Disclosure: Some links in this article are affiliate links. We may earn a commission on qualifying purchases at no extra cost to you.
Learning how to hover an RC plane, hold a torque roll and fly a harrier is one progression, not three separate projects: the harrier comes first, the hover second, and the torque roll last. All three are post-stall maneuvers, which means the wing has stopped producing useful lift, and the basic aerobatic maneuvers you already fly are the foundation rather than the goal.
What changes when the wing stalls is where the control authority comes from. In normal flight the wing does the lifting and the control surfaces bite clean air. In a harrier or a hover the wing is stalled, the aircraft is flying on thrust, and the tail only responds because the propeller is blowing air over it. That single fact explains most of the troubleshooting you will do on these three maneuvers.
If the tail feels soft, the fix is usually throttle rather than more elevator. If the model sinks in a hover at full power, no amount of stick work will save it, and the problem sits in the power system. If the aircraft oscillates, the inputs are arriving too late and too large for something that is balancing on its own propwash.
As an Amazon Associate, I earn from qualifying purchases.
What You'll Need Before the First Attempt
- A 3D-capable airframe with large elevator and rudder surfaces and a light wing loading
- A radio with separately switchable rates and enough channels for the airframe
- The manufacturer's high-rate throws dialed in, plus enough expo to keep the center of the stick precise
- A fresh, known-good battery that can hold current without collapsing under load
- A simulator, strongly recommended for the orientation work
- Altitude you are willing to give away, because every attempt should be recoverable
Before You Start: The Prerequisites Nobody Skips
These are not the maneuvers to learn early. AMA Park Pilot is explicit that an airplane can fall out of a hover in any attitude, so a pilot should already be able to recover from inverted and knife-edge flight before attempting one. The competency gate, in the order that makes sense:
- Conventional upright flight is automatic, not something you are still thinking about
- Flying toward and away from yourself feels the same
- Inverted flight and knife-edge recovery are reliable
- Slow flight and stall recovery are comfortable, and the rudder is a control you actually use
- You can identify and recover from an orientation mistake before the model hits the ground
Simulator time belongs here. RealFlight's hover challenge has been credited by pilots with a specific, immediate improvement in belly-in hovering and torque-roll orientation, which is exactly the skill that breaks first when the airframe rotates. Model Airplane News is careful to note that simulator 3D feels different from real flight, so treat sim hours as reflex and orientation training rather than a substitute. The simulator guide covers which packages are worth the install.
There is no defensible fixed timeline for any of the three maneuvers. Community reports range from a breakthrough within a few flights, once the correct throttle region clicks, to months of practice before feeling comfortable at low altitude. Work to skill gates, not to a calendar.
Safety altitude comes before technique. AMA guidance recommends new maneuvers at least one mistake high, meaning an unsuccessful attempt still leaves room to recover. Keep the model at least 25 ft from other people, and hold the same safety line between the aircraft and any spectators that applies to the rest of your flying.
Harrier: High-Alpha Flight That Never Stops Moving
A harrier is very-low-speed flight at an unusually high angle of attack, commonly around a 45° fuselage attitude. Both wings are deeply stalled, and control comes primarily from propwash over the tail. It is not simply slow flight with a lot of elevator.
How to Fly a Harrier
- Select the tested high-rate 3D flight mode.
- At safe altitude, slow the aircraft toward stall speed.
- Smoothly increase up elevator to raise the nose into high alpha.
- Add throttle at the same time, so altitude holds and propwash keeps the tail alive.
- Find the airframe's sweet spot, commonly near 45° rather than maximum elevator.
- Hold the angle with small, continuous elevator adjustments.
- Pulse throttle to maintain altitude and control authority without letting the model accelerate.
- Steer with small rudder bumps.
- In a straight harrier, correct wing drop by coordinating aileron and rudder in the same direction to counter adverse yaw.
- In a harrier turn, steer with rudder but add opposite aileron to keep the inside wing from dropping.
- Exit by smoothly lowering the angle of attack while adding enough power for conventional flight. If control departs, use power decisively while altitude remains.
Throttle Is a Control Surface
Holding too much power makes the aircraft climb or accelerate out of the harrier. Waiting until the tail has gone completely soft makes recovery too late. Model Airplane News treats throttle as a control surface in 3D, and its example airframe typically runs through roughly quarter-to-half throttle modulation. That percentage belongs to that motor, prop and airframe combination and does not transfer to another model.
