Skip to main content

Run a 10-minute data review between sessions

Generated from content/lms/race-car-engineering-and-operations/04-data-and-telemetry-ops/02-quick-look-between-sessions.md; edit the source file, not this page.

Source path: content/lms/race-car-engineering-and-operations/04-data-and-telemetry-ops/02-quick-look-between-sessions.md

Course: Run the paddock like a race engineer

Module: Make the data logger your crew chief

Estimated duration: 60 minutes

Principle: make the next session better, not the last session perfect.

A between-session data review is not a complete engineering analysis. You do not have the time, attention, or calm environment to explain every wiggle in every trace. The purpose of the 10-minute review is narrower and more powerful: turn the session you just drove into one clear objective for the session you are about to drive. If you leave the laptop with a better question, a cleaner target, and a more accurate picture of what you actually did, the review worked.

The mechanism is simple. The logger records far more numbers than you can inspect between sessions. Segers frames the problem directly: a logger can create megabytes of numbers in a practice session or race, so efficient analysis requires reducing those numbers into something you can understand quickly, either as graphics or as statistics. That is why the quick look lives on overlays, lap and segment reports, time or distance plots, histograms, and simple run charts. You are not trying to admire data. You are trying to reason from it fast enough to affect the next run.

For the driver, the best use of that information is awareness. Bentley's point is that the data system helps you see what you did, not what you thought you did. The classic example is the high-speed corner you were sure was flat until the throttle trace shows a tiny lift. That is not an insult and it is not a verdict on your courage. It is a calibration tool. The 10-minute review is where you compare memory, notes, and traces until your mental picture begins to match the actual lap.

This lesson assumes the logger is already configured and armed. That setup work belongs to the sibling logger lessons. It also assumes you have some driver comments. Turning comments into hypotheses is its own sibling lesson. Here, you are standing in the paddock or trailer after a session, sweat still drying, next run coming, and you need to decide what the data says you should do next.

Start before the laptop. In the first minute after you get out of the car, write the session context and your own first impression. Bentley recommends a driver record for every practice, test, qualifying, or race session, with objectives before the session and comments afterward about track conditions, changes made, changes needed, and results. That habit keeps the data review from becoming a disconnected screen exercise. The screen shows what happened. Your note says what you were trying to do and what you felt while doing it.

The 10-minute order is: note, compare, overview, detail, cross-check, objective. If you skip the order, you will either stare at a trace until time runs out or chase a detail that never mattered. The Data for Drivers process says to start with an overview, look for incongruencies, dig for details, use other channels if available to check, ask why, compare if you can, calibrate to your driving, imagine what ideal would look like, and set objectives for the next session. That is the spine of this lesson.

Minute 0 to 1: capture the raw driver record. Write the track condition, traffic level, tires or fuel state if relevant, any setup change made before the session, and one sentence about the lap. Keep it practical. Maybe the car felt hesitant on throttle exit. Maybe you thought you were carrying too much brake into the slow right. Maybe you believed the fast sweeper was flat. Do not write a novel. You are preserving the first useful subjective signal before the data changes your story.

Minute 1 to 2: choose the comparison. The best comparison is not always the single fastest lap. You want a lap or reference that answers the current question. If the goal is driver consistency, compare two clean laps from the same session. If the goal is finding lost time, compare your best clean lap to another clean lap with a different time loss profile. If a teammate or another driver in a similar car is available, that comparison can be valuable. If no other driver exists, compare with yourself: a previous lap, a previous session, or the data shape you mentally expected before driving.

Minute 2 to 3: look at the lap table and section report. You are asking where the review should spend its time. Segment or section times, fastest rolling lap, theoretical fastest, and simple run charts can show whether the problem is global or local. A global problem might be every lap degrading or every sector inconsistent. A local problem might be one corner entry or one throttle application costing the time. The quick look should not begin by zooming into a random corner. It begins by finding where the important portion is likely to be.

Minute 3 to 5: open the overview trace. Use speed over distance as the outcome channel, then add throttle, brake pressure or brake pedal, and steering angle if available. Distance-based plots are usually easier for a track review because the same corner appears at the same horizontal location. Multilap overlays and a time-difference plot are especially useful because they tell you where the compared laps separate. Your eyes should move from delta, to speed, to the controls that probably created that speed.

