26 replies to this topic

[Rajdeep]

I had an idea while in the shower this morning about how to read the testing times. Obviously, it's not going to give you the correct pecking order, nor if someone is faster or slower by 1-2 tenths. But I think it's pretty good to pick out if car A is more than 5 tenths faster or slower than car B. So, 5-tenths is probably the margin of error.

Here're the assumptions:

1. Tyres degrade very quickly, only the first lap is useful to evaluate any changes (Alonso)
2. Each 10 kg of fuel adds 4-tenth to the time; so each second is worth 25 kg fuel (James Allen)
3. Each flying lap takes about 2.5 kg fuel on average (James Allen)
4. In testing, everyone pushes as much as they can without risking a crash; running race stints is useless if you're trying to evaluate whether a component adds 1-tenth (race stints are usually about 3 seconds slower than the ultimate speed on the first lap)
5. Teams usually test with fuel loads of full tank, 2nd race stint, 3rd race stint and low fuel (at least 20 kg for top teams); so the fuel loads would be 20 kg, 50 kg, 100 kg and 150 kg (I think JA said something similar, but I don't remember the exact quote)

Based on the above, and assuming the top cars can do 1:22 on a pole attempt today in Barcelona, I will look for the following first lap times in their stints

20 kg : 1:22:7
50 kg : 1:24
100 kg : 1:26
150 kg : 1:28

Also, the degradation should be about 0.5 seconds par flying lap, to make sure that the driver is pushing.

Now, let's look at Rosberg's stint from yesterday. He starts with 1:22:6, does about 6 flying laps and with degradation of more than 0.5 seconds a lap. So, my theory would say that he started with 20 kg of fuel, pushed to 99% of the limit to get 1:22:6 and could possibly dip under 1:22 on a proper quali run at this time.

Compare that with Alonso's run. Starts with 1:22:95; 10 flying laps with degradation of more than 0.5 sec a lap. He'd have more than 20 kgs to to those many laps; but not 50 kgs (otherwise, the rest could go home now). I'd guess Alonso started with about 30 kg fuel, which would give about the same ultimate speed as Rosberg. At least it's within the margin of error and I can't say that one is faster than the other.

Comparing F138 and F2012 from the Barcelona race, Alonso did a 1:22:3 in Q3 with soft tyre and DRS everywhere. Assuming the 2013 spec medium is as fast as the 2012 spec soft, and the ability of using DRS adds 0.2 sec, I'd say that the Barcelona spec F2012 would have the ultimate speed of 1:22:5 with this years' rules and on medium tyre. Comparing that to my estimate of the current speed of F138, I'd say F138 today is about 5-6 tenths faster than the Barcelona spec F2012. Good news, but still a way to go to catch the leaders.

In summary, I'd be looking at stints of at least 4 laps with average 0.5 sec degradation per lap, that starts of at around 1:24 (50 kg), 1:26 (100 kg) or 1:28 (full tank). The difference in the first lap speed of the various cars is likely to be the difference in their current speed. I'll write an update to this at the end of the testing day.

By the way, I'm only talking about the top teams, i.e. McLaren, RB, Ferrari, Lotus and Merc. The rest are too difficult to estimate.


PS - sorry for the long post!

[zdzisio]

Everything very nice, up to the point where you realise that

5-tenths is probably the margin of error.

So if you look at the latest Spanish Grand Prix on the very same Circuit de Catalunya looking at best qualification times with 0.5s precision would give you very valuable information that Ferrari, McLaren, Mercedes, Lotus, Wiliams, Red Bull and Sauber are the seven fastest cars. In no particular order. Good to know!

Edited by zdzisio, 20 February 2013 - 12:22.

[Imperial]

It's reasonable up until point four in my opinion.

They don't all always push as hard as they can. They've got the system-check business out of the way already, at Jerez, but they're far from consistently pushing for speed alone.

There remain far too many factors to ever know what they're testing. Look at Esteban Gutierrez yesterday, Sauber didn't even bother disguising the fact that his runs were designed to get him used to running an F1 car at race distance. It's quite possible that car evaluation was secondary to his personal experience.

This debate will go on into eternity and nobody will be closer to the answer.

Edited by Imperial, 20 February 2013 - 13:38.

[joshb]

It's probably better than Andre Benson's logic of getting the average time for each car (excluding really slow laps)

or Gary Anderson's logic that one car has 10% more downforce than the rest yet is the 6th best car!

Edited by joshb, 20 February 2013 - 13:45.

[choyothe]

The best way to read testing times is to make or twist every piece of information into making the situation seem as good as possible for your favourite driver/team.

See: Every testing thread since the start of testing threads.

