The first half of the season has shown that, under the new rules, qualifying has become a part of the races itself, and strategy is more important than ever before. But despite the variety of tactics tried by the teams so far, there is only one that has proved the most efficient. Atlas F1's Richard Barnes analyses the 2003 strategic situation

At first glance, Sunday's Canadian Grand Prix was a mirror image of the same race in 2001. Two years ago, Ferrari's Michael Schumacher led the race from pole, with brother Ralf following closely for the first stint in his Williams-BMW. Crucially, Ralf was able to run five laps longer before pitting, and used the fast low fuel load laps to leapfrog his brother for the win. On Sunday, Michael returned the favour. However, this time a single lap worth of fuel was enough to do the job for the older Schumacher.

To the casual observer, the two races would have seemed almost identical - two talented brothers scrapping for the narrowest of margins in dominant cars, on a track at which both excel, with both winners having to use pitstop strategy (rather than on-track passing) to secure the victory from behind. Both were titanic tactical struggles. Yet even the engaging chess match of Canada 2001 was dwarfed by the strategic depth and consequences of Sunday's riveting four-car battle.

The new qualifying system has made the difference, and elevated F1 strategy to an unprecedented level. The midway point of the 2003 season is an apt time to reflect on the first half-season of the new rules, and judge the strategic winners and losers.

There can be no doubt that the new qualifying system has been the single most profound factor in the development of this year's championship battle. When the new rules were first announced, the most obvious implication was that of driver error. With the one lap shootout, a single missed braking point could send a hapless competitor from potential pole position back into the company of the Minardis and Jordans at the wrong end of the grid.

Kimi Raikkonen has made this mistake twice, and paid dearly both times. In Spain, he started from the back of the grid and never even made the start-finish line before his collision with Antonio Pizzonia's stationary Jaguar. In Canada, the Finn chose the pitlane instead of P20, and battled to a disappointing sixth, unable to influence the race in any meaningful way.

Renault's Fernando Alonso is the only other leading driver to have suffered from such a gross qualifying error, when a spin in Austria demoted him to the back of the grid. So, surprisingly, driver error has not been as big a factor as feared initially. Instead, the new qualifying format's most dramatic impact has been in the strategic area of fuel loads.

This was always going to be an important strategic element, although most of the pre-season speculation centred around the possibility of the slower teams qualifying on fumes, holding the lead for a sponsor-delighting lap or two, and then vanishing back into the pack after a forced early fuel stop. Happily, all the teams have been professional and sensible enough to avert this farcical approach. Instead, the fuel load factor among the genuinely competitive teams has become the sport's most strategically important and intriguing element.

Under the new system, qualifying can no longer be viewed as a separate event from the race. With all cars required to run the same set-ups, tyres and fuel loads for the start of the race, the qualifying lap should rather be seen as a controlled and staggered first lap of the race. That leaves team strategists with a dilemma - to run light in qualifying and try for pole, or to run heavy and sacrifice grid spots for a longer first race stint. The first half of the season has given us a categorical answer as to which option is better. And the answer is - neither.

If any driver has epitomised the 'hare' approach this season, it's Ralf Schumacher. On three occasions (San Marino, Monaco, Canada), the Williams driver ran light in qualifying and jumped to an early race lead. On all three occasions, he surrendered the lead, and consequently the race, by pitting before his major rivals - brother Michael in San Marino and Canada, and teammate Juan Pablo Montoya at Monaco. Discounting the chaotic races in Brazil and Australia, no driver this season has managed to win a race after being the first car (among the leading pack or otherwise) to pit.

That may be seen as a simplistic conclusion, as there are obviously other factors, such as tyre balance, in play. And naturally, all cars have slightly different race pace. Yet Canada illustrated perfectly that the underlying premise often holds good - the first man to blink (and pit) is invariably the loser. Ralf Schumacher didn't lose to Michael in Canada because he had a slower car, an unbalanced set of tyres during one stint, or any other mechanical shortfall. He lost because, in the Saturday gamble, he took on one lap worth of fuel less than his brother. It may have given him a deserved pole position, but it ultimately doomed his race chances.