Read speed as the result, not the instruction. A lower minimum speed may be a problem, or it may be the price paid for an earlier throttle and better exit. A higher entry speed may be useful, or it may be masking a late brake release and a delayed throttle. Between sessions, do not declare victory or failure from speed alone. Use speed to ask where to inspect the driver inputs.

Minute 5 to 6: inspect the brake trace where the time moved. The quick questions are the same ones from the Data for Drivers process: what is the shape of initial application, trail, and release, is the pressure consistent, and does it look light and long or hard and short. You are not proving a universal braking theory in the paddock. You are asking whether the brake trace matches your plan and whether the compared lap shows a different pattern. If you intended a decisive brake and clean release, but the trace shows a long hesitant tail, that becomes an objective. If the better lap used the same shape, the answer is somewhere else.

Minute 6 to 7: inspect the throttle trace. Look for coasting, hesitant application, early application followed by a lift, and lifts in fast corners. These are high-value tells because the throttle trace often reveals uncertainty. A driver who says the car would not go may actually be waiting on the throttle. A driver who says the corner was flat may have breathed out of it. A driver who says the exit was good may have gone to throttle too early and then lifted, giving up the exit twice. The quick look does not shame the driver for those traces. It names the exact behavior to improve.

Minute 7 to 8: add supporting channels only as needed. Steering, RPM, gear, GPS line, total steer angle, g-sum, front and rear lateral g, or understeer angle can help check the first interpretation. If the throttle hesitation appears only on laps with extra steering angle, the issue may be corner balance or line, not bravery. If RPM or gear differs, the throttle comparison may be contaminated by a shift or gear choice. If the GPS line shows a different path, a control trace comparison may be comparing two different corners. Supporting channels keep you from over-believing the first pretty line.

Minute 8 to 9: ask why and choose one cause to test. Data is strong at showing what happened. It is weaker at explaining why without the driver. This is why driver feedback still matters. Bentley emphasizes that data cannot replace driver feedback, and Segers describes logged data as objective measurement that should be combined with subjective driver comments to evaluate what is happening with the car. In the quick look, the useful sentence sounds like: the time loss begins before throttle pickup because I am coasting after brake release, or the fast-corner lift is real and happens when steering is still loaded, or the better segment has less brake tail and earlier full throttle.

Minute 9 to 10: set the next-session objective. The objective must be something you can actually drive. Do not write be faster in Turn 5. Write release the brake cleanly before asking for throttle in that corner, or hold the planned throttle through the fast sweeper if the car is settled, or compare two laps with the same gear and line before judging throttle timing. The next objective should also include what you expect the trace to look like. That mental prediction is the start of awareness calibration.

The best 10-minute reviews are narrow. A complete data system can show speed, throttle position, steering angle, brake pedal or pressure, RPM, gear, g forces, GPS line, understeer or oversteer indicators, engine data, histograms, and statistics. That does not mean every session deserves every channel. The Data for Drivers advice to keep it simple and focus on the basics is not beginner advice. It is how you protect the paddock review from turning into noise.

A practical screen template helps. Screen one is the session table: lap time, section times, fastest rolling, theoretical fastest, and notes. Screen two is the driver overlay: speed, throttle, brake, steering, and time difference against distance. Screen three is the location check: track map or GPS line when line might explain the input difference. Screen four is a summary screen when you have enough data: throttle histogram, total steer, g-sum, or a run chart for the one metric that matters today. If your software lets you save display templates, cursor markers, lap overlays, statistics per lap and section, and session notes, use those features so the review time goes into thinking rather than rebuilding views.

Comparison discipline is a sub-skill. A useful comparison has a reason. Comparing a traffic lap to a clean lap can reveal traffic cost, but it is a poor way to judge technique. Comparing a cold-tire lap to a later lap can reveal warm-up, but it may not answer a cornering question. Comparing your data to a faster driver in the same car can be very valuable, but only if you remember that the other driver may also be using a different line, gear, or risk level. The quick question is not who is better. The quick question is what behavior changed where the time changed.

