Madring and Baku: The Gap Between a Bad Race and Market Value

Madring and Baku: The Gap Between a Bad Race and Market Value

I. Four minutes and seventeen seconds

4:12 a.m. Melbourne time. On the left screen, Lando Norris's telemetry trace snaps at the eleventh braking zone of a circuit I have never seen with my own eyes. On the right screen, an exchange's odds board is moving. In four minutes and seventeen seconds, the quoted price on Norris to win the Azerbaijan Grand Prix travels from 3.20 to 5.40 and settles back at 4.60.

Not a single car ran a single extra metre in those four minutes. No new aerodynamic data appeared. No technical bulletin was issued from any factory. There was only a bad race that had already happened in Madrid a few days earlier, and thousands of people simultaneously deciding that the memory of it had to be priced in.

I sat there with a pencil in hand and drew a trapezoid on the back of a supermarket receipt. The top edge was Baku's top speed. The bottom edge was the grip of a street surface. The two side edges were the distance between the Madrid result and the Baku expectation. That trapezoid, not any number on the odds board, was what I wanted to take apart that morning.

A diagram does not lie, but the person reading it does.

This piece is not a recommendation to bet on anything. I am writing it because one headline — "Norris' Madring misery creates Baku betting value" — is both a news item and a test. It tests whether a reader can distinguish a sporting event from a market signal. Most cannot. And the gap between those two things is, to me, one of the most worthwhile areas of this sport to map.

II. An uncertain name and a real circuit

Madring — the shorthand the press uses for the new Madrid circuit, host of the Spanish Grand Prix from the 2026 season — is a variable the industry has not fully digested. Even the name is contested. In many official documents it does not appear at all. Elsewhere, it is read as a blend of "Madrid" and "Ring". I record that uncertainty in the first line, because my habit is never to convert an assumption about a label into a technical fact.

What I can verify is different. The Madrid circuit sits inside the IFEMA complex, runs roughly 5.47 km with 22 corners, and has a hosting deal spanning a decade. About a fifth of the lap is genuine city street; the rest is service road reworked inside a trade-fair campus. That is a hybrid structure — neither a pure street circuit nor a dedicated permanent track. And that hybrid structure generates a very particular kind of error for markets.

Baku City Circuit is the opposite. The Azerbaijan street track has spent almost a decade on the calendar, and every season adds another layer of meaning to the name. The main straight runs around 2.2 km, the longest on the entire calendar, where a modern car reaches velocities that no corner afterwards is wide enough to forgive. The braking zone into Turn 1 is one of the most compressed energy events of the season. The castle section mid-lap is narrow enough that two cars cannot run side by side without touching mirrors. And the whole circuit sits inside a city where wind, sand and surface oil can appear at any time.

These are two circuits that differ in their essential geometry. But the headline joins them with a verb: "creates". The Madrid race creates value in Baku. That verb is the entire claim of the original article. And it is a claim about people, not machines.

III. "Misery" is a word that does not diagnose

The English word "misery" in the headline is a notable editorial choice. It describes an emotion, not a cause. In the language of the paddock, there is a very large gap between the four situations that "misery" can cover.

The first is mechanical failure. A car stops at the side of the road because a gearbox or a cooling system has let go. In that case, the information needed to price Baku is a list of consumed components and the remaining margin under the cost cap — none of which anyone publishes.

The second is a setup error. The car runs, finishes, but sits in the wrong tyre-temperature window. This is simultaneously the most worrying and the most encouraging failure, because it can be fixed with a turn of a suspension tool, and its appearance in Madrid says nothing about whether it will appear in Baku.

The third is a strategy error. The team mistimes a pit stop or reacts slowly to a virtual safety car. This category is transferable — meaning it can repeat in Baku, where the safety-car probability runs above the season average.

The fourth is a pure driver error. Braking late, locking a wheel, running wide.

These four situations lead to four completely different conclusions about Baku. A headline that says only "misery" throws all four into the same bag. The reader of the odds board receives no information with which to tell them apart. They receive an emotion.

By my experience, emotion is the fastest-moving thing on an odds board. On a tactical map, emotion is the coordinate people forget to plot. But on a market map, emotion is the horizontal axis.

