solution. Logging it into your correction data actively poisons the well. Sort it out afterward and analyze only the middle band, and suddenly your dataset is honest.
An admission attached to that: of 405 rounds fired, our Garmin Xero C1 Pro captured 219. Just over half. The filtering method Scott is describing only works on rounds you actually measured, and we left 186 shots’ worth of information on the table. Closing that gap is one of the cheapest improvements available to us and I’d put it near the top of the list for next time.
The season of wind. I love this phrase and I hadn’t heard Scott use it before. In ELR, the conditions that make a hit possible are not continuously available. They arrive in windows — sometimes minutes long — and your total attempt is not really 405 shots, it’s however many rounds you can put into those windows. Which means every second spent carefully aiming a round you already know is out of node is a second stolen from a round that had a chance. It’s a time-budgeting insight disguised as a ballistics insight.
The temperature tracking. Fourteen degrees of swing from morning to afternoon, and Scott’s instinct is not to hand-wave it but to instrument it — build a per-shot record of temperature against velocity for this specific lot in this specific chamber. That’s not shooting. That’s building a dataset nobody else has.
Part Five: The communications architecture, and the acoustics
This part started as my complaint and turned into Scott improving my idea, which is roughly how our partnership goes.
Shepard: Something about our comm system that I love: when the person with the two-way radio back at the shooting position keys the mic, I can hear the shot through the radio. But just in case somebody drops something, whoever was on the mic — you or Mike — would also say “bang” or “shots away,” so I know that sound was in fact the shot.
Then I wait. About four seconds later I hear the actual muzzle report arrive — not on the radio, the real sound crossing the ground. Then I wait another six seconds or so and I hear the thud somewhere around me.
I know that even with a good chronograph there’s a little scrambled delay between the bang and the reading. But I think it would be handy for the forward spotter if the person at the firing line looked at the Garmin while the mic was still keyed and said “ignore this one” or “good shot.” Then I’d know whether to care where it landed. Because I really noticed out there — the few times I didn’t get a good warning that a shot was headed downrange, my attention wasn’t fully on it. Having all of my attention focused seemed to be a lot of help. That’s something I could have done differently: asked for that information.
Scott: Maybe one other thought on that — we would actually give you the chronograph reading. Because what’s really helping in all of this is, the more people that have the information, the better it is in the long run. You can start interpreting the data on your end as we’re seeing it on our end. If I say, for example — okay, our sweet spot was between like 1,101 and 1,104 feet per second.
And you’re under the speed of sound, and it’s interesting when you’re talking about hearing the bang. The bang is the initial crack. It’s not going supersonic, so you’re not hearing a supersonic crack. You’re hearing the initial concussion, and that is traveling at the speed of sound to you. All the while, the bullet started below the speed of sound and it’s slowing in its course toward you. So hence the reason you’re hearing that before the impact.
All of that, to me, is voodoo. That’s cool stuff. I mean, I get excited about that. And I’m still envious, because I have not been down at the bunker position, and I want to experience what you guys experienced.
Notice what he did there. I asked for a binary — good shot or ignore it. Scott’s counter was to give the spotter the raw number instead, on the theory that a spotter who knows the actual velocity can build his own model of what he’s hearing, rather than just executing someone else’s filter. That’s a systems-design instinct: push information to the edge of the network, don’t centralize the judgment.
And then the aside about the acoustics, which I think is the most quietly impressive thirty seconds in the whole recording. He explains, without being asked and without looking anything up, exactly why the report always beats the bullet in a subsonic engagement: the muzzle blast leaves at the speed of sound and stays at the speed of sound, while the projectile leaves at just under it and decelerates the whole way. Two different physical regimes, one arrival gap, correctly reasoned in real time.
And it checks out to the decimal. Sound crossing the roughly 1,493 yards from muzzle to bunker at our temperature and altitude takes about 4.02 seconds. Our measured time of flight was 9.694 seconds. Which leaves a gap of 5.67 seconds between the report reaching my ears and the impact reaching my ears.
What I said on the recording, from memory, sitting in that chair with no instruments: four seconds, then about six more. Any ballistician can look up a time-of-flight number. Fewer can tell you the mechanism cold, and fewer still have a spotter whose gut matches the solver to within a third of a second.
Shepard: That makes good sense, especially if we were doing an audio recording. That’d be a good backup to our scribbled notes — have someone on the shooting team say shots away, 1,103 feet per second at 79 degrees Fahrenheit. And then you’d have this multi-hour recording, unless we’re fortunate enough next time to get it in five minutes, that we could just have AI turn into a clean spreadsheet of every data point.
