Daniel's prompt this week is about a very specific kind of driving panic. He had to get someone to urgent care in the morning and then an emergency room in the afternoon, and the urgent care had no accessible parking. His point is that every navigation app assumes you want to arrive and stop. What he actually needed was to be routed to the door, drop the sick person off, and then be kept moving in some kind of holding pattern nearby while he figured out where to leave the car. He wants to know how to do that when you don't know the neighborhood, traffic is piling up behind you, and you have maybe five seconds to commit to a direction.
The phrase he used, "start holding pattern," is exactly the right mental model. Aviation has a formal version of this. A holding fix, a defined racetrack pattern, a published altitude. The pilot doesn't have to think about where to go next because the procedure already exists. Daniel is asking whether a driver can improvise one on the ground, in a city, under stress, with no chart.
And the honest answer is that most of the published defensive driving material barely touches this. There's a lot about escape routes at intersections, about leaving space in front of you, about scanning twelve seconds ahead. But the specific skill of orbiting a known point while you solve a secondary problem, that's mostly taught to executive protection drivers and police. Not to civilians.
I went looking for what actually exists. The closest formal thing is what close protection drivers call a "circulation plan." Before the principal ever arrives somewhere, the driver has already driven the route to the hospital, identified the drop-off lane, the ambulance bay, the two nearest parking structures, and at least one loop that brings them back to the door without reversing. The planning happens cold, days before. Daniel's scenario is the inverse. He's doing it hot, with a sick passenger, in a neighborhood he's never seen.
So the first real tip is to compress the planning window. You can't pre-drive the route to an urgent care you didn't know you'd be visiting. But you can decide, before you ever leave the house, what your default behavior is when the destination has no obvious parking. And that default should be: passenger out at the door, driver keeps moving, no exceptions.
That's the part most people get wrong. They treat the drop-off as a negotiation. They pull up, they see the curb is full, they start looking around, the person in the passenger seat says "just stop here for a second," and now you're blocking a lane while you have a conversation about whether you're blocking a lane. The professional move is to announce the plan while you're still moving. "I'm dropping you at the door, you go in, I'll find parking and text you." That's one sentence, said before you arrive.
And the sentence matters because it removes the passenger's instinct to wait for you. A sick person who thinks you're parking together will stand at the entrance holding their phone, not registering at the desk, because they're waiting for you to appear. If you tell them to go in, they go in.
Then the driving itself. The core geographic problem is that a holding pattern requires a loop, and most urban street grids are not loops. They're one-ways, dead ends, and turn restrictions. The app can't give you a holding pattern because it doesn't know which of those streets you can actually turn around on. It knows the legal map, but it doesn't know which street has a loading dock you can swing through, or which intersection has a wide enough mouth to do a three-point turn without holding up a bus.
So the skill is rapid loop identification. You're looking for any visual cue that a street will let you come back. A roundabout is the gold standard. A traffic circle. Even a large intersection with a left turn lane. Anything that lets you reverse direction without a three-point turn. In Jerusalem specifically, the older neighborhoods have these little stone circles at junctions that were never designed for modern traffic volume, but they work perfectly as turnaround points.
The second cue is a block with a through-street at both ends. You don't need to know the street names. You need to know that if you turn right here, you can take the next right, and the next right, and you'll come back to the same main road. A rectangular block is a holding pattern. The risk is the one-way that breaks the rectangle.
Which is why the third cue is signage. Not street name signs, those are useless when you're moving and stressed. You're reading the regulatory signs. One-way arrows, no-entry, dead end. A dead-end sign is actually useful information, because it tells you that street is out. You can eliminate it and never think about it again.
There's a triage framework here, and I think it maps to what emergency physicians do with a crashing patient. You don't try to solve the whole problem at once. You solve the thing that will kill you first. In the car, the thing that kills you first is being stationary in a travel lane. So the first decision is not "where do I park." The first decision is "which direction keeps me legal and moving." You commit to that before you even look for parking.
And commitment is the actual skill. Most drivers in this situation do what Daniel described, they slow down and start scanning both sides of the street while traffic stacks up behind them. That's the worst of all worlds. You're not parked, so you're still responsible for the car, and you're not navigating, so you're not building a loop. You're just drifting.
