Build a best possible lap you can actually drive
Generated from
content/lms/data-interpretation-for-drivers/04-comparing-laps/02-best-possible-lap.md; edit the source file, not this page.
Source path: content/lms/data-interpretation-for-drivers/04-comparing-laps/02-best-possible-lap.md
Course: Data Interpretation for Drivers
Module: Comparing Laps
Estimated duration: 55 minutes
The point of a best possible lap is not to make a pretty number on your laptop. It is to build a lap plan that survives contact with the next session. In data software, the theoretical or best possible lap can look clean: take the best sector here, the best sector there, add them up, and the result is faster than anything you drove. That number is useful only if you can explain how the pieces connect. If the best sector requires a brake release that destroys your next corner, or an early throttle that you only managed because you compromised the previous straight, the number is not a lap you can drive. It is a math artifact.
Your job in this lesson is to turn the idea of theoretical fastest into a driver tool. You will compare laps, find the sections where your own driving already proved a better version is possible, and decide which pieces belong together. Then you will convert that into one or two objectives for the next session. This lesson assumes you already know how to use delta time to find large losses and how to choose reference laps. We will use those skills, but this lesson is narrower: how to assemble a best possible lap without fooling yourself.
Principle: the best possible lap is a coherent lap, not a pile of best fragments.
The core rule is simple. A candidate best possible lap is only useful when every chosen section can be explained by channels that match the driving technique you intend to use. If a section is faster, you should be able to say what changed: brake timing, brake pressure shape, throttle commitment, lift behavior, steering demand, lateral g consistency, RPM, gear choice, GPS line, or another channel your system records. If the trace only says the section time was better but you cannot explain why, treat that section as a question, not as an instruction.
That rule comes from the driver-focused data process in the corpus. Start with an overview. Look for incongruencies. Dig for details. Use other channels when available. Ask why. Compare when you can. Calibrate the finding to your actual driving. Imagine what the better version would look like. Set objectives for the next session. That is the whole method in miniature. The best possible lap is the place where those steps matter most, because it is easy to take a software-generated theoretical fastest number at face value and skip the harder question: can you actually execute those pieces in one lap?
This is also why a fastest lap alone can mislead you. The supplied data lesson points out that if you only looked at the fastest lap, you could miss important information. One lap may contain your best brake approach to one corner and a weak exit from another. A slower lap may contain the exit you should keep. The useful question is not only which full lap was fastest. It is which pieces show repeatable, explainable driving that should become your next lap plan.
What data you need before you build one.
You can start with simple data, but you need more than a lap timer. The minimum useful package is speed plus longitudinal and lateral g. With those channels, you can see where speed changed, where the car was accelerating or decelerating, and how consistently you used cornering force. Ideally you also have throttle position and brake pressure, because those two channels tell you far more about the driver input that created the speed trace. Steering angle and engine RPM are the next layer, because they help explain line, scrub, gear choice, and whether the driver asked too much of the car.
Do not let the absence of every channel stop you, but do not pretend missing channels are present. If you have speed and g only, you can still compare section times and speed shapes, but you must be more cautious about declaring the cause. If you have throttle and brake, you can separate a lift from a brake brush, a hesitant throttle from a deliberate maintenance throttle, and a long light brake event from a shorter harder one. If you have steering, you can check whether the faster section used less total steering or whether the car was forced through the corner with extra wheel angle. If you have RPM and gear, you can check whether an apparent acceleration loss is actually a gear or shift issue.
The data source itself has limits. The corpus explicitly frames the webinar as aimed at drivers, not data engineers, and warns that data does not tell you everything. It also reminds you to know the limitations of your tool. That matters because a best possible lap tempts you to treat the screen as complete truth. The screen can show that the red lap lost speed, that the blue lap gained time, or that brake pressure appeared at a different place. It may not tell you whether there was traffic, whether your vision was late, whether your mental picture of the corner was wrong, or whether you made a courage decision rather than a technique decision. Build the lap from the channels, but keep the driving context in the room.
