Launch Monitor Numbers and What They Mean

A £450 radar unit sitting behind your ball will hand you fourteen numbers after every swing. Somewhere between four and seven of them are real measurements. The rest are modelled, estimated, or in one or two cases effectively invented by an algorithm that was fed two inputs and asked to guess a third. Nobody at Garmin or Rapsodo is trying to con you. The physics of a £450 box are just different from the physics of a £18,000 one, and the app doesn’t draw a line between “we saw this happen” and “we worked this out from something else we saw.”

That distinction is the whole game. A 14-handicapper who knows their unit measures ball speed honestly and guesses at spin will build a better practice session than a 9-handicapper who treats all fourteen numbers as gospel and spends six weeks chasing a spin figure that was never real.

What your budget golf launch monitor is physically capable of seeing

Two technologies, and they fail in opposite directions.

Doppler radar (Garmin Approach R10, FlightScope Mevo and Mevo+, Swing Caddie SC4, PRGR Portable) bounces microwaves off moving objects and reads the frequency shift. Radar is excellent at speed along the line of the beam. It struggles with anything requiring it to identify which part of a blurred mass it’s looking at, and it needs flight to work with. Most radar units want somewhere between 8 and 20 feet of ball flight before they have enough data to be confident.

Photometric systems (SkyTrak+, Bushnell Launch Pro, and the camera half of the Rapsodo MLM2PRO) take high-speed stills of the ball in the first few feet and measure it directly: dimple pattern rotation gives you real spin, stereo images give you real launch. They see almost nothing about the club unless you pay considerably more, and they’re fussy about lighting and ball position.

A budget golf launch monitor typically sits in one of those camps and then models everything the hardware can’t see. Here’s roughly how the popular units break down:

Print that. It’s the single most useful thing you can know about the box you already own.

Ball speed is the number you can build a practice session on

Ball speed is the honest one. Every device in the list above measures it, and they agree with a TrackMan to within about 1 mph when they’re set up properly. Everything else on a budget unit is either derived from ball speed or has a wider error bar.

The conversion that matters: in the range most club golfers live in, roughly 2 yards of driver carry per 1 mph of ball speed. A 130 mph ball speed at a sensible 13° launch and 2,600 rpm carries about 222 yards. At 140 mph it’s about 245. At 150, about 265.

That gives you a sanity check on any distance figure your app produces. If your R10 says you carried a drive 258 yards off a ball speed of 134 mph, the app is either running normalised altitude settings or it has decided you launched it perfectly. You didn’t carry 258. You carried about 228.

Iron ball speeds follow the same logic. A 7-iron ball speed of 105 mph at a 6,800 rpm spin rate and 18° launch carries about 150 yards. At 112 mph it’s about 162. If your range session produces a 7-iron ball speed of 108 mph and the app claims 168 yards of carry, check your settings before you change your yardage book.

Club head speed lies more often than ball speed, and it poisons smash factor

The R10’s club head speed reading is the most-argued-about number in amateur golf for good reason. The radar has to pick the club head out of a moving arm, shaft, and hands, at a distance, in around 4 milliseconds. Users comparing side by side with a TrackMan routinely report the R10 reading driver speed 3 to 5 mph low, with iron speeds slightly closer. It’s not random noise. It’s a systematic bias, which means it’s fine for tracking change over a winter and terrible for comparing yourself to a tour average.

Here’s why that matters more than it sounds. Smash factor is ball speed divided by club head speed, and it’s your strike quality score. Suppose your true numbers are 96 mph club speed and 138 mph ball speed:

TRUE:      138.0 / 96.0  = 1.4375   (off-centre, roughly 3/4" toward the toe)
R10 READS: 138.0 / 92.0  = 1.5000   (dead centre, tour-level strike)

A 4 mph reading error turned a mediocre strike into a perfect one. You’d walk away from that session thinking your contact is sorted and your problem is speed, when the opposite is true. If you’re going to lean on this metric at all, read Smash Factor Explained for Club Golfers first, because the number only means something once you know what your specific unit does to the denominator.

The workaround is cheap. Buy a can of Dr Scholl’s foot spray or a pad of impact tape, and calibrate your unit against reality for one session. Hit twenty 7-irons, mark the strike location on each, and note the smash factor your device reports. You’ll find your personal centre-strike value, which might be 1.31 or might be 1.39. That number, not the textbook 1.33, is your target from then on.

Spin is where budget units do their most creative work

Measured spin requires either seeing the dimple pattern rotate (photometric) or bouncing radar off something reflective attached to the ball. That’s why FlightScope sells metallic dot stickers and Rapsodo sells RPT balls at around £40 a dozen. Without them, the device takes your ball speed, launch angle, and club selection, and consults a model.

Estimated spin isn’t useless. It’s directionally right for irons hit reasonably well. It falls apart in three specific situations, and you should learn to spot all three:

Thin strikes produce wildly low spin estimates because the model assumes a normal launch-to-spin relationship that a thin strike breaks. Wedges produce absurd figures, with R10 users regularly seeing 12,000 to 14,000 rpm on a full sand wedge where the real number is around 9,500. Driver spin on a low-launching mishit gets guessed as low, when a low, spinny drive is one of the most common 18-handicap patterns there is.

