tjs282
Stone \ Cold / Fish
Yes, those screenshots you posted (and doing it like this was fine, by the way) do put a slightly different complexion on the situation you described earlier.I wanted to go on the offensive and pillage their resources and improvements, It's just that I couldn't at this point I had to protect the border city and It had barely enough units I also putted a wall and barracks, now that I think about it perhaps I should just left my others cities undefended with only a horseman and put all my military might in the border city to push for the iron.
When you roll a really cold, stringy, 'fjord-y' (80% water?) map like that, so that most/all of your cities end up being coastal (this problem is particular prevalent on Tiny/Small maps -- another reason to go Standard or Large-r), it does become more problematic/expensive to defend them all, since you have so many different points where the AI might be tempted to drop off boat-borne units. But if you found at C-x-x-C — as you did


Spoiler Another neat trick... :
The AI knows where all your (visible) units are, at all times, so if all your towns are equally well-defended, it's still a crapshoot to guess where the AI is going to show up (it likes Lux- and Strat-resource towns though). So something (else) you might want to consider for your next game(s), and especially if you ever roll another map like that, is the 'bait-city' trick, which basically allows you to 'tell' the AI where you would 'prefer' to be invaded!
By garrisoning all your coastal/border towns with (at least) one unit, but leaving an inland town ungarrisoned (if possible), this 'bait-town' will then act as an AI-unit magnet (instead of that coastal Mountain!): and it will tend to drop off all its seaborne invasion-forces outside that one town. Then all you need to do is to keep sufficient spare attackers/bombardiers (depending on your enemy's naval-transport capability) nearby but not in that town, to injure/kill all the landed units before they can attack your town(s) or pillage your tiles. Looking at your screenies, Nagoya is/was pretty much an ideal candidate for this trick: with its 'defenders' stationed on the adjacent Hills, you would have 'encouraged' the AI to (sail a long way, just to) land its units on the Tundra, where they get no real defensive bonus.
By garrisoning all your coastal/border towns with (at least) one unit, but leaving an inland town ungarrisoned (if possible), this 'bait-town' will then act as an AI-unit magnet (instead of that coastal Mountain!): and it will tend to drop off all its seaborne invasion-forces outside that one town. Then all you need to do is to keep sufficient spare attackers/bombardiers (depending on your enemy's naval-transport capability) nearby but not in that town, to injure/kill all the landed units before they can attack your town(s) or pillage your tiles. Looking at your screenies, Nagoya is/was pretty much an ideal candidate for this trick: with its 'defenders' stationed on the adjacent Hills, you would have 'encouraged' the AI to (sail a long way, just to) land its units on the Tundra, where they get no real defensive bonus.
Spoiler Regarding chokepoints :
Controlling chokepoints is good, but it's not always preferable to block them off completely, because the AI then just bypasses them by sea (although it does suck more at this, you may then be back in the situation where you can't be sure where it's going to drop its units). With reference to the above, if you have a town near but not blocking a chokepoint, with Walls (or Pop>6) plus multiple garrisons/bombardment-units, but you have another (bait-)town beyond it, then the AI will generally bypass the choke-town in favour of going after the softer target — even though it will take longer to get there, and take harassment all the way — but crucially, it will send all its units by land, allowing you to ping them with your Cats/Trebs, and then pick them off with your Archers/Horsemen as they (try to) retreat.
It's also not a good idea to build Fortresses/Barricades directly adjacent to your own (chokepoint-)towns, if you can possibly avoid it! As a general rule, I don't build Fortresses (or Barricades), but if/when I have the occasion to do so (or to steal one from the AI), I leave at least one tile between it and my nearest town (C-x-F/B) for 2 reasons: (1) Fortresses give Zone-of-Control to all units stationed there, so AI-units trying to bypass them get injured on the way in (and out); and (2) if I have extremely bad luck and lose that strongpoint, the AI units will still have to cross open ground before they can attack my town (rather than being able to attack directly from the Fortress).
It's also not a good idea to build Fortresses/Barricades directly adjacent to your own (chokepoint-)towns, if you can possibly avoid it! As a general rule, I don't build Fortresses (or Barricades), but if/when I have the occasion to do so (or to steal one from the AI), I leave at least one tile between it and my nearest town (C-x-F/B) for 2 reasons: (1) Fortresses give Zone-of-Control to all units stationed there, so AI-units trying to bypass them get injured on the way in (and out); and (2) if I have extremely bad luck and lose that strongpoint, the AI units will still have to cross open ground before they can attack my town (rather than being able to attack directly from the Fortress).