IV. The geometry of Baku: a three-layer polygon

When I redraw Baku on paper, it does not come out as a circle. It comes out as three stacked layers, each with its own mechanism, each with its own way of lying.

The first layer is the main straight. Around 2.2 km, starting at the final corner and ending at the braking zone into Turn 1. This is the layer of velocity. Here, differences in engine performance are magnified. A car with a power advantage or an aerodynamic efficiency advantage in low-downforce trim gains two to three tenths on this section alone. The key point: that advantage does not depend on whether the driver had a bad race in Madrid. It depends on the power unit's deployment map and the chief engineer's wing-level decision.

The second layer is the castle section. This is the layer of fear. Two corners less than eight metres wide, stone walls on both sides, and a surface that can turn slick as wet rock as temperatures rise. No overtaking is possible. There is no escape route. Here, the difference between a confident driver and an anxious one shows directly in lap time — but in a way telemetry does not fully capture. Braking loads in the castle are far lower than at Turn 1, yet I have seen intra-team gaps reach three tenths here. Those three tenths do not come from the car.

The third layer is the run from roughly Turn 16 onward — the technical zone, where the rear wing has to work continuously and the rear tyres have to endure high surface temperatures. This layer punishes teams that run a setup that is too stiff.

Together, the three layers form a paradox: Baku is a circuit where victory can come from the straight and defeat can come from the castle. Neither has anything to do with how a driver suffered at a different circuit.

V. The geometry of Madring and the transfer question

This is where I must be extremely careful, because this is where most commentary makes its error.

Madrid is new. Baku is old. When a car runs badly in Madrid, the natural reflex is to assume it will run badly in Baku. That reflex is correct in a world where every circuit demands the same kind of performance. It is wrong in the real world, where every circuit is a specific combination of corner radii, camber, surface roughness, ambient temperature and wing architecture.

I call this the geometric transfer problem. Mathematically, it is the problem of projecting a vector from a high-dimensional space onto a two-dimensional plane. You have a car whose model sits at point A on a performance plane. Madrid measures it along one direction. Baku measures it along another. No theorem says a failure in one direction implies a failure in the other, and none says the converse either.

The only thing a Madrid result tells you with certainty is this: on that Sunday in Madrid, the car-human-circuit combination operated below expectation, in a manner that has not been identified.

That phrase — "has not been identified" — is the entire problem. Because you do not know the cause, you know nothing about the probability of repetition. And when you know nothing about the probability of repetition, the only rational response is to hold your fundamental assessment fixed — which means staying out of the odds board.

Markets do not behave like a rational analyst. Markets behave like a crowd with a short memory and a strong aversion to loss. One bad race in front of your eyes overwrites five good ones before it. This bias has a name: availability bias. And it is precisely what the headline calls value.

VI. The mechanism of value: the expectation gap

I once wrote that transfers are not dry arithmetic but alchemy. The same holds for pricing. An odds board does not reflect true probability. It reflects the weighted average of the beliefs currently held in the heads of people with money to place. Those beliefs are fed by three sources: data, memory and story.

When a race ends badly for a big name, the last two sources take control. Memory becomes more vivid than data. Story becomes more compelling than a timing sheet. The result is a gap that opens between the quoted price and the fundamental assessment. People call that gap "value".

Mechanically, that is the entirety of what the headline means. It does not say Norris will win. It does not say his team has found a technical fix. It says only that the price is lower than it should be, according to some fundamental assessment the original article presumably lays out.

I say "presumably" because here I must confess something about my own method. When I harvested this headline, I did not capture the body. I have one line of text and seventeen navigation elements from a website. That means I am analysing a shadow on a wall. And if I wrote a piece about it without admitting that, I would be betraying my own rule of quantitative humility.

The pandemic taught me one thing: the silence of data also speaks. The absence of a body text is not a neutral blank. It is a datum. It tells me that the original article's genre is market commentary, not technical reporting. And that changes entirely how I should read it.

VII. A shock involving an eliminated German side

Let me tell an old story to make the point about genre.

In 2026 I wrote a piece on Germany against South Korea at the World Cup in Russia. Germany touched the ball more than 680 times, held over 70 per cent possession, and lost 0-2. South Korea used a truncated trapezoid pressing trap in midfield, forced Germany to circulate the ball in harmless directions, and counter-attacked at exactly two moments. My figures: only about 47 of Germany's possessions reached the final third in the second half. The piece drew 120,000 reads — thirty times my previous best.

