Driver debrief

LisboaKart, Pontinha · kart 19 · 16 October 2026, 15:20–15:55 · 45 flying laps, 30 of them alone · fourth session with GPS on the kart

Back in kart 19, your best laps matched 8 September almost to the hundredth — fastest five 43.782s against 43.787s — with a best of 43.662s, your quickest outside kart 28. Your typical lap did not stand still: once the slower karts had gone and the track was yours, laps 18–47 had a median of 44.093s against 44.264s, with a 0.247s spread, the most consistent run of that length in the log. In the same kart, two corners, S6 at 229–333 m (-0.208s) and S4 at 146–192 m (-0.136s), are quicker than on either earlier GPS day, and the time went back out a few hundredths at a time on the short straights.

A stop on lap 17 and a logger that lost count of its laps are covered in section 8, at the bottom; neither changes anything above it.

Best lap

43.662

lap 36; best outside kart 28, by 0.069s

Fastest 5 avg

43.782

8 Sep: 43.787s

Median lap, alone

44.093

8 Sep: 44.264s

Spread, alone

0.247s

over 30 laps averaging 44.116s

1 · The circuit, and what the section numbers mean

Every section number on this page refers to this map: 4 September's sections, as on the last three pages. Blue badges are corners, grey are straights, and the yellow ticks are the gates every lap is timed through. Numbering runs the way you drive it, from the start/finish marker.

Aerial imagery: Esri, Maxar, Earthstar Geographics and the GIS User Community. User visual calibration — provisional best visual fit: Esri imagery moved 3.356 m west, 1.854 m north. GPS coordinates and timing are unchanged; this is a subjective display alignment.

2 · The session as it ran

Kart 19 again, on a cloudless afternoon at 28–30 °C. At the start the track had very slow karts on it — your account; the sheet, which lists only you, does not show them — and after that you had it to yourself. Lap 1 began at the pit exit and lap 17 holds a one-minute stop, so both are left out of every pace figure; the 45 flying laps fall into a first stint in traffic and a second alone.

Stint	Sheet laps	Recording	Laps	Mean	Spread	Best	Slowest
Before the stop	2–16	`a_0007`	15	44.757s	0.528s	43.996s	45.902s
After the stop	18–47	`a_0008`	30	44.116s	0.247s	43.662s	44.664s

The comparison that is fair to 8 September — solo throughout — is the second stint, driven alone:

Measure	4 Sep, solo	8 Sep	16 Oct, alone	vs 8 Sep	vs 4 Sep solo
Best lap	43.940s	43.731s	43.662s	-0.069s	-0.278s
Fastest five	44.072s	43.787s	43.782s	-0.005s	-0.290s
Fastest ten	44.169s	43.874s	43.862s	-0.012s	-0.307s
Median lap	44.336s	44.264s	44.093s	-0.171s	-0.242s
Mean lap	44.399s	44.446s	44.116s	-0.330s	-0.282s
Spread, all laps	0.300s	0.785s	0.247s	-0.538s	-0.053s
Spread, laps under 45.5s	0.300s	0.328s	0.247s	-0.081s	-0.053s

All three in kart 19, on an empty track. 4 Sep: its last 21 laps, driven alone at 29–30 °C. 8 Sep: its 47 laps of the circuit, at 24–27 °C. 16 Oct: laps 18–47, after the slower karts had gone. Over all 45 flying laps, traffic included, the median is 44.219s.

Measured The same best laps as 8 September. Fastest five 43.782s against 43.787s, fastest ten 43.862s against 43.874s, best ten in a row 44.041s (laps 30–39) against 44.055s. All of them came after the stop. The best lap, 43.662s on lap 36, is 0.069s under 8 September's 43.731s.

