Defensive riding in mixed motor traffic: lane positioning, primary vs secondary position, door zone, right hook + left cross at the intersection, SMIDSY / look-but-failed-to-see — how to avoid conflicts with cars

Scootify 13 min read

Braking technique, cornering, night riding, riding in the rain, and emergency maneuvers are technical rider competence. They answer the question “what do I do with the scooter itself?” But that leaves a separate safety layer — the strategy of interacting with motor traffic: where to position yourself in the lane before you need to brake or swerve; how to read drivers approaching the intersection; where “as far right as possible” stops being safe and starts being the door zone; and why, statistically, the intersection rather than the open road carries the bulk of serious car-on-rider incidents. This guide is about positioning strategy and active conflict avoidance before the moment an evasive maneuver becomes necessary.

Why this is a separate layer. The NACTO Urban Bikeway Design Guide 3rd ed. 2025 reports that just over 40 % of urban bike fatalities in 2022 happened at intersections, not on straight road segments (NACTO — Don’t Give Up at the Intersection). The picture for e-scooters is no better: the UK DfT National Evaluation of E-Scooter Trials (August 2023) puts e-scooter casualty rate at over three times that of pedal cycles in the same trial areas (UK DfT — National Evaluation of e-scooter trials findings report, PDF). Meanwhile, the Helsinki TBI cohort 2022-2023 shows that 52 % of e-scooter injuries are solo falls (news-medical.net) — so intersections plus solo cornering falls together account for the majority of serious cases. Traffic-flow strategy is the second-most-important safety layer after helmet + traffic rules (Safety, gear, and traffic rules).

The principles below carry over to the e-scooter the classical vehicular cycling rules systematised by John Forester in Effective Cycling (MIT Press 1976, 7th edition 2012) and developed into modern pedagogy by League of American Bicyclists Smart Cycling (bikeleague.org — Smart Cycling Education) and CyclingSavvy (American Bicycling Education Association). The engineering layer is in Lighting & visibility engineering, where light output, retroreflectivity, EN 1150 hi-viz, and UN R148 are treated separately.

1. Why “as far right as possible” is the worst strategy

A beginner’s intuition: “I am slower than the cars, so I will hug the curb to stay out of their way.” That is inverse-dangerous for three reasons.

First, edge effect. The closer the rider is to the curb, the less lateral buffer there is for maneuver. The cubic curb or pavement edge is a barrier you cannot ride into; only one side has any room to dodge. In the middle of the lane both sides have a buffer; you can move left or right.

Second, the driver “shy zone.” Coleridge-Bristol Cycling 2014 and later Cycling UK guidance show: when the rider hugs the curb, the driver perceives this as a “gap” rather than as a vehicle — and overtakes with minimum lateral clearance, because the rider has effectively “given permission” to be passed. When the rider takes the middle of the lane, the driver subconsciously treats them as a legitimate vehicle, leaving a larger lateral offset or waiting for a safe gap (Cycling UK — Road positioning). UK Highway Code Rule 213 (2022 revision) explicitly states that a rider may take the centre of the lane where it is safer to do so — and drivers must leave ≥1.5 m when passing at speeds up to 30 mph, ≥2 m above (UK Highway Code Rule 213).

Third, longitudinal hazards. Curbs, pavement-edge joints, gravel accumulations, open stormwater grates, the door zone of parked cars — all of these are concentrated along the edge of the lane. The middle of the lane has guaranteed smooth pavement (this is the automobile’s tracking path).

Conclusion: “as far right as possible” is the posture of being overtaken with minimum clearance in the zone with maximum longitudinal hazards and with minimum room to maneuver. It is the worst strategy, and it is the one beginners most often choose.

