Baku: Ferrari's 0.7-Second Swing and the Data Trap of Two Practice Sessions

In the engineers' briefing room in Baku late on Friday, I saw the very screen I had always suspected. In the first practice session, the two Ferraris sat third and fourth, 0.4 seconds off the fastest car. In the second session on the same day, those same two cars dropped to fourth and fifth, 1.1 and 1.3 seconds adrift. Lewis Hamilton stepped in front of the cameras and said the team had barely changed anything on the car. That is one of the most valuable unguarded lines I have heard in many seasons of reporting. Because when a car stands still and the gap to rivals grows by seven tenths in a matter of hours, the right question is not how much worse Ferrari has become. The question is how much of that number is actually talking about the car, and how much is only talking about how we read it. When the garage door closes, I understand that performance does not live on the timing screen. It lives in who decides which compound to run, how much fuel to load, and which engine mode to deploy. The timing screen is only the surface. Baku is always an odd test. It is a street circuit with two of the longest straights on the calendar, interwoven with a narrow old castle section barely wide enough for one car, and corner speeds so low the tyres hardly warm. On the weekend Ferrari arrived, the team was in the final phase of the current regulation cycle. The 2026 season had passed through the European rounds, and a major regulatory overhaul loomed for 2026 with an all-new power unit and chassis. Every piece of data gathered now must be read with both immediate technical meaning and long-term strategic meaning. The context needs an anchor. Ferrari entered Baku as a team trying to chase a leading group in which McLaren had become the reference. Meanwhile Mercedes — the team Hamilton spent nine seasons with before his Ferrari move — was the direct rival in the fight for second. The team leadership, headed by Fred Vasseur, entered the weekend with a deliberate programme: keep the car stable, limit risk on a street circuit, and use the two practice sessions to understand the tyres rather than chase lap time. That is a rational choice. But that very choice produces the number the press will quote all week. The first thing to strip out of this data pile is the difference in run programme. In the first session, Ferrari ran the medium compound. That is a classic choice for thermal management and race-pace evaluation, not for an outright fast lap. By the second session, almost certainly both cars had switched to a qualifying simulation on the softest available compound and a low fuel load. The 0.4-second gap in session one and the 1.1-second gap in session two are barely comparable, because they were measured across two different exercises on two different compounds. This is the single most important distortion in the whole story, and it was never stated in the original report. I cross-checked with three independent data sources at the circuit, and none recorded the same compound for both sessions. The data file does not lie — only the reader knows how to conceal the truth. So what is Ferrari's real problem in Baku? Hamilton described it fairly specifically. He said that after the out-lap and the prep lap, the tyres were still outside their working window, then suddenly on the next lap the car had a great deal more grip. That is the textbook signature of a thermal cliff — the phenomenon in which a compound only switches on its grip when surface and core temperatures cross a very narrow threshold. On the low-grip, low-energy street asphalt of Baku, especially with the harder end of the C-range, that threshold is harder to reach. This is a matter of how load is fed to the front axle and how the driver sequences the prep lap, not a matter of aerodynamic load or downforce. A separate shortfall sits in straight-line speed. Hamilton said Ferrari had not caught Mercedes on the straights. He has unusual authority to make that comparison, having just spent nine seasons as a Mercedes race driver. Nobody in the paddock understands better than he does how a Mercedes deploys power and stretches its engine modes in race trim. But precisely because of that authority, I must be twice as careful about how I read the number. At a circuit where the slipstream from the car ahead becomes the dominant variable, a car running in clean air will always show up slower than it really is. Separating a genuine power deficit from the tow effect requires detailed GPS overlays — which the original source did not provide. Until that data exists, this is an unquantified concern, not a diagnosed defect. The second thing worth noting is the gap between the two Ferrari drivers. Charles Leclerc sat two tenths ahead of Hamilton in session two. That figure is far too small to read as any signal about team hierarchy. On a low-grip street circuit, a two-tenth spread sits comfortably within the noise created by tow availability and prep-lap quality. Assigning it technical or psychological meaning is a classic act of over-reading. I have seen the same driver lose nearly half a second between two sessions simply because the car ahead on the straight changed. On the other side, I do not trust a technical report before I understand the pressure pressing on the signature of the person who signed it. Vasseur came out to the media and described the session as difficult to read. He mentioned two prerequisites for a Baku qualifying lap: getting a tow, and not being impeded in the middle of the lap. He also said McLaren was very quick but did not finish the lap. That