The working definition.
For this lesson, a best possible lap you can actually drive has four tests.
First, it uses your own proven sections, or comparison sections from a reference that you can adapt to your car and skill level. If you have never made that corner entry, never released the brake that way, and never held that throttle commitment, it may still be a target, but it is not yet your personal best possible lap.
Second, each selected section has a cause you can see in the data. The better section was not just better. It was better because the throttle trace was cleaner, the brake trace was shorter and more decisive, the lateral g was more consistent, the GPS line avoided a delay, the steering trace showed less total steering, or the segment report exposed a section that rolled better into the next one.
Third, the sections do not fight each other. The best entry to one corner cannot be selected if it only happened by ruining the exit before it. The best exit cannot be selected if it required a midcorner correction that would be dangerous or non-repeatable in normal traffic. The lap must have a shape you can picture before you drive.
Fourth, the lap produces a small number of next-session objectives. If the analysis ends with six major changes, it has not been turned into a driving lesson yet. The project is to get faster the next time the car hits the track, not to admire a complex analysis.
Step 1: start with the overview before you chase the number.
Open the session and begin at the full-lap level. Look at the fastest lap, but do not stop there. Also look at the lap that has the best rolling section, the lap with the cleanest repeated corner sequence, and any lap that looks unusually strong in a section where the fastest lap was weak. If your software provides fastest rolling, theoretical fastest, and segment or section reports, use them as a map of where to inspect, not as the final answer.
A useful first pass is to list the sections where the theoretical lap takes time from your actual fastest lap. You are not yet deciding that those sections belong in the plan. You are only locating the evidence. For each section, ask what the time gain could be made of. Is it earlier throttle? Less coasting? Later or harder braking? Better release? Higher minimum speed? More consistent lateral g? Cleaner steering? Better RPM placement? Less line distance? The next steps will test those possibilities.
This first pass should be fast and plain. Data analysis texts in the corpus emphasize extracting metrics and visualizing lap metrics so important portions are quickly detectable and conclusions can be made efficiently. That is exactly the goal here. You are trying to reduce a whole session into a short list of candidate places where a better lap may exist.
Step 2: use delta or compare time to prioritize, then leave delta behind.
Delta time is a ranking tool. It tells you where the largest differences happened. It does not, by itself, tell you what to do. The supplied data lesson says to overlay two laps, look for differences, and use delta or compare time to find the biggest differences and prioritize. That is the correct order. Let delta tell you where to look first.
Once you find a major delta movement, switch from scoreboard thinking to mechanism thinking. If the red lap loses time over a section, identify what changed in the channels. Did GPS speed fall earlier? Did the throttle trace show a lift? Did brake pressure appear? Was there a brake release problem? Did steering angle rise, suggesting scrub or line compromise? Did lateral g spike or fall away? Was RPM wrong for the exit? Delta points at the leak; the channels explain the leak.
This distinction protects you from a common trap. Suppose one lap gains a tenth before corner entry but loses two tenths on exit. If you only hunt the local delta gain, you may select the wrong piece for your best possible lap. You must evaluate the whole corner or section, not just the moment when the delta line temporarily moved in your favor. The useful candidate is the section that improves the lap when driven as a connected sequence.
Step 3: split the lap into sections that match how you drive.
Software sectors are useful, but they are not always the right driver sections. A track is driven as braking zones, turn-in commitments, midcorner balances, exits, and linked sequences. Build sections that match decisions you make in the car. A section might be brake release through apex. Another might be apex to track-out. Another might be two corners treated as one because the first corner exists to set up the second.
Use the segment or section report as the starting point. The corpus lists segment and section report times beside fastest rolling and theoretical fastest, which is exactly the tool family you need. But after the report identifies a section, define the driver task inside it. If the candidate gain occurs from brake application to apex, the task is likely braking shape, release, and rotation. If it occurs from apex to exit, the task is throttle application, steering unwind, and line. If the gain begins before the braking zone and continues after the exit, the task may be the whole approach and not a single input.