Measured A quicker, much steadier typical lap. Alone, the median lap was 44.093s: 0.171s under 8 September's, 0.148s under its own laps 18–47 (44.241s), and 0.242s under 4 September's solo laps. The 30 laps had a 0.247s spread against 0.917s over the same laps on 8 Sep, or 0.321s leaving out its 2 laps over 45.5s, the offs. None was slower than 44.664s — lap 18, the first one back — and the last twenty averaged 44.067s with a 0.162s spread. No earlier session has thirty consecutive laps this tight; 28 August's best thirty, at 0.290 s, are the nearest.

Conditions The heat moved little. Air 29.9 °C at 15:00 falling to 28.3 °C at 16:00, cloud 0%, humidity 38–40%, wind 13–17 km/h from the north gusting 33–41, nine dry days. That is 4 September's heat with full sun on the track, and about three degrees warmer than 8 September. The three kart-19 days span that range and their solo median laps lie within 0.24s of each other; the two hot days are the slowest and the quickest of the three, so temperature does not order them. The weather is a model grid point a few kilometres away, and air temperature is not track temperature.

Pending driver input What you were working on is not recorded.

3 · Kart 19 over time

Every session in the log, from the printed sheets: the fastest five laps, which an out-lap, a stop or traffic barely moves.

Session	Kart	Best	Fastest five	Fastest ten
11 Aug	19	45.644s	45.941s	46.241s
14 Aug	19	45.043s	45.128s	45.249s
18 Aug	19	45.046s	45.125s	45.190s
20 Aug	19	44.725s	45.094s	45.169s
28 Aug	21	44.885s	44.977s	45.030s
4 Sep	19	43.873s	43.980s	44.073s
8 Sep	19	43.731s	43.787s	43.874s
7 Oct	28	43.239s	43.286s	43.337s
16 Oct	19	43.662s	43.782s	43.862s

Measured In kart 19 the best laps have levelled off. Its fastest five fell from 45.941s on 11 Aug to 43.980s on 4 Sep, a further 0.193s to 8 Sep, and 0.005s since. Kart 28 on 7 Oct (43.286s) is the one quicker session; by your account that is the kart, and it is left out of the comparisons below.

Supported interpretation A plateau at the front, not underneath it: section 2's typical solo lap is quicker, and the next section shows the lap being driven differently.

4 · A different lap in the same kart

Each of the 13 sections, with every lap slid into 4 September's positional frame first and timed through the same gates (section 8 has the method): this session's 30 laps alone against 8 September's 47 and 4 September's 21 solo laps, all in kart 19. Down and blue is quicker.

Sec	Type	Distance	4 Sep solo	8 Sep	16 Oct alone	vs 8 Sep	vs 4 Sep	Slowest Δkm/h	Fastest Δkm/h	Route Δ
S1	straight	0–44 m	2.537s	2.554s	2.588s	+0.033s	+0.051s	-0.4	-0.9	+0.3 m
S2	corner	44–104 m	5.171s	5.153s	5.081s	-0.073s	-0.090s	+0.3	-0.2	-0.4 m
S3	straight	104–146 m	2.840s	2.797s	2.853s	+0.057s	+0.013s	-1.0	-0.3	-0.0 m
S4	corner	146–192 m	2.886s	2.889s	2.753s	-0.136s	-0.134s	+2.8	+0.5	-0.1 m
S5	straight	192–229 m	2.293s	2.294s	2.226s	-0.068s	-0.067s	+0.3	+1.9	+0.1 m
S6	corner	229–333 m	8.790s	8.793s	8.584s	-0.208s	-0.206s	+1.1	+0.4	+1.6 m
S7	straight	333–364 m	2.288s	2.276s	2.332s	+0.056s	+0.044s	-0.5	-1.2	-0.1 m
S8	corner	364–415 m	3.276s	3.257s	3.252s	-0.004s	-0.023s	+0.4	-1.4	-0.1 m
S9	straight	415–463 m	2.929s	2.902s	2.971s	+0.069s	+0.043s	-0.4	-1.1	+0.1 m
S10	corner	463–497 m	2.020s	1.976s	2.012s	+0.036s	-0.008s	-0.4	-0.7	+0.1 m
S11	straight	497–572 m	4.357s	4.341s	4.305s	-0.037s	-0.052s	+3.4	+0.1	+0.1 m
S12	corner	572–610 m	2.816s	2.839s	2.885s	+0.046s	+0.069s	+0.6	+3.2	+0.4 m
S13	straight	610–641 m	2.135s	2.143s	2.212s	+0.069s	+0.077s	-1.8	-1.2	+0.1 m

Section medians. Speed and route changes are 16 Oct alone against 8 Sep.