The working rule is simpler than a number. If the tail feels soft or the model begins sinking, a brief power increase restores propwash. If it accelerates or climbs, reduce the sustained throttle level. If it departs controlled post-stall flight entirely, apply substantial recovery power while you still have altitude.
Diagnosing Wing Rock
Wing rock happens because the two deeply stalled wings are not behaving identically. The correction is usually counterintuitive:
| Symptom | Likely cause | Correction |
|---|---|---|
| Continuous side-to-side rock | Large or late aileron/rudder corrections, or the wrong angle of attack | Smaller, earlier inputs; test a slightly deeper high-alpha attitude instead of instinctively lowering the nose |
| Inside wing drops in a turn | Rudder-only turn | Opposite aileron to hold the wings, plus recovery power if the drop develops |
| Tail stops responding | Propwash too weak, throttle too low | Immediate brief increase in throttle |
| Aircraft accelerates out of the harrier | Too much sustained power, or the nose is too low | Reduce sustained throttle and restore the high-alpha attitude |
| Aircraft climbs excessively | Throttle too high for the selected angle of attack | Modulate down rather than freezing the stick at one setting |
| Strange behavior across several maneuvers | Excessive elevator-to-spoileron mix | Reduce or disable the mix and retest |
Deeper elevator can sometimes reduce rock rather than worsen it, by keeping both wings similarly stalled. An optional elevator-to-spoileron mix, raising both ailerons approximately 20 to 30% with up elevator, can reduce wing rock on some airframes. Excess mix produces abnormal responses, so treat it as a tuning tool for a specific model after the mechanical setup is clean, not a default 3D mix.
One caution on rudder direction. Model Airplane News tells pilots to use right rudder in its example, to counter the conventional left-turning tendency produced by propeller wash and P-factor. That direction depends on the propulsion rotation and the individual airframe's thrust and coupling. Write your corrections in airplane-relative terms and check which way your model actually pulls.
Hover: Zero Airspeed and Continuous Correction
In a hover, forward airspeed approaches zero, the fuselage is nearly vertical, thrust carries the aircraft's weight, and the propeller slipstream keeps the tail controls effective enough to balance pitch and yaw. The wing is fully stalled in the conventional sense. A successful hold is a continuous series of fine corrections rather than one fixed stick position.
Ailerons can be noticeably less effective here, because a large part of their area sits outside the propeller slipstream.
How to Enter and Hold a Hover
- Establish altitude and recovery space, and learn the control responses in a simulator first.
- Approach slowly in straight, controlled flight.
- Pull the nose smoothly and firmly toward vertical.
- In the traditional learning orientation, quarter-roll so the canopy or top of the airplane faces you.
- As forward speed decays, adjust throttle until the aircraft neither sinks nor climbs.
- Begin continuous small pitch and yaw corrections immediately.
- Use throttle for vertical movement, remembering that every throttle change also alters propwash and torque.
- Exit by adding enough power to establish a vertical upline, or transition the nose forward into a controlled harrier or forward-flight recovery when altitude allows.
The Hover Control Map
| Aircraft behavior | Immediate correction |
|---|---|
| Sinking while still vertical | Increase throttle smoothly |
| Climbing | Reduce throttle slightly while keeping propwash and control authority |
| Nose falls toward the wheels | Apply up elevator until vertical is restored |
| Nose falls toward the canopy | Apply down elevator |
| Nose or yaw moves airplane-left | Apply airplane-right rudder |
| Nose or yaw moves airplane-right | Apply airplane-left rudder |
| Conventional motor torque starts rolling the aircraft left | Apply enough right aileron to arrest it, if the propulsion produces that torque direction; reverse for opposite prop rotation |
| Corrections grow into oscillation | Reduce the size and duration of inputs, correct earlier, confirm adequate expo and no excessive gyro gain |
| Orientation becomes uncertain | Abort rather than stirring the sticks, add recovery power, and fly out in the orientation currently presented |
The pitch and yaw corrections are airplane-relative, not pilot-relative. Once a hovering aircraft rotates 180°, an elevator input produces the opposite motion on screen even though the aircraft-relative control has not changed. Right rudder stays right for the airplane while appearing left from the pilot's viewpoint.