Ranges worth knowing, for a male club golfer swinging a driver in the 85 to 100 mph band:

Above about 3,000 rpm with a driver, every extra 1,000 rpm costs you roughly 5 yards of carry and another 8 to 10 yards of roll, because the ball lands steeper. If your unit is estimating rather than measuring spin, you cannot chase this number. What you can do is chase the things that cause it: attack angle and strike height on the face.

Attack angle is the cheapest 15 yards in your bag

This one is measured reasonably well by radar units, because it falls out of the ball’s initial flight and the club’s approach in a way the hardware genuinely sees.

A worked example. Take a golfer at 94 mph driver club speed, hitting down on it at -3°, which is extremely common among players who learned to sweep irons and never adjusted for the tee shot. Their typical numbers look like 13° launch, 3,400 rpm, 138 mph ball speed, carrying about 218 yards.

Move that same golfer to a +2° attack angle by teeing the ball higher, moving it forward of the left armpit, and tilting the spine away at address. Club speed doesn’t change. Dynamic loft goes up slightly, spin loft comes down, and the numbers become roughly 15° launch, 2,500 rpm, 140 mph ball speed. Carry goes to about 240.

Twenty-two yards, no gym, no new driver. This is the single most reliable win available from a budget golf launch monitor, because attack angle is one of the few club-delivery numbers the cheap hardware reports with enough accuracy to act on. Watch the trend across ten shots rather than any single reading.

Face angle, club path, and why your unit probably can’t diagnose your slice

Start line is roughly 80% face angle and 20% club path for a driver, closer to 75/25 for irons. Curvature comes from the gap between the two: face-to-path.

If your ball starts 4 yards right of target and curves 25 yards further right, the face was about 3° open to target at impact and the path was somewhere around 3° to 4° left of the face. You don’t need a launch monitor to know that. You need a target line and an honest look at where the ball starts.

Budget radar units do report club path and face angle, and the reported values wobble by 2° to 3° on repeated identical-feeling swings. That’s larger than most of the corrections you’re trying to make. Treat them as a tiebreaker, not a diagnosis. If you want real face data on a club golfer’s budget, a HackMotion wrist sensor (around £300) measuring wrist flexion through the downswing will tell you more about your face control than any radar in this price bracket, because it measures the thing that actually causes the face angle.

Carry distance is a simulation, and the settings matter in Britain

No budget unit measures carry indoors. It measures launch conditions and runs a ball flight model. Outdoors it may track part of the flight and still model the rest.

That model has settings, and the defaults are often normalised to sea level at 21°C. Most UK golf does not happen at 21°C. Carry drops roughly 2 yards per 10°F of temperature drop, so a driver that carries 235 on a July evening carries about 229 in early February. Add a two-club wind and a soft fairway and your “230-yard driver” is a 215-yard driver for six months of the year.

Three more corrections nobody makes:

Hitting off a mat inflates your numbers. The club bounces into the back of the ball instead of digging, which rescues fat strikes and raises smash factor. A 7-iron that reads 155 off a mat is often 148 off winter turf. Range balls spin 800 to 1,200 rpm less than a Pro V1 with a 7-iron, and considerably less with wedges, so wedge gapping built on range balls will be out by 8 to 10 yards per club. Indoors into a net, a radar unit with only 6 feet of flight to work with is extrapolating aggressively, and spin estimates get noticeably worse.

Which AI tools actually read phone footage correctly

Now the part that determines whether any of this becomes a swing change.

Start with the camera, because no software recovers from bad capture. A driver head at 100 mph travels 1,760 inches per second. At 60 fps, which is what your phone shoots by default, the head moves 29 inches between frames, so there is no frame at impact and there never will be. At 240 fps it moves 7.3 inches per frame. That’s the minimum usable rate, and every iPhone from the 11 onward and most recent Samsungs will shoot 1080p at 240.

Frame rate isn’t the only problem. Shutter speed is. Indoors under LED garage lighting, the phone opens the shutter as wide as the frame rate allows, roughly 1/240s, and the club head smears across several inches in a single exposure. Outdoors in bright light the shutter drops to 1/2000s or faster and the club is crisp. If your impact frames look like a grey ghost, the answer is more light, not a better app.

Camera position, once, properly: down-the-line, phone on a tripod at hand height (about three feet), lens directly on the extension of your toe line pushed back, 10 to 12 feet away, landscape, ball and full club in frame at the top of the backswing. Face-on: same height, perpendicular to the target line, level with the ball. Mark the tripod feet with tape so next month’s video is comparable.

On the tools themselves, the field splits cleanly.

Sportsbox 3D Golf is the only consumer product that does what it claims at a meaningful level. It estimates a 3D skeleton from single-camera 2D video and reports pelvis sway, thrust, lift and turn, chest bend and side bend, and hand path depth. Against lab motion capture it’s good on large body segments, within a few degrees on pelvis and thorax rotation. It gets progressively less trustworthy the closer you get to the hands and the club through impact, where the pose estimation has the least to work with. Use it for the big stuff: early extension measured as pelvis thrust toward the ball, side bend at the top, hand depth in transition. Do not use it to argue about 2° of shaft lean. Around £15 a month at the player tier.