The advice from police pursuit driving, oddly enough, applies here. They teach officers to make a decision early and execute it fully, because a hesitant driver at speed is more dangerous than a wrong decision made cleanly. If you turn right and it's a dead end, you've lost thirty seconds. If you drift down the middle of the road for thirty seconds, you've created a hazard and you still don't know where to go.
Let's talk about the navigation app for a second, because Daniel asked specifically about a feature that doesn't exist. The reason Google Maps doesn't have a holding pattern button is that it optimizes for a single destination state. It wants you to arrive and stop. A holding pattern is a non-arrival state. It's the opposite of what the app is built to do.
But there's a partial workaround. If you have a passenger who can operate the phone, or if you can do it hands-free, you can drop a pin on the destination, then ask for directions to a point half a kilometer away, then ask for directions back. That's a manual holding pattern. Two destinations, alternating. The problem is it requires you to know the name or address of that second point, which you won't in an unfamiliar neighborhood.
The better audio cue is to leave the navigation running to the original destination and just ignore it when it tells you to turn. The map stays on screen, you can see the shape of the streets around you, and the blue line keeps pointing back to the hospital. That's your visual anchor. You don't need the app to route you in circles. You need it to keep showing you where the center of the circle is.
Which brings up a useful feature that does exist in some apps, and it's the compass heading. If you know the hospital is east of you, and you keep turning in a way that keeps you roughly east, you're in a loose orbit. The moment the compass flips and you're heading west, you know you've broken the loop and you need to correct.
That's the poor man's inertial navigation. Before GPS, that's how pilots flew holding patterns at night. They had a heading, a timer, and a known fix. One minute outbound, a standard-rate turn, one minute inbound. You can do a sloppy version of that in a car with a compass and a clock.
And the clock part is real. If you're circling a block, you want to know how long the loop takes. Twenty seconds? Forty? Because that's your decision budget. If the loop takes forty seconds, you have forty seconds to scan for parking, check the phone, make a call. Then you're back at the hospital door and you have to commit to the next loop or take a spot.
The communication side is where I think Daniel's instinct about triage is sharpest. In the moment, there are three conversations happening. The sick person in the passenger seat, who needs reassurance. The person on the phone, if there is one, who needs an update. And yourself, trying to hold the geography in your head. You cannot do all three well.
The clinical version of this is called a "sterile cockpit" in aviation, or in medicine, it's the difference between a quiet trauma bay and a loud one. The person running the code announces what they need and everyone else shuts up. In the car, the driver is running the code. So the driver gets to say "I need quiet for the next minute while I figure out where to park." That's not rude. That's task saturation management.
And the passenger's job, if they're well enough to have one, is to be the second set of eyes. Not to navigate, because they don't know the neighborhood either, but to read signs. "Dead end on the left." "No entry ahead." "Parking garage sign, two hundred meters." That's the kind of raw data the driver can use without having to look away from the road.
I want to get specific about where you actually can stop, because Daniel asked about the legal boundary. In most places, and Jerusalem is no exception, the hierarchy of acceptable stops goes like this. A legal parking space is best. A loading zone is next, if you can plausibly claim you're loading or unloading a person, which you are. An ambulance bay is for ambulances, full stop, do not. A bus stop is a moving violation and you'll get a ticket, but sometimes the ticket is the cost of doing business.
I think that last part needs to be said plainly. If the choice is between blocking a bus stop for ninety seconds while a sick person gets out, or circling for another ten minutes while they get sicker in the passenger seat, you take the ticket. That's not reckless driving. That's a defensible decision made under pressure. The fine is the price of the faster outcome.
And this is where having cash or a payment app ready matters. If you're going to risk a ticket, you should know that before you do it. The time to figure out whether you can pay a Jerusalem parking fine online is not while you're circling an unfamiliar block with a feverish passenger.
There's also the hospital-specific knowledge. Emergency departments in big cities usually have a drop-off lane that's separate from the parking entrance. It's marked, but it's often around the side or behind the main entrance. Urgent care clinics in strip malls or mixed-use buildings often don't have one at all, which is exactly what Daniel hit. The building was designed as retail or office space, and the clinic moved in without changing the parking layout.