[GodHimself]

The best way to read testing times is to read nothing into them.

/thread

[choyothe]

The best way to read testing times is to read nothing into them.

/thread

See, He has spoken.

/thread

[Seanspeed]

The best way to read testing times is to read nothing into them.

/thread

There's people who get overexcited about testing times and then there's people who say the times mean nothing.

Both are wrong.

[mattferg]

There's people who get overexcited about testing times and then there's people who say the times mean nothing.

Both are wrong.

Testing times mean close to nothing, but true, not nothing. Unless a car is way faster than the field or way slower, then the times mean nothing.

[Seanspeed]

Testing times mean close to nothing, but true, not nothing. Unless a car is way faster than the field or way slower, then the times mean nothing.

If you know what to look for, the times can be interesting for many reasons. Especially when they start doing race simulations.

[zdzisio]

If you know what to look for, the times can be interesting for many reasons. Especially when they start doing race simulations.

Not really unless you're inside the team and know all the parameters like amount of fuel, pressures, ride hights, cambers...

You can tell much more standing in some tricky fast corner watching and comparing how particular cars are handling it. Or watching suspension work in some tight chicanes.

[ethirtyfour]

Top 10's for all three test sessions last year (2012)

Jerez (1st test)
1. Nico Rosberg Germany Mercedes-Mercedes 1m 17.613s **
2. Romain Grosjean France Lotus-Renault 1m 18.419s
3. Michael Schumacher Germany Mercedes-Mercedes 1m 18.561s **
4. Fernando Alonso Spain Ferrari-Ferrari 1m 18.877s
5. Mark Webber Australia Red Bull-Renault 1m 19.184s
6. Sebastian Vettel Germany Red Bull-Renault 1m 19.297s
7. Lewis Hamilton Britain McLaren-Mercedes 1m 19.464s
8. Daniel Ricciardo Australia Toro Rosso-Ferrari 1m 19.587s
9. Jean-Eric Vergne France Toro Rosso-Ferrari 1m 19.597s
10. Kimi Raikkonen Finland Lotus-Renault 1m 19.670s
(** 2011 cars for Rosberg & Schumacher)


Barcelona (2nd test)
1. Kamui Kobayashi Japan Sauber-Ferrari 1m 22.312s
2. Pastor Maldonado Venezuela Williams-Renault 1m 22.391s
3. Nico Hulkenberg Germany Force India-Mercedes 1m 22.608s Weds
4. Sergio Perez Mexico Sauber-Ferrari 1m 22.648s Weds
5. Sebastian Vettel Germany Red Bull-Renault 1m 22.891s Weds
6. Paul di Resta Britain Force India-Mercedes 1m 23.119s
7. Fernando Alonso Spain Ferrari-Ferrari 1m 23.180s Weds
8. Jenson Button Britain McLaren-Mercedes 1m 23.200s
9. Michael Schumacher Germany Mercedes-Mercedes 1m 23.384s Thurs
10. Felipe Massa Brazil Ferrari-Ferrari 1m 23.563s


Barcelona (final test)
1. Kimi Raikkonen Finland Lotus-Renault 1m 22.030s
2. Sergio Perez Mexico Sauber-Ferrari 1m 22.094s Sat
3. Jenson Button Britain McLaren-Mercedes 1m 22.103s Sat
4. Daniel Ricciardo Australia Toro Rosso-Ferrari 1m 22.155s Sat
5. Fernando Alonso Spain Ferrari-Ferrari 1m 22.250s
6. Bruno Senna Brazil Williams-Renault 1m 22.296s
7. Nico Hulkenberg Germany Force India-Mercedes 1m 22.312s
8. Kamui Kobayashi Japan Sauber-Ferrari 1m 22.386s
9. Felipe Massa Brazil Ferrari-Ferrari 1m 22.413s Sat
10. Lewis Hamilton Britain McLaren-Mercedes 1m 22.430s

It's pretty much useless to draw conslusions. As someone has already mentioned outside of direct track observations and maybe a bit from long runs.. that's about it.

[Seanspeed]

Not really unless you're inside the team and know all the parameters like amount of fuel, pressures, ride hights, cambers...

Yes really. You dont need to know what camber settings are being run to be able to gain some perspective on a race simulation, especially if its being run side by side with another driver/team.

You can tell much more standing in some tricky fast corner watching and comparing how particular cars are handling it. Or watching suspension work in some tight chicanes.

Just like with times, if you know what to look for, you can find some interesting things standing trackside. But thats no more useful than looking at a sheet of laptimes.

The trick is to look at everything. And that includes laptimes. People go on and on about how laptimes dont mean anything, but thats usually just people responding to others that are going overboard with reading into the laptimes. As I said, the truth is somewhere in the middle.