Setup Requirements for Hover
- Start at the manufacturer's recommended 3D-capable center of gravity, not aggressively tail-heavy
- Set enough elevator and rudder to command the model while nearly stalled, using the manufacturer's high rates where supplied
- Add enough expo to make small center-stick corrections possible despite large maximum throws
- Carry enough static thrust to sustain the aircraft plus clear reserve for bailout
- Use a manufacturer-approved propeller that produces strong static thrust and propwash
- Check that servo and linkage free play is minimal, since a hover exposes weak or sloppy tail controls
- No mix is fundamentally required to learn a basic hover
- A rate gyro is optional. Rate stabilization can help, but an angle-limited self-leveling mode must not be allowed to prevent the near-vertical attitude
Torque Roll: Let Motor Torque Do the Work
A torque roll is an extension of a hover. The aircraft stays approximately stationary and vertical while motor and propeller reaction torque rotates the airframe around its longitudinal axis. Flex Innovations states the mechanism directly: establish the hover, then neutralize the aileron that was holding the rotation, allowing motor torque to roll the aircraft.
That is why a true torque roll is not taught as holding left aileron while hovering. Aileron can modulate or assist the roll rate, but the defining rotation is torque-driven. Getting this backwards produces a vertical aileron roll with the model drifting off its axis.
How to Fly a Torque Roll
- Establish a stable vertical hover in a known orientation.
- Stabilize altitude, pitch and yaw before anything else.
- Reduce or neutralize the opposite-aileron correction you were using to stop natural rotation.
- Let motor torque begin rotating the airframe.
- Maintain altitude with throttle.
- Continue elevator and rudder corrections as the aircraft rotates.
- Re-map the apparent control directions through each quarter turn.
- If the roll rate becomes too fast, a small amount of opposite aileron slows or arrests it.
- Exit by stopping rotation in a known orientation, then add power for a vertical upline or transition into a harrier or forward flight.
Orientation Remapping Through the Rotation
| Aircraft attitude | What changes for the pilot |
|---|---|
| Canopy or top toward the pilot | Familiar hover orientation, corrections read normally |
| Side-on at 90° | Elevator and rudder corrections appear as motion in an unfamiliar axis |
| Belly toward the pilot at 180° | Up elevator moves the airplane away rather than toward the pilot; right rudder remains aircraft-right but appears left |
| Opposite side-on at 270° | Corrections read as a mirror of the 90° position |
| Back to canopy | Return to the familiar reference |
That remapping is the real difficulty. Maintaining throttle is not the hard part. Maintaining pitch and yaw while the visual meaning of elevator and rudder changes continuously is, which is why AMA Park Pilot describes full torque-roll control as harder than learning a basic hover.
Common Torque-Roll Failures
| Failure | Cause | Corrective action |
|---|---|---|
| Aircraft spirals away instead of rotating in place | Pitch or yaw was not vertical before rotation began | Stop or slow the roll, re-establish vertical pitch and yaw, then resume |
| Pilot loses control at belly-in | Insufficient orientation training | Arrest rotation with opposite aileron and bail; practice the belly-in stationary hover separately |
| Roll too fast | Too little roll-arrest aileron, or strong torque | Add a small amount of opposite aileron |
| Torque roll will not start | Pilot is still unconsciously holding the roll-cancelling aileron | Relax toward neutral rather than forcing the roll |
| Altitude oscillates | Overactive throttle inputs or a narrow usable throttle range | Smaller, smoother throttle changes; calibrate the ESC and evaluate throttle response |
| Model sinks at maximum power | Insufficient thrust, battery sag, or the wrong prop and power system | Abort and diagnose the power system before more attempts |
| Pilot starts stirring the sticks | Orientation has been lost | Stop trying to save the maneuver blind; power out while altitude remains |
No special torque-roll mix is required. The prerequisite is a stable hover in more than one orientation.
Radio, CG and Power Setup That Makes All Three Possible
Flying the maneuvers is the visible part. The setup underneath decides whether they are possible at all.
Radio and Mechanical Checks
- Verify control directions in normal flight and in every stabilization mode
- Mechanically center the control surfaces before relying on large transmitter sub-trim
- Check servo arm and control-horn geometry, and eliminate loose linkage and free play
- Confirm high-rate throws against the actual aircraft manual
- Add expo to achieve precise center-stick behavior, without copying another aircraft's percentages
- Check full-stick movement for servo binding and current overload
- Ensure each surface reaches its commanded deflection without linkage flex
- Verify throttle calibration and the throttle cut function
- Confirm propeller orientation, tightness and condition
- Verify the battery can deliver the required current without excessive sag
- Start at the manufacturer's center of gravity
- Fly normal upright, inverted and knife-edge trim flights before judging 3D behavior, using the process in the trimming guide
- Test stabilization gains at normal flying speeds, and reduce any mode that produces oscillation
- Keep self-leveling or angle-limited assistance from fighting intentional extreme attitudes
- Do not install harrier mixes until the airplane has been flown mechanically clean
Servo selection and linkage setup matter more than they look. Extreme Flight's current build documentation warns that inadequate servos or linkages can produce flutter and the loss of the aircraft, which is a fair summary of what a hover does to a marginal tail installation. The servo guide covers sizing and installation.
