The AI borders on hopeless when it comes to naval invasions. For that reason, clearing your continent goes a long way towards ensuring survival. (Though you can still lose to a space shuttle launch or a cultural victory.)
It's nice that this topic helps not just me.This is a great learning thread.
The math of stacking/not stacking workers in roading was well worth the read.
You're proabably aware, Ionize, that railing irrigation and mines improves their bonus. this becomes especially vaulable in specialist farms. With the extra food from railed, irrigated tiels you can feed a large population of specialists.
I find replacable parts is so useful (artilary, faster workers, infantry) that it makes sense to research it.
Thanks for the advice Spoonwood. Here are some comments to that:Alright, looking at your latest save:
1. You can't get the benefit of railroads on a particular square until you have a mine or irrigation up already. So, the workers roading unmined mountains would have done better to mine first, and then railroad.
2. I may have mentioned this idea before here, but let's take a look at Sumer:. Do you see all that food on the right side of the box? Actually, if you put your cursor over it, it'll say "food stored". Anytime you have "food stored" and you have your city at maximum desired size, you can often transform that food into shields via mining/foresting *before* railroads. So, if you mined some of those plains, you could get 4 turn infantry instead of 5 turn infantry.![]()
3. You still have barracks in some cities on your home continent where you have Sun Tzu's Art of War (zoom to say Ur and right-click on barracks, and you'll get an option to sell). You can sell these barracks for some gold.
4. I don't understand what you're doing with your ships. Take a look here:
![]()
I've circled the spots where you would optimally (for warring) have cities. If you have cities in those spots, your galleons could start at the northern spot, and more quickly reach the southern spot to unload your units to your new areas, and then come back to unload more units also.
5. You don't need to do research, so go with tax farms instead of scientist farms.
6. Mobilize for war.
Just some suggestions. As Scott Johnson says, if we're not having fun, we're doing it wrong!
It only workers out as slower when you consider an individual tile. It's NOT slower when you consider things empire wide. If you road a single tile with say 2 industrious workers, and you stack them you spend 2 turns in movement, and 2 turns in roading. So, you spend 4 worker turns doing this, and 2 game turns to build that one road. If you just use one industrious worker here, you spend 1 turn in movement, and 2 turns in roading. Thus, you spend 3 worker turns and 3 game turns in building that road. So, you do get that particular road slower in terms of game time if you don't stack, but if you do stack workers, you've basically "payed" more for getting that job done quickly. If you want to build 2 roads, and stack the entire time, it takes you 2 game turns to build the first road and 2 game turns to build the second, for 4 total turns. However, if you don't stack them, it takes you 3 total game turns, since two workers move to two distinct squares on the first turn, and then both road on the 2nd and 3rd turns.
Now, what does that mean more practically in terms of an empire-wide game?
Well, let's say you have 10 cities, and 80 tiles to road (say you have 12 workable squares per city minus some coast, and some roads already built). You have 20 workers. Now you can
1. Build 10 roads every 2 turns if you stack those workers. In this case, you spend 16 total game turns building those 80 roads, since (10*8)=80 the number of roads needed and (2*8)=16.
2. Or if you don't stack those workers you can build 20 roads every 3 game turns. So, it takes you 12 total game turns to build those 80 roads, since (20*4)=80 for the number of roads needed, and thus (3*4)=12 for the number of game turns needed.
So, empire wide, using workers to build roads doesn't pay off in terms of stacking workers in general.
How does that happen again? Well, if you stack the workers you spend 2 worker turns to move into a square, and 2 worker turns to road, for 4 worker turns for every road built. Thus, you need 320 worker turns to build all those roads. If you do not stack those workers, it takes you 1 worker turn to move into a square, and 2 worker turns to road. So, you need 240 worker turns to build all those roads.
Things get even worse with non-industrious workers, since if you stack them in 3s, then you spend 3 worker turns in just movement instead of 1. And even if you spend only one turn moving on roads to get to an unroaded spot, with regular workers, that only takes 2 worker turns in terms of movement instead of the 3 needed if you stack those workers. The problem gets even worse when you start talking about slaves.
How does it come as a problem to get to unroaded tiles in the early game especially? Just build just enough of a network between your cities initially, and you'll usually have plenty of open unroaded spots that your workers can get to without wasting movement. Or build such a network, and don't put all that many mines/irrigation down early, just put roads and then have other workers that move a turn or two on those squares, mine/irrigate them, and then move to the unroaded square after that. Also, you lose even more worker turns by roading anything other than a flatland, hill, or mountain square before clearing it.
Fiddlin_Nero said:I Recently did some testing of my own with PTW, found I could road with 3 workers 3 tiles in 5 turns working each tile with a single worker, 1 turn to move, 4 turns to road. Stacking, it took 3 separate turns for moves and 3 single turns to road for a total of six. Not a significant improvement in my mind.
Huh? First off, you sure about this? I know Cracker's numbers differ from yours. I've also checked this myself in PTW and what you say simply doesn't hold. A single non-industrious worker on flatland needs 3 turns to complete a road if he's already there, and an industrious worker needs 2 turns.
Second, you've only thought about 3 tiles. You have far more many squares to road than 3 tiles. And basically the difference here grows wider and wider. If you have 6 tiles to road, the unstacking method saves 2 turns, 9 saves 3 turns, and in general n turns saves floor((n/3)) turns, or exactly (n/3) turns if a multiply of 3.
Try doing a test for 50 or 60 or 100 tiles.