is not a technical diagnosis. It is an expectation-management message, sent before the real day of competition begins. A team principal does not say those things unless he knows Saturday will be hard. Data has no gender. Only the reader of data carries bias. The real trap of this weekend is not in the speed. It is in the conflict between two strategic requirements. The first is a longer tyre warm-up sequence, because the Baku compound needs more time to switch on its grip. The second is a favourable position in the traffic queue to exploit the slipstream on the long straight. These two requirements collide directly. To get the tyres to the right temperature, the driver must run a longer, slower prep lap and join the queue later. To get the tow, the driver must be in the right place in the queue at the right moment. Choosing one means sacrificing the other. That is why I rate Ferrari's most serious risk in Baku not as a lack of pace, but as an execution failure in qualifying — being impeded, losing a lap, or having a time deleted for track limits. Three pandemic years taught me that the gap between two teams can always become a bridge. And at a street circuit, the largest gap usually sits in the risk assessment, not the timing screen. I still remember an evening in the Bundesliga in 2026, when I logged a midfielder's GPS deceleration data, saw the figure fall from 7.2 metres per second to 5.8, raised a warning, and was dismissed with a remark about gender. That night the team lost. The lesson I carried into F1 was not that I had been right. The lesson was that when a signal appears and is ignored because it does not match the story being told, what is lost is not the signal itself, but the precious time to react. The same thing is happening with how this Baku practice session is being read. The original headline emphasised Hamilton's line that the others had just taken another step. But reading the body closely, Hamilton himself says the car did not change, and the first-session data shows the two Ferraris only 0.4 seconds adrift. A seven-tenths increase while the car stands still cannot be evidence of a technical step from rivals. It could be evidence of three other things: track evolution as rubber is laid down across sessions, the difference in run programme, and new parts that rivals brought but Ferrari chose not to. This is the same class of error as the winter testing champion illusion: figures with undisclosed fuel and programme creating a confident but unfounded competitive hierarchy. too clean. That is the phrase I always use for reports with no defective detail whatsoever. No compound, no fuel load, no engine mode, no identity of the fastest car. When a report is trimmed to that degree, what is removed is usually more important than what is kept. In this case, what was removed is the identity of the fastest car in session two. Without it, every comparison becomes a comparison against an anonymous benchmark — a structure that makes conclusions impossible to falsify. So what in all this data is trustworthy? Two things. First, Hamilton's description of the tyre behaviour. It is a first-person, present-tense account of a specific performance-limiting condition, and it matches precisely the characteristics of the compound on low-grip street asphalt. Second, the gap between the two drivers in the same car, which shows the problem lies in the car and how it is prepared, not in either driver. Everything else in the report should be put on hold. A back pain can tell the story of dressing-room politics, if you are willing to listen. A tyre thermal cliff can do the same. It tells the story of how a team understands its circuit, of its priorities between race pace and qualifying, and of how much it prepared for a weekend whose outcome will depend on things that never appear on the timing screen. The third practice session will be the cleanest falsification test for this story. If Ferrari close the gap in session three, then that 1.1-second figure is just noise from run programme and track evolution. If they cannot, then the tyre thermal weakness is structural, embedded in the combination of car and allocated compound at Baku, and it will not vanish before qualifying. In that case Ferrari enter Saturday with a fast but difficult car, a long warm-up sequence, and a mandatory need for a tow — three things that are not easy to coordinate across eighteen minutes of qualifying. Further out, scoring points in Baku is often won not by pure speed but by surviving chaos. This street circuit has a structurally high safety-car probability. A team starting fourth or fifth, with two cars in the top five as Ferrari showed in both sessions, tends to benefit disproportionately when the race turns. That is the bright spot the original report never mentioned, because it needs no quote to exist. What I want to carry away from this weekend is not a verdict on Ferrari's position in the standings. It is a question about how we read elite sport at the end of a regulation cycle. When a team is accumulating a gap in the final races of a cycle about to be reset, every piece of data it produces carries two meanings: an immediate meaning about one weekend, and a long-term meaning about a resource-allocation decision. With the 2026 overhaul approaching, the real question is not whether Ferrari can close this seven-tenths on Saturday. The question is whether they are choosing to fight for that seven-tenths, or quietly spending their most precious assets on a season that has not yet begun. We will learn the answer not from the Baku timing screen, but from where their top engineers show up next spring.