Zoom discipline is another sub-skill. Start wide, then zoom. If you begin zoomed in, you will make the first interesting shape feel important. At the overview level, find the place where the delta opens, where the speed trace separates, or where a lap metric looks odd. Then zoom to the braking, throttle, or steering input around that location. The smaller your time box, the more important this discipline becomes.

Channel triangulation is the third sub-skill. One trace can mislead. Throttle alone may show hesitation, but steering may explain why. Brake pressure alone may show a long release, but speed may show that the release created a better exit. GPS line alone may show a wider path, but throttle and speed show whether it mattered. The Data for Drivers process specifically says to use other channels if available to check. That line should live in your head during every quick look.

Driver-awareness calibration is the fourth sub-skill. Before you look, say what you expect to see. After you look, compare the trace to the mental picture. Bentley suggests taking a session with a mental picture of what the data will look like, then comparing that picture with the actual data after getting out. Over time, the two get closer. That is one of the most valuable outcomes of data work because it improves your driving while you are still in the car. The driver who can accurately feel a small lift, a coasting gap, or a long brake tail can fix it sooner than the driver who only discovers it at the laptop.

Do not let data replace the driver. Data can tell you what the car and driver did, and it can show the location of a handling problem more objectively than memory alone. But driver feedback tells you whether the car felt nervous, whether traffic changed the line, whether a curb unsettled the car, or whether the driver intentionally gave up entry to test exit. The quick review uses both. If the data disagrees with the driver, that is not a fight. It is the review's most useful moment.

Do not turn every review into a setup meeting. Vehicle-performance analysis can help an engineer decide setup changes before the next session, but this lesson is a driver quick look. If the objective evidence and driver comments point to a handling issue, record the location and symptom. Then decide whether it belongs to a longer engineering review. In most HPDE and club-racing contexts, the fastest next action is still a driver objective, not touching the car because one trace looked different.

The output of the review is a written next-run line. A good line includes location, behavior, and evidence. For example: in the fast sweeper, compare throttle trace to memory and aim for no unplanned lift if steering and balance feel settled. Or: in the slow corner after the back straight, reduce the coast gap between brake release and throttle pickup and check whether the time delta stops opening before exit. Or: hold the same gear through the comparison segment so the throttle trace is easier to interpret. Each line is small enough to execute and specific enough to review.

The lap-time signature of a good quick-look habit is not always immediate fastest lap. Early improvement often shows as consistency: less lap-to-lap scatter in the target segment, fewer unexplained lifts, more repeatable brake pressure shape, less random coasting, and a closer match between your predicted trace and the actual trace. Over time, the time follows because you are removing uncertainty and replacing it with deliberate inputs.

The instructor signature is similar. A good instructor reviewing your data between sessions will not drown you in channels. They will ask what you were trying to do, find where the time changed, check the control trace that matters, compare it to your comment, and give you one thing to drive. That is the model you are learning to run for yourself.

Cross-reference this skill with the lesson on turning driver comments into data hypotheses when your notes are vague or emotional. Cross-reference the logger configuration lessons if the channels you need are missing, mislabeled, or hard to find. This lesson begins after the data exists and after the comment exists. Your job here is to turn both into a useful next lap.

Worked example: the fast sweeper that was not actually flat

This is the cleanest driver-development example in the bonded corpus. You come in after a session convinced that a fast sweeper was flat. The data says otherwise: the throttle trace shows a small lift. The wrong reaction is to argue with the logger or decide that the system must be wrong because your memory felt clear. Bentley warns that this defensive reaction wastes the tool. The right reaction is to use the mismatch as awareness training.

Run the 10-minute process. First, write the driver note before opening the data: believed fast sweeper was full throttle, car felt secure, no obvious correction. Then choose a comparison. Use another clean lap from the same session, your previous session, or a similar driver's lap if available. Open speed and throttle over distance with a time-difference plot. Find the sweeper by distance or track map. Look at the exact point where the lift appears.

Now cross-check. Add steering angle and lateral g if you have them. If the lift happens while steering angle is still high and the car is loaded, the behavior may be a confidence or balance response. If the lift happens at the same distance every lap, it is a habit or a planned protection move even if you did not consciously name it. If the lift happens only once, ask about traffic, line, or a moment in the car. Add GPS line if the path changed. Add RPM or gear if a shift or limiter event might explain a throttle change.