Throws and Expo Are Model-Specific
There is no generic "3D rate." Extreme Flight's 67-inch Extra 260 shows the shape of a real setup, where enormous maximum throw is paired with substantial expo so the center of the stick stays precise:
| Surface | Rate | Deflection | Expo |
|---|---|---|---|
| Elevator | Low | 8 to 10° | 15 to 20% |
| Elevator | 3D | 45 to 50° | 60 to 65% |
| Elevator | XA / tumbling | 55° | 65 to 70% |
| Aileron | Low | 15 to 20° | 40 to 45% |
| Aileron | High | 38 to 40° | 70 to 75% |
| Rudder | Low | 20° | 40 to 45% |
| Rudder | High | 45 to 50° | 80 to 90% |
Those numbers belong to that airframe. Flex's QQ Extra uses substantially different percentages for the same purpose. Use magnitudes only as a starting reference and follow your model's manual. Expo sign convention also differs across transmitter brands, so check your radio's definition rather than assuming a positive number means the same thing everywhere. The dual rates and expo guide walks through the setup order.
Center of Gravity
Begin at the manufacturer's recommended CG and move it only incrementally, after conventional flight is proven. Extreme Flight places the 67-inch Extra 260's preferred CG at the rear edge of the wing tube, a datum that belongs to that aircraft alone.
Moving the CG aft reduces pitch stability and can make rotation and high-alpha response more immediate, but it also makes the model less forgiving. Copying another airplane's millimeter measurement, or deliberately going far aft on a new 3D model before it is trimmed, is not a shortcut. The CG balancer guide covers how to measure it properly.
Power and Propeller
A true stationary hover needs vertical thrust that balances the aircraft's weight, and climbing out of a failed hover needs thrust in excess of it. That establishes 1:1 static thrust-to-weight as the theoretical equilibrium threshold, not a desirable endpoint. A capable 3D setup should carry meaningful reserve so the pilot can accelerate away instead of holding full throttle just to stay in place.
There is no authoritative universal "ideal" ratio of 1.2:1 or 1.5:1 that applies to every 3D aircraft. Those figures circulate as hobbyist rules of thumb and should be treated as such. The useful targets are strong static thrust, immediate throttle response and propwash over the tail.
Propeller choice follows the airframe, not a generic "3D prop" prescription:
| Aircraft class | Propeller approach |
|---|---|
| 32-inch EPP profile | Large slow-fly prop for a tiny airframe, for example a 9-inch prop on the official 2S recommendation |
| 67-inch full-fuselage 3D | Model-specific size paired with a matched motor, ESC and battery, for example an 18x8 prop on a 6S 4000 to 5000 mAh system |
Stay within the manufacturer's approved diameter, pitch and current limits. Increasing prop diameter or pitch blindly can overload the motor or ESC, and if a hover needs everything the pack has, the underpowered troubleshooting guide is the place to start diagnosing.
What a Gyro Can and Cannot Do
A gyro can buy reaction time in gusts, and AMA Park Pilot notes that 3- or 6-axis stabilization can make a hover easier to stabilize. The limits matter:
- No universal gain percentage exists, and excessive gain produces oscillation
- Self-leveling and angle-limited modes can oppose the extreme pitch and bank attitudes that unrestricted post-stall aerobatics require
- Throttle and orientation remain pilot skills regardless of stabilization
- Transmitter rates and expo must be coordinated with any receiver-generated rates rather than duplicated
Flex's Aura documentation is a useful illustration. The QQ Extra's 3D mode ships with model-specific gains, rates and expo already programmed, and the manufacturer warns users not to blindly duplicate its internal reference values in the transmitter. Even with assistance, a pilot transitioning from a small Crack Yak to a 67-inch aerobat can still struggle to hold a hover at around 10% gyro gain. Stabilization is not the same thing as self-leveling, and neither one flies the maneuver.