I learned two things. First, geometry beats abstraction. Readers do not need an equation. They need a shape they can see. Second, genre determines the kind of truth you find. Had I written that piece as a results report, I would never have seen the truncated trapezoid.

With the Norris-and-Baku headline, the original genre is market commentary. The truth it seeks is a truth about pricing, not a truth about aerodynamics. A reader who mistakes the genre will draw a technical conclusion from a market premise. That is the most common editorial error of them all.

VIII. Baku's historical data and so-called structural volatility

Now let me talk about what I can actually measure.

Baku is a circuit with an above-average probability of safety cars and virtual safety cars. This is not an opinion. It is a consequence derivable from geometry. A narrow street track with walls on both sides, an extremely heavy braking zone at Turn 1, a tight blind section mid-lap, and a long straight producing high approach speeds — that combination raises the probability of an accident. Accidents produce debris. Debris produces neutralisation. Neutralisation produces variance.

That structural variance has a very specific consequence for the Norris story. In an ordinary race, the performance gap between teams expresses itself as a finishing-time gap. In a race with three neutralisations, that gap is compressed. Strategy becomes more important than raw pace. And luck becomes an independent variable with large variance.

In other words: Baku is the kind of circuit where a team in trouble can finish above its true level, and a team in full flight can be sunk by a piece of carbon fibre. That makes Baku attractive to those who believe the market has overreacted. But it also makes Baku the most dangerous circuit on which to draw conclusions about quality.

This is where most market commentary shoots itself in the foot. It says: "Baku is volatile, so bet on the underdog." But if Baku is volatile, then the outcome itself is volatile, and the expected value of a bet on the underdog does not automatically rise. Variance is not value. Variance is only the space in which value can exist.

IX. Tyres, temperature and the surface problem

There is a technical dimension of Baku I want to draw as its own diagram, because it is one of the things that decides races and is one of the least discussed.

Baku is a street track. Street asphalt has lower roughness than permanent-circuit asphalt. Lower roughness means less friction generated at the same load, which means tyres warm more slowly on the opening lap and cool faster when the car slows. This creates two opposing effects fighting continuously across a lap.

The first effect shows at the start and on pit exit. Before the tyre reaches its working temperature, a car locks a wheel in the Turn 1 braking zone. A lock creates a flat spot. A flat spot creates vibration. Vibration destroys the performance of the entire stint. A single mistake on a pit-exit lap can ruin the fifteen laps that follow.

The second effect shows in the castle. As the car slows, air is not pushed through the tyre quickly enough, surface temperature drops, grip drops, and the driver must slow further. This is a negative feedback loop. It explains why the gaps between drivers in the castle are rarely stable from one lap to the next.

For McLaren, a team that in recent seasons has been known for its tyre-temperature management, Baku is a fascinating test in both directions. If the team finds the optimal temperature window, it can achieve above-average performance. If not, it can fall back at the start of every stint.

That is the kind of analysis a headline about "value" never supplies, because it requires long-run data, and long-run data only exists after practice sessions.

X. Pit stops and the geometry of a time polygon

A sequence of pit stops, drawn on paper, is a polygon of time. The x-axis is the lap. The y-axis is the time cost. Each stop is a sharp vertex. The height of the vertex depends on the stop time, the pit-lane transit, and the loss from cold tyres.

At Baku, the total loss of a pit stop lands around twenty seconds, plus roughly one to two seconds for cold tyres early in the stint. On a circuit with a high neutralisation probability, the pit window blurs compared with an ordinary circuit. That creates two opposing strategies.

The first is to stop early, accepting lost time in the middle phase to gain better tyres in the final phase, insuring against a safety car. The second is to stop late, extending the first stint to exploit tyres already at temperature and to avoid running cold rubber in traffic.

Both strategies' win probability depends on a variable nobody controls: the timing of neutralisation. Statistically, this is an enormous confounding variable. And any analysis that tries to infer car quality from a Baku race without isolating that confounder is an analysis fooling itself.

XI. The relationship between result and expectation: a small model

I like to build miniature models to test whether an argument stands. Here is one for this headline.