2. Primary vs secondary position — the vehicular-cycling vocabulary

Standard terminology comes from UK Cycling UK Bikeability + CyclingSavvy + League of American Bicyclists Smart Cycling:

Primary position (control the lane / take the lane). Rider in the middle of the lane. Used when:

  • The lane is not wide enough for a car to overtake with a safe lateral clearance of ≥1.5 m without leaving the lane (this describes most urban lanes, where lane width is under 4 m in many EU cities).
  • Approaching any intersection (sections 4-5-6 below).
  • At chokepoints, debris, or road works.
  • On descents where rider speed is high (the closing-speed differential with traffic vanishes).
  • Among groups of parked cars — the door zone wipes out the secondary position.

Secondary position. Approximately 1 m from the edge of the carriageway (not from the curb — from the line of parked cars or the road edge). Used when:

  • The lane is wide enough for a safe overtake without the car leaving the lane.
  • On straight segments with no conflict zones ahead.
  • The permitted traffic speed is low (≤30 km/h in traffic-calmed zones).

The primary ↔ secondary transition. Made well in advance (30-50 m before the conflict point), with a shoulder check and a hand signal, never abruptly. This is a fundamental vehicular-cycling rule: behave like another vehicle — declare intent, act predictably.

“Declare intent and be predictable” is the single principle that resolves most conflicts. The SMIDSY phenomenon discussed in §7 below is not primarily about visibility — it is about predictability.

3. The door zone — the most underrated longitudinal hazard

The door zone is a 1.2-1.5 m wide strip from the line of parked cars, in which open driver/passenger doors intersect the rider’s path. Wikipedia systematises the data: dooring accounts for 12-27 % of urban bike collisions across various municipal datasets (Wikipedia — Dooring). Boston PD 2009-2012: 7-13 % of collisions; Chicago 2011: 344 reported doorings, 1 in 5 bike crashes; London 2010-2012: 3 fatal; Great Britain 2011-2015: 3,108 injured, 8 fatalities with “vehicle door opening as a contributing factor” (UK Department for Transport — reported road casualties summarised by Dutch Reach Project).

The number that explains the mechanism: a door swings from “closed” to “fully open” in 0.2-0.3 seconds. The rider’s typical perception-reaction time is 1.0-1.5 seconds (AASHTO PRT for cyclists). At 20 km/h (5.5 m/s), the rider therefore travels 5.5-8.3 m before they even register the situation. If the rider is already in the door zone, the door leaves no room to maneuver — it is a guaranteed collision or a panicked swerve into traffic.

Countermeasures:

  1. Never ride in the door zone. Full stop. Not “where possible” — never. If the lane is so narrow that the only riding line falls inside the door zone, the rider takes primary position in the adjacent lane (not the door-zone strip).
  2. Dutch Reach is the behavioural norm for drivers (open the door with the far hand, which forces a body twist and a glance back over the shoulder), codified in the Massachusetts driver’s manual 2017+, Illinois 2019+, and the Dutch driver’s test since 1962 (Dutch Reach Project — Dooring Statistics). The rider cannot influence this, but knowing the norm helps — in jurisdictions with Dutch Reach the door-zone risk is lower.
  3. “Parked-car scan”: the rider scans parked cars 2-3 vehicles ahead for an illuminated cabin, a driver’s head, brake-lights / turn-signals that have “just parked,” or pre-departure dialogue visible through the glass. Any of these signals = the secondary line is not safe.
  4. Tail vehicle: if a car is behind you and you want to move out of the door zone — this is your chance to accelerate to the speed of traffic and take primary, not to slow them down with gas-and-brake.

4. Right hook — the most common intersection conflict

Right hook — a vehicle moving in the same direction as the rider turns right across the rider’s path, having failed to notice the rider going straight. NACTO 2022 data: >40 % of urban bike fatalities are at intersections, with right-hook + left-cross the largest subcategories (NACTO — Don’t Give Up at the Intersection).

Why it happens. A driver preparing to turn right looks back over the shoulder for pedestrians in the crosswalk (their primary check) and into the mirror for cars in the adjacent lane. A rider in the bike lane to the right of the driver’s lane is in the A-pillar + side-mirror blind spot; a rider in secondary position at the lane edge can also be ignored, because the driver scans car lanes, not the curb.