Spoiler That was the good cop, now here's the bad cop... :
Your terrain-improvement and building-assignment schedules leave a lot to be desired, though. Irrigating Grass (without Wheat or Cows) is not usually a good idea, especially non-shieldGrass tiles, and especially if you started doing that while still under Despotism: Grass-tiles should almost always be mined in the early game, because (under Despotism) you'll still only get the 2 food per turn from it, so you might as well get a shield at the same time (I try to get at least 1 shield per worked tile in the early game, and preferably 2). Once you go to Republic, you might want to consider irrigating some shieldGrass-tiles, because that way you still harvest a shield per tile, while getting faster growth.
Under Republic, getting your towns — especially core-towns — to Pop7 or more should always be highest priority, to prevent your economy from getting tanked by excess unit-support costs. So WHY is Ise stalled at Pop5 (despite its Harbour)?!? It's a core-town, with massive (water-trade) commerce-potential, so it needs an Aqueduct + Lib and/or Market, to take advantage of that; conversely, its shield-potential is minimal, so it certainly doesn't need a Barracks! Kagoshima should also have got an Aqueduct and Pop12 long since (with its Grass-tiles mass-mined not irrigated! How much food is it wasting?!). Sapporo needs a Harbour to grow (and possibly also a Courthouse), to make it worth even thinking about building a Lib there (I don't usually bother building Libs before Pop5-6 in any town). And you control 3 Luxes, so you shouldn't need any Temples anywhere to keep order (Marketplaces would have given you the same Happiness, but earned you income at TAX%=80, instead of costing you!).
Also, it is generally a bad idea to tie up all your towns with building improvements simultaneously, especially when a war is imminent (e.g. no more free space on your Continent left to Settle) — or already happening! Because when war breaks out, you'll start losing units (even the best-conducted campaigns have casualties), and if you haven't got enough gold to rush the remaining shields on all those builds-in-progress, you will then face a choice of 'waste shields everywhere by switching to unit-production, or start losing ground/towns' — which is actually no choice at all, since your towns are always much more valuable than any shield-wastage. That doesn't make eating all that wastage any less painful, though (and you'll have to start the project(s) all over again after the war's over).
So it's usually better to keep most (Barracks-)towns building (attack)-units, and only if you're approaching your unit-maintenance limits, then fit your (useful!) buildings into the schedule, a couple of towns at a time, ideally starting with the cheapest (needed!) improvements in your most important towns. That way, when war breaks out you should have plenty of towns available to switch over to (or continue) producing a constant stream of replacement units immediately, with minimal wastage — and you're more likely to be able to keep those towns in the first place!
Under Republic, getting your towns — especially core-towns — to Pop7 or more should always be highest priority, to prevent your economy from getting tanked by excess unit-support costs. So WHY is Ise stalled at Pop5 (despite its Harbour)?!? It's a core-town, with massive (water-trade) commerce-potential, so it needs an Aqueduct + Lib and/or Market, to take advantage of that; conversely, its shield-potential is minimal, so it certainly doesn't need a Barracks! Kagoshima should also have got an Aqueduct and Pop12 long since (with its Grass-tiles mass-mined not irrigated! How much food is it wasting?!). Sapporo needs a Harbour to grow (and possibly also a Courthouse), to make it worth even thinking about building a Lib there (I don't usually bother building Libs before Pop5-6 in any town). And you control 3 Luxes, so you shouldn't need any Temples anywhere to keep order (Marketplaces would have given you the same Happiness, but earned you income at TAX%=80, instead of costing you!).
Also, it is generally a bad idea to tie up all your towns with building improvements simultaneously, especially when a war is imminent (e.g. no more free space on your Continent left to Settle) — or already happening! Because when war breaks out, you'll start losing units (even the best-conducted campaigns have casualties), and if you haven't got enough gold to rush the remaining shields on all those builds-in-progress, you will then face a choice of 'waste shields everywhere by switching to unit-production, or start losing ground/towns' — which is actually no choice at all, since your towns are always much more valuable than any shield-wastage. That doesn't make eating all that wastage any less painful, though (and you'll have to start the project(s) all over again after the war's over).
So it's usually better to keep most (Barracks-)towns building (attack)-units, and only if you're approaching your unit-maintenance limits, then fit your (useful!) buildings into the schedule, a couple of towns at a time, ideally starting with the cheapest (needed!) improvements in your most important towns. That way, when war breaks out you should have plenty of towns available to switch over to (or continue) producing a constant stream of replacement units immediately, with minimal wastage — and you're more likely to be able to keep those towns in the first place!