Decide repair, adjust, or park

Generated from content/lms/race-car-mechanic-reference/07-manage-the-trackside-window/03-decide-repair-adjust-or-park.md; edit the source file, not this page.

Source path: content/lms/race-car-mechanic-reference/07-manage-the-trackside-window/03-decide-repair-adjust-or-park.md

Course: Service the race car that has to finish

Module: Manage the trackside work window

Estimated duration: 55 minutes

The trackside window is not a laboratory day. It feels like one because the car is hot, the driver is emotional, and everyone can name three things they would like to change before the next session. But the real job is narrower: decide whether the next action should restore the car, tune the car, or stop the car.

That is the whole skill. Repair means return the car to a known intended condition. Adjust means make a deliberate, reversible change to a known good car because the balance or response needs to move. Park means do not send the car back out until the question that matters has been answered. Those are not three moods. They are three different standards of evidence.

The governing principle is baseline discipline. Paul Van Valkenburgh's testing chapter gives the clearest mechanic's rule: a change cannot be judged unless there is a fixed basis of reference, and a negative change may matter even more because you need to get back to the original condition. The same passage warns that scattered lap times make one fast lap meaningless. That matters at the trackside scale because the between-session decision is usually made from incomplete evidence. If the evidence is weak, the answer is not to become more confident. The answer is to choose the door that protects the baseline.

You should think of every complaint as arriving in one of three states. The first state is a known physical fault. Something about the car has stopped being the car you intended to run. That belongs in the repair path, not the tuning path. The second state is a repeatable handling or response complaint on a car that still appears to be in its known condition. That belongs in the adjust path only if the driver report is clear enough and the team can return to the prior setting if the change is wrong. The third state is uncertainty that touches control, confidence, or mechanical failure. That belongs in the park path until the uncertainty is reduced. Parking is not a punishment. It is the choice to stop collecting risk when you cannot collect useful information.

This lesson is deliberately not another lesson on diagnosing under the clock or running the full between-session turnaround. The sibling lessons cover the broader workflow. Here you are learning the decision gate. You are learning how to say, with evidence, this is a repair, this is an adjustment, or this car should not run yet.

The first trap is assuming the car is the problem just because the driver feels a problem. Lopez makes the uncomfortable point that the car could have a problem, but it might also be the driver, especially early in a racing career when the driver can suddenly find one or two percent in lap time. Bryan Herta's prompt from the same instructional tradition is the question you should keep alive in every debrief: is there something different to do with the car, or something different in the approach to the corner? That is not a polite philosophical question. It is a triage filter. If the driver's braking point, release, turn-in, throttle application, or line changed, a handling complaint may be real in the driver's hands but not caused by a changed car.

The second trap is treating adjustment knowledge as optional. Bentley is direct that understanding chassis and suspension adjustments, and what they mean to the driver, is a critical part of the driver's job. That statement also tells the mechanic what to demand from the driver. A useful report is not just faster, worse, pushy, loose, or weird. A useful report links the driver's input to the car's response. Jenkinson praises the driver who can make clear, concise observations that help the designer, and gives the example of a driver wanting the rear end, not the front end, to respond to throttle movement. For this lesson, translate that into a trackside standard: the driver does not have to be an engineer, but the driver does have to tell you what the car did in response to a control input, in which phase of the corner, and whether it repeated.

The third trap is believing a setup change can rescue bad evidence. It cannot. Van Valkenburgh's baseline rule is harsh on paddock guessing. If you cannot compare the proposed change to a known condition, and if you cannot go back, you are not testing. You are altering the car and hoping the next session explains what the last session did not. Sometimes the right choice is to make no setup change, tighten the observation task, and send the car only if the physical car is known safe. Sometimes the right choice is to park because the next lap could produce more consequence than information.

Use this five-question gate before you touch the car.

Question one: what is the symptom, stated as a car response rather than an opinion? A response sounds like the rear does not follow the throttle, the front takes longer to take a set, the car rotates too much after brake release, or the driver cannot repeat the same entry without extra correction. An opinion sounds like it feels bad. Opinions matter because they tell you the driver is struggling, but they are not enough to choose an adjustment. You need the response.

Question two: when did it happen? The trackside window compresses time, so you need the corner phase. Entry, brake release, mid-corner, throttle pickup, exit curb, straight-line brake zone, and shift event are not interchangeable. A repair may be required if the symptom appears outside the expected phase or follows an obvious mechanical event. An adjustment may be warranted if the symptom appears in the same phase repeatedly and matches the type of balance change the adjustment is meant to influence. Parking becomes more likely when the phase is unclear and the consequence could be loss of control.