Countermeasures:

  1. Before an intersection with a turn conflict — take primary position. 30-50 m out, with a signal and a shoulder check. In primary position a driver cannot overtake and turn in front of you; they must merge in behind.
  2. If the bike lane runs to the intersection and you are in it — unequal situational risk. NACTO’s “Don’t Give Up at the Intersection” recommends an infrastructural countermeasure: a “protected intersection” with an island buffer that keeps the bike lane separated all the way to the crossing. If the infrastructure is absent, the rider has to choose: stay in the bike lane and accept the right-hook risk, or move into primary and ride with traffic.
  3. Eye contact with the driver who is about to turn (turn signal, gentle deceleration, head turned). If eye contact is not made, the driver does not see you, no matter how conspicuous you are. Decelerate; be ready to fall in behind.
  4. Do not overtake on the right alongside a car that is about to turn. This is the canonical right-hook scenario: a rider in the bike lane passes a stopped or slowing car on the right, the car moves and turns, the rider goes under it. Stop.

5. Left cross — the second-deadliest

Left cross — a driver from the opposite direction turns left across your path while you are going straight. This is the classic motorcycle SMIDSY scenario that IIHS and the Hurt Report (Hurt Report 1981 — Motorcycle Accident Cause Factors, NHTSA HS-805 862) anchored in the literature with the 1981 baseline (75 % of motorcycle crashes involve a passenger car; in two thirds of those the car driver violated the motorcyclist’s ROW). For e-scooters the picture is the same: the driver sees you (sometimes) as a “small, slow, irrelevant” unit and underestimates closing speed.

Why it is particularly bad for the e-scooter:

  • Small silhouette area. At 30-40 m the rider on a scooter looks like “small something” — below the threshold for the driver’s classification attention.
  • Misjudged closing speed. A driver is used to automobile closing-speeds (20-30 km/h to the car ahead) — a scooter rider at 25 km/h looks slow because the visible rate of size change (looming rate) is small, even though the objective closing speed can be 50+ km/h.
  • Headlight height. The scooter’s front lamp sits at handlebar height (≈1 m) rather than at car-headlight height (≈0.6 m) — the driver scans the band of car-headlight heights and your lamp can fall outside the active-scan zone.

Countermeasures:

  1. Always assume an oncoming driver with a turn signal on does not see you. Ride the intersection on the assumption that “this car is about to turn through my path.”
  2. If the option exists, let the oncoming car go before you, even where you formally hold the ROW. The law does not raise the dead.
  3. Raise your closing-rate signal: front light in flashing mode during daytime — prescribed for motorcycles in the MAIDS Final Report 2009 (a daytime running light reduces left-cross risk by 13 %); the same effect is documented for bicycle commuters by Madsen 2013 (Trafitec Denmark — Bicycle daytime running lights).
  4. Lane positioning before the intersection — primary, centre, not edge. The closer to the edge, the more an oncoming driver thinks “there is nothing there.”

6. The full intersection toolkit

Beyond right hook + left cross, intersections have a handful of additional “typology” patterns worth knowing:

T-bone (perpendicular sideswipe). A driver from a perpendicular direction runs a stop sign / red light and hits the side. A rider in secondary position is 1 m further from the trajectory than a rider in primary — a positive of secondary on some crossings. But in most T-bone cases the rider is non-faulted; primary vs secondary is largely irrelevant.

“Stale green” phase. An intersection with a green light that has been green for a long time — amber is imminent. Entering on a stale green = the risk that an opposite turning vehicle starts moving on your amber. Better to stop on a fresh red and wait for a fresh green.

ASL / bike box. Advance Stop Line (UK), bike box (US) — the marked area in front of the automobile stop line, between the intersection and the queue of cars. If you enter the ASL first (on red), the rider is visible to all drivers, out of the blind spot, with start priority. Always enter the ASL where possible.