Edited by Seanspeed, 20 February 2013 - 19:45.

[zdzisio]

Again and again I hear those magic words "race simulations". That the teams run them and you can somehow watch the results and compare to other teams.

Well, they do run them, but they hardly ever run them in ONE run. It is really not a rocket science to slice the race simulation it into two or three stints and do some other work in-between. Watching from the sidelines you really are in no position to draw any valuable conclusions.

Not to mention that the results of such simulations can vary greatly simply by the choice of tires used, whether they were brand new or heat-processed, or scrubbed, or whatever...

Edited by zdzisio, 20 February 2013 - 21:31.

[Seanspeed]

Again and again I hear those magic words "race simulations". That the teams run them and you can somehow watch the results and compare to other teams.

Well, they do run them, but they hardly ever run them in ONE run. It is really not a rocket science to slice the race simulation it into two or three stints and do some other work in-between. Watching from the sidelines you really are in no position to draw any valuable conclusions.

The results of such simulations can wary greatly simply by the choice of tires used, whether they were brand new or heat-processed, or scrubbed, or whatever...

Anybody who talks about 'conclusions' during testing is talking nonsense. Nobody is talking about proving anything for certain. But race simulations are certainly one of the most useful and indicative signs of competitiveness.

Some people think that just because its not 100% certain, its useless and thats also nonsense. Like I said, its about putting everything together into a clearer picture. Looking at laptimes is part of that picture.

[zdzisio]

Anybody who talks about 'conclusions' during testing is talking nonsense.

That is precisely my point, thank you.

Some people think that just because its not 100% certain, its useless and thats also nonsense.

No. Most of the contemporary science deals with the fact that we do not know anything for certain. But usually we can estimate some precise error bars.

So in situations like this we should start with estimation of the precision of our observation. The original poster estimated this to be half a second. IMHO that's bit too little but nevertheless even such level of uncertainty would mean that we can conclude hardly anything from given data.

Look at the example I gave in my first post: at race weekend at the same circuit as today tests are conducted - when nobody was sandbagging and everybody were giving their best - 7 best cars were in our given margin of error. What kind of information that gives us? Virtually none.

Edited by zdzisio, 20 February 2013 - 21:59.

[Seanspeed]

No. Most of the contemporary science deals with the fact that we do not know anything for certain. But usually we can estimate some precise error bars.

So in situations like this we should start with estimation of the precision of our observation. The original poster estimated this to be half a second. IMHO that's bit too little but nevertheless even such level of uncertainty would mean that we can conclude hardly anything from given data.

Look at the example I gave in my first post: at race weekend at the same circuit as today tests are conducted - when nobody was sandbagging and everybody were giving their best - 7 best cars were in our given margin of error. What kind of information that gives us? Virtually none.

I didn't say this specific method was some tried and true way of determining the order. There is none. I dont know why you keep arguing with me as if I said there was.

My only point was that times can be useful and indicative of performance. You just have to know what to look for and to incorporate other bits of evidence. There is no guarantee of getting things right, and you probably wont get everything pegged, but its not too difficult to get a somewhat good idea.

[Owen]

I had an idea while in the shower this morning about how to read the testing times.
.........
PS - sorry for the long post!

Could we see your findings?

[zdzisio]

I didn't say this specific method was some tried and true way of determining the order. There is none. I dont know why you keep arguing with me as if I said there was.

Aaaa, so you are not talking about pecking order, so what are you talking about?

My only point was that times can be useful and indicative of performance.

Aaaaa, now I get it, what you are saying is "I don't know whichever of them is the best, but damn those F1 cars are fast", right?

You just have to know what to look for and to incorporate other bits of evidence.

Exactly. My point being that not being inside of a team you do not know were to look, and have virtually no other bits of evidence. Or, at least, you can't in any reasonable way distinguish the valuable pieces of information from the noise.

[Seanspeed]

Aaaa, so you are not talking about pecking order, so what are you talking about?

What?

This obviously isn't going anywhere if you still aren't getting what I'm saying.

[jrg19]

So how do we look at Rosbergs last stint?

[Imperial]

Maybe this belongs in the BBC/Sky coverage thread, but one thing disappointing about the Sky testing coverage is their complete lack of analysis or attempts to draw any conclusions.

As with virtually all of the above, it's hard to reach a comprehensive conclusion, however it is possible to look at lap times from throughout the day for each driver and talk about that a bit. So far the F1 test round-up seems to focus solely on a drivers fastest lap, while Ted Kravitz seems to want to focus on Sauber's buffet selection.

[zdzisio]

This obviously isn't going anywhere if you still aren't getting what I'm saying.