The reason to use driver-shaped sections is that your next-session objective must be executable. You cannot drive a spreadsheet sector. You can drive an earlier initial brake, a cleaner release, a deliberate no-coast handoff, a later throttle pickup only after the steering begins to open, or a line that lets you reduce total steering. The data must become an action in the cockpit.
Step 4: classify each candidate section by confidence.
Do not put every best sector into the same bucket. Sort candidates into high-confidence, medium-confidence, and low-confidence pieces.
A high-confidence piece is visible in multiple channels and does not damage the next section. For example, the lap is faster from apex to exit, the throttle trace shows earlier and smoother commitment, brake pressure is already gone, steering angle is unwinding rather than increasing, GPS speed builds earlier, and the following straight remains better. That is a strong candidate. It shows a technique change and a result.
A medium-confidence piece is real but incomplete. Maybe the speed trace and delta are better, but you do not have throttle or brake data. Maybe the brake pressure shape looks better, but the GPS line is missing. Maybe the section is faster, but the following section is neutral and you need another session to verify repeatability. Medium confidence can still become an objective, but it should be framed as a test.
A low-confidence piece is a number without a cause, a gain followed by a larger loss, or a section that depended on circumstances the data cannot explain. The corpus gives examples of questions that data may generate: throttle lift, braking, steering angle, line, vision, mental image, bravery, and traffic. If the candidate gain may be traffic or a one-time bravery moment rather than technique, do not build the lap around it yet. Mark it for investigation.
Step 5: use channel checks to decide what the better section teaches.
The speed trace shows the outcome. It tells you whether the car slowed more, slowed less, accelerated earlier, carried a higher minimum, or gave up speed in a fast corner. Start there, but do not stay there. A speed trace without input context is a symptom list.
The throttle trace answers several driver questions. Are you coasting where you thought you were on throttle or brake? Is the application hesitant? Did you get on throttle early and then lift, which usually means the first application was not actually usable? Did you lift in a fast corner where the better lap stayed committed? For a best possible lap, a throttle-based candidate should show commitment you can repeat, not just a momentary stab that creates a later correction.
The brake pressure trace answers a different set of questions. Look at the shape of the initial application, the trail, and any long tail. Look for inconsistent pressure. Compare light and long braking against hard and short braking. If the faster section uses a cleaner brake shape, decide exactly which part matters. Was the initial brake more decisive? Was release more progressive? Did the driver avoid dragging pressure too long? Did the trace show a pressure event that explains a speed loss? Do not summarize it as brake later unless the trace really supports that.
The lateral g trace tells you whether the car was using cornering capacity consistently. Look at peak g, whether peak usage is consistent lap to lap, and whether spikes appear in either direction. A best possible section that contains a lateral g spike may not be a clean target. A smoother section with similar or better speed may be more drivable than a jagged one that happened to score well in a small sector.
Steering, if available, helps you check whether the lap is being made with the front tires or against them. The corpus lists steering angle and total steer angle as useful analysis channels, and Bentley's cornering material emphasizes that less steering generally belongs with more speed. In best possible lap work, this means you should distrust a gain that requires a big steering increase unless the section context explains it. If the better lap used less steering, cleaner lateral g, and earlier throttle, that is a more teachable section than a lap that simply forced a higher minimum speed with extra wheel.
RPM and gear help you keep the analysis honest on exits and straights. If the candidate section gains or loses acceleration, check whether it is really a driving line or throttle issue, or whether the gear and RPM trace explain the difference. The data process list includes RPM and gear because they keep you from blaming corner technique for a power delivery or shifting difference.
GPS line and g-sum add the vehicle path and combined-load picture when your tool supports them. GPS line helps answer whether the car was placed differently. G-sum can help you see whether the car was being asked to combine braking, turning, and acceleration differently. Treat those as confirmation channels. They should strengthen or challenge the story you already see in speed, throttle, brake, steering, and g.
Step 6: reject the fantasy lap.