Part Six: Two men, two priors, and an honest zero
I wanted to talk about probability, because I think how a person estimates their odds before an attempt tells you more about their expertise than how they explain the result afterward.
Shepard: Going back to 2022, when we went for the world record with centerfire — I think my number was 1 in 1,000, or 1 in 10,000, for our probability of success that day. And I believe you were really cocky, and were saying something like you thought it’d be closer to 1 in 300. Am I remembering that accurately?
Scott: Yeah. I think if I’m not mistaken, I think you were 1 in 100,000. I think I was 1 in 400-ish, if I remember correctly.
Shepard: And we ended up, from a 69-shot data point, saying it was 1 in 69. We didn’t keep shooting 400 or 100,000 times, so we don’t really know — but from the small dataset we have, it would in fact be 1 in 69. Which was also the hit ratio for Paul Phillips’ team at Global Precision some years before. Theirs was 1 in 69 too. They got it on their 69th shot.
Scott was off by roughly a factor of six. I was off by roughly a factor of fifteen hundred.
That is not a small thing. Calibration — the ability to say in advance how likely something is and be approximately right — is the hardest and rarest skill in any technical field, and it is the one that cannot be faked, because the world scores it for you afterward. Scott’s prior was closer to reality than mine by more than two orders of magnitude, and he had never made that shot before.
Which makes the next answer more interesting, not less.
Shepard: So for this .22 Long Rifle shoot — what did you go into the day thinking our hit probability would be?
Scott: I’m kind of ashamed to say this, but zero.
I really did not — I really did not think we were even going to hear impacts. And I think part of that is based on, you know, you and Lynn had been out three different days and it was a struggle finding impacts, hearing impacts, seeing any impacts. That was quite the challenge.
And from, if I’m not mistaken, the first or second shot, you’re like, okay, I heard that in such and such a location. And the pessimism in me wanted to say, yeah, right. That was like a chiseler.
Shepard: A ground squirrel dropped a nut or something.
Scott: Yeah — because it’s such a huge difference. Here’s a bullet that’s traveled over 1,500 yards, it’s impacting at 7 foot-pounds of energy versus, if I’m not mistaken, 290, almost 300 foot-pounds for the 4.4-mile shot. So a significant difference in terms of hearing it, or lack thereof.
And I just really didn’t think we were going to do well. Initially I was all excited about this, thinking just like the world record 4.4 — it’s just math. Well, of course, once you get into the weeds, it’s more than just math. There were a lot of learning curves along the way, for certain.
And with this .22, I just thought — and again, I’m ashamed to say this — there’s no possibility we’re going to be anywhere near it. And I would say probably within 10, 15 shots, I turned around to Frank and I said, we’re going to do this. We’re going to do this today. I could not believe it.
Seven foot-pounds against roughly three hundred. That’s the number I want people to sit with. The 4.4-mile centerfire shot arrived with something on the order of forty times more energy than this bullet did, and that one was hard to detect. What Scott is saying, in effect, is that he expected the experiment to be unmeasurable — not that we’d miss, but that we wouldn’t be able to tell where we’d missed.
Shepard: I was amazed too. I thought there was hardly any possibility I’d actually hear the thud. In some of the testing before this we’d shot into a pond, and I was down at the far end while my wife was shooting, and I could hear it hit the water once, and hit the ground once or twice, but I couldn’t triangulate at all. I thought there was no chance of success.
And then the first shot I heard — and yeah, I agree with you, it was the first or second — I thought, oh my gosh. After that I could hear essentially every shot landing. I think I only lost 20 or 25 the whole day. I heard almost every single impact out of 405 shots fired.
Scott: That’s awesome.
Shepard: That’s a 95 percent data capture rate, near enough. Which really made me think — hey, this is actually something that can happen. And now I think it will happen, whether we or someone else sets it at 1,500 yards. Somebody’s going to set it at 1,600. It might not be us. But it just made the whole thing a lot more possible.
I’d note what a good scientist does with a badly wrong prior: Scott revised inside fifteen shots, said so out loud to the shooter, and then said so again on the record nine days later with the word “ashamed” attached. He didn’t retroactively claim he’d known. That’s the behavior you want from the person doing your math.
Part Seven: How close is close
Because a miss is not a single thing, and this particular miss deserves specifics.
