Build and read a useful G-G diagram
Generated from
content/lms/data-interpretation-ii-advanced/03-friction-circle/01-gg-plot-fundamentals.md; edit the source file, not this page.
Source path: content/lms/data-interpretation-ii-advanced/03-friction-circle/01-gg-plot-fundamentals.md
Course: Read the data your hands can't feel
Module: Map the grip envelope and find the holes
Estimated duration: 55 minutes
The job of the G-G diagram
A G-G diagram is a compression tool. It takes the lateral acceleration and longitudinal acceleration channels from a lap and puts them into one picture so you can see how the car was being asked to use grip. It does not tell the whole story by itself. The data packet that supports this lesson shows speed, throttle, brake pressure, steering angle, RPM, lateral acceleration, and longitudinal acceleration together, and the point is important: the G-G picture is the beginning of a question, not the end of the answer.
For this lesson, you are not trying to trace G-sum around the lap. That is the sibling lesson. You are also not comparing grip envelopes between two drivers. That is another sibling lesson. Your job here is narrower and more useful at the intermediate level: build a clean G-G picture from one driver, one car, and one session, then read the major shapes without making claims the chart cannot support.
The core principle is simple. Lateral acceleration shows how hard the car is cornering. Longitudinal acceleration shows how hard the car is braking or accelerating. When you plot them against each other, each point is one instant in the lap. A cluster near the center means the car was not asking much from the tires. A point far from the center means the car produced more acceleration in at least one direction. A diagonal movement away from the axes means the car was combining braking or acceleration with cornering. That combined-use region is where the friction-circle idea becomes practical: you are looking at how much of the tire's available work went into turning, how much went into slowing or accelerating, and where those jobs were blended.
The first rule is restraint. A G-G diagram does not know why a point exists. It can show that longitudinal deceleration was weak, but it cannot by itself say whether you braked too softly, braked too late and released in a panic, lacked confidence, had a brake issue, entered with the wrong mental picture, or were protecting traffic. It can show that the exit acceleration shape is messy, but the chart needs the throttle trace, speed trace, steering trace, GPS line, and your own feel before the cause becomes credible. The Data for Drivers material is explicit about the method: look at the G channel, ask why, and confirm the issue with other channels when available.
That is the operating mindset for the whole lesson. You do not read a G-G diagram as a scoreboard. You read it as a map of questions.
Build the plot from channels that can answer questions
Start with a clean lap. Clean does not mean fastest. A fastest lap can hide important driving information if one part of the lap was strong and another was weak. The supplied data lesson uses that exact warning: if you only look at fastest lap, you can miss the useful information. For your first G-G diagram, choose a lap without traffic, major mistakes, obvious data dropouts, cooldown behavior, or a tire warm-up lap. You want a lap that represents the driving you are trying to understand.
Use lateral acceleration on one axis and longitudinal acceleration on the other. The exact sign direction depends on your logger and software, so label the plot before you read it. Then keep the supporting channels open beside it: GPS speed, throttle position, front brake pressure if you have it, steering angle, RPM, gear, and GPS line. If your logger can show a synchronized cursor, use it. The useful moment is not the static scatterplot alone; it is the ability to click or scrub through a point and see what the car and driver were doing at that instant.
Build only as much plot as you can explain. A full-lap plot is useful as a first look, but it can be too dense for learning. After the first view, isolate one corner or one repeated type of corner. If your software lets you select by distance, highlight the braking zone, turn-in, apex region, and exit separately. If it only gives you full-lap scatter, use the cursor and supporting traces to identify where major groups of points came from. A plot you can phase-label is worth more than a pretty full-lap blob you cannot explain.
Keep the lap distance or time trace visible. The bonded data examples show speed, throttle, brake pressure, steering, RPM, lateral acceleration, and longitudinal acceleration over a common distance axis. That common axis matters because the G-G plot removes location. Once location is removed, you must add it back through synchronized traces. Otherwise you may confuse a good braking point from one corner with a weak braking point from another corner, or a messy exit in a slow corner with normal behavior in a fast transition.
Do not start by hunting for the largest number. Start by identifying the phases: straight-line braking, brake release and turn-in, mid-corner, throttle pickup, and exit. Bentley's cornering material separates begin-of-braking, end-of-braking, and begin acceleration as meaningful points. Performance Driving Illustrated also frames cornering around turn-in, apex, and exit, with the goal of straightening the corner enough to accelerate early. Your G-G plot becomes teachable when each part of the scatter can be connected to one of those phases.