A fantasy lap is a theoretical lap made from incompatible truths. Every chosen fragment happened somewhere, but the assembled lap cannot be driven as one connected lap. You reject fantasy laps by asking five questions.
Did the chosen section require a different entry than the section before it would provide? If yes, it is not a plug-in piece. You need to include the setup before the gain or leave it out.
Did the gain create a loss immediately afterward? If yes, inspect the whole sequence. The better piece may still be useful, but not in the way the segment report first suggests.
Was the gain caused by an input you cannot repeat safely? A one-time lift avoidance in a fast corner may show possibility, but if the rest of the channels show instability or a large correction, the next-session objective should be understanding the corner, not matching that trace.
Is the gain supported by enough channels? Speed plus delta may be enough to point you toward a candidate, but a best possible lap plan needs an input story. When the tool lacks channels, frame the plan as a test, not a conclusion.
Can you picture the lap before you drive it? Bentley's introduction makes a useful learning point: understanding the theory and picturing it before you drive makes you more sensitive to the experience. If your best possible lap is just a table of sectors, you cannot picture it. Convert it into a lap narrative you can rehearse.
Step 7: turn the assembled lap into a driver narrative.
The best possible lap plan should read like a sequence of intentions, not a data report. For each selected section, write the driver action and the evidence. Keep it short enough to carry into the next session.
A usable entry might be: in the first priority section, keep the stronger brake application from lap 6, but copy the earlier release from lap 9 only if the steering trace stays clean. The evidence is that lap 9 gained time through the middle without a following exit loss, while lap 6 had the better approach speed. The test is whether the combined brake shape produces the same lateral g without a steering spike.
A usable exit entry might be: delay throttle pickup until the wheel starts to open, then commit without the small lift that appears on the weaker lap. The evidence is that the better lap's throttle trace was cleaner, speed built earlier, and the next straight stayed better. The test is whether the throttle trace shows one committed application rather than early throttle followed by a lift.
A usable fast-corner entry might be: remove the confidence lift only after confirming the line and steering angle match the better lap. The evidence is that the lift explains the speed reduction, but the data cannot say whether vision, mental image, bravery, or traffic caused it. The test is not simply more courage. The test is whether the car placement and steering demand make the no-lift version reasonable.
Notice the pattern. You are not telling yourself to go faster everywhere. You are giving yourself a small number of conditional instructions that came from data and still respect what the data cannot know.
Step 8: set next-session objectives that match the size of your attention.
The overall process in the corpus ends with setting objectives for the next session. That is where best possible lap work either becomes useful or collapses into desk racing. You should leave the analysis with one primary objective and, at most, one secondary objective.
The primary objective should target the section with the largest explainable gain. Largest means it mattered in delta or section time. Explainable means the channels point to a specific driver action. Drivable means it does not require you to change the entire lap at once.
The secondary objective should be a monitor, not another full project. For example, if the primary objective is throttle commitment on corner exit, the secondary monitor might be steering angle at throttle pickup. If the primary objective is brake release, the secondary monitor might be lateral g smoothness. If the primary objective is avoiding a lift in a fast corner, the secondary monitor might be GPS line or steering demand. This keeps you from improving one channel by damaging another.
After the session, compare the target section first, not the lap time first. Lap time can be affected by many things. The objective was to change a specific driver behavior and see whether the section improved without an obvious penalty. If the section improved and the channels look cleaner, you made progress even if the full lap was not your fastest. If the lap time improved but the target behavior did not, the data is telling you that something else happened and you need to keep asking why.
Calibration cues: how you know the best possible lap is becoming real.
The first cue is repeatability. The corpus repeatedly points to consistency lap to lap, especially in lateral g and comparisons across laps, drivers, cars, or sessions. If the candidate section only appears once and never again, it may be a clue, but it is not yet a dependable skill. If the same cleaner trace begins to appear in multiple laps, the best possible piece is becoming part of your driving.