The closest positively identified impact struck the ground approximately one to four inches in front of the target’s right-front corner. We have video confirming it. Four inches. After 1,500-plus yards, 1,030 feet of vertical travel, and nine and a half seconds of flight.
A red jacket we had staged near the target was hit three times. One of those projectiles was recovered lodged in the left front pocket, and it showed remarkably little deformation. The bullets broke through the first layer of Cordura and stopped. Seven foot-pounds is enough to punch a nylon shell and no more.
The spotter’s bunker — which is to say, me — took three impacts. More on that below, because it turned into the day’s cleanest piece of physics.
Roughly five rounds landed within a foot of the target.
I mention all of this not to make our failure sound like a success but because it’s the honest measure of where the project stands. We were not throwing rounds into the desert hoping. We had a solution that put bullets into a small piece of Wyoming repeatedly, and the thing standing between us and the record was a fifty-cent cartridge with too much variance in it.
Part Eight: The bunker geometry
Scott: Question — how far was the actual distance that you were in your bunker from the target itself?
Shepard: Seven yards back. And I would have been seven yards back directly under the target, which would have been ideal for listening. But the height of my bunker — it was a four-by-eight sheet of half-inch plywood on two stands, leaning back about thirty degrees, and even slanted down like that it blocked the shooters’ view of the target. So I couldn’t be there. I ended up seven yards closer and slightly left, which meant doing geometry in my head all day.
I think next time: ten yards back, directly in line, and either lower the bunker or raise the target. But having the target on the ground was actually a good idea. It would have helped me as the forward spotter to be directly under it and not have to do the geometry in my head — though for someone who isn’t from Tennessee, maybe the math would’ve been easy.
A small item, but it’s the kind of thing that costs you a record. The spotter’s position was compromised not for acoustic reasons but because his shelter occluded the shooters’ sight line. Every element of an ELR setup is coupled to every other element. Next iteration: get the spotter on the axis and solve the occlusion some other way.
Part Nine: Why the centerfire playbook doesn’t scale down
Shepard: So what’s physically different about pushing a subsonic .22 Long Rifle bullet to this range versus a centerfire? Why doesn’t the playbook we used for the Wyoming state record at 3.06 miles, and then years later the world record at 4.4 miles — .375 CheyTac and then a .416 Barrett chambering — why doesn’t that just scale?
Scott: I’m not sure I completely understand what you’re looking for. In terms of scaling, what do you mean?
Shepard: Well, some of the things I think about — when you’re shooting centerfire, you’re going to go from supersonic to subsonic. And you’re going to be able to handload the cartridge precisely for your barrel.
Scott: Got it, got it, got it. Okay. I understand it now.
Well, certainly with centerfire we would set up a pet load that would specifically be tailored to that specific barrel. And I could get my velocity nodes down. Generally speaking, when we’re doing those type of shots at a mile and beyond, we’re typically getting under 10 feet per second. Although the world record — again, that was different, but it was a different temperature than I had, or than Jake Muschany had, initially set that up for. So that changed things.
But with a .22 Long Rifle, you don’t have that option. Not only do you not have that option, but it’s going to change from lot to lot of ammunition. I can’t handload .22. You can, technically, but you’ll never get your extreme spreads anywhere close to even basic factory.
So you are relegated to utilizing the best ammunition that you have been able to find, shooting shot-to-shot, lot-to-lot. The ones that we found worked the best in our Vudoo .22 are the Lapua Center-X. When I initially tested that — and I don’t remember the specific lot number — I was getting around 12 feet per second, and I was pretty stoked.
But of course, as you can see from the other day, that particular lot, which arrived two days before our shoot, was considerably different. And here’s the thing: it might’ve done better at a warmer temperature, or a cooler temperature.
Shepard: Or a different barrel.
Scott: Yeah. Or a different — yeah. For sure, a different barrel. But there are so many more variables shooting rimfire than there are centerfire, in terms of being able to have control over your cartridge and over your ballistics. Your environmental ballistics, of course you don’t have control over that; you’re basically just chasing that for the day. But your cartridge ballistics, you try to maintain or control as much of that as you can — and you just can’t with a .22.
This is the structural finding of the whole experiment, and I want to state it plainly for anyone planning an attempt:
In centerfire ELR, ammunition is a variable you engineer. In rimfire ELR, ammunition is weather.
You do not tune it. You do not load for your barrel. You sample the market, you find a lot that your specific rifle likes, and then you pray that the next case you buy behaves like the last one. Scott had validated Center-X at around 12 fps. The lot that showed up two days before the attempt delivered 57. Same headstamp, same manufacturer, same rifle, roughly five times the spread.
