#F1 Forensics: The all new F1 Data Table

Brought to you by TheJudge13 contributor Tourdog

It started with ICE’s, then gearboxes, now things have expanded a little. Lap times are where it’s at, and it was high-time somebody sorted through them.

So I did.

24,945 laps have been run in the 2015 F1 season from Australia through Hungary.

How do I know that you may ask?  Because I have inputted every single one of them into a spreadsheet.

Yes. Every. Single. One.

But it goes way beyond just lap times, because what good are lap times if you don’t know what tyre they were on? Or how much fuel they had in? or which turbo was in the car?

So I have compiled a much larger dataset for us to play with.

Here is an overview of the dataset:

dataset example

So for each lap logged, the following information is logged with it:

  • PU Manufacturer
  • Gearbox Manufacturer
  • Team
  • Driver
  • Race
  • Session
  • Lap #
  • Lap Time (in seconds. example: a lap time of 1:12.237 is displayed as 72.237)
  • Average speed (calculated as lap distance/time in Kph)
  • Fuel Load
  • Classification
  • Points
  • ICE #
  • Turbo #
  • MGU-H #
  • MGU-K #
  • Energy Store #
  • Control Electronics #
  • Gearbox #
  • Chassis #
  • Lap Distance
  • Lap Type

Most of these are self explanatory. The options for the “Driver” category are limited to the drivers that have participated in at least one session in 2015, for example.

However, once I got things together, and I started to crunch numbers, I realised that the lap times were kind of useless. I had a whole lot of numbers, but there were some laps listed with a time stamp, instead of a lap time, some lap times were off the charts too long, there were lots of problems. This forced me to really look at HOW the times are compiled, and how I could use these times.

The first thing I needed to do was differentiate between “useful” laps, and worthless ones, and to do this we need to know how the FIA live timing works.

Let’s start with the mechanicals of it. There are timing beams laid out along the entire circuit length. These beams break the circuit into sectors (a maximum of 18 I believe) and allows the FIA to keep track of the cars. Each car has a unique transponder in the nose. When that transponder breaks the beam, a computer somewhere registers a time stamp in a database.

All we are really concerned with however, is the timing beam at the start finish line. Technically there are TWO timing beams at the start finish line. One beam spans across the track. The second beam spans the pit lane, and should be in the same plane as the circuit beam.

In practice and Qualifying, things are pretty simple. The clock starts the timing. Once the clock starts, the cars are free to leave the pits.

If a team’s garage is at the beginning of the pit lane, they will break the Pit timing beam, and the computer will register a time stamp.

If a team’s garage is at the end of the pit lane, they will not break the Pit timing beam. The first time stamp for one of these cars will occur once they either cycle back into the pits, or cross the start/finish timing beam.

Once the timestamp happens, what we will call the “Zero” lap time, each subsequent entry into the timing log will be a simple calculation of subtracting timestamps to come up with a lap time. This is how the FIA views their lap times, and then dumps the list of these times into a PDF file to post on their Website.

Unfortunately this makes it difficult to know what is actually the first lap a car runs the early session.

Example 1:

Let’s say Sauber is at the end of the pit lane, beyond the pit timing beam. We will use a random Lap time of 1:30.000.

FP1 starts, Nasr leaves the pit, circles the track, and immediately re-enters the pit. When he breaks the pit beam upon his return, the first timestamp happens. He now pulls the car in the garage for a 5 minute checkup.

Nasr leaves the pit again, and circles the track, this time not going into the pits, but continuing on across the circuit timing beams.

Nasr will now get his second timestamp, giving us a lap time for LAP 1, as 6:30.000.

But Nasr has done two laps.

Example 2:

Williams garage is at the beginning of the pits. At the end of FP2, Massa is running hot laps out on track. The session ends, and 30 seconds later he crossed the start/finish line.

This will register his last lap time. Massa then continues around the track, circling back into the starting boxes and does a practice start, circling the track one more time before parking in the garage.

The last two times Massa circled were not registered anywhere in the timing.