The next-session objective should not be reckless. It is not simply keep your foot down no matter what. A good objective is conditional and observable: enter on the same line, hold the planned throttle only if the car is settled, and predict the trace before the run. After the session, the success criterion is whether your memory matches the actual throttle trace and whether the lift was planned, reduced, or correctly preserved for balance. The win is not bravado. The win is that your awareness became accurate.

Worked example: same-car comparison after a lost segment

Suppose two drivers use the same car, or you have data from a teammate or similar car, and the segment report shows that you lose most of the lap in one section rather than everywhere. This is exactly the kind of comparison the corpus supports: data can compare different laps by one driver or compare style and performance among multiple drivers, especially when the same car is used. The quick look should stay local. Do not compare the whole lap as a moral judgment. Compare the section where the time moved.

Start with the section report and time-difference plot. Find where the delta begins to open. Then display speed, throttle, brake, and steering over distance. If the faster trace brakes later but also carries a long brake tail and delays throttle, do not assume later braking is the lesson. Look at where the delta actually improves. If the faster segment gains on exit, the useful behavior may be earlier throttle pickup, less coast, or a cleaner brake release. If the faster segment gains before apex and then holds, the useful behavior may be entry speed or brake shape. The data tells you where to ask, not what to copy blindly.

Now add checks. If the faster driver uses less steering angle, the difference may be line or rotation rather than pedal courage. If the faster driver uses a different gear, normalize that before judging throttle percentage. If the GPS line differs, the two traces may not be answering the same question. If the comparison driver has a different risk tolerance or car state, keep the objective conservative.

A good next-session objective from this example might be: in that segment, repeat the same gear and line for three laps, then reduce the coast gap after brake release while keeping the throttle application progressive. That objective came from the data but remains driveable. It also sets up the next review: did the coast gap shrink, did the section time improve, and did the line remain comparable.

Common mistakes and what good looks like

Mistake 1: starting zoomed in. You open the file, find a squiggle, and spend the whole break explaining it. What it costs is focus. You may solve a shape that had no lap-time or consistency consequence. Good looks like starting with lap and section overview, finding where the time or metric changed, and only then zooming into the relevant controls.

Mistake 2: treating speed as the whole answer. Speed is an outcome channel. It tells you where to look, not what to do. A higher minimum speed can be good or bad depending on exit, throttle, line, and control overlap. Good looks like reading speed together with throttle, brake, steering, and delta.

Mistake 3: arguing with the data when it contradicts memory. The common version is the fast-corner lift you did not feel. What it costs is awareness. If you reject the tool, you lose the chance to calibrate what you feel in the car. Good looks like writing your prediction first, comparing it calmly to the trace, and treating the mismatch as the next training target.

Mistake 4: comparing bad references. A traffic lap, cold-tire lap, or different-line lap can answer a specific question, but it can also contaminate a technique review. Good looks like choosing the comparison because it matches the question, then checking gear, line, traffic, and session conditions before copying a behavior.

Mistake 5: adding every channel because it is available. More traces can make the review feel more serious while making the decision worse. Good looks like starting with basics: speed, throttle, brake, steering, lap or segment time, and only adding RPM, gear, GPS, g-sum, total steer, understeer angle, or engine channels when they test a specific interpretation.

Mistake 6: leaving with five objectives. A driver cannot execute five new behaviors at speed. The result is usually vague effort and no clean data in the next session. Good looks like one objective with a location, a behavior, and an expected trace.

Mistake 7: using data to replace the driver comment. Logged data is objective, but the driver knows intent, traffic, confidence, and feel. Good looks like combining subjective comments with objective traces. If they conflict, the conflict becomes the question for the next run.

Mistake 8: turning every finding into a setup change. Data can support setup decisions, but the quick look is usually too thin for a confident car change unless the evidence is clear and repeatable. Good looks like recording the location and symptom, checking available channels, and deciding whether the next action is a driver objective or a deeper vehicle review.

Drill: three-session 10-minute quick-look loop

Run this drill at your next event over three consecutive sessions. The count is three reviews. The duration is 10 minutes each, timed. The success criterion is not that you set a personal best. The success criterion is that each review produces one written next-session objective and that your predicted trace becomes closer to the actual trace by the third review.