There is no amount of shooting skill that recovers from that. Frank could have been a machine and the outcome would have been the same. Which reframes what a rimfire ELR record attempt actually is: it is a procurement and lot-validation problem wearing a shooting problem’s clothes. The team that breaks 1,500 yards will be the team that solved ammunition sourcing, not the team with the best trigger press.
Scott: There were some shots that just didn’t make sense whatsoever, and I’ve researched this a little bit. For example, we would be right in that 1,101 to 1,104-ish foot per second range, and of course as soon as we see the number we’re like, okay, this is the one, this is the one. And we would get the communication back that, oh, you were 10 yards long, or 10 yards short. And we’re just scratching our head going, what in the world?
There is a good possibility, because of the time of flight, the distance, and the height that it’s going — at that different layer of atmosphere — that we were dealing with some issues with thermals, pushing it up or pushing it down. Again, that’s a much lighter bullet than we were shooting for the 4.4-mile shot. So any type of thermal activity is going to have more of an effect on it than it would a 422-grain bullet.
Shepard: Right. And how many grains was the .22 pill? Forty?
Scott: Yeah. Forty grains.
Forty grains versus 422. A bullet roughly one-tenth the mass, spending nine and a half seconds in air that is itself moving vertically in ways no Kestrel on the ground can measure.
In his written report Scott lists the candidates for this residual dispersion as aerodynamic jump, thermal vertical movement, or something not yet identified, and explicitly declines to draw a conclusion until there’s more testing. I’d flag one complication with the thermal hypothesis that he’d be the first to raise: it was overcast all day. Cloud cover suppresses the surface heating that drives strong thermals, which makes them a less comfortable explanation than they’d be on a bluebird afternoon. Something moved those rounds ten yards while the muzzle velocity said they should have been dead on, and right now we have candidates rather than an answer.
That’s the honest state of it. This is the single most important open research question in rimfire ELR, and if you have data on it, we want to hear from you.
Part Ten: What actually mattered, and what was hiding in the weeds
Shepard: We’ve talked about this with the 4.4-mile shot, but I’ll ask about this one too. Spin drift, Coriolis, transonic effects. What came into play that really mattered? Not just technically — maybe technically they all did — but what actually moved the impact?
Scott: If I’m not mistaken, Coriolis was under two minutes of angle. Like a minute and a half. I’d have to look through my data and I’ll send you the exact. So that wasn’t much of an issue.
The subsonic realm — you have the advantage that you’re not dealing with a destabilization of your projectile. And that was proven in the fact that those holes on your bunker were perfectly round. I mean, they were going in point-first, round-nose in. They weren’t tumbling in any way, shape, or form.
Which would be the difference between trying to tailor a load for shooting extreme distance in a centerfire. If you have too long of a bullet with not enough twist, you’re going to end up — that bullet’s going to tumble at some point. We also know on our 4.4 that that was a stable bullet. We actually have the video and the picture of that bullet in flight. And we could see the holes in the ground, which were nice, small holes.
So there are some advantages in shooting a subsonic. And of course the disadvantage has to do with the weight that you’re shooting. The weight of the bullet is so light — so environmentals have much more of an effect on it.
His memory was good and his data was better. The report puts Coriolis at 1.56 MOA left, with the shot oriented primarily south.
The stability finding deserves its own paragraph, because it’s the cleanest result of the day. My bunker was half-inch plywood. All three impacts left a circular, point-forward signature with no evidence of keyholing, and each buried roughly a quarter inch into the wood before stopping.
That is direct physical evidence of a gyroscopically stable projectile arriving nose-first after more than 1,500 yards, and it settles a question that otherwise would have taken a lot of arguing: stability was never our problem. A 1:16 right-hand twist over a 40-grain bullet at subsonic velocity is entirely adequate at this range. Cross it off the list. And note the energy budget while you’re there — a quarter inch into softwood is the sum total of what seven foot-pounds will do.
I love this as a piece of methodology. The plywood wasn’t instrumentation. It was a bullet barrier and failed sunshade. Scott read it as instrumentation anyway.
Spin drift was the surprise
Scott: Well, spin drift was significant. We were initially holding pretty far left in order to make our hits — we were pointing into the wind a little bit — but spin drift was, and I don’t know the numbers off the top of my head, 14-ish rings a bell in my head. And 20 is the most that we ever held throughout that day.
So we were basically dealing with five minutes of angle left and right for our windage holds. Which tells you the wind was really mild.