Measured Two corners are quicker than on either earlier GPS day in this kart. S6, the long complex at 229–333 m: 8.793s on 8 Sep and 8.790s on 4 Sep's solo laps, 8.584s now. S4 at 146–192 m: 2.889s and 2.886s, 2.753s now. The slowest point of each is 1.1 and 2.8 km/h faster than on 8 Sep. Together they are 0.345s quicker.

Measured Some of it went back out, a few hundredths at a time. 7 sections are slower than on 8 Sep, none by more than 0.069s, 0.366s between them. 5 of them are the short straights, S1, S3, S7, S9 and S13: 0.033–0.069s each and 0.284s together. On every one of them the kart arrived 0.4–1.8 km/h slower and topped out 0.3–1.2 km/h lower, and the same straights are 0.228s slower than on 4 Sep's equally hot solo laps. The long straight, S11, is not among them (-0.037s): top speed 68.0 km/h against 67.9. The sections add up to -0.161s against 8 Sep; the median lap through the gates is 44.084s against 44.280s.

Supported interpretation Faster in the middle of two corners, slower out of most of them. A tired engine would lose most where the kart accelerates longest, on S11, and S11's top speed has not moved; what has moved is the speed at which every short straight begins. That points at the corner exits before them rather than at the kart. With no RPM, throttle or brake channel this is a reading of speed traces, not a measurement of what your feet did.

5 · S6 alone, and in traffic

S6, the complex at 229–333 m, is the biggest change on the lap and still the least settled part of it. Alone, its 30 passes split clearly by route.

Measured Alone, the shorter way through S6 is the quicker way. Route length and time through it go together across the 30 solo laps (correlation 0.81). The quickest quarter of passes, 7 laps, took 8.460s on a 103.1 m route; the slowest quarter 8.751s on 105.4 m, 2.4 m further, and they kept 39.8 km/h at the slowest point against 37.1: a shorter way, taken faster. They are 0.17s up by 262 m, 33 m into the corner, give 0.05s back through its middle, and take another 0.15s over its last 31 m. None of it comes back after: S7, the straight out, takes 2.315s after the quick passes against 2.353s.

The first stint, among the slower karts: its 6 laps over 45 s against its other 9, section by section.

Lap	Time	vs the stint's other laps	Where it went
2	45.902s	+1.473s	S12 +0.47s and S11 +0.37s
3	45.053s	+0.624s	S6 +0.23s and S4 +0.13s
5	45.219s	+0.790s	S4 +0.41s and S2 +0.39s
7	45.452s	+1.023s	S6 +0.30s and S4 +0.29s
11	45.246s	+0.817s	S6 +0.62s and S2 +0.13s
13	45.069s	+0.640s	S6 +0.35s and S11 +0.13s

“Where it went”: the two sections that lap lost most in, against the median of the stint's other laps.

Measured In traffic, S6 is where the time went. The slow laps lost 0.263s there at the median, 29% of their deficit and more than twice any other section, taking its slowest point at 36.5 km/h against 38.5. The losses move around the lap from one slow lap to the next, as being held up would make them.

Supported interpretation The long complex is where a slower kart holds you up longest and where passing is hardest, which fits the first stint's slow laps. Nothing else on this page uses them.

Two corners drawn, each the same way: blue is the quicker way through, orange the slower, and the pale dashed line, where there is one, 8 September for reference. Beside each map, the time gap shows where the quicker line gains, and the speed and line traces show how; under it, a short list says why, stretch by stretch.