This of course doesn’t really matter in terms of useful information the Teams and Drivers can get out of the car, but an extra 2 laps per session on a PU in 2015 could add up to a statistically significant amount of distance when your PU only lasts 1000Km!

The race timing is a different ball of wax.

Think about how the races start. Unlike the Practice and Quali sessions, the actual “Lights Out” time for the race is random. Charlie decides when the race starts, not the clock.

So, the cars all form up in the start boxes, and once Charlie is satisfied, he will start the timing sequence.

Charlie presses and holds the start button, the lights count down, 5, 4, 3, 2, 1 and at 0, the computer registers a time stamp we will again call LAP 0.

Charlie, holds the start button until all the cars cross the start/finish line. Holding this button essentially “turns off” the start/finish timing beam, so that no cars register a time as they cross it.

The cars circle the track.

As each car crosses the start/finish, it registers a time stamp of that moment as LAP 1.

The second time a car crosses the start/finish (or pit) timing beam, the computer does a subtraction, and spits out a lap time.

The reason that this matters, is that NOWHERE could I find an official FIRST LAP time for any driver ever. The PDF files that the FIA makes public, have the time stamp as LAP 1, but since we do not know what the actual moment Charlie pushed the button was, we cannot calculate a lap time.

Even on the F1 app, and in live TV coverage, we never see a lap time for the first lap.

Go back and look, I didn’t realize it either.

So, in order to figure out what a driver’s lap time is for the first lap, we need 2 things, a sum total of all known lap times, and the official race classification time. Simply subtracting these two gives us a driver’s lap time for LAP 1. This is what I have done in our data for the races.

As far as the earlier sessions are concerned, I have not bothered to try and calculate a lap time. These times really don’t matter anyway. I have kept the time stamp as “Lap 1” in all of these sessions simply to signify that at least one lap was run, thereby getting us a little closer to being accurate in terms of distances.

I challenge you to find a lap time, anywhere online, for LAP 1 in a 2015 F1 race.

So now we know how to count laps, but the lap times given for some of these laps is way out out bounds, like  Example 1, I gave earlier. In order to be able to sort through these laps, I have given each and every lap a “type”, in the “lap type” field.

There are several different classifications of lap type, and they are as follows:

Pit: a pit-in or pit-out lap that is run during FP1, FP2, FP3, and Quali. These are always warm up or cool down laps and their times are useless.

Pit-in: Used to signify a pit-in lap during races only.

Pit-out: Used to signify a pit-out lap during races only.

Yellow: Any race lap run under a Yellow Flag.

Safety Car: Any race lap run under a safety car.

VSC: Any race lap run under a Virtual safety car.

Parade: every driver who makes it to the starting box before the formation lap gets 1 lap added to their race distance as a “parade” lap

Formation: Every driver that successfully makes it around the track and into their starting box for the green flag gets one lap added to their race distance as a “formation lap” (in Hungary all drivers got 2 formation laps added)

Podium: Any driver that crosses the finish line, gets one lap added to their race distance as a “podium” lap.

Green: is all Green flag laps in all sessions.

Scratch: This is my new catch-all for unwanted laps. Any lap whose time is beyond “normal” is called a scratch lap and thrown out of lap time calculations.

So what is a scratch lap you may ask?

As an example, after looking at all of his laps, I can see that Romain, in the early FP sessions, will often do alternating hot laps over a long run. So even numbered laps are often 8-10 seconds slower than odd numbered ones. This means Romain is alternating between hot laps, and cool-down laps, most likely doing practice quali runs. If we were to keep all of Romains lap times in our table as “green” laps, when we do our averaging calculations, Romains times are going to look comparatively slower. SO what I have done is filtered all of those slower laps out by calling them “scratch laps”.

Another example:

Sometimes the timing system freaks out, believe it or not. I have only seen it occur during FP sessions, but there are several occasions in the data, where anyone running out on track at a specific moment, will get 250 or so seconds added to a random lap time. Why? who knows, but it’s in the data, and the FIA PDF files do not call them out as Pit laps, so I have called those types of laps “scratch, as well.