The second cue is agreement between channels. A good candidate does not require you to ignore half the screen. Speed improves, throttle or brake explains why, lateral g looks reasonable, steering does not show a panic correction, and the following section does not pay the bill. When channels agree, confidence goes up.
The third cue is a cleaner mental image. Before the next session, you should be able to describe the target without looking at the graph. If you cannot, the analysis is too abstract. The best possible lap must be translated into what you will see, when you will release pressure, when you will commit throttle, or what line you will place the car on.
The fourth cue is a smaller task list. Beginners in data often come away with too many notes. Intermediate drivers should learn to reduce the analysis. The supplied process says to keep learning, keep it simple, focus on basics, and ask why. A best possible lap that produces one clear correction is more valuable than one that produces ten vague intentions.
The fifth cue is that the theoretical gap changes for the right reason. If the best possible lap number gets closer because you improved the target section, that is evidence. If it changes because the software found a new isolated fragment but your repeatable driving did not improve, treat it as a new question.
How to handle contradictions.
Data often gives you an uncomfortable split. One lap shows a better entry. Another lap shows a better exit. One driver carries more speed but uses more steering. Your lap has cleaner brake pressure but lower minimum speed. The theoretical lap wants all of it.
Resolve contradictions by asking which piece teaches the next skill. For an intermediate driver, the best answer is usually the section with the clearest input-to-result connection. If the faster entry cannot be explained but the better exit clearly comes from removing a hesitation in throttle, work the exit. If the lower minimum speed produces a better straight because it allows earlier throttle and less steering, do not reject it just because the speed trace dips lower at the apex. If a high minimum speed comes with a later lift or a steering spike, it may be less useful than a calmer version.
This is where the data process becomes a driver process. You are not trying to crown one lap as morally correct. You are trying to understand what the car and driver did, what can be learned from it, and how that information can be used the next time the vehicle hits the track. That is the practical promise of data acquisition in the corpus.
Common failure mode: fastest-lap worship.
The fastest lap is important, but it is not automatically the best teacher. The supplied lesson material explicitly warns that looking only at the fastest lap can miss important information. If your fastest lap contains a great first half and a weak second half, it may hide the better second-half driving from another lap. If your slower lap contains the cleanest throttle trace of the day in one key section, that slower lap may be the source of your next improvement.
Good looks like this: use the fastest lap as one reference, then use segment reports, fastest rolling, theoretical fastest, and overlays to find better sections inside other laps. The final plan can borrow from the fastest lap, but it should not be trapped by it.
Common failure mode: treating theoretical fastest as a command.
A theoretical fastest number can be seductive because it is precise. Precision is not the same as truth. The number may combine sections that happened under different setups, different traffic, different driver risk, or different line choices. The corpus gives fastest rolling and theoretical fastest as analysis tools among many others, not as stand-alone answers.
Good looks like this: treat theoretical fastest as a searchlight. It tells you where to inspect. It does not tell you what to copy until the channels explain why the selected pieces were faster and whether they connect.
Common failure mode: channel shopping.
Channel shopping happens when you find one trace that supports the answer you want and ignore the others. You want to brake later, so you point to a later brake marker even though the speed trace collapses afterward. You want to go to throttle earlier, so you point to the first throttle movement even though the same trace shows a lift. You want to carry more speed, so you point to minimum speed while ignoring steering angle or lateral g spikes.
Good looks like this: use other channels to check the claim. The corpus says to confirm issues with other channels when available. If the channels disagree, slow down and ask why. The disagreement is often the lesson.
Common failure mode: no cockpit translation.
Some drivers produce a good analysis and then drive out with no clear instruction. They know the theoretical lap is faster by a certain amount, but they do not know what to do at turn-in, brake release, apex, or throttle pickup.
Good looks like this: every selected section becomes a cockpit action. Not improve sector two. Instead, remove the coast before throttle in the exit section. Not gain two tenths in the fast corner. Instead, verify line and steering demand, then avoid the confidence lift if the approach matches the better lap.
Common failure mode: asking data to answer a human question it cannot answer alone.