Read the main regions before reading the extremes
The center of the plot is low demand. You will see center points on straights, during gentle transitions, and during moments when you are neither braking hard, accelerating hard, nor cornering hard. Do not obsess over the center unless there is too much of it in a place that should be a committed corner phase. A long, lazy drift through the center during a braking zone may mean you did not brake with much force. A long coast through the center between brake release and throttle may mean you gave the tires neither a clear braking job nor a clear cornering job. But again, the plot only starts the question. Confirm it with speed slope, brake pressure, throttle position, and steering.
The vertical or longitudinal ends of the plot show your strongest braking and acceleration events. The Data for Drivers notes ask you to look at peak G-loads, whether you are using them consistently, and whether there are braking or acceleration issues. That is your first reading pass. Are the braking points repeatedly reaching a similar deceleration range, or does one corner have a much weaker braking signature? Are acceleration points strong and smooth on exits, or do they appear broken up by throttle changes? The answer is not automatically good or bad. A low deceleration peak might be correct in a short brake zone, wrong in a major brake zone, or simply the result of a corner that should not be braked hard. That is why you go back to location and phase.
The lateral sides of the plot show cornering load. A strong lateral point means the car produced more cornering acceleration at that instant. That does not automatically mean the corner was driven well. You can produce a high lateral value in a way that compromises exit speed. You can also drive a lower lateral peak in a way that sets the car up for earlier throttle and better speed down the following straight. Performance Driving Illustrated emphasizes driving the corner so you can begin accelerating early, and the page on braking and acceleration phases shows why the begin-acceleration point matters more than many drivers admit. On the G-G diagram, high lateral numbers are interesting, but the exit and speed trace decide whether they helped.
The diagonal regions are where the lesson becomes most valuable. When braking and turning overlap, the points move away from pure braking toward a combined braking-and-cornering region. When accelerating and unwinding overlap, the points move toward a combined acceleration-and-cornering region. The Data for Drivers material gives a direct example of using a speed trace to suspect trail braking, then using brake pressure to confirm it. The G-G diagram helps you see that same idea as a shape: a driver who truly blends brake release into corner entry should not have only straight-line braking points and then a disconnected lateral-only cornering patch. There should be a transition between them. But you still need the brake trace to prove what your foot did.
The outer shape is not a magic tire limit. It is the outer shape of what happened in that run. Maybe it represents the tires near their current grip. Maybe it represents your confidence limit. Maybe it represents an input habit that kept the car from using more. Maybe it represents one corner type but not another. The data method is to ask why, compare consistency, and confirm with channels. Treat the outer shape as evidence of produced acceleration, not as an official declaration of available grip.
The five sub-skills
The first sub-skill is phase labeling. Before you make a claim, name the driving phase that produced the point you are reading. Was it begin-of-braking, end-of-braking, turn-in, mid-corner, begin acceleration, or exit? Without that label, you are reading dots without context. With that label, the same dot becomes a coaching question.
The second sub-skill is channel confirmation. The G-G plot tells you what acceleration the car produced. Inputs tell you what you asked for. Speed tells you the result of that request. Steering tells you how much hand input was involved. Throttle and brake pressure tell you whether the longitudinal part of the point came from the pedals. RPM and gear help you understand whether the engine was in a useful range. GPS line shows whether the car was positioned to make the corner easier. The Data for Drivers channel list is long for a reason: each channel can prevent a false conclusion.
The third sub-skill is consistency reading. Do not ask only whether you touched a high value once. Ask whether similar corners produce similar patterns, whether braking peaks repeat, whether acceleration shape repeats, and whether a lap-to-lap change is real. The supplied material specifically points at consistency in the longitudinal G trace and at comparisons with other laps, sessions, cars, and drivers. In this lesson, keep the comparison mostly inside your own session. Before you compare yourself to another driver, learn to read your own repeatability.
The fourth sub-skill is sensory calibration. The driver is a sensor. Bentley's material says you gather information through vision, kinesthetic feel, and hearing. Ultimate Speed Secrets pushes the same direction by asking you to improve speed sensing and traction sensing through vision, balance, touch, G-forces, vibration, pitch, roll, and hearing. When you review the G-G plot, ask what you felt at the same moment. Did the tire sound build and then taper? Did the car feel loaded and settled, or abrupt and overloaded? Did the steering feel like it took a set, or did you add hand after hand? The purpose is not to replace the data with feel. The purpose is to make your feel more accurate next session.