Question three: did the driver repeat it, or was it one lap? This is where scattered lap times and single fast laps matter. A one-lap complaint can be a real warning, but it is a poor tuning target. One superfast lap among scattered times is not proof that a change worked, and one unsettled lap among inconsistent driving is not proof that the car needs setup. If the driver is inconsistent, the repair path still handles known defects, but the adjust path should slow down. You may assign a driving task for the next session instead of changing the car.

Question four: is the car still at baseline? Baseline means more than the setup sheet. It means the car still represents the condition you think you are evaluating. If the car has suffered a mechanical failure or might have done so, the adjustment conversation is premature. Lopez's car-control chapter includes the blunt reminder that spins and wall contact can result from mechanical failure. That is enough to justify a conservative rule: if a suspected mechanical fault can turn the next loss of control into an incident, you park until you can either repair and verify the car or honestly say the suspicion was resolved.

Question five: can you reverse the action? If the proposed action is an adjustment, you must be able to return to the previous condition. If you cannot go back, you have converted a short trackside decision into an uncontrolled test. A reversible change is not automatically a good change, but irreversibility raises the evidence standard. In a short window, the best adjustment is usually the one that is tied to a clear symptom, small enough to compare, and recorded well enough that the next session can teach you something.

Now apply the gate to the three paths.

The repair path starts when the car is no longer the known car. Do not tune around that. A repair is not a performance experiment. Its success criterion is restoration. The car should leave the window closer to the prior known condition, not merely different. If a repair changes the baseline because you had to replace, resecure, reset, or otherwise restore something, record that fact before the next session. The next driver report must be interpreted against the repaired condition, not against the morning's setup sheet as if nothing changed.

The mechanic's discipline in the repair path is to stop adding theories. Fix the known problem first. If the driver's complaint was vague and you find a real physical fault, the first answer is repair and verify. Do not combine that repair with a handling adjustment unless the adjustment has its own separate evidence. Otherwise the next session becomes unreadable. If the car improves, you will not know whether the repair solved the complaint or the adjustment masked it. If the car gets worse, you will not know which action to reverse first. Van Valkenburgh's baseline rule is not just for formal test days. It is what keeps the trackside window from becoming noise.

The adjust path starts only after the car is still the known car and the complaint is repeatable enough to tune against. Bentley's chassis-adjustment warning belongs here. If the driver and mechanic do not understand what an adjustment means to the driver, the adjustment is not ready. You do not need to teach the whole book in the paddock, but the driver must know what the change is supposed to feel like. If you make a balance change for entry, the driver should not judge it on exit only. If you make a change meant to affect throttle response, the driver should pay attention to what the car does as throttle is applied, not merely whether the lap felt calmer.

An adjustment also needs a rollback. Before the wrench turns, write down the current condition, the proposed change, the expected response, and the decision rule after the next session. The decision rule can be simple: keep it if the same driver can repeat the target corner with less correction and similar or better lap-time consistency; reverse it if the symptom moves, worsens, or the driver cannot tell what changed; stop and inspect if a new mechanical suspicion appears. This is not paperwork for its own sake. It is how you prevent an adjustment from becoming a permanent drift away from the only condition you understood.

The park path starts when the next session cannot answer the question safely or usefully. Parking is easy to say after a failure and hard to say when the symptom is ambiguous. The intermediate mechanic has to learn that ambiguity is not neutral. If the car might have a mechanical problem that can produce loss of control, ambiguity leans toward parking. If the driver cannot produce a clear report and the lap times are scattered, ambiguity leans away from adjustment. If you cannot verify the repair before the next session, ambiguity leans toward parking. The goal is not to maximize track time at any cost. The goal is to preserve the car, the driver, and the evidence so the next action is meaningful.

There is a special parking case for the car that technically runs but has lost your trust. You may not have a named broken component. You may only have a pattern: the driver reports a sudden response that does not match prior behavior, the issue appears in a high-consequence phase, and you cannot reproduce or inspect enough in the window. The temptation is to send the car for one gentle lap to see. Sometimes that is appropriate in a controlled test context, but this lesson's corpus does not support treating guess-and-see as a default. The supported principle is that testing requires a baseline and consistency. If the outing cannot produce trustworthy information and could produce damage, park.

Calibrate your decisions by what the next session teaches. A good repair decision makes the complaint disappear or converts it into a clearer, lower-risk setup question. A good adjustment decision produces a driver report that names the intended response, not just a mood. A good park decision prevents the team from adding consequences while evidence is weak. The signature of improvement is not that you always choose the aggressive option. It is that fewer sessions become unreadable.

Listen for better driver language. Early reports often sound like the whole car is bad. Better reports sound like a corner phase and a control input. The driver says the rear does not respond to throttle the way it did before, or the car will not take the same entry with the same brake release, or the front only gives up when the driver asks for rotation too early. That is the language Jenkinson is pointing toward: observations that help the people responsible for the car. When the report improves, your decision can become sharper.