That's a real failure mode in healthcare facility design. A clinic will lease space in a building with a parking ratio that was calculated for a law firm, where visitors are spread across the day. Then the clinic brings in forty patients an hour, each one arriving by car, and the lot was never sized for that. The result is exactly what Daniel described. No accessible parking, no drop-off lane, and a line of frustrated drivers.
So the defensive driving advice for this specific scenario starts before you leave the house. If you know you're going to an urgent care or an ER, assume the parking will be a problem. Put the passenger's insurance card and ID in their hand before you start driving. Put your phone on a mount where you can see the map without holding it. And decide the drop-off plan before you pull out of the driveway.
The insurance card thing is important. The last thing you want is to be standing at the ER registration desk, the clerk asks for the insurance card, and you realize it's in the glove box of a car that's parked six blocks away. That's a ten-minute round trip on foot at the worst possible moment.
And if you're the driver and the passenger is the patient, the walk from the car to the door matters. You want to park as close as you legally can, but you also want to park somewhere you can actually get out of. A tight underground garage with a spiral ramp might be physically close, but if it takes you five minutes to exit and pay, you've lost the time you saved.
The exit strategy is the part everyone forgets. You're not just parking. You're parking with the intention of leaving later, possibly in a hurry, possibly at night, possibly with a discharged patient who's still weak. So the ideal spot is on the street, facing out, near a corner, where you can pull away without a three-point turn. The worst spot is the back of a dead-end lot behind a dumpster.
Let me bring this back to Daniel's specific question about visual cues for circling. He said he'd have to rely on visual cues to go around in circles intentionally and make sure he finds his way back. The strongest visual cue in an unfamiliar urban area is the landmark you can see from multiple streets. A tall building, a water tower, a distinctive sign. In Jerusalem, it might be a hospital tower itself. You keep the landmark on your left, or your right, and you know you're orbiting.
That's the same technique used in orienteering and in open-water swimming. You pick a fixed point on shore and you keep it in a consistent position relative to your body. In the car, you pick the hospital and you keep it consistently on one side of the car. The moment it crosses over to the other side, you've reversed direction and you need to correct.
There's a subtlety here about one-way streets. If you're orbiting a hospital and you accidentally get pushed onto a one-way that takes you away from the landmark, you can't just turn around. You have to complete the block and come back. The instinct to slam on the brakes and reverse is wrong, and it's also illegal in most places. Take the wrong turn cleanly, finish the block, and rejoin the loop.
That's the same principle as the pursuit driving advice. A wrong decision made cleanly is recoverable. A panicked correction is how you end up facing the wrong way on a one-way street with a bus honking at you.
Let's talk about the radio, or the phone, or whatever else is making noise in the car. When you enter the holding pattern, the first thing you do is kill the audio. Music off, podcast off, navigation voice down to a whisper. You need the cognitive bandwidth for the map and the signs. Any sound that isn't the passenger or the navigation is stealing cycles.
And the navigation voice itself can be a problem. If it's still routing you to the original destination and you're deliberately ignoring it, it'll keep saying "make a U-turn" every fifteen seconds. That's not helpful. Mute it. Keep the map on screen, but mute the voice. The visual is what you need, not the nagging.
I want to address the idea of a "start holding pattern" button more directly, because I think it's actually a good product idea and someone should build it. The app would need to know three things. The drop-off point, which is the destination. The radius you're willing to work within, which is probably user-set or defaulted to half a kilometer. And the direction of the next loop, which it could calculate from real-time traffic and one-way restrictions.
The output would be a series of turn-by-turn instructions that form a closed loop. "Turn left in two hundred meters. Turn left again. Turn left again. You are approaching the destination." That's a holding pattern. The app would just keep generating loops until you tell it you've parked or you've handed off the passenger.
And the reason it doesn't exist yet is that navigation apps are built around the concept of a final state. You arrive, the navigation ends, the app shows you a parking icon. A holding pattern has no final state. It's a continuous process. The app would have to be comfortable saying "I'm not taking you anywhere, I'm keeping you near somewhere." That's a philosophical shift for a product team.