This isn't going anywhere because it just became clear what you are talking about.

"It's not like times mean ****. I mean they do, but they don't. But they do. You cant tell the order by reading into lap times, but you can tell a lot about the performance. Not that kind of performance that would give us a clue about the order though, but still if you know where to look you will see it"

Man, this is getting silly.

[zdzisio]

Maybe this belongs in the BBC/Sky coverage thread, but one thing disappointing about the Sky testing coverage is their complete lack of analysis or attempts to draw any conclusions.

As with virtually all of the above, it's hard to reach a comprehensive conclusion, however it is possible to look at lap times from throughout the day for each driver and talk about that a bit. So far the F1 test round-up seems to focus solely on a drivers fastest lap, while Ted Kravitz seems to want to focus on Sauber's buffet selection.

And now we are going to Barcelona where at Circuit de Catalunya is our reporter:

"The cars were running around the track all day. Looking at lap times what I can tell you is that they were running pretty fast, although talking about actual pecking order would be a bit premature at this point. Back to studio. Jim?"

[Rajdeep]

Ok, let's try to see what today's times mean.

Alonso: His laps 71-75, 80-84, 89-93 are nearly identical sequences. He mentioned that he's on the harder tyres in the afternoon. Not sure if it's medium or the hard. But if he used different types, it looks like there's not much difference in times between them. For now, I'm assuming that it was all on the medium tyre. According to my assumptions, he probably had about 50 kg of fuel in each of those runs. So, I'd say that he can do 1:22:0 on the medium. Given that his best time was on a 4 lap run on the soft, he could've done 1:21:5 if fuel was taken out. Putting the above 2 together, along with the difference between soft and medium tyres, my take is that the Ferrari's ultimate potential speed in a Q3 simulation would be about 1:21:2 or so. Could possibly dip under 1:21, but more likely it's going to be just outside.

Rosberg: His lap 16-24 stint was the most relevant for my analysis. He did 1:22:6 on the medium tyre, which fuel corrected would be just under 1:22:0 on fumes. Again, taking the difference from medium to soft, the Merc's ultimate potential is very similar to the Ferrari's, in Rosberg's hands. His 37-43 stint is also very similar to the above. Rest of his runs were for tyre deg – not really relevant for my analysis.

Button: Only relevant stint was the 13-18. Don't know what tyre he was on. If to compare to Alonso's starting times on similar stints, Button starts of 1 tenth faster. But it's too tiny a difference to say anything.

Webber and Grosjean: Didn't do the type of stint that I'm looking for. So no comments.

Perez (yesterday): Looking at his 29-35 stint, it must have been done on the soft tyre. Even then, it looks too fast to fit into my assumptions. Either McLaren can easily dip into 1:20s, or Checo had not more than 40 kg fuel at the time.

Which brings me to Hamilton's stint which everyone was drooling about.

Hamilton: 1:22:7 on the hard tyre on a 9 flying lap stint. He must have had about 30 kg of fuel, which means he could have been well into 1:21 if on fumes, on the hard tyre. And applying the delta from hard to soft, this is the only evidence that I can think of that any of the 2013 spec cars can get to 1:20s. So, it was a very good laptime!

Vettel: Hard to draw any conclusion as he was doing 1-2 lap stints. But none of it looked super impressive. Either Redbull is sandbagging big time, or they're not that fast. Everyone commented that Webber's long run wasn't as good as Grosjean's today.

So, to conclude, I think Merc has surprised me on the positive side and Redbull on the negative side. Ferrari is there or thereabouts. I'd think McLaren is also in that group and Lotus is close as well. If I've to put a tenner, I'd say that at least one of the Merc guys (possibly both) will be in the front 2 rows in Melbourne and at least one of the Redbull guys will be outside the front 3 rows.

Disclaimer: I'm an Alonso and Ferrari fan.

[Seanspeed]

This isn't going anywhere because it just became clear what you are talking about.

"It's not like times mean ****. I mean they do, but they don't. But they do. You cant tell the order by reading into lap times, but you can tell a lot about the performance. Not that kind of performance that would give us a clue about the order though, but still if you know where to look you will see it"

Man, this is getting silly.

Exactly - you dont know what I'm saying.

[boldhakka]

Exactly - you dont know what I'm saying.

Maybe because you aren't really saying anything. Since you say some conclusions can be drawn if one knows what to look for, perhaps it would be useful to give a couple of examples of those things that one should look for.

I would say reliability and balance are good examples, but the topic is about testing times, so I'll let you offer examples of what to look for in testing times that help draw conclusions about performance.

Edited by boldhakka, 22 February 2013 - 16:13.