Shepard: Yes.
Scott: And 14-ish minutes on that, which is significant. That was a surprise to me.
His recall was closer than it sounds. The report puts spin drift at 12.71 MOA left and Coriolis at 1.56 MOA left — which sum to 14.27, almost exactly the number he pulled from memory. What he was remembering was the combined non-wind horizontal correction, and he was off by a quarter of a minute of angle.
Back that out of the 20 MOA maximum hold and you’re left with roughly 5.7 MOA of actual wind correction across the entire day — and because the wind ran left to right, against a right-hand twist, there were stretches where it cancelled the drift and holds were made directly on the target.
Two findings there, both useful to anyone attempting this:
Spin drift dominated windage. Nearly 13 MOA of gyroscopic drift against under 6 MOA of wind correction. The bullet’s own spin pushed it more than twice as far sideways as the atmosphere did. Most shooters treat spin drift as a rounding error. At nine and a half seconds of flight it becomes your primary horizontal input — and Scott says plainly that the magnitude surprised him. Anyone modeling a rimfire attempt should plan for spin drift as a first-order term, not a refinement.
Coriolis is the hood ornament. Which brings me to the mail I get.
Shepard: After some of our other long-distance shooting, people got in touch trying to use our results to prove the earth isn’t spherical. We are not the people who can settle that for you. But here’s how much Coriolis mattered:
Picture your old Lincoln going down the highway, and it’s not getting the gas mileage you want, and you’re wondering whether it’s the hood ornament. Is there a microscopic amount of drag from the hood ornament? Sure, technically. Is the hood ornament why that car is getting the mileage it’s getting? No. That’s this. A minute and a half of Coriolis is the hood ornament. Our problem was the engine.
Part Eleven: The rifle, and what he’d change with an unlimited budget
Shepard: Tell me about the rifle you used. And if you could start over with an unlimited budget, what would you do?
Scott: Honestly, with the rifle, I don’t know that I would do much different.
So it’s a Vudoo Gun Works. The V3-RH-V22T 360 — so the three-lug version. 22-inch barrel, MTU contour, so it’s a much heavier barrel. One-in-sixteen right-hand twist. TriggerTech Diamond, two-stage, straight shoe.
Much of the specifics and the details on what we chose for this rifle came directly from Greg Roman, former Director of Compliance at Vudoo. He’s no longer with Vudoo, and Vudoo is no longer in St. George. But he was instrumental in how we built that rifle up. He had a number of other guys that were shooting extreme distances with a .22, and he gave me thoughts on what they were doing, and if he was going to build it, what would he do. He was the man I was looking to. And I still trust that to this day.
I’d add that Scott toured the Vudoo facility in St. George while it was still there, and Greg’s recommendations shaped this rifle specifically for this attempt. That’s worth saying plainly: the configuration wasn’t assembled from forum consensus. It came from the compliance director of the company that builds the actions, informed by the handful of people on earth already shooting rimfire at these distances.
Shepard: And the chassis?
Scott: It’s a KRG Bravo chassis. It has the night vision Picatinny rail forward of the scope so that we can mount our Charlie on it. It has adjustability in the stock, so I can fit that to the shooter specifically. I like that aspect of it, of course.
But I don’t think I would change any of it. A Vortex Razor HD Gen II, 4.5 to 27 by 56 with the EBR-7C reticle in MOA — that has been a workhorse for us for a lot of our precision work — using Spuhr mounts. LRA — Long Range Accuracy — Light Tactical bipod with the long legs, seven to ten inch, which is my favorite of bipods. Accuracy First anti-cant. And then of course the Charlie TARAC to give us that extra 900 minutes of angle more elevation.
Shepard: What did we have the Charlie set on, and what was our total MOA?
Scott: Our Charlie was set for 700 minutes of angle. And if I’m not mistaken, it was 738 total minutes of angle.
And we utilized the Kestrel for a ballistics calculator — which, you know, you and I go back and forth on, electronics or no electronics. In this case, it worked out really well.