Why should you care? Well, once we filter out any crazy, out-of bounds lap times, we can get a much clearer picture of exactly how fast a driver is. It also allows us to compare lap times on the same tire across multiple sessions.

At this point, I believe the largest hurdle to us getting right down to the meat of the data, is fuel loads. Fuel loads are something we just don’t have much data on. In an effort to narrow our focus, I have given one of several different types of fuel loads to each lap.

In FP1 for all races, fuel loads are “unknown”.

In all other sessions, fuel loads are either

  • Unknown
  • Heavy
  • Medium
  • Low
  • Quali

We know Mercedes runs a heavy fuel load in FP2, and I figure Williams does as well.

We know Ferrari was running Medium fuel loads in FP2 up until June, since then I have them as running heavy fuel in FP2.

I have all teams as running Low Fuel in FP3

Quali” fuel loads are listed for all qualifying sessions, and for the last 5-7 laps of all races.

The races are broken into chunks, with the first group of laps signified as “Heavy”, the next group as “Medium”, the next as “Low”, and, as I said, the final group as “Quali”.

It is by no means as exact as I would like it to be, but it is a start.

I would like to come up with a formula to calculate at least a rough percentage of load, but I am going to need someone with a little more expertise to help me come up with it.

Anyway. What all this work has done, is allowed me to spit out all of these wonderful graphs. We are working on a method for you to be able to search this data on your own, right here within the website, so keep your fingers crossed. Until then you can enjoy the tables I have started for us. If you have a suggestion as to a certain table or graph you would like to see, please leave a comment and I will see what I can do.

You can find all this good stuff in the CHANCERY’S ARCHIVE

Advertisements

12 responses to “#F1 Forensics: The all new F1 Data Table

  1. I challenge you to find a lap time, anywhere online, for LAP 1 in a 2015 F1 race.

    You haven’t looked at the FIA Race History Chart!

    eg: Hamilton’s first lap today was 1:57.421.

  2. I’m not sure if there is sufficient information available in the public domain for you to be absolutely certain on the allocation of mileage to individual PU elements during FP1 and FP2.
    The regulations allow teams to build a PU using any elements which have a homologation approval seal, irrespective of whether it was attached at the current event or any previous event. This is what McLaren want to exploit by having two engines sealed concurrently at the SPA event.
    Additionally Ferrari have two low mileage ICE’s which it is reported on European sites they are using in FP1 & 2, to preserve their 3 token Canada engine for FP3, qualifying and race day, in an attempt to avoid penalties later on.
    The FIA opened a can of worms by their ‘missing 2015 homolagation date’ faux pas.

    • Hey Engineer, you are absolutely correct.
      There is a major element of speculation involved, in trying to figure out who is using what when. I am making my best guess in most cases, but the issue is up for debate and if we can reach some kind of consensus, I am happy to adjust the data.
      As far as Ferrari is concerned, they did use 2 ICE’s early, but their 3rd ICE utilized token upgrades. According to the regulations, a manufacturer cannot run two different specifications of PU at the same event. So TECHNICALLY, they should only be able to use ICE number 3 right now. But as I discussed in a previous article, Ferrari were in fact running two different specs for several races anyway. In England and Hungary, Sauber was running ICE #2 under the old homologation, and Ferrari was running ICE #3, using token upgrades. I don’t know how, but Ferrari seems to be able to do it.

      Mercedes I can say with some certainty, is pretty accurate. The 8 mercedes powered cars are all only running on their 2nd PU, so they only have PU #1 to go back to for practice anyway. Hulkenberg is the exception to that, but it is apparent his First ICE, TURBO, MGUH, and MGUK are all trash, so he only has #2 and #3 to play with.

      McLaren is a mess, and its anybody guess as to which elements are really being used, so I have tried to keep it logical, however Mclaren in general is running so little mileage in practice that it hardly matters.