S6 · 229–333 m · corner quicker by 0.291s

Quicker — 16 Oct alone, quickest quarter: 8.460sSlower — 16 Oct alone, slowest quarter: 8.751s8 Sep, all laps: 8.793s

Overall: 0.291s quicker, on a route 2.4 m shorter (105.4 → 103.1 m), with the slowest point 2.7 km/h faster (37.1 → 39.8 km/h).
entry → A (0–34 m): gains 0.17s. The quicker line is 1.3 m shorter, 1.2 km/h faster on average and up to 0.9 m tighter at 10 m.
A → B (34–72 m): gives back 0.05s. The quicker line is 0.3 m shorter and 0.6 km/h slower on average.
B → exit (72–104 m): gains 0.17s. The quicker line is 0.7 m shorter, 1.1 km/h faster on average and up to 1.4 m tighter at 84 m.

Each line is the median of its set at every 2 m of the 4 September lap — 7 and 7 laps — after the slide into one frame. The time gap falls where the quicker line pulls ahead; circled letters mark where it turns, at the same places on the map and the charts, and the numbers along the map's edge are metres into the section. The line chart is blank where the track runs straight and neither side is the inside.

S4 · 146–192 m · corner quicker by 0.136s

Quicker — 16 Oct alone: 2.753sSlower — 8 Sep, all laps: 2.889s

Overall: 0.136s quicker, on the same route, with the slowest point 2.8 km/h faster (55.2 → 58.0 km/h).
entry → exit (0–46 m): gains 0.14s. The quicker line is the same distance and 2.1 km/h faster on average.

Each line is the median of its set at every 2 m of the 4 September lap — 30 and 47 laps — after the slide into one frame. The time gap falls where the quicker line pulls ahead; circled letters mark where it turns, at the same places on the map and the charts, and the numbers along the map's edge are metres into the section. The line chart is blank where the track runs straight and neither side is the inside.

6 · Where the remaining time is

The 30 laps alone: the median through each section against the fastest pass recorded there.

Measured The median-to-best gaps sum to 0.93s. The largest are S2 (0.160s), S6 (0.159s) and S12 (0.127s), holding 48% of the total. The fastest section times add up to 43.124s against an actual best circuit of 43.651s. Each “best” is the quickest of 30 passes that each carry a few hundredths of slide error, so these rank the sections; they do not promise the time.

7 · Findings, and the next session

The same best laps as 8 September, in the same kart. Fastest five 43.782s against 43.787s; best 43.662s, your quickest outside kart 28 by 0.069s.
Alone, a quicker and far steadier typical lap. Median 44.093s against 44.264s; 30 laps with a 0.247s spread, the most consistent run of that length in the log.
The lap changed underneath. S6 at 229–333 m (-0.208s) and S4 at 146–192 m (-0.136s) are quicker than on either earlier GPS day in this kart; S1, S3, S7, S9 and S13, the short straights, gave 0.284s back, each starting slower.
S6 is still the least settled corner. Alone, its quickest passes are its shortest; in traffic, it is where the slow laps lost most.

One experiment for the next session in kart 19: S6, the complex at 229–333 m.

The change: Drive S6 the way your quickest solo passes did on every lap: the shorter line, which pays in its first 33 m and again in its last 31 m, and the extra speed at its slowest point that comes with it. Keep the exit onto S7.
Why there: It is already the biggest gain on the lap (8.793s → 8.584s), one of the two largest median-to-best gaps alone, and within it route and time go together. Your quickest quarter of solo passes took 8.460s; the median was 8.584s.
The measure: S6's median over at least 10 laps driven alone, in this page's frame: target 8.50s, from 8.584s, with S7 no slower than 2.332s. Change nothing else, and stay in kart 19.
If it fails: If S6 reaches 8.50s and the lap does not move, the exits are the limit, and the next page looks at them corner by corner.