Before session one, write one objective and one predicted data signature. Keep it simple. For example, you might predict no coasting between brake release and throttle in a chosen corner, or no unplanned lift in a fast corner. Drive the session normally with that target in mind.

After session one, start the timer. Minute one is notes. Minute two is comparison choice. Minutes three through five are overview and section location. Minutes five through seven are brake and throttle detail. Minute eight is supporting-channel check. Minute nine is why. Minute ten is the next objective. Stop when the timer ends. If you did not finish, your review was too broad.

Before session two, drive only the new objective. Do not add extra targets. After session two, repeat the same 10-minute structure. This time, compare not only lap time but also mental prediction versus actual trace. Did the thing you thought you changed actually change? Did the trace improve but the time not follow? Did the time improve for a different reason than expected?

Before session three, refine the objective rather than replacing it too early. If the data showed a real fast-corner lift and you planned to make it deliberate, the objective might become identify whether the lift is car balance or driver uncertainty. If the brake trace showed a long tail, the objective might become make the release repeatable before changing entry speed. After session three, the drill is successful if you can state what the data showed, what you thought you did, where those agreed or disagreed, and what one behavior you would keep practicing.

When to stop the quick look and ask for a deeper review

A 10-minute review has limits. Stop the quick look when the conclusion depends on channels you do not have, when the same symptom appears in several unrelated places, when driver comments and data conflict in a way you cannot explain with line, traffic, gear, or conditions, or when the evidence points toward a vehicle behavior rather than a single driver input.

That is not a failure. Segers separates driver-performance analysis from vehicle-performance analysis, and the latter combines objective data with subjective comments to identify handling problems and where they occur. Between sessions, you may only have enough time to mark the issue and avoid making it worse. A deeper review can then use more laps, more reliable statistics, more channels, and a calmer comparison.

A good stop statement is precise: the car feels unstable at the same location where steering and throttle traces change, but the quick look cannot separate line, balance, and confidence. Or: brake pressure shape is inconsistent in several corners, so the next session should prioritize repeatability before judging setup. The point is to leave with a safe, honest next step rather than a dramatic conclusion built from one trace.

Author Review

No quiz questions are attached to this lesson.

Sources

#DocumentChunkPagesScoreCollection
1Data for Driverscabda699642b26311b0a7ef998da2c71151uio_books_raw_v1
2Inner Speed Secrets - Ross Bentley02822a18-86f0-c082-0db4-b817146db0d91531uio_books_raw_v1
3Analysis Techniques for Racecar Data Acquisitionbe2d8954-f3d6-95a0-e682-04b9cf37b69861uio_books_raw_v1
4Analysis Techniques for Racecar Data Acquisition9998c72b-304d-0767-6517-dc3b82cea9fe61uio_books_raw_v1
5Analysis Techniques for Racecar Data Acquisition41138569-fa56-a0a4-38c5-301475e4131a211uio_books_raw_v1
6Speed Secrets Professional Race Driving Techniques Ross Bentleya009c9a4-cb8d-b3b5-063d-33e44ea0b5cb761uio_books_raw_v1
7Ultimate Speed Secrets - Ross Bentley7212e525-6587-a46d-1fab-5d027a6e940e5531uio_books_raw_v1
8Analysis Techniques for Racecar Data Acquisition7ae7b884-5466-cf01-8e1a-333086305e8551uio_books_raw_v1
9Analysis Techniques for Racecar Data Acquisition15474906-387d-234d-cb57-341d5efc4d3a51uio_books_raw_v1
10Analysis Techniques for Racecar Data Acquisition66088a66-7d06-8e55-03eb-967374239bec61uio_books_raw_v1
11Analysis Techniques for Racecar Data Acquisition52e7d5ab-412b-acc5-fb49-cb0e8d5511b161uio_books_raw_v1
12Analysis Techniques for Racecar Data Acquisition5eeea298-6191-0fb2-1054-b10fe574a80421uio_books_raw_v1
13Data-for-Drivers-PRINTb80dc634-a0a7-d6de-d470-353aed47e2a6171uio_books_raw_v1
14Speed Secrets Professional Race Driving Techniques Ross Bentleyac8d24bf-596a-c55e-02f9-aded61bf903a761uio_books_raw_v1