The fifth sub-skill is restraint under uncertainty. If the plot says the exit acceleration pattern is messy and the throttle trace shows corrections, the data packet itself suggests possible causes: line, vision, mental image of track-out, too much mid-corner speed, or throttle use. Notice how many possible causes remain. A good reader does not jump straight to the most dramatic answer. You pick the most testable cause, go back to the next session with one change in driving, and see whether the shape and speed trace improve.
What good reading sounds like
A weak reading sounds like this: my max lateral G is low, so I need more grip. That may be true, but the chart has not proven it. A useful reading sounds more like this: in this corner, the car never built much lateral acceleration, the steering trace shows I turned in slowly and kept adding steering, the speed trace shows I slowed more than needed, and my notes say I was looking at the apex instead of through the exit. That is a driver-technique hypothesis you can test.
A weak reading says: I am not braking hard enough. A useful reading says: in the main braking zone, longitudinal G peaks are lower than my other clean laps, the speed trace deceleration slope is shallower, and brake pressure confirms I did not apply the pedal as firmly. That is now a braking execution issue, not a vague feeling.
A weak reading says: I need to trail brake more. A useful reading says: the speed trace shows a change in deceleration slope near corner entry, the G-G plot shows a transition from braking toward lateral load, and the brake pressure trace confirms whether pressure was actually still present. If the brake trace does not confirm it, you do not claim trail braking just because the G-G plot looks diagonal.
A weak reading says: this exit is bad. A useful reading says: exiting onto the front straight, the speed trace changes, the throttle trace shows adjustments, and the G-G plot does not show a clean move from lateral load toward acceleration. Now the question becomes whether the line, vision, track-out picture, mid-corner speed, or throttle timing caused the interruption.
The useful reading is longer because it names evidence. It connects G to inputs, inputs to car response, and car response to the corner phase.
Calibration cues on track
You know this skill is improving when the plot stops surprising you. After a session, you should be able to predict the major G-G shapes before opening the software. You should know which corner had the strongest brake event, which exit had throttle hesitation, and which corner felt like you were asking the front tires for too much steering. Then the data either confirms your feel or corrects it.
You also know you are improving when your questions become smaller. At first you may ask whether you used the tires well. That is too broad. Later you ask whether your brake release into one corner produced a continuous transition, whether your exit throttle was interrupted, or whether your steering input was larger than it needed to be for the lateral acceleration achieved. Smaller questions are easier to test.
Your instructor would notice the same trend from the right seat. They would hear less vague language in your debrief. Instead of saying the car felt weird, you would say the car felt loaded during the first half of the corner, then I had to pause the throttle before track-out. Instead of saying you need more confidence, you would say the brake trace may show I released too early and coasted to the apex. That is how a G-G diagram changes your driving: it makes your next on-track question precise.
Look for a better relationship between your hands and the lateral side of the plot. Bentley's cornering chapter warns that you can slow your steering inputs without slowing the car, and that less steering wheel generally supports speed. If you repeatedly see big steering input without the cornering result you expected, the next practice target may be vision, line, speed selection, or smoother hands. The G-G plot shows the car's acceleration response; steering shows what you asked with your hands.
Look for a cleaner relationship between begin acceleration and the exit shape. Performance Driving Illustrated emphasizes shaping the corner so you can accelerate early. If the plot and throttle trace show that exit acceleration is repeatedly delayed or interrupted, do not chase the highest mid-corner lateral number first. Ask what would let you begin acceleration from a higher speed and carry it cleanly.
Failure modes and recovery
The first failure mode is treating the G-G plot as a verdict on the car. The chart may show under-use, but it does not explain whether the limitation came from driver inputs, line choice, vision, mental image, brake release, throttle timing, or mechanical grip. Recovery is to write a one-sentence hypothesis and require at least two confirming channels before you believe it.
The second failure mode is chasing peak numbers. A single high lateral point can belong to a corner that still loses time because it delays throttle. A single high braking point can belong to a brake zone that unsettles the car or forces a poor release. Recovery is to read the phase after the peak. What happened to speed, steering, throttle, and begin acceleration afterward?
The third failure mode is reading the plot without your body. Bentley's sensory material matters because the driver must eventually act before the data screen exists. If the G-G plot says the tire was near a limit but you cannot remember what the tire sounded like, what the wheel felt like, or where you were looking, the learning loop is incomplete. Recovery is to add one sensory note immediately after the session for the corner you will review.
The fourth failure mode is reading the plot without the track. A G-G diagram removes location. That is useful for seeing the acceleration envelope, but dangerous if you forget that different corners demand different jobs. Recovery is to keep the GPS line or distance trace open and phase-label every claim.