Two-stage left turn (US “Copenhagen left,” UK “turn box”). Instead of moving to the left-most lane for a left turn (which is typically stressful and dangerous for a scooter rider because of cross-traffic), the rider continues straight to the right-far corner of the intersection, stops in the marked turn box, waits for the next phase, and completes the left turn as a straight movement. This is safer and often legally required in NL/DK/DE; in the US MUTCD 2009 introduced the two-stage turn box (MUTCD 2009 — § 9C).

Roundabout. On a small roundabout (<25 m diameter), the rider must take primary position in the lane, no exceptions. Small roundabouts produce motorist swept-volume in which a rider at the edge sits in a dead zone throughout. Large multi-lane roundabouts — use bicycle infrastructure (a segregated bypass) or take primary of the innermost lane and exit accordingly.

7. SMIDSY / look-but-failed-to-see is not about “visibility”

“Sorry mate, I didn’t see you” is the British label for the phenomenon known in the US as LBFTS — Looked But Failed To See. Hurt Report 1981 NHTSA HS-805 862 and the MAIDS Final Report 2009 (European motorcycle accident study) qualify it as the causal factor in roughly 30 % of intersection collisions involving motorcycles or bicycles.

Key insight: SMIDSY is not about hi-viz clothing or brighter lamps. The Science Of Being Seen project (former Royal Society for the Prevention of Accidents consultant, UK) argues that the driver looks toward the rider but the brain does not register their presence, through three perceptual mechanisms (Science Of Being Seen — Looked but failed to see):

  1. Saccadic masking. Between saccadic eye movements (300-700 ms) the human visual system does not register information — the brain “stitches” the scene from fixation points. If the rider happens to fall within a saccade transit, the driver literally did not see them.
  2. Motion camouflage. If the rider is moving directly toward the intersection, the driver’s visual system does not see a parallax change — the rider looks like a stationary point in the field of view. This is particularly bad for the left cross: a rider on a closing course has nearly zero angular velocity in the driver’s field of view.
  3. Inattentional blindness. A driver scans categories of objects expected at the intersection — other cars, pedestrians at the crossing, the traffic light. A scooter as a category is not on the expectation list at a standard intersection, especially in cities with low e-scooter density.

Conclusion: conspicuity by itself does not solve SMIDSY. Hi-viz clothing, daytime running lights, and reflectors reduce the risk by 5-15 % (Trafitec / Madsen 2013), not by 50 %. What does work is predictable movement + eye contact + assuming that the driver does not see you.

A SMIDSY-resistant riding strategy:

  • Do not listen to music in earbuds / use a phone in motion — keep a reaction reserve for the driver who starts to move.
  • Vary the lateral position. Sage van Wing in CyclingSavvy: a rider who actively varies lateral position in the lane (slightly left, slightly right) does not look like a stationary point in the driver’s field — this breaks motion camouflage.
  • Daytime running light in flashing mode at the front — in water-tight cases. A daytime flashing front light adds an angular-velocity signal that the driver’s brain cannot filter out.
  • Anticipate turn points.

8. Hand signals, eye contact, shoulder check

Hand signals. UK Highway Code Rule 67 + US Uniform Vehicle Code: left turn — left arm horizontal; right turn — right arm horizontal (or left arm vertical up as an automotive turn-signal proxy); stop — left arm down. On a scooter this is awkward, because both hands are on the bars (unlike a bicycle); realistically — shoulder check first, then release a hand for 2-3 seconds to signal, then back on the bars for the maneuver.

Eye contact. Before entering a conflict zone (a right-turner behind you, an intersection conflict, a driver emerging from a parking lot) — look the driver in the eyes. Not “toward the car” but into the eyes. If the driver has seen you, they will give a micro-nod or a change of facial expression (the non-verbal confirmation); if not, they are looking through you at something else.