Choosing an Airframe to Learn On
The progression runs from a light EPP profile, to a turnkey flat foamy, to a full-fuselage 3D aircraft. The 3D aerobatics model roundup covers the wider field; these three map directly onto the learning curve.
Twisted Hobbys TH 32-inch EPP Crack Yak-55
- 32-inch wingspan, 31-inch length, approximately 204 g / 7.2 oz
- Printed EPP profile construction with carbon-fiber truss reinforcement
- Pre-hinged control surfaces and adjustable control-rod end links
- Recommended battery 2S or 3S, 450 to 600 mAh, with a 9-inch prop on the official 2S setup
- The related Lite version runs a 2203-1800Kv motor, 15A ESC, three small digital servos, an 8x4.3 slow-fly prop and a 2S 350 to 600 mAh pack
- Airframe kit only; electronics and power system are sold separately
- Manufacturer lists hovering, harriers, rolling maneuvers and aggressive aerobatics among its intended capabilities
Check the Twisted Hobbys 32-inch EPP Crack Yak-55 on Amazon
Pros: EPP absorbs the imperfect recoveries that are part of learning, large control surfaces stay effective at low speed, and repairs are cheap. Cons: kit only, so the power system and servos are a separate decision, and the very low mass means gusts displace it easily.
Verdict: the strongest learner airframe of the three, and the one that matches how these maneuvers are actually learned. Perfect for a pilot who wants to practice without worrying about the repair bill.
E-flite Eratix 3D FF 860mm BNF Basic
- 860 mm / approximately 34-inch flat foamy
- 3S-compatible brushless power system with a factory-installed Spektrum Avian 15A Smart ESC
- Three installed Spektrum sub-micro digital servos
- Spektrum AR630 receiver with AS3X and SAFE Select in the BNF Basic version
- Requires a full-range Spektrum DSMX/DSM2-compatible transmitter and a 3S 600 to 850 mAh LiPo
- Manufacturer positions it for slow flight, sport flying and full 3D aerobatics
Check the E-flite Eratix 3D FF on Amazon
Pros: factory power and radio remove the setup uncertainty that slows down a first 3D build, and the low inertia keeps crash energy down. Cons: AS3X is a rate gyro rather than a self-leveling system, and the included SAFE Select mode must be switched off for unrestricted extreme attitudes rather than relied on.
Verdict: the sensible turnkey step between an airframe-only EPP kit and a full-fuselage model. Perfect for an intermediate pilot who wants to spend bench time flying rather than wiring.
Extreme Flight 67-inch Extra 260 ARF
- 67-inch span, 66-inch length, approximately 8 to 8.5 lb, 860 sq in wing area
- Preferred CG at the rear edge of the wing tube
- Recommended power: T-Motor AM670, T-Motor AM116A ESC, 6S 4000 to 5000 mAh battery, T-Motor 18x8 propeller
- Recommended servos include the Theta 989, Savox SV-1261MG or MKS HV-747
- Carbon fiber and genuine Ultracote covering, carbon landing gear, sealed aileron and elevator hinge gaps, removable SFGs
- Manufacturer specifically describes strong upright and inverted harrier performance, low coupling and low, slow 3D capability
Check the Extreme Flight 67-inch Extra 260 on Amazon
Pros: purpose-built for exactly these maneuvers, with the rigidity and hinge sealing that make high-alpha flight repeatable. Cons: at roughly 8 lb it carries far more inertia than a 200 g foamie, costs considerably more to repair, needs better servo and linkage setup, and demands a larger controlled flying site.
Verdict: a mastery and transition aircraft, not a first 3D trainer. Perfect for a pilot who can already hover a foamie and is moving technique into full-fuselage flying.
RealFlight Evolution RFL2001
- PC-based RC flight simulator
- More than 100 aircraft beyond the base models, plus dozens of 3D fields and PhotoFields previously sold as expansions
- Supports airplanes and helicopters, with online multiplayer and VR-headset compatibility where supported
Check RealFlight Evolution on Amazon
Pros: the cheapest way to attempt a hover a hundred times, and specifically useful for belly-in orientation and torque-roll work. Cons: simulator 3D feels different from real flight, and hours do not map to a fixed number of real-world flights.
Verdict: reflex and orientation training, not a substitute for airtime. Perfect for the winter before a first serious 3D season.
Common Mistakes to Avoid
- Learning low. Hovering is easier to read close in, but the ground is the worst place for a first attempt. Low-altitude 3D is a later proficiency exercise, and it should be kept inside an appropriate club or controlled flying area.
