Testing analysis: Has F1 finally got a three-way title battle?

F1 pre-season testing is over! (2:10)

Week two of pre-season testing is complete, so here's the highlights of what happened, in tweet form. (2:10)

MONTMELO, Spain -- Although there are no winners in pre-season testing, a clear competitive order started to emerge during the second week at the Circuit de Catalunya. It was no surprise to see Formula One's top three teams -- Mercedes, Ferrari and Red Bull -- unchanged for 2018, and some estimates put the gap between "the big three" and the midfield at over a second. But with just one team able to win the opening race in Melbourne, who is in the best position after a fortnight of testing in Spain?

Single-lap pace

Prior to this week's test, Felipe Massa held the track record at the Circuit de Catalunya. Back in April 2008 his Ferrari was fitted with a set of slick tyres being developed for Bridgestone's switch from grooved rubber to slicks in 2009 and the never-to-be-raced combination of 2008 downforce levels and 2009 tyres resulted in an unofficial track record of 1:18.339. Nearly a decade later and Formula One finally caught up with itself when Red Bull bolted Pirelli's new hyper-soft tyres on Daniel Ricciardo's car and the Australian set a 1:18.047 on Wednesday. That record lasted just 24 hours before Sebastian Vettel switched to the hyper-softs for the first time and gave a clear indication of Ferrari's pace by setting a 1:17.182. Had he hooked up his three best sectors on Thursday morning he would have been even faster with a 1:17.026.

The Italian team was not shy about stretching its legs in the second week of testing and took to the track on both Thursday and Friday to focus on a series of performance runs on Pirelli's three fastest tyres -- the super-soft, ultra-soft and finally the hyper-soft. There's little doubt that these were low-fuel runs, typically sticking to five or six laps and alternating between push laps and cool-down laps. In terms of outright pace, Vettel's fastest lap left Ferrari 0.6s quicker than its nearest rival, McLaren's Fernando Alonso, 0.9s faster than Ricciardo's best effort in the Red Bull and 1.2s faster than Mercedes' quickest time. But don't rush down to the bookies just yet...

GPS data indicates that despite the 0.9s difference between Vettel's lap and Ricciardo's, the Red Bull was the quicker car in all of the Circuit de Catalunya's 16 corners. That's impressive given that Ricciardo's time was set with a day's less rubber on the track and also indicates that Ferrari had its engine turned up in order to make up the difference on the straights.

A better indication of Red Bull's true pace came on the final day when Ricciardo set a 1:18.327 on the super-soft tyres. Based on data amassed by Pirelli, the gap in performance between the super-soft compound and the hyper-soft was estimated to be around 1.3-1.4s. Therefore, Ricciardo's tyre-corrected lap time comes out somewhere between a 1:16.9 and a 1:17.0.

But Ricciardo wasn't alone in setting a quick lap on the super-softs on the final day. Kimi Raikkonen also clocked an impressive lap on the red-striped rubber on Friday that went largely unnoticed once it was usurped by his hyper-soft run later in the morning. The Ferrari registered a 1:18.238 on the super-soft which, tyre-corrected, puts him in 1:16.8-1:16.9 territory. Perhaps that was the extra pace Raikkonen was referring to when he spoke to the press on Friday.

And then there's Mercedes. The world champions didn't run on the hyper-soft tyres during the two weeks of testing and its best lap was only a 1:18.400 set by Lewis Hamilton on the ultra-softs. But as with Red Bull and Ferrari, the most telling lap time came on a harder compound; in this case the medium tyres used by Valtteri Bottas on the final day. Using the white-striped tyre, he set a 1:18.825, putting him sixth fastest on Friday afternoon. It was nothing special on the face of it, but based on Pirelli's estimated 2.5s-2.6s performance delta between the medium and hyper-soft compounds, Bottas may have been able to dip down to a 1:16.2 had Mercedes used the hyper-soft.

Much of that theoretical advantage hinges on Mercedes successfully banishing some of the vices of last year's W08 from the new W09. When it came to bolting on Pirelli's softest compounds in 2017, Mercedes routinely failed to find the same performance gain as its rivals. It was also always quick relative to its rivals on the 2017 soft tyre, which formed the basis for this year's medium compound. Pirelli's estimated performance deltas referred to above are based on a spread of data across all teams during the two weeks of testing, and if Mercedes sits at one end of that data it may be unable to find the same gains as Ferrari and Red Bull when it comes to bolting on the softer rubber in a competitive environment.

Of course, the standard F1 testing caveats should also be considered when looking at the fastest laps towards the end of the week. All of the above could have been skewed by one of the three teams carrying a few extra laps of fuel to mask their pace -- a practice referred to as sandbagging in the F1 paddock.

Long-run pace

To get a better idea of how the cars stack up on similar fuel levels, we need to look to the race simulations completed by the top three teams towards the end of testing. In order to complete a race distance at the Circuit de Catalunya without returning to the garage to refuel (as all three teams did on Thursday), the cars would have needed to be fully fuelled -- or close to it -- when they embarked on their race sims.

Comparing the race sims of the top three teams, Bottas in the Mercedes was on average 0.55s per lap faster than Vettel in the Ferrari and as much as 0.8s per lap faster than Max Verstappen in the Red Bull. The Mercedes and Red Bull stuck solely to the medium tyres across three stints whereas Ferrari used a more realistic strategy of super-soft, medium, medium. This may have been a factor in skewing the opening stint in favour of Mercedes and Red Bull, but in the final stint -- when all cars were on mediums -- Mercedes still held a sizable advantage.

Ferrari's pace plateaued in the 1:21s during the final third of the race while Bottas was able to pump in consistent laps in the 1:19s. Vettel set a single quick lap in the 1:19s on lap 65 of 66, which suggests something beyond tyre degradation was limiting the Ferrari towards the end of the race sim. One theory is that Vettel was forced to fuel save towards the end of the race sim, and if true that could prove problematic in Australia given that Albert Park is one of the more demanding tracks for fuel consumption on the calendar.

But Mercedes may also have a weakness over longer runs. Earlier in the week there were signs that the W09 was struggling to keep Pirelli's softer compounds in one piece. A number of long runs on the soft tyres resulted in significant blistering on the rear tyres, which in turn manifested in high levels of degradation in performance. Ferrari did not appear to suffer the same amount of blistering, suggesting Mercedes could be vulnerable over a race distance in hot conditions. What's more, Pirelli's announcement on Thursday that it plans to bring the super-soft, soft and medium compounds to the Spanish Grand Prix in May, suggests the tyre supplier will not be afraid to go aggressive in its tyre choices this year.

So who's quickest?

All the data suggests Mercedes has an advantage in both short-run and long-run pace ahead of Melbourne. A question remains over the depth of that advantage on Pirelli's softer compounds, especially in hotter conditions, but for now the four-time world champions are the favourites as the paddock packs its bags to go down under. Ferrari and Red Bull appear to be more closely matched, with GPS data suggesting the Red Bull could put its nose ahead if Renault can provide its promised Q3 qualifying mode in time for the start of the season.

But with an upgrade arms race expected throughout the opening races, expect the order to remain fluid as F1 visits different track layouts and races in different conditions over the first four rounds. Even the results of Melbourne, a parkland circuit with the potential for mixed weather, will not necessarily translate to the next three rounds in Bahrain, China and Azerbaijan. Each of the top three cars appear to have strengths and weaknesses and much will depend on how the teams play to these when racing gets underway later this month.