The full kit
| Category | Item |
|---|---|
| Action / rifle | Vudoo Gun Works V3-RH-V22T 360 |
| Barrel | 22 in, MTU contour, 1:16 RH twist |
| Trigger | TriggerTech Diamond, two-stage, straight |
| Chassis | KRG Bravo, with NV Picatinny rail forward of scope |
| Optic | Vortex Razor HD Gen II 4.5–27×56, EBR-7C (MOA) |
| Mount | Spuhr |
| Anti-cant | Accuracy First |
| Bipod | Long Range Accuracy Light Tactical, 7–10 in legs |
| Elevation device | Charlie TARAC (approx. 900 MOA capacity, set to 700) |
| TARAC calibration | Digi-Pas DWL3500XY-B two-axis digital machinist level |
| Ballistic solver | Kestrel 5700 Elite w/ Applied Ballistics and Link |
| Chronograph | Garmin Xero C1 Pro |
| Ammunition | Lapua Center-X |
| Spotting scopes | Leupold Gold Ring 20–60×80 and 20–40×60, both with Impact reticles |
| Tripod | Really Right Stuff |
| Rangefinders | Leica Geovid 3500.com and 3200.com |
| Digiscoping | OLLIN Snapshot, mirrored live to a Smart TV |
| Spotter’s bunker | 7/8 in plywood |
Two entries in that table are more interesting than they look.
Seven hundred thirty-eight minutes of angle. For readers who don’t live in this world: a typical hunting rifle scope offers, at best, somewhere between 50 and 90 MOA of total internal elevation. We needed roughly eight to fifteen times that much just to point the muzzle high enough for the bullet to arrive. You cannot get there with a scope. You get there by stacking systems — the scope’s own travel plus a prism device forward of the objective that optically shifts the incoming image, letting the rifle point absurdly high while the shooter still sees the target. The Charlie TARAC will give you roughly 900 MOA if you need it. We used 700 of that and 38 from the scope.
The Digi-Pas. This is the entry I’d point to if someone asked me to prove Scott is a laboratory guy rather than a range guy. Setting 700 MOA of optical offset isn’t something you eyeball or trust to a factory index mark — an error of even a few minutes at that magnitude puts the bullet in a different zip code. So before firing, the Charlie was calibrated to exactly 700 MOA using a two-axis high-precision digital machinist’s level. A machinist’s level. On a rifle. In the desert.
That’s the “mortar” comment from earlier made literal. And the answer to “what would you change with unlimited money” being essentially nothing is worth noting on its own. The hardware was not the constraint. Scott knew where the money mattered before he spent it, and the failure came from the one component in the system that money genuinely cannot fix — a fifty-cent cartridge.
Part Twelve: What the cameras couldn’t do
One finding from the day belongs to the equipment rather than the ballistics, and it’s the one I’d most want other teams to know.
We ran an OLLIN Snapshot digiscoping system, recording every shot on video and mirroring it live to a Smart TV so the team could watch for impact signatures in real time. It’s a system that has performed exceptionally well for us on centerfire.
It could not do this job. The dust signature a 40-grain subsonic bullet raises at 1,500-plus yards — carrying seven foot-pounds — was too faint and too inconsistent for reliable shot calling. We could see some impacts. We could not see enough of them, or see them dependably enough, to correct from.
Which is why the forward spotter was indispensable. Not helpful. Indispensable. Of 405 rounds, the optics gave us 25 to 35 visual confirmations. My ears gave us roughly 380 detections and about 350 usable location estimates.
Anyone budgeting for a rimfire ELR attempt should understand the implication: the expensive glass is not what solves your feedback problem. A person in a hole downrange is. Plan for that person, protect that person, and put them on the target axis.
Part Thirteen: Trueing, and what the Kestrel was actually for
Shepard: Did you true your ballistic solution in the Kestrel against the impacts we were seeing? Or was it more like, that’s just to get us close, and then we use real-world impact points to walk in?
Scott: We basically used the impact marks to walk us in. And the Kestrel got us within, I want to say, 10 or 15 yards from the initial shot.
I don’t know if you remember the first shot — whether it was a visual or an auditory.
Shepard: It was auditory, and I believe it was 20-ish yards to the right, as I recall. But it was my first time in weeks hearing an audio bullet thunk, so I don’t know how accurate I was then compared to toward the end of the day.
Worth being precise about which component failed here, because the temptation is to blame the model.
The solver was good. Ten to fifteen yards of error on a cold first round at more than 1,500 yards, with a subsonic rimfire, over a thousand feet of drop, is genuinely excellent work. Scott’s model wasn’t the problem. The model got us into a fifteen-yard circle — and then the ammunition scattered rounds across a 158-yard corridor straight through the middle of it.
Then the impacts walked us the rest of the way, because the bullet doesn’t get to vote.
Part Fourteen: Who certifies this?
Scott’s field notes flag a question the recording never got to, and since it’s my end of the business I’ll answer it.
