This is a guide to playing Tiny Island maps in Civ 4. These are the most "extreme" forms of Archipelago map. Most of this guide will also be appropriate to Archipelago maps. This is written for the (patched) vanilla game. Hopefully the general strategy won't change significantly in other versions.
This guide aims to help intermediate players adjust to Tiny Island maps, by highlighting how and why conventional play styles need to change. Arguably the absolute optimum strategy for such maps is primarily Cultural. Culture is discussed as an advanced strategy, towards the end, because Cultural players still need to understand the quirks of Tiny Island maps.
Tiny island gameplay is genuinely different from standard large land masses. Tiny island maps:
Veterans may find these maps too easy, because the AI isn't a strong opponent. Newer players, especially those inclined to follow "standard" strategies, may be frustrated by the extra economic and naval management. However, for intermediate players, I think tiny islands can be a great place to learn: A chance to understand aspects of Civ IV that don't seem important to the beginner, but are important to understand on harder difficulty settings.
Guide structure:
1. MAP CONCEPTS
Before you play, consider what makes tiny island maps unusual, likely challenges, and optimum starting traits.
Most larger-land mass maps are production rich, and initially food and commerce poor. Tiny island maps are production poor, but food and commerce rich in the early-to-mid game. A biologically-farmed or fully-cottaged (towned) conventional map city will ultimately produce more food or commerce, but won't be doing that until quite late in the game. Tiny island maps allow cities to develop strong food and (especially) commerce earlier, with less need for worker improvements.
You'll need that financial advantage to offset increased maintenance costs: Your second or third cities may not even be on the same island as your Palace. Oceans may force expansion in one direction away from your Palace, rather in an area geographically around your Palace. Early-to-mid game expansion will be far more expensive than on a larger-land mass map.
Hills are almost as important in settlement location as resources, but still need to be carefully balanced against food, so that a city has enough excess food to be able to work those hills. Every settlement will be on the coast, with likely half of its productive tiles being water. With a Lighthouse, every one of those tiles will produce 2 food. Working those tiles will sustain the citizen, but won't ever grow the city. Fresh water is rare, so almost the only tiles to produce more than 2 food will be improved food resources. With 1 improved food resource (and especially with 2), excess food will be available, which is essential for serious production: Initially either by working mines, running a Specialist production economy, or using Slavery to complete production. Later Workshops or Lumber Mills can also contribute.
Since likely every city is a port, trade does not need roads (except to connect land-based resources to the nearest port city). This makes it much easier to share resources between your cities, and trade with AI civs. Religions also seem to use easier travel to spread more virulently across the map. While profitable for the religion's founder, rapid spread can make it harder to control religions for diplomacy.
Land-based resources are naturally confined by the lack of land. This creates unexpected quirks in geography. Normally common resources, like Corn, may now be rare. Other resources may be clustered on only 1 or 2 islands - it is not uncommon to find 3 or 4 identical plantation or camp resources (Sugar, Furs, etc) on a single tiny island. Copper tends to appear on the ice-caps, which often means that civs do not have access to it until after they discover Iron. All these quirks open up interesting diplomatic and miltary strategies. And start to demonstrate the importance of boats:
Boats are your scouts, your best defenders, your settler and invasion transports, maybe your workers. On other maps boats are almost irrelevant. Here, the ability to manage boats well is likely to be what wins the game. Naval technology has an especially high priority, because of the very dramatic improvements in the strength (and movement) of subsequent naval units: Frigates devastate Galleys, Destroyers sink Frigates. Promotions, strategy and sheer volume of ships are irrelevant. Even a defensive player will need ships to protect their economy.
Unfortunately for your enemies, the AI doesn't handle boats well. It settles land much slower than in comparative difficulty games. On Monarch or Emperor difficulty, perfectly good city locations can remain untaken until late in the late. The AI also has no effective strategy for defending its empire: Remote colonies are defended by small garrisons, often allowing a well-planned war to be executed and peace sought before the AI has managed to find a boat to transport a counter-attacking army. Indeed, even the most aggressive AIs have a habit of launching half-hearted wars that finish with the sinking of their first Galley! While such combat may sound easy, it is only easy because you are managing your boats better.
2. SETUP AND TRAITS
I dislike rigid recommendations on starting Civs - a good player should be able to make the most of any start, adjusting their strategy accordingly. That said, some starts are better than others. And in the original game, Washington was almost perfectly suited to the core strategy outlined in this guide: The 2 best traits, decent starting technology, and an ideal special unit...
Technologies are unlikely to be critical. Fishing is ideal because of the immediate Work Boats, guaranteed coastal start, and immediate Sailing research. Starting technology is most likely to influence early Wonder choice (discussed later).
Special units are not important. Pre-Medieval special units are difficult to use effectively on island maps, because you can't rush without boats. Gunpower units may be slightly more useful, especially the American Navy Seal (improved Marine). Even India's (non-military special) Fast Worker isn't as useful as it might seem, because there isn't much land to be improved.
The 2 easiest traits to play (based both on personal play, and from watching the AI play) seem to be:
Other traits are potentially advantageous, but less universally useful than the first 2:
Philosophical and Spiritual are interesting to play, but only fit advanced strategies (described towards the end):
I don't see that the expansions would change this analysis dramatically (although they clearly nerf Washington, and therefore make tiny island maps more balanced).
Finally, longer length games will allow more time to move boats around the map. Settlement and warfare generally take longer on island maps, however ships always move at the same speed.