There's a military analogy here, and I think it's the right one. Close air support aircraft don't just fly to a target and leave. They establish an orbit at a safe altitude and they hold there, sometimes for hours, waiting for the ground commander to call them in. The orbit is the mission. The strike is the exception. Navigation apps are built for the strike. Daniel needed the orbit.
And the orbit is where the thinking happens. That's the part I want to underline. The holding pattern isn't wasted time. It's the time you use to solve the parking problem without the pressure of being stationary in a travel lane. You're buying yourself decision time at the cost of fuel and distance. That's a good trade.
The fuel cost is trivial, by the way. Circling a city block at thirty kilometers an hour for ten minutes burns less fuel than idling in a parking lot for the same ten minutes. And it keeps the air conditioning running, which matters if you're in Jerusalem in September with a sick passenger.
Let me give the concrete triage sequence, because Daniel asked for it and I think it's the most useful thing we can hand him. Step one, announce the plan to the passenger while you're still moving. Step two, drop them at the closest legal point to the door, even if it's not a parking space. Step three, get moving immediately, any direction that's legal. Step four, kill the audio and mute the nav voice. Step five, identify your landmark and start orbiting. Step six, scan for parking on each loop. Step seven, when you find a spot, commit to it and text the passenger where you are.
I'd add a step zero, which is before you leave the house. Know your destination's address, know the cross streets, and know whether it's a hospital with a drop-off lane or a clinic in a mixed-use building. That one piece of knowledge changes the whole calculus. If it's a clinic in a strip mall, you plan for the holding pattern before you start driving.
And step eight, which is after you park. Take a photo of the street sign or the parking spot number. Send it to the passenger. Then walk to the door. The photo is your memory. You will not remember where you parked after three hours in an ER waiting room.
That's the part of the scenario that gets overlooked. The arrival is stressful, but the departure is hours later, in a different mental state, possibly in the dark. The photo is the bridge between those two states.
There's a question Daniel raised about whether this is covered in advanced defensive driving courses. The answer is mostly no, for civilians. Executive protection courses cover it, police pursuit courses cover the decision-making under pressure, and some ambulance driver training covers the drop-off-at-the-door maneuver. But the specific combination, circling an unfamiliar urban area while a passenger needs medical care, that's a gap in the civilian curriculum.
Which is a shame, because it's a scenario that happens to ordinary people all the time. A spouse with a kidney stone, a kid with a fever, a parent who fell. You're not a professional driver, you're just the person with the car keys, and suddenly you're doing task-saturated navigation under time pressure. The skills are teachable, and they're not being taught.
The closest civilian equivalent is what's taught in some Scandinavian countries as part of the licensing process. They have a module on "driving under stress" that includes exactly this kind of scenario. You're given a destination, you're told there's a passenger in distress, and you have to navigate while managing the pressure. It's not universal, but it exists.
The Israelis, honestly, have a cultural version of it. The driving here is aggressive, the parking is scarce, and everyone has done some version of the urgent-care drop-off. The skill is learned socially, not formally. Daniel is asking for the formal version of what most Jerusalem drivers learned by doing it badly once.
That's the thing about this prompt. It's not a hypothetical. It's a post-incident review. Daniel did it, it worked out, and now he's reverse-engineering what he actually did and asking what the professional version would look like. That's a good instinct.
The professional version, to summarize, is a circulation plan executed hot. You pre-commit to the drop-off, you use a landmark as your fix, you orbit until you find a spot, and you document the spot when you take it. The holding pattern is the tool that buys you the time to do all of that without becoming a hazard.
The communication discipline is the part that makes it work. One sentence to the passenger. Quiet in the car. A text when you park. That's the whole protocol.
Hilbert: The problem with the holding pattern is the same problem we had with the courier vans in oh eight. The dispatcher would send you to a building on a one-way street and the building had no loading bay, so you'd circle the block and every time you came back around there was a different car in the fire lane. The trick was to find the alley behind the building. Every building in that part of Boston had an alley, and the alley had a dumpster, and next to the dumpster there was always a spot that wasn't legal but the building super wouldn't tow you if you left the hazards on.
