It's summer break time in Formula 1. That means it's time for the traditional "there are too many races" articles from the journos. I quite like the BBC's choose your races for a 16 race season thing because it's something a little different to the usual.
I like "different to the usual", and this is my contribution to that.
What if we don't get rid of races, but, instead, find a better order in which to race them to save time, distance and cost.
There is a shiny app dedicated to solving the travelling salesman problem. Simplifying horribly, the travelling salesman problem asks "what is the most efficient route round x number of points?" The app, here, does the hard work for you. It's really good because it copes with cross-ocean travel, which not all apps for the travelling salesman problem do.
To find the most efficient route to cover all the races of the 2018 F1 season, I tried to put in the locations of the races.
That was when the first problem occurred. Completely understandably, the app only includes cities, and several of the racetracks are not in or near cities. Therefore, I've gone with the nearest city to the locations of the races. The races affected by this are:
Race - closest city
Bahrain GP - Manama
French GP - Marseille
Austrian GP - Ljubljana (stop laughing in the back)
British GP - Leicester
German GP - Stuttgart for Hockenheim
Belgian GP - Brussels
Italian GP - Milan
Japanese GP - Nagoya
This was the map that the app produced:
In case you can't read the labels, the order the program puts the locations in is Melbourne, Singapore, Nagoya, Shanghai, Abu Dhabi, Manama, Baku, Sochi, Ljubljana, Budapest, Milan, Barcelona, Marseille, Monte Carlo, Stuttgart, Brussel, Leicester, Montreal, Austin, Mexico City, Sao Paolo.
That's start in Australia, fly over to Asia and race the Asian and Middle Eastern tracks. The route then moves across Europe from East to West, from the UK go across to Canada, down the Americas and finishes in Brazil.
The one downside to the solution is that it has to return to its starting point, because it solves the travelling salesman problem. That makes it more likely that an outlying point, in this case Melbourne, will be the starting point.
But, and this is the important thing, doesn't the order that it proposes better from a logistical point of view? Doesn't it make more sense than jumping from Australia to Bahrain, only to jump back to China. Following that, in the real calendar, there's a slow drift across from Asia to Europe, only for the circuit to leap from Monaco to Montreal, then back to France followed by the rest of the European races bar one. Then, in the real world, we go from Europe to Singapore, back to Europe, then out to Japan, which is two long flights the smaller teams could probably do without. From Japan, it's over to the US, then down, through Mexico to Brazil, which makes sense. Going from there to Abu Dhabi does not.
The app's more efficient plan avoids a number of weeks where the teams barely have time to land before they have to take off again. The simplified route only crosses the Atlantic once (as opposed to four times in the actual calendar) and doesn't cross the Indian Ocean at all (as opposed to twice in the real calendar). It avoids a trip to Singapore and back in the middle of the season. It also avoids a 15+ hour trip from Japan to Texas. Those changes should make logistics easier for the teams and mean that the pit crew (and other, even less heralded team members) might have chance to see their families once in a while.
Of course, the app is only bound by geography and the information entered. In reality, the layout of the season is determined by money and politics. Certainly, the US Grand Prix is at that point in the season just so there's a chance that the Championship might be decided on US soil so there might finally be major US commercial interest. It also wouldn't surprise me if Abu Dhabi's contract is to host the final race of the season, not any old race.
Someone with more mathematical / coding skill than me would probably be able to come up with a way of running a similar algorithm with a fixed final point. That would mean you could find the most efficient route is if it has to finish at Abu Dhabi. Whatever that calendar is, it won't be the one that F1 is using at the moment. Going all the way over to the Americas only to come back to the Middle East really adds mileage.
It will be interesting to see how Liberty deal with this when Abu Dhabi's contract is up for renewal because I don't think anyone likes where Abu Dhabi is on the calendar.
It's not just because I'd rather see titles decided in Brazil. Given the way the last few seasons have gone, they title is more likely to be wrapped up before the last race. My objection to Abu Dhabi having the last race is that I'd rather the last race be somewhere where we might get actual racing action.