- Flying over people, vehicles or structures. AMA safety code principles keep powered models at least 25 ft from other people except in defined pilot, helper, takeoff and landing situations, with an established safety line to spectators.
- Copying another model's CG. "Tail-heavy for 3D" is an oversimplification, and every airframe publishes its own datum and range.
- Treating full-throttle hover as normal. If the model needs everything the pack has to stay up, there is no reserve to climb out.
- Assuming 3D rates are universal. High throws are paired with high expo, and the numbers differ by airframe and by radio brand.
- Steering a harrier turn with rudder alone. The inside wing drops without opposite aileron.
- Correcting late and large. Big inputs after the fact turn into oscillation and wing rock.
- Stirring the sticks after losing orientation. Power out in the orientation the model is currently presenting.
- Expecting a gyro to fly the maneuver. Stabilization buys reaction time; it does not manage throttle or orientation.
- Installing harrier mixes before the airframe is mechanically clean. Tune after the linkage and setup are right.
- Chasing a hover that sinks at full power with more stick input instead of diagnosing battery sag and the power system.
Frequently Asked Questions
Q: Why does my plane rock side to side in a harrier?
Wing rock occurs because the two deeply stalled wings are not behaving identically. Smaller, earlier aileron and rudder inputs usually help more than larger ones. Avoid instinctively lowering the nose; on some airframes a slightly deeper high-alpha attitude settles the rock by keeping both wings similarly stalled. Locate the model's high-alpha sweet spot, and only then consider an elevator-to-spoileron mix of roughly 20 to 30%, which can reduce rock on some aircraft but causes abnormal responses when it is overdone.
Q: My plane sinks in a hover even at full throttle. Is that a skill problem?
No. A stationary hover requires vertical thrust that at least balances the aircraft's weight, and climbing out of a failed hover requires thrust above weight. Sinking at full power points to insufficient thrust, battery sag, or a prop and power combination that does not match the airframe. Diagnose the power system before making more attempts, because no stick input will fix an aircraft that cannot hold itself up.
Q: Will a gyro or a SAFE-style mode hover the plane for me?
No. A rate gyro such as AS3X reduces disturbances and can buy reaction time in gusts, but it does not manage throttle or orientation. Angle-limited, self-leveling modes can actively oppose the near-vertical attitudes a hover needs, so they have to be switched out for unrestricted 3D. There is also no universal gain setting: too much gain produces oscillation, and model-specific gains should never be duplicated blindly in the transmitter.
Q: How long does it take to learn to hover?
There is no reliable fixed timeline. Some pilots report a breakthrough within a few flights once the correct throttle region clicks, while others spend months on the simulator before feeling comfortable at low altitude. Use competency gates instead of a schedule: harrier before hover, and a stable hover in more than one orientation before allowing a torque roll.
Q: Is a torque roll just a vertical aileron roll?
No. A torque roll starts from a stable hover and lets motor and propeller reaction torque rotate the airframe around its longitudinal axis. The aileron that was holding the aircraft straight is relaxed toward neutral so the rotation can begin, and aileron is used only to slow or stop the roll afterward. Teaching it as holding left aileron while hovering misses the mechanism entirely.
Conclusion
Harriers, hovers and torque rolls are one progression built on a single requirement: enough power reserve to keep the tail alive in propwash, and enough orientation control to correct before the aircraft departs. Learn them in that order, fly them high, and get the airframe mechanically clean before chasing low-altitude passes. A model that is fighting loose linkage, a poor throttle feel or a battery that sags under load will never hold a hover no matter how much stick time goes into it.
Foam and EPP airframes remain the consensus learning platform for a reason that goes beyond purchase price. Low mass means lower kinetic energy, large control surfaces stay effective in the propwash, the model can fly slower, and repeated imperfect recoveries are survivable. Larger full-fuselage 3D aircraft track better and handle wind once the pilot is competent, but the cost of a single orientation mistake is far higher. Crashed foam repairs are cheap; a folded 8 lb airframe is not.
If you want a starting point, a 32-inch EPP profile teaches all three maneuvers for the least money and the fewest repairs. Step up to a turnkey flat foamy once orientation stops being the bottleneck, and only then consider a full-fuselage aerobat. For related ground, see the basic aerobatic maneuvers guide, the pattern and aerobatic aircraft roundup, and the wind flying guide for keeping small, light 3D models in one piece on breezy days.
Check the current price of the E-flite Eratix 3D FF on Amazon
