The nearer to the top of the ladder you are, the more difficult it is to take the step to the next rung. The same is true of Formula 1 car development.
When the regulations change significantly - as they did for the 2014 season and, before that, in '09 - someone always gets the jump on the others. For the current set of rules that team is Mercedes, which had, previously, underperformed after taking over the Brawn team that had so much success the last time the rules changed.
Everything points to the Mercedes power unit package being the best in the pitlane, but I have no doubt that the chassis has backed this up. The car is competitive on all types of circuit and it usually gets the best out of the tyres over one lap. It's also been as good as any other chassis with regard to degradation, so overall it's a solid all-round package that will take some beating.
If any team can do that, it has to be Ferrari. On a few occasions in 2015 it took the fight to Mercedes but didn't have the last bit of performance needed to do so consistently. So there were a few great days in Malaysia, Hungary and Singapore, but the championship was never on.
From 2014 to '15 Ferrari made up a lot of ground, finishing second in the constructors' championship after the previous year's very distant fourth. I believe most of that gain came from power unit improvements as, at the start of 2014, Mercedes caught everyone with their trousers down on that side.
During the 2015 season, Ferrari didn't really close the gap to Mercedes. Some tracks suited one chassis a little better than the other, but based on average fastest laps of the weekend, there was still about 0.8 per cent between them in terms of pace. So around a theoretical 90-second lap, that equates to 0.72s.
Ferrari was able to spring the odd surprise in '14, but couldn't sustain a challenge against Mercedes © LAT
Ferrari will have been pretty confident about which areas needed to be worked on. It will have drilled deep into every bit of data on Mercedes from the 2015 season and compared it to its own performance. But what is key is deciding what direction to go in to close that gap.