Suppose the market's expectation for a driver at a race equals a weighted average of two components: a fundamental assessment of car capability, and the most recent result. The weights depend on how fresh the most recent result is, and freshness is an increasing function of how emotionally intense that result was.

When a result carries high emotional intensity — meaning it makes a strong impression, positive or negative — its weight rises. As that weight rises, market expectation drifts from the fundamental assessment. That drift is what people call "value".

This model has an important and rarely recognised consequence. Value does not live in the result. Value lives in the distance between the result and the way the market remembers it. If a catastrophic defeat in Madrid is remembered too long, the price distorts one way. If a resounding victory in Madrid is remembered too long, the price distorts the other way. In both cases, the distortion is the entire opportunity.

This is also why this kind of analysis is often written without needing a single technical datum. It needs only two things: a fundamental assessment, and an observation about crowd psychology. No engine is taken apart in the process.

XII. The blind spot: when "value" is read as "truth"

This is the part I consider most important in the whole piece, and the reason I decided to write it.

There is a logical error that appears everywhere in sports commentary, and it appears especially often in commentary about pricing. The error is reading a statement about a market as though it were a statement about quality.

The headline says: the bad race at Madring creates value in Baku. That is a statement about the relationship between two numbers. It says the first number has moved the second further than it should have moved. It does not say the car will be faster in Baku. It does not say the team has found the problem. It does not say the driver has regained form.

But the reader — and this is the largest blind spot — will read it as though it says all of those things. Because the memory of a bad race is still hot, and because the story of a driver seeking redemption is extraordinarily seductive, the reader automatically fills in the blanks the article leaves.

That automatic filling-in is the very mechanism by which value operates. Without it, the market would not overreact, and there would be nothing to exploit. Put differently: value exists precisely because people misread the genre. This is not a comforting conclusion, and it applies equally to those who believe they are exploiting value and those who believe they understand the sport.

XIII. A second blind spot: the absence of data is not neutral data

I need to state this clearly, because it concerns how I work.

In my source analysis of this headline, most cells in the analytical table were left blank for "insufficient information". That is a methodologically correct conclusion, but it is not a cognitively neutral one. The fact that a technical article lacks technical data is one event. The fact that a market-commentary headline contains no technical data is a different event. The two events share a shape but not a meaning.

Back when I sat on the coaching bench in Melbourne, I made a similar mistake with an A-League player. He had very low deep-recovery pressing numbers, about two actions per match. I concluded he did not fit the system. The club signed him anyway. By season's end he had seven assists and helped the team to a semi-final. My data was right about the number and wrong about the meaning. I had measured the absence of a behaviour and called it the absence of value.

Since then, every analysis of mine has a section I call the human factor, and I write it before the technical part. For Baku, that section would contain the following: the sound of an enclosed grandstand when a car hits the wall in the castle section. The way a driver relaxes his shoulders on exiting the tightest blind section. The silence on the team radio after an eliminated qualifying lap. None of those three things appear in any telemetry table, and all three affect the weekend's result.

Data is a shelter, but story is home. I sheltered in data for years, and I know exactly the safety it provides. But every time I step out of that room, I realise the thing I am trying to explain was never fully inside it.

XIV. A counter-intuitive angle: maybe there is no constriction at all

Now I want to invert this entire piece.

Everything so far assumes a notable constriction exists: that the Madrid race over-moved the Baku pricing, and that the move is observable. But there is another possibility I must put on the table.

That possibility is: the odds board reacted entirely reasonably, and there is no value at all.

Consider it. If the Madrid race exposed a repeatable problem — say, a structural weakness in managing rear-tyre temperature at low ambient temperature — then the market marking that team down is a rational information update. Baku, with qualifying in the late afternoon and surface temperatures shifting rapidly, may be a similar environment. In that case there is no expectation gap. There is only a correct assessment being corrected.

I have no data to distinguish these two possibilities. And this is precisely the point I want the reader to carry away: a headline saying value exists is not evidence that value exists. It is only evidence that somebody believes so, and that the belief was written down and published.

This is why I always keep a counterfactual scenario at the end of my analyses. If I am right, what must happen in Baku? If I am wrong, what must happen? And if both scenarios are possible, am I analysing a phenomenon or telling a story?