The fifth failure mode is using the chart to justify more aggression. More G is not automatically better driving. Better driving is the right grip use at the right phase so the car is positioned to brake, turn, and accelerate effectively. Recovery is to connect every proposed change to a phase outcome: stronger initial brake, smoother release, more stable mid-corner, earlier throttle, cleaner exit, or less steering for the same result.
How this lesson connects to the rest of the module
Once you can build and read a basic G-G diagram, the next step is to follow total grip use around the lap. That is where the G-sum sibling lesson belongs. G-sum can help you see where the car is close to or far from the combined acceleration it produced elsewhere, but it is easier to misuse if you cannot already explain the lateral and longitudinal components.
After that, comparing envelopes between drivers becomes useful. But do not skip the single-driver skill. If you cannot explain your own plot with your own speed, brake, throttle, steering, line, and feel, comparing your shape to another driver may only make you copy symptoms instead of understanding causes.
The G-G diagram is a teaching picture. Build it cleanly, phase-label it, ask why, confirm with other channels, and take one testable driving question back to the car.
Worked example: the front-straight exit that will not clean up
Use the exit-onto-front-straight situation from the supplied data material as the model. You come off the corner and the speed trace should begin climbing cleanly. Instead, the speed line changes shape and the throttle trace shows adjustments. On the G-G diagram, you would expect a clean exit to move from lateral load toward acceleration as the wheel unwinds and the throttle comes in. If the exit cloud looks broken or hesitant, do not declare a power problem first. The bonded material points to the right question set: line, vision, mental image of track-out, too much mid-corner speed, or throttle timing.
Here is the reading process. First, locate the exit on the distance trace. Second, mark the moment where you expected begin acceleration. Third, look at the throttle trace. If the throttle rises, pauses, backs up, or jitters, the G-G plot is showing the result of a driver request that was not committed. Fourth, look at steering. If steering is still large when throttle is requested, you may be asking the tires to keep cornering while also accelerating. Fifth, compare speed. If you carried too much mid-corner speed but delayed throttle, the high middle of the corner may have cost the straight.
Your next-session test is not to mash the throttle earlier everywhere. It is to choose the same corner, look farther through track-out, slightly prioritize the exit shape, and ask whether the throttle trace becomes cleaner and the G-G movement from lateral load toward acceleration becomes less interrupted. Success is not a prettier chart alone. Success is cleaner throttle, better speed growth, and a feel note that says the car was pointed before you asked it to go.
Worked example: suspected trail braking in one corner
The supplied Data for Drivers material describes a speed-trace clue: a change in the slope of deceleration can make you suspect trail braking, and brake pressure can confirm whether it was really there. This is exactly the kind of job a G-G diagram can help with, as long as you do not let it overclaim.
Start at the braking zone. In straight-line braking, the plot should show mostly longitudinal deceleration. As you approach turn-in and release the brake, the point path should move away from pure braking and toward lateral acceleration. If the brake is still present while the car is taking lateral load, that is a combined-use phase. The G-G picture can show the blend, but it cannot prove the pedal trace by itself.
Now confirm. Open brake pressure. If pressure remains as steering and lateral acceleration build, you have evidence of trail braking. If the brake trace is already zero while the G-G plot moves diagonally, the diagonal shape may be coming from timing, filtering, track shape, data sync, or another effect, and you should not claim trail braking from the G-G chart alone. Then open speed. If the speed trace shows a smooth transition in deceleration and the car reaches the apex without a coast gap, the entry may be disciplined. If speed falls, flattens, and then waits before throttle, the problem may be release timing or entry speed, not lack of bravery.
The intermediate lesson is this: use the G-G plot to notice blending, use brake pressure to prove the pedal, and use speed to judge whether the blend helped the corner.
Common mistakes
Mistake one is reading the biggest dot as the best driving. Good looks like reading what happened after the dot. If the highest lateral point delays begin acceleration, it may not be the fastest way through the corner.
Mistake two is diagnosing the car from the G-G plot alone. Good looks like pairing every G-G claim with at least one driver-input channel and one result channel. For example, brake pressure plus speed, or throttle plus speed, or steering plus lateral acceleration.
Mistake three is forgetting that the plot removed location. Good looks like keeping distance, GPS line, or a synchronized cursor open so each group of points belongs to a named corner phase.
Mistake four is confusing suspected trail braking with confirmed trail braking. Good looks like using the speed trace to notice the possibility, the G-G plot to see the blend, and brake pressure to confirm whether the pedal was actually present.
Mistake five is using the plot to chase aggression. Good looks like chasing a cleaner phase: a more decisive brake application, a smoother brake release, less unnecessary steering, earlier and cleaner begin acceleration, or fewer throttle adjustments on exit.