Nobody certifies this. There is no sanctioning body for extreme long range record shots, no governing federation, no ratified rulebook. What exists instead is reputation — a small community in which everyone knows roughly who everyone is, and in which a claim is worth exactly what the claimant’s history is worth.
Which means documentation is the whole of it. For the 4.4-mile shot in 2022 we brought independent witnesses, multiple camera angles including video of the bullet in flight, surveyed distance, recorded environmentals, and a shot-by-shot log — and then we published all of it, including the 68 misses. We maintain the record listings on the same principle: publish the data, publish the failures, and let people audit you.
That’s also why this article exists. A team that only publishes its hits is not a laboratory, it’s a marketing department. The 405 rounds we’re reporting here didn’t earn us a record. They earned us the standing to be believed the next time we say we hit something.
Part Fifteen: Garbage in, garbage out
Shepard: Anything I haven’t asked that you think matters to someone trying to have more success than we did?
Scott: A couple of thoughts.
Kestrel always uses this phrase: garbage in, garbage out. What you put into your ballistic calculator is going to get you as close as you’re going to get. Understand those inputs. There are lots of good YouTube videos to research your specific ballistic calculator. I love Applied Ballistics. I love the Kestrel as a tool, for sure. That would be one huge recommendation.
The other one is — it’s not an inexpensive road to go down, when you start adding up everything, as you well know. But there are some hacks to that, and maybe at some point we could come up with a way for the much more budget-conscious. I know our first book had some ideas in there with respect to the budget-conscious, because that’s where I was when I started. I didn’t have the funds to be able to do the long range, because of lack of equipment. And you learn the hacks around that. There are certain things you can’t get around, certain things you have to have. But anyway — it’s a worthy goal. I love it.
Shepard: Yes. So much fun compiling the data, doing the actual research and testing and experimenting on this.
And by the way, folks — we are not encouraging you to go hunting with a .22 at over 1,500 yards to get your moose this year. This is a recreational thing. We’re pushing the edges here.
Well — thank you, Scott. And thank you for figuring out all the ballistic stuff. This is not something I would have been able to work out. So thank you so much for being the brains behind this. And maybe we’ll get out and try it again sometime.
Scott: Well, thanks for making it happen. Really, really appreciate it.
What we learned: the summary sheet
For anyone planning an attempt, here is the whole experiment compressed.
1. Ammunition is the entire ballgame. A 57 fps extreme spread produced roughly 158 yards of longitudinal dispersion — about 8.3 feet of impact shift per single foot per second. Scott’s threshold for a shot that can be made on demand is an extreme spread inside 3 to 4 fps. Championship centerfire is 8 to 12. You cannot handload your way there in rimfire. Find the lot, buy the lot, validate the lot.
2. Never accept a case that arrives two days before the attempt. Our Center-X had validated near 12 fps in earlier testing. The delivered lot ran 57. Same product, same rifle. That single procurement decision cost us the day.
3. Chronograph every round. We captured 219 of 405 — barely half. Every unmeasured shot is an impact you can’t interpret and can’t safely correct from.
4. Use a discard band. Log muzzle velocity on every round, then analyze only the rounds that fell inside your node. Out-of-node impacts are noise and will corrupt your corrections if you feed them back in.
5. Fire faster when you’re out of node. A season of wind is finite. Don’t spend it carefully aiming rounds you already know are dead.
6. Instrument temperature against velocity, per shot. Fourteen degrees of ambient swing across a shooting day is normal, and its effect on a rimfire velocity node is not something to guess at.
7. Push data to the forward spotter. Give him the actual velocity reading over the radio, not just a verdict. He’ll build a better model of what he’s hearing.
8. Record the whole day as audio. “Shots away, 1,103 feet per second, 79 degrees.” Six hours of recording transcribes into a clean dataset later. Handwritten notes in a bunker do not.
9. Do not rely on cameras for shot calling. Our digiscoping setup, excellent on centerfire, could not resolve .22 LR dust signatures reliably. Optics gave us 25–35 confirmations; ears gave us roughly 380.
10. Put the spotter on the target axis and solve the sight-line occlusion some other way. Ten yards back, on centerline, low enough that the shooters can see over him.
11. Subsonic stability is a non-issue. Three circular, point-forward impacts in half-inch plywood with no keyholing, after more than 1,500 yards. A 1:16 twist over 40 grains is fine. Spend your worry elsewhere.