The Lime Rock example in the corpus shows the correct humility. The data can ask what led to a speed reduction and can point at throttle, braking, steering angle, line, and brake pressure. But it may also raise human causes such as vision, mental image, bravery, or traffic. If you ignore those possibilities, you may prescribe a technique fix for a perception problem.
Good looks like this: when a channel points to an input but the cause is unclear, write the next objective as a test. For example, test whether the lift disappears when your eyes and line match the better lap, rather than commanding yourself to lift less in isolation.
How this lesson connects to the sibling lessons.
Use the delta-time lesson to find the lap's biggest leak. Use the reference-lap lesson to choose the comparison that is fair and useful. Then use this lesson to assemble the drivable version of your lap. The sequence matters. Delta tells you where the time moved. Reference choice tells you what comparison deserves attention. Best possible lap building tells you which pieces become your next-session plan.
The finished output.
At the end of the process, you should have a short written best possible lap plan with three parts. First, the current honest lap: the fastest or most representative lap you can actually repeat. Second, the candidate gains: the sections where another lap or comparison shows a better version. Third, the next-session plan: one primary objective and one monitor channel.
A strong plan is specific enough that another instructor could understand it without opening the whole file. It names the section, the input change, the confirming channels, and the success criterion. It also names what you are not changing. That last part matters. A best possible lap you can actually drive is not an invitation to chase every trace on the screen. It is a disciplined way to decide which better piece of your own driving deserves to become normal.
Worked example: Lime Rock Park red-blue overlay
The bonded data packet includes a Lime Rock Park overlay with red and blue laps, GPS speed, throttle position, front brake pressure, and time lost plotted over distance. The lesson text around that image gives the right instructor question: when the red lap loses speed, what caused it? The possible causes listed are throttle lift, braking, steering angle, line, and brake pressure, followed by the reminder that the deeper cause may be vision, mental image, bravery, or traffic.
For best possible lap work, you would not simply choose the blue section because it is faster. You would inspect the moment where time lost begins to change, then walk down the channels. If speed drops and throttle also drops, you have evidence of a lift. If speed drops and brake pressure appears, the cause is different. If the speed shape changes without throttle or brake explanation, steering angle or line becomes more important. If the channel story still does not explain the difference, you mark the section as uncertain instead of forcing a conclusion.
Now turn that into a drivable lap plan. If the blue lap stays on throttle where red lifts, and the GPS line plus steering trace show the car was placed similarly, the candidate objective might be to remove the lift only when the approach line is correct. If the red lap uses brake pressure where blue does not, the objective might be to decide earlier whether the corner needs a brake or a committed maintenance throttle. If the red lap has more steering angle before the speed reduction, the lift may be a consequence rather than the original problem. In that case the next-session task is line and steering demand, not bravery.
The success criterion is not just a faster Lime Rock lap. The first success criterion is a cleaner target section: less unexplained speed reduction, a throttle and brake trace that match the intended plan, and no obvious time payback immediately afterward. If that target section improves and the following straight or next section holds steady, the best possible lap piece has become more real.
Worked example: two laps that each contain half the answer
The corpus includes a simple but important warning: if you only looked at the fastest lap, you would miss information, and if the driver had put the red and blue laps together, the result would have been better. That is the heart of best possible lap construction.
Imagine your fastest lap is red. It has your best opening section and a strong braking approach in the middle of the lap, but it contains a hesitant throttle application on one exit. Blue is slower overall, but in that exit section the throttle trace is cleaner, speed builds earlier, and the following straight is better. The wrong answer is to say red is the reference because red is fastest. The better answer is to treat red as the base lap and blue as evidence that one section can be driven better.
You then test whether the blue exit can be attached to the red approach. Look at the entry speed, brake trace, steering angle, and lateral g before the exit. If blue's better exit came from a much slower or different entry, the piece may not attach cleanly. You may need to copy more of the blue corner, not just the exit. If blue's approach is similar and the main difference is less coasting or a cleaner throttle pickup, that is a strong candidate. The best possible lap is not red plus a wish. It is red plus a specific blue behavior that the data can explain.