Mistake six is ignoring your senses. Good looks like writing one feel note for the same point you review in data: tire sound, steering load, body pressure from G-force, vibration, pitch, roll, or where your eyes were pointed.
Drill: three-session G-G readback
Run this drill over three track sessions at your next event. The count is three clean laps per session, one chosen corner per session, and one written hypothesis after each download. Do not use traffic laps, warm-up laps, cooldown laps, or laps where you had to abandon the corner.
Session one is the build session. Pick one corner with a clear braking zone and exit. After the session, build the G-G plot for one clean lap and keep speed, throttle, brake pressure if available, steering, and lateral and longitudinal acceleration open. Your only task is to phase-label the plot: braking, brake release and turn-in, mid-corner, begin acceleration, and exit. Success means you can point to the supporting traces for each phase.
Session two is the confirmation session. Use the same corner. Before opening the data, write what you think happened in one sentence. For example, you may write that you released the brake too early and coasted before the apex, or that you delayed throttle because you did not see track-out soon enough. Then open the G-G plot and supporting traces. Success means the data either confirms your sentence or forces you to rewrite it with better evidence.
Session three is the test session. Choose one driver change only. If the issue was a throttle hesitation, work on vision through track-out and a cleaner unwind before throttle. If the issue was weak braking, work on a more decisive initial brake application without changing the whole corner. If the issue was a missing entry blend, work on release timing rather than simply entering faster. Success means the G-G shape, the supporting input channel, the speed trace, and your feel note all move in the same direction.
The drill is complete when you can produce one evidence sentence after session three: in this corner, changing this one input changed this channel and produced this result. That sentence is more valuable than a screenshot of the plot.
When to stop believing the plot
Stop believing the plot as soon as you cannot connect it to phase, input, and result. If you do not know which corner created a group of points, do not diagnose it. If the throttle or brake trace contradicts your interpretation, do not force the G-G plot to win. If your notes say there was traffic, a missed shift, a cool tire, or a compromised line, do not treat the shape as representative.
Also stop before making setup claims. The bonded corpus for this lesson supports data-channel confirmation and driver-technique questions, not a setup diagnosis from G-G shape alone. If the chart suggests under-use or inconsistency, your first responsibility is to check driver inputs and repeatability. A setup conversation comes later, after the driver-side evidence is controlled.
Finally, stop if the lesson turns into a hunt for a perfect round envelope. The goal is not to make a pretty picture. The goal is to understand how the car was slowed, turned, and accelerated, then choose one better action for the next session.
Author Review
No quiz questions are attached to this lesson.
Sources
| # | Document | Chunk | Pages | Score | Collection |
|---|---|---|---|---|---|
| 1 | Data-for-Drivers-PRINT | 95d759bb-9d50-5ef8-90ea-2e92ab2c47b4 | 9 | 1 | uio_books_raw_v1 |
| 2 | Data-for-Drivers-PRINT | ba11efce-fa7b-39cf-f1bf-2bbfddb3acaa | 1 | 1 | uio_books_raw_v1 |
| 3 | Data-for-Drivers-PRINT | 8b3d290e-b0d8-f5d5-aee7-3cb2aa93adda | 5 | 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-PRINT | 849f6d32-91c8-10c7-d758-d545a8a31713 | 1 | 1 | uio_books_raw_v1 |
| 6 | Performance-Driving-Illustrated-Ross-Bentley | 8b67a9e8-4248-eb95-b368-7432c4da6d4b | 30 | 1 | uio_books_raw_v1 |
| 7 | Ultimate Speed Secrets - Ross Bentley | f00b98f3-ede1-c1d8-e803-d6c896a65ecc | 478 | 1 | uio_books_raw_v1 |
| 8 | Performance-Driving-Illustrated-Ross-Bentley | 010579d2-336f-078b-c6e0-55dc487ccbbe | 41 | 1 | uio_books_raw_v1 |
| 9 | Performance-Driving-Illustrated-Ross-Bentley | 70f49d32-218c-93c6-9585-c000a6657440 | 19 | 1 | uio_books_raw_v1 |
| 10 | Ultimate Speed Secrets - Ross Bentley | 47f6de8d-9d56-5b6d-547a-f1e7bb92faaf | 152 | 1 | uio_books_raw_v1 |
| 11 | Performance-Driving-Illustrated-Ross-Bentley | 3c5e70bc-2026-8b6d-1e97-33f866559ee6 | 7 | 1 | uio_books_raw_v1 |