The holding pattern is a stopgap, but the real solution is local knowledge of the informal parking layer.
Hilbert: The informal layer is the whole game. The apps know the legal spots. They don't know which loading dock is empty after eleven, which church lot doesn't tow on weekdays, which alley the super doesn't care about. You only learn that by doing the route once when you're not in a hurry.
Which is the argument for a dry run. If you know someone with a chronic condition, and you know which urgent care they're likely to end up at, drive the route once on a Sunday morning when there's no pressure. Find the drop-off point, find the two nearest parking options, find the loop. Then when it happens for real, you're not improvising.
Hilbert: My brother-in-law did that for his wife's dialysis clinic. Drove it three times before her first appointment. Knew exactly where the wheelchair ramp was, where the elevator was, where the parking garage entrance was. When the day came, he was in and out in four minutes. The other husbands were circling the block looking for a spot. He just knew.
That's the cold planning version of what we've been describing. The hot version works, but the cold version is better.
The cold version is something you can do for any high-stakes destination. The school, the hospital, the airport. One dry run, no pressure, and you've built the mental model that will hold up when the pressure arrives.
Hilbert: The other thing about the dry run is you learn which streets have the speed bumps. You don't want to be circling a block with a passenger in pain and hit a speed bump on every loop. That's a detail the map won't tell you.
Speed bumps as a factor in route selection. That's the kind of thing that only comes from local knowledge.
Hilbert: It's the kind of thing that matters when you're doing it for real. The map says the loop is four hundred meters. The map doesn't say it has six speed bumps and a blind corner where the school bus stops at eight in the morning.
That's the limit of the "start holding pattern" feature. It can give you a loop, but it can't tell you whether the loop is a good loop. The good loop is smooth, has a turnaround point, and keeps the landmark visible. The bad loop is a legal rectangle with speed bumps and a left turn across two lanes of traffic.
The feature needs a human layer. Either you've driven it before, or you're reading the cues in real time. The app can suggest, but the driver has to confirm.
Hilbert: The other thing we learned in the courier days is that the first loop is always the worst. You're learning the streets, you're checking the signs, you're trying to keep the building in sight. The second loop is easier. By the third loop, you've got it. The mistake people make is they bail after the first loop because it felt chaotic. But the chaos is just unfamiliarity, and it goes away fast.
That's a useful insight. The first orbit is the learning orbit. The second orbit is the working orbit. Don't judge the plan by the first lap.
That maps to the neuroscience of spatial learning. The first pass through an unfamiliar environment, your hippocampus is building the map. The second pass, you're reading the map you just built. The third pass, you're predicting. The holding pattern is a spatial learning exercise compressed into ninety seconds.
Hilbert: My brother-in-law said the same thing about the dialysis clinic. First time he drove it, he was lost. Second time, he was fine. Third time, he could do it in his sleep. The dry run is just doing the learning before the emergency.
The advice to Daniel, in the end, is a three-layer thing. The cold layer, which is the dry run and the pre-committed plan. The hot layer, which is the landmark orbit and the triage sequence. And the informal layer, which is the alley behind the building and the loading dock that doesn't tow.
The cutting-room floor detail I wanted to mention is that some hospital systems in the US have started putting live parking availability on their websites, so you can check before you leave the house whether the garage is full. It's not universal, but it's spreading. That's the kind of data that would make the holding pattern unnecessary, if it were reliable and real-time.
The forward-looking question is whether the navigation apps will ever build the holding pattern as a first-class feature. I think they will, eventually, because the use case is real and the demand is there. But it'll probably come from a startup first, not from Google or Apple. The big players are too invested in the arrival paradigm.
Until then, the skill is yours to learn. The landmark, the loop, the quiet car, the photo of the parking spot. It's not a feature. It's a discipline.
Thanks to Hilbert Flumingtop for producing, and for the courier van wisdom.
This has been My Weird Prompts. Email us at show at my weird prompts dot com with your own weird navigational emergencies.
We'll be back soon.