The driver instruction might become: keep the red lap's braking reference, but use the blue lap's throttle discipline after apex. The monitor is whether the new lap avoids early throttle followed by lift. If the new trace shows one deliberate throttle build, similar steering unwind, and better speed onto the next straight, the splice worked. If it shows early throttle, lift, and a steering correction, the blue fragment was copied at the wrong time or without the setup that made it work.
Common mistakes: what wrong looks like and what good looks like
Fastest-lap worship is the first mistake. Wrong looks like deleting every lap except the fastest one and assuming it contains the best version of every corner. It costs you learning because a slower lap may contain a cleaner section. Good looks like using the fastest lap as a base while scanning segment reports, fastest rolling, theoretical fastest, and overlays for better sections hidden in other laps.
Theoretical-lap worship is the second mistake. Wrong looks like taking the software's best possible number as your next target without checking whether the pieces connect. It costs you focus and can push you toward incompatible inputs. Good looks like using theoretical fastest as a map of candidate gains, then requiring each candidate to pass the channel, connection, and cockpit-translation tests.
Delta-only diagnosis is the third mistake. Wrong looks like seeing where time was lost and immediately deciding what to do. It costs you accuracy because delta shows where the difference appears, not necessarily what caused it. Good looks like using delta to prioritize, then checking speed, throttle, brake pressure, steering, RPM, lateral g, longitudinal g, and GPS line when available.
Single-channel certainty is the fourth mistake. Wrong looks like proving a conclusion from one trace while ignoring other channels. A driver may say the answer is earlier throttle even though the same throttle trace shows a lift. Another may say the answer is higher minimum speed even though steering angle spikes and exit speed suffers. Good looks like making the channels agree before a section becomes part of the plan.
No next-session objective is the fifth mistake. Wrong looks like ending analysis with a long list of interesting findings. It costs you track time because you cannot drive a long list at speed. Good looks like one primary objective, one monitor channel, and a defined section where you will check whether the behavior changed.
Ignoring the human cause is the sixth mistake. Wrong looks like assuming every lift, brake touch, or speed loss is purely a technique choice. The corpus reminds you that vision, mental image, bravery, and traffic may sit behind the trace. Good looks like writing uncertain findings as tests, especially in fast corners or places where context may have changed.
Drill: the three-candidate best-lap build
Do this drill after your next event or between sessions if you have enough time to think clearly. Use two to four laps from the same session or comparable sessions. You need speed and longitudinal or lateral g at minimum. Throttle and brake pressure are strongly preferred, and steering, RPM, gear, GPS line, and g-sum make the drill better.
First, spend ten minutes on the overview. Identify your fastest lap, the software's theoretical or best possible lap if available, and the three sections where the theoretical lap gains the most over your actual lap. Do not solve anything yet. Just name three candidates.
Second, spend fifteen minutes on channel checks. For each candidate, write the section, the apparent gain, and the likely cause. Use speed first, then throttle and brake if available, then lateral g, steering, RPM, gear, and GPS line as confirming channels. Mark each candidate high, medium, or low confidence. High confidence means the channels explain the gain and the following section is not damaged. Medium confidence means the gain is real but some evidence is missing. Low confidence means the gain is unexplained, isolated, or probably context-dependent.
Third, spend five minutes rejecting at least one candidate. This is required. The purpose is to train honesty. If all three look tempting, reject the one with the weakest channel support or the poorest connection to the next section. A best possible lap you can drive is built as much by leaving bad fragments out as by collecting good ones.
Fourth, spend five minutes writing the next-session objective. Choose one candidate as the primary objective and one channel as the monitor. The objective must be phrased as a cockpit action. Examples: clean up the throttle application after apex, shorten the long brake tail, remove the coast before throttle, or verify line before deleting a fast-corner lift. The monitor might be throttle shape, brake pressure shape, steering angle, lateral g consistency, or exit speed.