Rapsodo’s Impact Vision on the MLM2PRO is the best value pairing on this list, because it films impact at high frame rate and syncs the clip to the shot data. Seeing a 1.29 smash factor next to a video of your strike halfway out the toe closes the loop in a way that a number alone never does.

OnForm and V1 Golf are video tools with automatic club tracing and drawing tools. The tracing works. Neither is doing biomechanics, and neither pretends to.

Swing Profile auto-detects swings in a long recording and clips them, which sounds trivial and saves about twenty minutes per session.

Then there’s the category to avoid: apps that take a single frame or a low-frame-rate clip and return several paragraphs of confident coaching prose. These are language models describing a picture. They hallucinate faults. You can prove it in four minutes with two tests.

Test one, the mirror test: take a video of your swing, flip it horizontally so a right-hander appears left-handed, and submit both. If the analysis of the two clips describes different faults, the tool is generating text rather than measuring anything.

Test two, the repeat test: submit the exact same video file twice, a few hours apart. If the diagnosis changes, you have your answer.

Test three, if you want to be thorough: film a tour player’s swing off a screen and submit it. A tool that tells Rory McIlroy he’s casting the club is not a tool.

A 20-minute repeatability test for the box you own

Before you trust any number, find out how much it wobbles. Hit ten 7-irons with a deliberately repeatable, three-quarter tempo, off the same lie, with the same ball. Export the data, or write it down.

SHOT   CLUB   BALL    SMASH   LAUNCH   SPIN     CARRY
 1     81.9   107.4   1.311    18.4    6,740     151
 2     83.2   109.8   1.320    17.9    6,110     155
 3     82.6   108.2   1.310    18.8    7,450     152
 4     84.0   110.1   1.311    17.2    5,280     157
 5     82.1   107.0   1.303    19.1    8,020     149
 6     83.4   109.4   1.312    18.0    6,390     154
 7     82.8   108.6   1.312    18.3    6,880     153
 8     83.1   109.0   1.312    17.6    5,940     154
 9     82.3   107.8   1.310    18.7    7,210     152
10     83.6   109.6   1.311    18.1    6,520     154

mean   82.9   108.7   1.311    18.2    6,654     153
sd      0.6     1.1   0.004     0.6      810       2.3

Read that the way a scientist would. Ball speed standard deviation of 1.1 mph across ten swings that felt identical: that’s the device being accurate and you being human. Smash factor holding to ±0.004: usable. Spin with a standard deviation of 810 rpm on shots whose ball speed varied by 3 mph: that spin figure is being generated, not observed, and you should stop looking at it.

Run the same test with a driver. Run it again in three months. The comparison between your own sessions is where a budget golf launch monitor earns its money, because the systematic errors cancel out.

Turning trustworthy numbers into a practice plan

Work in this order, because each step makes the next one measurable.

Strike first. Foot spray on the face, 30 balls, 7-iron. You are not trying to hit it further. You’re trying to get twenty of thirty marks inside a 20p coin in the centre. Your smash factor mean should rise and its standard deviation should fall. A 16-handicapper who moves centre-strike rate from 30% to 60% typically gains 6 to 8 yards per iron and cuts their dispersion nearly in half, without any change to club speed.

Low point second. Put a tee in the ground two inches ahead of the ball and try to clip both. Off a mat, lay a towel an inch behind the ball. Watch the carry standard deviation in your ten-shot sets: the fat-and-thin golfer has a carry sd of 8 to 12 yards, the clean striker is at 3 to 4.

Attack angle third, driver only, using the setup changes above. Target +1° or better. Check it across ten shots, not one.

Face control fourth, using start line rather than the reported face angle. Put an alignment stick in the ground 15 feet ahead, two feet right of target. Hit 20 balls and count how many start left of the stick. That’s a more reliable face measurement than anything a £450 radar will give you.

Speed last, and only once the first four are stable, because speed added to a poor strike pattern just makes the misses longer.

Here’s what to aim at, based on what club golfers actually produce rather than tour averages:

That last row is the one nobody tracks and the one that shows up on your card. Two golfers with an identical 150-yard 7-iron average score very differently if one has a 4-yard spread and the other has a 14-yard spread, because the second player is short-siding themselves twice a round.

The one habit worth building this winter

Pick a fixed ten-shot protocol: same club, same ball, same tempo, first thing in the session before fatigue sets in. Log the mean and the spread of exactly three numbers, ball speed, smash factor, and carry standard deviation. Ignore everything else the app shows you until those three move.

Then, once a month, film the same two angles at 240 fps with the tripod on your taped marks, and put the down-the-line clip through whatever biomechanics tool you’ve decided passes the mirror test. Compare it against last month’s clip rather than against a tour player. Your swing in February versus your swing in November is a question the technology can answer honestly. “What’s wrong with my swing” is not, and the tools that claim otherwise are the ones to switch off.