10 things we learned from the final F1 test

Eight days, well seven really once you factor in the Wednesday washout (snowout?) at the first test, have gone by in a flash. The preparation is over and now teams are loading up to head to Australia and the first race of the year.

Of the 7978 laps officially logged, the ones during the second test are the key ones and have set the key storylines for the early races of the season.

Autosport has been there throughout, and a combination of following the race runs, qualifying sims and run programmes of the 10 teams, talking to drivers and key paddock figures and trackside observation has provided a picture of where the running so far points for the start of the season.

This is what we have learned.

Ferrari's record is an illusion...

Yes, Ferrari blitzed the unofficial lap record at Barcelona with both cars. Yes, Kimi Raikkonen said it could have gone faster. And yes, no other team got within half a second. But it does not have the fastest car.

Sebastian Vettel's 1m17.182s was 1.2s faster than the previous fastest lap of the Spanish circuit set by Felipe Massa (also in a Ferrari) in 2008. Raikkonen followed that up with a time 0.039s slower, reinforcing the Scuderia's pace on the hypersoft tyres.

But while Ferrari was doing this, Mercedes wasn't even trying. The German manufacturer's best time on day three, when Vettel set his time, was 2.1s slower, and it was 1.6s slower on day four, as it focused on race simulations and other long runs.

Vettel sounded cautious after his best day: "We still need to work on the performance and the feeling. It's the wrong conclusion to look at the time sheet - there's more to it than a good lap.

"Today the track was quite fast [and] we ran a little bit different programme to others."

...Mercedes is the team to beat

Mercedes opted for evolution rather than revolution, and the result was a car that is seemingly an improvement in every aspect over its predecessor. As we've become used to in testing, it racked up the miles without any major problems, and despite no attention-grabbing headline lap time it looks to be the fastest overall package by a healthy margin.

Glance at the video Mercedes released with technical director James Allison comparing this year's W09 to last year's car, and you can see what evolution really means. At a glance, the cars are similar, but the 2018 version just looks that bit more refined, with improved sidepod packaging among the details that become obvious.

And the drivers seem to like it too, with Hamilton declaring it to be a significant step forward. Watching the car on track, it turns in well (something last year's car didn't always do in testing), while in the chicane Hamilton was able to attack on the brakes and rotate the rear on entry with precision.

In a word: ominous.

Red Bull's threat is real

The Red Bull RB14 looks beautiful on track, certainly in the fast and medium speed corners. But it perhaps isn't the step ahead of the Mercedes chassis that the first half of testing suggested. Add to that the limitations of the Renault engine package, which is being run conservatively for reliability reasons, and you have a combination that's not too far off - but still decisively behind. Based on the race simulations, it could be nip and tuck with Ferrari even though other indicators suggest Ferrari is ahead.

But Red Bull is the team Mercedes needs to watch out for. The concern is that the Renault engine is still 40-50bhp down and that, if pushed, it may jeopardise reliability. There's a major ERS package upgrade on the horizon, including the long-delayed new MGU-K that was due to be introduced at the start of last season, but worryingly it's on the distant horizon with no clear schedule.

Red Bull might still be close enough to give Mercedes a headache on Sundays, and given Ferrari's battles to get on top of its car it heads to Melbourne as the team that Mercedes will be most concerned about.

Haas the dark horse amid midfield mess

Renault has been impressing throughout the test but a bad final day, in which gearbox gremlins limited its running significantly, makes its position tricky to judge. That only makes a murky midfield picture more difficult to decipher.

Force India's big update package that is planned for Melbourne will be expected to keep last year's fourth-best team firmly the mix, while Toro Rosso has now revised its expectations and thinks it would be "crazy" not to target points on its debut with Honda engines. Then throw in the confusion of Williams's decision not to perform qualifying simulations while also not completing full race runs.

Almost every team outside the big three has an unknown variable heading to Melbourne, whether it's reliability, untested updates or a lack of representative running. Haas is the exception, and a third season of F1's newest team starting strongly looks like a decent bet.

Applying the laptime delta between each Pirelli compound throws up a bizarre order that has Haas on top and while clearly that is not the case, its testing pace should not be completely dismissed.

Haas has been quietly impressive in testing. It ended the test fifth-best on supersoft tyres and more encouragingly its long-run pace looked a match for Force India. If you're looking for a midfielder to back, the American team might well be it.

Standing restarts could cause problems

Romain Grosjean has a reputation for complaining, but it would be wrong for a director of the Grand Prix Drivers' Association not to be outspoken on things he disagrees with: and he clearly disagrees with the idea of standing restarts.

Grosjean revealed he thought the FIA was just experimenting with the procedure as it roped in teams to carry out practice runs at the end of the morning and afternoon sessions through this week of testing.

But it's a fact that F1 races this year may feature standing starts after red-flag periods, with the cars being led by a safety car before forming up on the grid.