Watch for tighter evidence, not heroic laps. The useful next session is not necessarily the fastest session. Van Valkenburgh's warning about a single fast lap among scattered times is the anchor. If a change produces one flyer and nine confused laps, it has not proved much. If a repair restores repeatability, or an adjustment lets the driver repeat the target phase with fewer corrections, you have better information even before the stopwatch flatters you.

Finally, remember the driver is part of the system. Lopez is blunt that amateur athletes are often far from perfect and should get used to looking inward for speed. That does not let the mechanic dismiss the driver. It makes the decision more precise. Sometimes the next action is not repair, not adjust, and not park. It is run the same car with a clearer driving task because the evidence says the variable is the driver. In this lesson's three-door model, that is closest to staying on baseline, not tuning. It preserves the car's known condition while asking the driver to answer Herta's question about approach.

Cross-reference this lesson with diagnose under the trackside clock when you need the fault-finding workflow, run the between-session turnaround when you need the physical order of operations, and end the event with better evidence when you need to turn the day's notes into a development plan. This lesson sits at the decision point between them. It teaches you not to confuse motion with progress.

Worked example: the driver reports a throttle-response balance problem

The driver comes in saying the rear of the car is not helping on throttle. That report is useful only if you turn it into a response statement. Ask where it happens, whether it happens every lap, and what the driver is doing with the throttle when the response appears. Jenkinson's example gives the shape of a useful observation: the driver may not know the engineering solution, but the driver can know whether the rear end or the front end responds to throttle movement.

If the car is physically sound and the report repeats in the same phase, this can be an adjust decision. The adjustment is justified not because the driver used a familiar complaint word, but because the report links a control input to a car response. Before changing anything, write the present condition and the expected response after the change. The next session's task is narrow: the driver must judge the same throttle phase, not the whole lap as a mood.

If the report is new, sudden, or paired with suspicion that the car is no longer in its known condition, the same words may become repair or park. Do not tune around a car you no longer trust. Inspect what can be inspected in the window. If you find a concrete physical fault, repair and verify rather than adding setup. If you cannot inspect enough and the symptom could create loss of control on throttle, park. The difference is not the vocabulary of the complaint. The difference is the evidence behind it.

Worked example: the driver says the car is slow, but the laps are scattered

A driver returns frustrated because the car feels slow and the last session produced no clean improvement. The emotional pressure points toward adjustment. Van Valkenburgh's testing rule points the other way. One standout lap among scattered times does not prove a setup direction, and scattered laps do not give you a stable target for tuning.

Start by separating the driver question from the car question. Lopez warns that the car could have a problem, but the driver may also be the source of the missing speed. Herta's question is the right debrief tool: what might need to change on the car, and what might need to change in the approach to the corner? If the driver cannot identify a repeatable phase or response, the adjust path is weak.

The practical decision is usually to hold baseline unless inspection finds a real fault. Give the next session a narrower driving task. Pick the corner or phase that created the strongest complaint, ask the driver to repeat the same approach, and collect a cleaner report. If the car develops a physical symptom during inspection, repair that. If the driver describes a control-risk symptom that cannot be verified, park. But do not make a blind chassis change just because the session felt disappointing. That turns driver variability into car variability, and the next session will be harder to read.

Worked example: a suspected mechanical failure after a loss-of-control moment

The car has a spin or a major correction, and the driver reports that it felt unlike prior mistakes. Lopez's car-control material includes the reminder that mechanical failure can be the reason a car spins and hits the wall. You do not need to overstate that into a rule that every spin is mechanical. The driver can still be wrong. But you must treat credible mechanical suspicion as different from ordinary balance dissatisfaction.

The decision gate becomes conservative. If the car shows a known physical fault, repair is the only useful first answer. After the repair, the car is not automatically cleared by optimism; it needs enough verification to make the next outing meaningful. If the fault cannot be found and the symptom remains plausible in a high-consequence phase, the park decision is justified. Parking is the cleanest way to avoid converting an unknown into an incident.

If inspection clears the car and the driver's report becomes more like an approach problem, you can return to baseline and assign a driving task. That is still not an adjustment. The important move is that you did not confuse a loss-of-control event with a setup opportunity until the mechanical question was handled.

Common mistakes

Mistake one is tuning the driver. This happens when a vague driver complaint becomes a car adjustment before anyone asks whether the driver's approach changed. Good looks like asking for the phase, the control input, and repeatability, then holding baseline when the evidence points to driver variation.

Mistake two is repairing and adjusting in the same breath. A known physical fault already changes the question. Good looks like restoring the car first, recording what changed, and making the next session prove the repair before adding a separate setup experiment.