The count is three candidates, one rejected candidate, one primary objective, and one monitor channel. The duration is about thirty-five minutes. The success criterion is that after the next session you can answer three questions from the data: did the target behavior change, did the target section improve, and did the following section avoid an obvious penalty? If you can answer those, the drill worked even before you set a new personal best.
When the principle breaks down
This method breaks down when the bond between the data and the driving action is too weak. If you only have lap time and no speed or g, you do not have enough to build a responsible best possible lap. If you have speed and g but no throttle, brake, or steering, you can still form hypotheses, but you should be careful about claiming causes. If the tool has known limitations, respect them.
It also breaks down when comparison context dominates the section. Traffic is the obvious example from the corpus. A lift or brake touch may have nothing to do with your normal corner technique. The same caution applies when comparing other drivers, other cars, or other sessions. The corpus allows those comparisons, but it also puts compare inside a larger process of asking why, checking channels, and calibrating the finding to your driving.
Finally, it breaks down when the assembled lap cannot be imagined as a driving sequence. If you cannot picture the approach, input, release, balance, and exit, you have not finished the work. The fix is to go back from the theoretical number to the traces, from the traces to the driver action, and from the action to one objective you can actually attempt.
Author Review
No quiz questions are attached to this lesson.
Sources
| # | Document | Chunk | Pages | Score | Collection |
|---|---|---|---|---|---|
| 1 | Data for Drivers | cabda699642b26311b0a7ef998da2c71 | 15 | 1 | uio_books_raw_v1 |
| 2 | Data-for-Drivers-PRINT | c76265b3-e38e-d9dd-45b1-10446beb0a9b | 11 | 1 | uio_books_raw_v1 |
| 3 | Data-for-Drivers-PRINT | bbb02386-778f-20ec-ad16-b9c016921743 | 16 | 1 | uio_books_raw_v1 |
| 4 | Data-for-Drivers-PRINT | c493f39d-9ba1-5829-3168-d38e471cc061 | 9 | 1 | uio_books_raw_v1 |
| 5 | Data for Drivers | d631abbb-2f0e-2c19-352a-be07deb00c4d | 1 | 1 | uio_books_raw_v1 |
| 6 | Data-for-Drivers-PRINT | 36e668ca-5bb7-9f4d-1af8-a1eac0343197 | 1 | 1 | uio_books_raw_v1 |
| 7 | Data-for-Drivers-PRINT | a6f319de-d741-e8b4-12ab-b099ed06bbc8 | 1 | 1 | uio_books_raw_v1 |
| 8 | Data-for-Drivers-PRINT | 27a86c4f-e78b-a3ac-5011-0bb82408d23d | 1 | 1 | uio_books_raw_v1 |
| 9 | Data-for-Drivers-PRINT | 849f6d32-91c8-10c7-d758-d545a8a31713 | 1 | 1 | uio_books_raw_v1 |
| 10 | Data-for-Drivers-PRINT | 95d759bb-9d50-5ef8-90ea-2e92ab2c47b4 | 9 | 1 | uio_books_raw_v1 |
| 11 | Analysis Techniques for Racecar Data Acquisition | 41138569-fa56-a0a4-38c5-301475e4131a | 21 | 1 | uio_books_raw_v1 |
| 12 | Analysis Techniques for Racecar Data Acquisition | d0db9128-dc9a-aec3-14a8-5f101654753f | 3 | 1 | uio_books_raw_v1 |
| 13 | Analysis Techniques for Racecar Data Acquisition | 5eeea298-6191-0fb2-1054-b10fe574a804 | 2 | 1 | uio_books_raw_v1 |
| 14 | Ultimate Speed Secrets - Ross Bentley | 0237a5bd-e2d4-724e-bc2e-ba13db924f66 | 11 | 1 | uio_books_raw_v1 |
| 15 | Ultimate Speed Secrets - Ross Bentley | 47f6de8d-9d56-5b6d-547a-f1e7bb92faaf | 152 | 1 | uio_books_raw_v1 |