8 · Technical notes

Measured Two recordings, every flying lap accounted for.

Recording	Sheet laps	Began	Logger clock	Began before the line	First crossing	Line crossings it counted	GPS fixes	Gates vs sheet, median / worst
`a_0007`	2–16	15:19:57	—	14 m	3.6 s, 21 km/h	1 of 17	19,265	19 / 44 ms
`a_0008`	18–47	15:33:46	restarted	74 m	8.7 s, 53 km/h	31 of 31	34,636	21 / 134 ms

The gates found every flying lap in both recordings and agree with the sheet to a median of 20 ms, 55 ms at worst apart from lap 18 (134 ms), the first after the restart. Lap 1 began with the kart crossing the line at 21 km/h out of the pits; the gates time it 50.629s against the sheet's 51.047s.

Measured The logger lost count of its laps, and was restarted on lap 17. a_0007 counted the true line once, then switched to the return stretch — 130 m into the lap, 11 m from the stored point, 9.3–9.5s after the true line — as a_0005 did on 7 October. It closed 7 laps, 6 of them on the return stretch, for the 16 driven, and none after 15:26:17, 6.5 minutes before the stop. The kart then stood still for 60s, 57 m from the stored point, and a_0008 starts on a fresh logger clock within 3 m of where the first recording ended: the logger was restarted there. Afterwards it cut every lap on the straight, a median 0.5 m from the point. Across the two recordings GPS reproduces lap 17 as 112.098s against the sheet's 112.274s. The stop cost about 68s of track time and nothing in the analysis, which never uses the logger's lap line.

Pattern, not mechanism Where a recording begins seems to decide which line it counts. a_0007 began 14 m before the line and counted the displaced one; a_0008 began 74 m before it and counted the true one. Every earlier recording fits the same split: 4 September (31 m before) and 7 October's a_0006 (34 m) counted the true line; 8 September (past the point), a_0004 (10 m) and a_0005 (on the return stretch) the displaced one. If the lap time on the dash matters to you, letting the logger start well back from the line may be enough; the reason is not known.

Measured The printed total is short, not the laps. The 47 transcribed laps average 45.9184s against a printed 45.918s and sum to 2158.164s; the printed total is 2132.739s, 25.425s less. GPS puts 2157.8s between the first crossing of the line and the last, so the session clock started about 25.425s into lap 1. The sheet's own stamp, track time 14:55:20, is 25 minutes before lap 1 began by GPS (15:20:01); it is not used.

Method One frame, fixed gates. Every lap is timed through 4 September's 13 section starts, a point on the ground and a direction of travel each, after its trace has been slid east and north — no rotation, no stretch — until its sideways position at the gates best matches 4 September's solo laps.

Session	Recording	Laps	Typical slide needed	Size	Range across laps
4 September	`a_0002`	46	0.0 m north, 0.3 m west	0.3 m	0.2–1.3 m
8 September	`a_0003`	47	1.9 m south, 0.3 m west	1.9 m	0.8–3.9 m
16 October	`a_0007`	15	1.1 m south, 2.2 m east	2.5 m	1.8–3.7 m
16 October	`a_0008`	30	1.6 m south, 2.2 m east	2.8 m	1.8–3.6 m

Sliding each recording as one block instead of lap by lap moves no section median by more than 0.033s, and differences below about 0.05s are treated as nothing. After sliding, the three days' typical lines sit within 0.6 m of each other at every gate, and their routes through every section agree to within 2.4 m: the resolution of anything said about lines here.

Measured Clean recordings, and still no engine data. 53,901 GPS fixes at nominal 25 Hz over 35.9 minutes and 30.8 km, none lost to the stock parser's guard. Satellites used ranged 5–16, most often 13; the receiver's accuracy estimate had a median of 81 cm (61–126 cm). RPM recorded zero throughout, for the fourth session running, and the logger has no brake or throttle channel.