      Renalt is the tough one, and now that I have built a table that helps me visualize things a little better (it will be posted soon), I am going to go back and revise some of FP1 & 2 PU element use.

      I am open to suggestions.

      T

      • Hi tourdog
        I don’t think that I am able to make a better guess than anyone else, however the probability of making a correct guess is dependant on using the correct assessment criteria!
        As far as what is or isn’t allowed to be used by Ferrari or for that matter any other team, the regulations are quite simple and concise: –
        Appendix 4.3 states in respect of homologating a PU component
        The supplier of an homologated power unit and/or the team using the homologated power unit must take and/or facilitate such steps as the FIA may at any time and in its absolute discretion determine in order to satisfy the FIA that a power unit used at an Event is indeed identical to the corresponding power unit delivered to and held by the FIA.

        Note the specific use of ‘corresponding’ and not ‘current’, so it is the homologated PU at the time the component was ‘sealed’ that is relevant.

        Section 28.4(e) states for PU elements
        … Each driver will therefore be permitted to use four of each of the above six components during a Championship season and any combination of them may be fitted to a car at any one time.
        and also 28.4(h) states
        After consultation with the relevant power unit supplier the FIA will attach seals to each of the relevant components within the power unit prior to them being used for the first time at an Event in order to ensure that no significant moving parts can be rebuilt or replaced.

        There is nothing in the regulations that states an approval seal can/will be invalidated by a later PU being homologated – what would be the point to prevent the use of a viable older (presumably worse performing) component variant?

        Personally, I would expect that when a team introduces an newly homologated PU it would be checked for early life failure by running it in FP1 & 2. In later events I would expect teams to preserve PU’s for only FP3, qualy and race if they had any viable other engines that could be used in FP1 & 2.

        • “… Each driver will therefore be permitted to use four of each of the above six components during a Championship season and any combination of them may be fitted to a car at any one time.”

          This is true, but does not take into account the changes made for the Token system…

          “There is nothing in the regulations that states an approval seal can/will be invalidated by a later PU being homologated”

          I beg to differ:
          2015 sporting regulations Appendix 4
          2. A manufacturer may homologate no more than one specification of power unit.

          That seems pretty clear to me.

          “After consultation with the relevant power unit supplier the FIA will attach seals to each of the relevant components within the power unit prior to them being used for the first time at an Event in order to ensure that no significant moving parts can be rebuilt or replaced.”

          So if this is true, and teams are not allowed to break the seals on old engines to update the parts, then I read it as Ferrari would not be able to use a previously homologated PU in practice.

          Believe me, I would love to have a definitive answer on this..

          Maybe I am reading it wrong, there are several rules that seam to contradict each other…

          • Sorry – I cannot see any ambiguity or contradictions in the regulations.

            Homologation is the process of providing the FIA with a reference PU for storage – no more no less.

            There were/are no changes made for the token system – they are just introduced progressively. When any ‘token’ upgrades are introduced, a new PU must be lodged with the FIA which becomes the new reference.

            If your component has an FIA seal already applied that verified its’ compliance with that reference PU at the time when the seal was applied, then you can use it as and when you wish.

            The ‘first time in an Event’ refers to the first time the component is ever used, not the first time per Event that the component is used.

            So if this is true, and teams are not allowed to break the seals on old engines to update the parts, then I read it as Ferrari would not be able to use a previously homologated PU in practice.

            The whole point is that they are not allowed to break the seals to upgrade the parts – they just fit the part and use it ‘as-is’, until it fails or is no longer reliable.

            With respect, I believe that you are trying to introduce interdependencies between regulatory items which were not intended to exist.

  3. Just thinking, you say to calculate the first lap time you subtract race time from total of the other laps.

    Does that work? Would race time not start from the point the lights go green? If so then anyone on the back row (Hello Messrs Button and Alonso) would have several seconds before they passed the start/finish line so therefore their first lap time would be longer than it actually was.

    Unless you count the first lap as lap + grid offset which will mess up any calculations anyway….

Leave a Reply