XV. Zooming out: Baku is a node, not the whole network

I have a bad habit I am trying to break: getting stuck in the constriction. When I find a constriction in a network, I tend to stand there too long, dissect it, draw it, measure it, until I forget it is only one node in a larger network.

So let me zoom out.

Baku is one race in a long season. It accounts for roughly one twenty-fourth of a season. Its points value is bounded by the reward structure: twenty-five points for the winner, eighteen for second, and one point for the fastest lap if set inside the top ten. In a season where the lead is decided by a few points, a bad race in Baku can cost more than a bad race anywhere else, purely because of its structural volatility.

But here is the part I want to stress: the meaning of a race only becomes visible after the season ends. At the moment of writing, Baku has not happened. We are discussing a race not yet run, based on a race already run, based on a headline whose body we do not have. Three layers of indirection. At each layer, a little information is lost. By the third layer, what remains is only a blurred shape.

And I draw that blurred shape, because it is all I have.

XVI. Back to the trapezoid on the receipt

Back to that morning in Melbourne.

The trapezoid I drew on the back of the receipt has four vertices. The first is Baku's top speed. The second is the grip of a street surface. The third is the expected value the market assigns to a driver. The fourth is the fundamental assessment I believe to be correct.

The distance between the third and fourth vertices is what the whole piece is about. But when I look closely at that trapezoid, I realise two of the four vertices — the first and the second — are things I know reasonably well, and the other two are things I am only guessing at. Half my trapezoid stands on hard ground and half on sand.

That is not a failure. It is the normal state of analysis. What I need is not to forget which vertex stands on what.

I folded the receipt, put it in my coat pocket, and went to make another coffee. Outside the window, Melbourne was beginning to brighten. In this apartment, some twelve thousand kilometres from the circuit in Azerbaijan, I was trying to understand a race not yet run by drawing a trapezoid on scrap paper.

Perhaps that is what I have always done. Perhaps it is what all of us do when we love a sport with so many variables that it cannot be loved rationally.

XVII. What I will verify at Baku

I always end with things that can be verified, because analysis without verifiability is just opinion decorated with numbers.

First, I will watch tyre temperatures across the first three to five laps of every stint. If one team consistently falls back in that window, that is a structural signal, not an accident.

Second, I will record castle-section times for two drivers in the same team. A gap above two tenths in that section, while gaps elsewhere sit below one tenth, will be a signal about the human, not the machine.

Third, I will record the timing of the first neutralisation. If it arrives before lap ten, the entire strategic architecture of the race is rewritten, and any conclusion about pure performance becomes void.

Fourth, and perhaps most important, I will compare the final result with the pre-race quoted price. Not to see who was right or wrong, but to measure how much the crowd misread. If the error is large, I know I am observing a network with real constrictions. If the error is small, I know I am reading a compelling story and calling it data.

Every race is a network; I only look for the constriction. At Baku, the constriction I suspect most does not sit on the two-kilometre straight, nor in the castle. It sits in the moment between the pit-exit lap and the first flying lap of a stint — where temperature, decision and error meet inside about twelve seconds.

That is where I will look. Not at the odds board.

XVIII. An unended ending

I have written close to five thousand words about a headline whose body I do not have. It sounds wasteful. But there is a reason for it.

The first shock taught me to listen, the second shock taught me to write. The first shock was a meeting in Melbourne, where I explained a spatial concept to fourteen players and received blank stares in return. The second shock was a wrong conclusion about a player I had publicly opposed, who then played the best football of his career. Both taught me the same thing: the distance between data and people is the distance my work lives in.

The headline about Norris, Madring and Baku sits exactly in that distance. It speaks of a driver through a market number. It speaks of one circuit through the memory of another. It speaks of the future with evidence from the past.

When the Baku race ends, some will be right and some wrong. Some will say value existed and was realised. Some will say it never existed. And very few will go back and reread what they wrote beforehand to see where they erred.

I will be one of those few. I always keep what I have written, including what is wrong, because it is the only way to know whether I am progressing or merely talking louder.

And if you have read this far, hold on to one question for the Baku weekend: when the result arrives, do you see it confirming something about the car, or only confirming something you already believed before the race began?

The answer to that question is not in any telemetry table. It sits where the data goes silent, and there you have to draw your own shape.