12. Spin drift is a first-order term. 12.71 MOA for us — more than twice our total wind correction. Model it properly.
13. Coriolis is the hood ornament. 1.56 MOA. Account for it, then stop thinking about it.
14. Residual dispersion is the open question. Rounds squarely inside the velocity node still landed ten yards long or short. Candidates: aerodynamic jump, thermal vertical movement, something unidentified. It was overcast, which complicates the thermal story. Nobody has answered this yet, and whoever does will own rimfire ELR.
15. Bring roughly 740 MOA of elevation, and calibrate the device that provides it with a precision level rather than a factory mark. Ours was 700 from a Charlie TARAC plus 38 from the scope.
16. Your solver is probably better than you think. Ours put a cold first round within 10 to 15 yards. Then let the impacts walk you in, because the bullet doesn’t get to vote — and it doesn’t lie.
A closing note about the humble man across the table
I’ve been doing this a long time and I’ve met a lot of people who can talk about ballistics. The number who can do what Scott Austin does is very, very small.
Build a first-round solution that lands inside fifteen yards at nearly a mile with a subsonic rimfire dropping a thousand feet. Diagnose the failure down to the specific lot of ammunition. Read a bullet barrier as a stability instrument. Recall spin drift and Coriolis from memory to within a quarter of a minute of angle. Explain the acoustic physics of why the report beats the bullet without looking anything up — and be right to within a third of a second of what the spotter’s ears actually reported. Calibrate 700 minutes of optical offset with a machinist’s level. And then say, on the record, “I’m ashamed to say this, but I thought our odds were zero.”
Anybody can be right in public. Scott’s distinguishing habit is that he keeps score on himself, publishes the score, and adjusts. In 2022 he called our odds at 1 in 400 when I called them at 1 in 100,000, and reality came in at 1 in 69. In 2026 he called them at zero and revised inside fifteen shots. Both times he told us the number before he knew the answer, which is the only version of that exercise that counts.
That’s what a field laboratory is. Not a place where you’re right. A place where you find out.
We’ll be back out there.
Before you try any of this: best practices
This article is a field report, not an instruction manual. Everything described here was done on private land with a known, controlled impact area, by a team that has been building and shooting extreme long range rifle systems for years. Read the following as the conditions under which what we did was defensible — not as a checklist that makes it safe by default.
The forward spotter is the part to think hardest about. A man sat seven yards from the target while 405 rounds were fired toward it, and three of them hit his shelter. That arrangement is only acceptable with all of the following in place at once: a purpose-built barrier rated well beyond the impact energy involved, continuous two-way radio contact with a hard stop that either end can call and that stops the firing line instantly, an agreed no-fire condition whenever contact drops, and a spotter who chose the position knowing exactly what was coming. Take away any one of those and it becomes something we would not do and would not want you to do. Do not put a person downrange to save yourself a walk.
Know where the bullet stops. At these distances your backstop is terrain, not steel. That means surveying well past the target, accounting for rounds that land a hundred yards long — ours did, routinely — and confirming nothing sits behind that impact area, including roads, structures, livestock, fences, and neighboring property lines you cannot see from the firing point. A .22 Long Rifle bullet arriving with seven foot-pounds is not harmless; it broke Cordura and buried itself in plywood.
Get the land right first. Written permission or ownership, and enough of it in the right direction. Discharge distances, shooting-hours rules, and rules about shooting across property lines or public land vary by state, county, and parcel, and federal land generally has its own restrictions. Verify locally before you go; nothing in this article is legal advice.
Basics still apply at 1,500 yards. Eye and ear protection for everyone on the line, muzzle discipline, an unloaded and open action any time anyone moves downrange, one person calling the range hot and cold, and a first aid kit plus a plan for getting to help — we were a long way from a hospital.
And the obvious one. This is recreational research at the edge of what a cartridge can do. It is not a hunting technique. Do not shoot at an animal at a distance where you cannot guarantee a clean, immediate kill, which for a subsonic rimfire is a small fraction of the range discussed here. Nothing above should be read as encouragement to try it.

Scott Austin is the lead ballistician of Nomad Rifleman and developed the ballistic solution for the documented 4.4-mile (7,744-yard) rifle shot in September 2022. Shepard Humphries is a Wyoming-based entrepreneur, author, and co-holder of that record. Their companion piece on the forward spotter’s experience during this attempt — six hours in a plywood bunker, 405 rounds, and the surprising accuracy of the human ear — is here.
If you shoot .22 Long Rifle past 1,000 yards, or centerfire past two miles, we hope this helps! And if you have data on longitudinal dispersion that muzzle velocity can’t explain, we especially want to hear from you.