Though drivers will have the option to fit new tyres in the pitlane while preparing for the restart, they may not have fresh rubber left in their allocation - for instance if the restart comes very late in the race - and that would leave some contending with poor grip.

Grosjean suffered from exactly that while performing a practice run of the restart procedure on old rubber in testing on Wednesday, saying it was so bad "I didn't go above fourth gear" and claiming he would usually reach eighth by Turn 1.

"To me it could be carnage," he said. "You could lose the car in a straight line."

Leclerc's brilliance can't hide Sauber's struggles

Charles Leclerc is one of Formula 1's most exciting rookies in some time, but even his clear talent hasn't been enough to tame a tricky Sauber this week.

The rookie was angered by the spin that contributed to almost five hours of lost running on the final test day before his debut, beaching his C37 into the Turn 12 gravel and making light contact with the tyre wall.

The reigning Formula 2 champion said it was an "error that has cost me quite a lot" and admitted he felt a bit angry with himself afterwards. "It's a bit stupid from my side to do it so early in the day," was his verdict.

It was the third time this week a Sauber caused a red flag by ending up in the gravel. Leclerc had already done it in the same spot on Wednesday, while team-mate Marcus Ericsson also dropped it on Thursday. The drivers have spoken positively about the new Alfa Romeo-badged, Ferrari-assisted package, but when both are chucking it off it suggests a machine that at the very least is proving tricky to handle.

McLaren's last minute validation

With six minutes remaining of pre-season testing, Fernando Alonso leaped to the top of the timesheets with a time of 1m16.720s. There was an audible gasp when that happened, only for the fairytale of a late stunner to be snatched away when it became clear Alonso had cut out the chicane while getting past Charles Leclerc's Sauber.

But even so, Alonso's 1m17.784s lap set earlier on that run confirmed what was suspected - that the judgement McLaren had built a dog was incorrect. Its pace is certainly good enough to be there for best-of-the-rest, a step behind the big three, subject to the improvement delivered by what is expected to be a big upgrade package in Melbourne.

But McLaren was only half-vindicated. The pace was OK, but what wasn't acceptable was the reliability. With 599 laps completed, 58% of the tally racked up by Mercedes, and myriad problems including oil and hydraulic leaks, turbo problems and burning to the engine cover around the exhaust the reliability was not good enough. But McLaren did at least complete one full race simulation.

Tyre unknowns could spice up opening races

"I am sure there are some details we don't know," said Pirelli boss Mario Isola at the end of the test. The tyre supplier has found collecting meaningful data as tricky as the teams at Barcelona, where it also had a lot to learn about new products.

In addition to bringing a new pink-walled softest tyre (the hypersoft), Pirelli has also revised almost all its other compounds. The only one that isn't new is the medium, which is last year's soft tyre.

Unsurprisingly last week's bad weather wasn't conducive for evaluating the changes in the specifications, so learning about five tyre compounds - the hard can be dismissed as it is a back-up tyre - would have to be mostly completed in these four days.

No wonder several teams and drivers are predicting a few surprises at the start of the season, and maybe not just in Melbourne. Isola said we will not "really know" the compounds until mid-season.

Still, we also learned there was some benefit to last week's bad weather - the wet running on Friday has led Pirelli to revise its wet-tyre choice for the Spanish Grand Prix. The data collected on the softer version of the wet tyre has made it change its mind on bringing the harder compound to the May race.

Williams has work to do - but there is hope

The Williams looked skittish and tricky to drive throughout most of testing, with corner-entry instability the main problem. That, combined with challenges getting the tyres switched on made life tricky for Williams.

But progress was made during the final afternoon of running, reflected in the hints of pace shown by Sergey Sirotkin. This gives some hope that a car that's a dramatic shift from its predecessor and that technical supremo Paddy Lowe believes represents a step change for Williams is being understood.

Still, given the team spent a lot of testing working to get on top of its troubles, that means it's a little behind in terms of the race simulation work others were able to crack on with. Even so, the Williams completed a healthy, if unremarkable, number of laps during testing and crucially Sirotkin was able to get a good amount of running in on the final day to make up for his losses elsewhere in testing.

For much of testing, things looked very difficult for Williams, but the late step suggests the car is being understood and made to work.

The halo is now invisible

The halo cockpit head protection system is still the cumbersome halfway house that it always was, but an interesting thing happened during the second test. When first watching trackside for an extended period of time, it didn't register. It shows how quickly something new - even something as grating as this - can become normal.

That doesn't mean it's without its problems. As well as the difficulties of recognising the driver shrouded within, for the shorter drivers in particular a lot of care needs to taken not to stand on the intricate aero detailing in front of the sidepods when getting out!

So for those predicting the end of days thanks to the halo being grafted onto grand prix cars, it's not going to be as bad as you think. Spend a few hours watching the television coverage and it soon will just become part of the furniture.

That doesn't make it any less of an inelegant solution and F1 must work to find a better way to do this, but it does at least mean that it is possible to watch 2018 grand prix cars without being distracted by the halo.