Shoulder check (head turn). Not to be confused with a mirror. The mirror shows what crosses its visible cone at the moment of glance — but the blind spot rear-left (UK/AU) or rear-right (US) is exactly where the mirror does not show an overtaker. Before each lateral maneuver, turn your head fully (90°) for one second and verify the zone is clear.

Mirror. At higher speeds (>25 km/h) the benefit of a mirror outweighs its limitations. A bar-end mirror, helmet mirror, or frame-mounted mirror with a field of view ≥30° gives continuous awareness of the situation behind you, which reduces the cognitive load of frequent shoulder checks. It does not replace the shoulder check, but complements it.

9. A bike lane is not always safer than the road

A bike lane painted on asphalt without a buffer to parking is a door-zone trap. A bike lane between parked cars and the car lane without a physical separator — class III in NACTO classification — is statistically no safer than a primary position in the lane. If the bike lane is ≥1.5 m wide outside the door zone, fine; if not, the rider has the legal right not to use it (most jurisdictions with a bike-lane mandatory clause exempt unsafe conditions).

The NACTO 3rd ed. 2025 distinguishes:

  • Class I — protected bike lane with a physical barrier (curb, parking, planters): the highest safety.
  • Class II — buffered bike lane with a painted buffer (no physical barrier): medium.
  • Class III — conventional bike lane painted on asphalt with no buffer: low, often worse than primary in the lane.
  • Sharrow (shared lane marking): an indicator that “riders are legally in the lane,” not infrastructure — it functions as public education.

For an e-scooter, Class I is the optimum. Class II is fine in a traffic-calmed context. Class III and sharrows demand a critical assessment of the door zone and a switch to primary if the risk is high.

10. Practice drill — 30 min/week in neutral conditions

Vehicular cycling is a skill, not an idea. Knowing the theory is not enough; the body and peripheral awareness must execute positioning subconsciously in the moment of conflict.

Drill (repeat weekly):

  1. 15 min positioning: pick a segment with several intersections and trial-and-error change position 30-50 m before each. Hold primary consciously; watch how drivers react (larger lateral clearance? overtake and turn in front of you? wait?).
  2. 5 min shoulder check + signals: at every one of 10 intersections, deliberately shoulder check and signal, even if there is no one behind you. This automates muscle memory.
  3. 5 min door-zone scan: on a section with parking — deliberately assess each of 5-10 parked cars as “active / inactive” (illuminated cabin, recent brake-lights, heads visible inside).
  4. 5 min eye contact: at every 3 intersections where a car has a turn signal — achieve eye contact before entering the conflict zone. Do not enter if contact has not been established.

After 4-6 weeks of regular drill these behaviours become the default rather than an option in the inevitable reactive moment.

11. Recap

  1. “As far right as possible” is the worst strategy. Edge effect, driver shy zone, longitudinal hazards along the edge.
  2. Primary position in the lane is the norm at intersections and on narrow lanes; secondary is fine only on wide lanes with no conflict zones ahead.
  3. The door zone is the critical longitudinal hazard. 12-27 % of urban bike collisions. Never ride in the door zone.
  4. Right hook + left cross at intersections account for >40 % of urban bike fatalities (NACTO 2022). Primary position + eye contact + readiness to stop.
  5. SMIDSY is not about visibility. It is about predictability and about assuming the driver does not see you. Saccadic masking, motion camouflage, and inattentional blindness are perceptual phenomena that hi-viz alone does not solve.
  6. Class III bike lanes (painted, no buffer) are often less safe than primary in the lane. NACTO 3rd ed. 2025 classifies infrastructure across four levels.
  7. Hand signals + shoulder check + eye contact are three rider-driver communication tools — do not skip any of them.
  8. A 30-min/week practice drill in neutral conditions — positioning skill must be subconscious, not intellectual.

Companion context: Braking technique, Cornering and lean technique, Night riding visibility, Emergency maneuvers and obstacle avoidance, Safety, gear, and traffic rules, Lighting & visibility engineering, Human factors and ergonomics engineering.