Mistake three is chasing one lap. A single quick lap or a single ugly lap can feel persuasive, but Van Valkenburgh's standard is consistency. Good looks like judging the pattern, not the anecdote. If the lap times and reports are scattered, your next action should create cleaner evidence.

Mistake four is making an adjustment with no rollback. If you cannot return to the prior condition, you have weakened the baseline. Good looks like recording the starting point and choosing a change that can be reversed if the next session says no.

Mistake five is accepting driver language that does not help the car. Jenkinson's standard is useful, clear, concise observation. Good looks like helping the driver convert feeling into response: what the car did, where it did it, and what input produced it.

Mistake six is treating parking as failure. Parking can be the best decision when the car cannot be verified and the next session would add risk without adding trustworthy information. Good looks like documenting the unresolved question and carrying it into a proper inspection or test instead of pretending the window solved it.

Drill: the three-door decision log

Use this drill at the next event for three consecutive sessions. The count is three trackside windows. The duration is five minutes of disciplined decision work inside each window, not including normal mechanical service. The success criterion is that every window ends with exactly one primary decision: repair, adjust, park, or hold baseline for a driver task. If you choose hold baseline, record it as a no-adjust decision, not as indecision.

Before session one, write the known baseline in the briefest useful form: setup state, recent changes, and any unresolved concerns. After the car comes in, ask the driver for one response statement tied to a corner phase and control input. Then answer the five gate questions: what is the response, when did it happen, did it repeat, is the car still at baseline, and can the proposed action be reversed?

In window one, you are not allowed to make a setup adjustment unless the complaint repeats and the car is still the known car. If you find a physical problem, repair only. If the car cannot be verified, park. If the report is vague and inspection is clean, hold baseline and give the driver a narrower task for session two.

In window two, use the previous decision as the baseline. If you repaired, ask whether the original complaint changed. If you adjusted, ask only about the intended response of that adjustment. If you held baseline, ask whether the driver can now produce a clearer repeatable report. Do not let the second window accumulate unrelated changes.

In window three, audit yourself. You succeed if the car's story is clearer than it was at the start: a repaired fault stayed repaired, an adjustment was kept or reversed for a stated reason, a parking decision preserved an unresolved question, or a driver-task session separated approach from car behavior. You fail the drill if you cannot explain why each action was repair, adjust, park, or hold baseline.

When this principle breaks down

The principle does not break down because the paddock is busy. It breaks down only when the corpus no longer supports the level of specificity required. The supplied material strongly supports baseline discipline, reversible testing, driver-versus-car skepticism, adjustment literacy, clear driver observations, and caution around mechanical failure. It does not give a detailed trackside repair manual for specific components, torque procedures, fluid systems, or safety inspection thresholds.

That boundary matters. Use this lesson to choose the correct decision path. Do not use it as the final technical authority for a component-level repair. Once the decision is repair, you need the car's preparation manual, class rules, component data, and the team's inspection standards. Once the decision is park, you need a real diagnostic plan. Once the decision is adjust, you need the team's setup knowledge and a reversible change. This lesson teaches the gate, not every wrenching procedure beyond it.

Author Review

No quiz questions are attached to this lesson.

Sources

#	Document	Chunk	Pages	Score	Collection
1	Race Car Engineering Mechanics Paul Van Valkenburgh	4a0085b1-a5b6-20ef-c288-ff092fa3e4d9	116	1	uio_books_raw_v1
2	Going Faster Mastering the Art of Race Driving - Carl Lopez	ef9ea5d6-92b2-e60a-d6d0-5adac150482c	234	1	uio_books_raw_v1
3	Ultimate Speed Secrets - Ross Bentley	149c4d5c-d228-0358-acc0-8a92ac07ec7c	50	1	uio_books_raw_v1
4	The racing driver The theory and practice of fast driving Denis Jenkinson	c6871b03-6f4b-38fb-4f07-644ea98f124a	63	1	uio_books_raw_v1
5	Going Faster Mastering the Art of Race Driving - Carl Lopez	f2410e4f-42d0-24db-af78-3d9940ff312d	75	1	uio_books_raw_v1
6	Going Faster Mastering the Art of Race Driving - Carl Lopez	ea871b36-1445-a23a-54de-3d5c063243da	76	1	uio_books_raw_v1
7	Speed Secrets Professional Race Driving Techniques Ross Bentley	26bc8e35-76a6-4f72-ea86-df10ba43a636	14	1	uio_books_raw_v1
8	Going Faster Mastering the Art of Race Driving - Carl Lopez	b25a5abe-55f5-bfe9-c7d7-d89151314400	47	1	uio_books_raw_v1