There are just too many risky downsides to the hare approach. Pitting earlier than the pack will often put the leader back into midfield traffic, who are running longer (and slower) first stints - at the very time that his closest rivals are released into clear air. The driver who pits early also has the burden of a heavy fuel load immediately after the stop, again at the very time that his main rivals are running on fumes and setting fastest race laps. Under such circumstances, even one lap is often enough to reverse the order, as happened on Sunday. Often, races are described as being won or lost by inches or fractions of a second. Sunday's Canadian Grand Prix was won and lost by a few litres of fuel.

At the other extreme, the tortoise approach of David Coulthard has not paid off either. Ever since hooking up with Mercedes in 1996, McLaren have used their oversized fuel tanks to run super-long stints and let the race come back to them. Under the old qualifying format, it was a viable gambit. They could qualify well up in the first three rows of the grid, stuff the car full of fuel before the start, and afford to let Ferrari or Williams race off into the distance. It didn't matter if they lost some time during the first stint, they'd make up the ground when their rivals were forced to pit earlier. As long as the silver and black cars were running in clear air and able to maintain their maximum pace, they were in the hunt.

That is no longer viable. All the teams have been tempted to run slightly light for qualifying in order to gain grid spots, and the result is that Coulthard is often stuck behind slower cars for the entire first stint of a race. Unable to run at his optimal pace, the advantages of the longer stint are eradicated. Even if Coulthard is able to get past the slower cars when they pit earlier, the damage has already been done.

The Scot's qualifying performances have been dismal under the new rules. After a promising start in Malaysia (4th) and Brazil (2nd), it's been downhill ever since. In four out of the eight races so far, Coulthard hasn't even made the top half of the grid. Team boss Ron Dennis' assertions that 'it's the right strategy for the race' are sounding less and less convincing.

So if neither the hare nor tortoise qualifying extremes are ideal for the race, then what is the ideal strategy? The answer is the sweet spot - light enough to make the first three rows of the grid, heavy enough that the driver's closest rivals have to pit one to five laps before he does. By now, it is predictable that Ferrari's strategic brains trust of Michael Schumacher and Ross Brawn would be the combination to find and define that sweet spot most often. And that is how it has played out.

In a season marked by some lopsided grid formations, Schumacher has been the model of consistency. Only once has the German failed to make the first three rows of the grid, and that by the smallest of margins (7th) in Brazil. Additionally, his consistent choice to run slightly longer first stints has led to four key 'pitstop passes'. Two were for the lead, against brother Ralf in San Marino and Canada - passes that would have proved near-impossible on the track.

The other two passes, against Ralf and Renault's Jarno Trulli at Monaco, elevated the German to an eventual podium finish - on a track where passing is impossible. While getting past Jarno Trulli's Renault in a superior F2003-GA might not be rated as much of a racing achievement (even at Monaco), it's worth noting that David Coulthard couldn't pass Trulli in an equally superior McLaren - thanks entirely to the unfortunate timing of his pitstops. Just two laps more fuel on either of his first two stints would have seen the Scot comfortably past the slower Renault and into a higher points finish. In a competitive season made even closer by the new points system, such incremental gains could prove vital by season's end.

It would be churlish to blame Coulthard's failure to pass Trulli on poor strategy, just as Schumacher's pass cannot be credited to strategic brilliance. In both cases, pure luck played a part. With teams not knowing what fuel loads the others are running in qualifying, it's no more than educated guesswork. Nevertheless, in a sport where teams and drivers often make their own luck, it's not coincidence that Schumacher and Brawn's educated guesswork is so often more accurate than their rivals'.

With Schumacher now leading the 2003 Championship for the first time, the launch of McLaren's MP4-18 cannot come fast enough if Kimi Raikkonen is to sustain his Championship challenge. Although, even with a new and faster machine at his disposal, the Finn cannot afford any more qualifying mistakes - especially not if Schumacher keeps hitting the strategic sweet spot.