I like "different to the usual", and this is my contribution to that.
What if we don't get rid of races, but, instead, find a better order in which to race them to save time, distance and cost.
There is a shiny app dedicated to solving the travelling salesman problem. Simplifying horribly, the travelling salesman problem asks "what is the most efficient route round x number of points?" The app, here, does the hard work for you. It's really good because it copes with cross-ocean travel, which not all apps for the travelling salesman problem do.
To find the most efficient route to cover all the races of the 2018 F1 season, I tried to put in the locations of the races.
That was when the first problem occurred. Completely understandably, the app only includes cities, and several of the racetracks are not in or near cities. Therefore, I've gone with the nearest city to the locations of the races. The races affected by this are:
Race - closest city
Bahrain GP - Manama
French GP - Marseille
Austrian GP - Ljubljana (stop laughing in the back)
British GP - Leicester
German GP - Stuttgart for Hockenheim
Belgian GP - Brussels
Italian GP - Milan
Japanese GP - Nagoya
This was the map that the app produced:
In case you can't read the labels, the order the program puts the locations in is Melbourne, Singapore, Nagoya, Shanghai, Abu Dhabi, Manama, Baku, Sochi, Ljubljana, Budapest, Milan, Barcelona, Marseille, Monte Carlo, Stuttgart, Brussel, Leicester, Montreal, Austin, Mexico City, Sao Paolo.
That's start in Australia, fly over to Asia and race the Asian and Middle Eastern tracks. The route then moves across Europe from East to West, from the UK go across to Canada, down the Americas and finishes in Brazil.
The one downside to the solution is that it has to return to its starting point, because it solves the travelling salesman problem. That makes it more likely that an outlying point, in this case Melbourne, will be the starting point.
But, and this is the important thing, doesn't the order that it proposes better from a logistical point of view? Doesn't it make more sense than jumping from Australia to Bahrain, only to jump back to China. Following that, in the real calendar, there's a slow drift across from Asia to Europe, only for the circuit to leap from Monaco to Montreal, then back to France followed by the rest of the European races bar one. Then, in the real world, we go from Europe to Singapore, back to Europe, then out to Japan, which is two long flights the smaller teams could probably do without. From Japan, it's over to the US, then down, through Mexico to Brazil, which makes sense. Going from there to Abu Dhabi does not.
The app's more efficient plan avoids a number of weeks where the teams barely have time to land before they have to take off again. The simplified route only crosses the Atlantic once (as opposed to four times in the actual calendar) and doesn't cross the Indian Ocean at all (as opposed to twice in the real calendar). It avoids a trip to Singapore and back in the middle of the season. It also avoids a 15+ hour trip from Japan to Texas. Those changes should make logistics easier for the teams and mean that the pit crew (and other, even less heralded team members) might have chance to see their families once in a while.
Of course, the app is only bound by geography and the information entered. In reality, the layout of the season is determined by money and politics. Certainly, the US Grand Prix is at that point in the season just so there's a chance that the Championship might be decided on US soil so there might finally be major US commercial interest. It also wouldn't surprise me if Abu Dhabi's contract is to host the final race of the season, not any old race.
Someone with more mathematical / coding skill than me would probably be able to come up with a way of running a similar algorithm with a fixed final point. That would mean you could find the most efficient route is if it has to finish at Abu Dhabi. Whatever that calendar is, it won't be the one that F1 is using at the moment. Going all the way over to the Americas only to come back to the Middle East really adds mileage.
It will be interesting to see how Liberty deal with this when Abu Dhabi's contract is up for renewal because I don't think anyone likes where Abu Dhabi is on the calendar.
It's not just because I'd rather see titles decided in Brazil. Given the way the last few seasons have gone, they title is more likely to be wrapped up before the last race. My objection to Abu Dhabi having the last race is that I'd rather the last race be somewhere where we might get actual racing action.