3. EARLY GAME
This section describes your early moves and settlement, roughly until conflict breaks out - typically near the start of the medieval period. The fourth section below describes Wonders and technology during the early game.
Archipelago maps guarantee a coastal start. Indeed, with tiny islands it is often physically impossible to build an inland city! You will almost always start on a separate continent (islands) from other civilizations, but this is not guaranteed. This guide assumes separation - the early land grab and military would become considerably more important if there was no separation. Similarly, isolationist starts are theoretically possible, but unlikely. This guide assumes early contact is possible with most other civilizations via a combination of coastal tiles and territorial borders.
Before you engage autopilot and build a Warrior or Worker or Scout or Settler or whatever you normally do, stop and take a look around: Chances are you starting island is almost explored already. Surrounded by water, with no (or almost no) fog-of-war tiles on your island, you are perfectly safe until the Barbarians start spawning in boats. There may be no tiles a Worker can improve until you've research a few techs, so why prioritise the Worker? Don't be afraid to start building something that doesn't get completed until much later (like a Barracks) while you let the population grow and wait for the first technology to research.
On this map scouting means either Work Boats or Galleys used to explore. You will normally be able to reach most areas of the map while remaining in coastal (or territorial) waters, but the route won't be direct. Work Boats can be used to scout, but can't defend themselves. Galleys take slightly longer to build, but can often survive a Barbarian Galley encounter. Barbarian Galleys will not appear immediately, so Work Boats make excellent Scouts at the very start, when you need them the most.
The most important scouting objective is to be the first to circumnavigate the globe, which gains the permanent +1 naval movement speed. This is a huge boast in the early game, when otherwise ships only move 2 tiles per turn. It remains useful later - from faster troop travel to chasing down enemy shipping. Another civilization will beat you to this bonus if you delay. Ideally, build 2 Work Boat scouts as soon as you have researched Fishing, sending one east and one west. With luck, each will navigate half of the globe, which is good enough (the 2 boats don't even have to meet, merely reach the same longitude). Theoretically circumnavigation might not be possible until Caravels, but at least you'll discover that fact quickly, and know to prioritise Optics research.
While circumnavigating, you are likely to contact most of the other civilizations. Such early contact allows you to plan diplomatic alliances before the AI civs have contacted one another. This helps survive a potentially hostile map. At worst it gives you an advantage in any conflict: Often you will know where the enemy's cities are on the map, but the enemy will not have discovered your cities. Early contact and scouting (of their city locations) also helps establish trade routes. Trade routes are especially profitable on these maps, because almost every city is on another "continent".
Depending on the map, you may need to prioritise Writing research, for Open Borders agreements, allowing easier exploration and mapping trade routes. Of course, negotiating Open Border can be difficult (even impossible) if you have already converted to a "heretic" religion. Note that unless you have a relatively isolated starting position, it is rarely possible to stop other civilizations exploring and settling land by "blocking off" territory and denying Open Borders agreements. Also be aware of locations that can only be reached by moving Galleys through ocean tiles within territorial waters: Suddenly losing an Open Borders agreement can cut off supply lines in ways you probably weren't expecting.
Ideal settlement sites generally have a double-food resource, or a food resource and hills, or a food resource and another useful early resource (such as Iron). Food resources are important: Without at least one, you're likely construct a city that can do nothing except feed itself - a poor choice when so many thriving locations are available to be settled.
However, the distance maintenance penalty can become crippling. 90%, 80%, maybe even 50% research might be viable if it means you are expanding. But it's also easy to get into such terrible financial difficulty that the first genuine financial shock (like a war) might kill you. Colonising land close to the AI and then back-filling later settlements (in the gaps around your Palace) doesn't work well here: Water means that even cities which are on islands close to your Palace gain significant maintenance burdens. And if you leave those sites empty, the AI will often sail right round and settle them: Even well-developed cultural borders tend to leave gaps on a tiny island map.
Build order will vary by location. But Lighthouse (to quickly maximise food output) and Courthouse (once available) are recommended. A Granary is especially important in cities where production will be based on Slavery, since it speeds population growth. Harbors also have great health and trade benefits (once researched). Beyond those 4 will probably depend on what role the city is intended to perform.
In a larger-land mass games you might aim to build 1 or 2 Workers to improve every city. Here, 1 Worker will comfortably serve 2 cities. Indeed, if you wage war early you'll probably capture all the Workers you need. Instead of Workers, you should prioritise Work Boats. Ideally these can be sent out with new Settlers, and used to improve seafood tiles the same turn the city is founded. Remember, each Work Boat is a single-use improvement, but they cost half the price of a Worker and don't stop population growth while being produced.
On a tiny island map you can get away with almost no city defenses at the start. A single (ideally City Defense-promoted) Archer is normally adequate in the early game, mostly to avoid the unhappiness created by having no garrison! This minimalist approach to defense works because the AI doesn't scout water effectively (so is unlikely to discover many of your cities in the early game), nor does it launch war effectively over water (early wars tend to involve a single Galley), nor does it have the economy to produce enough units to contemplate serious warfare. Ultimately you will need slightly stronger defenses, and don't completely neglect garrison upgrades. But ideally use boats to intercept invaders before they land.
Lumber or Workshops? Unimproved forests provide an extra hammer when worked. Instead of chopping, it is tempting to work these tiles unimproved, and much later add Lumber Mills to them. Unfortunately, Lumber Mills appear relatively late in the game. When first researched (Metal casting) Workshops are only for the desperate (-1 food, +1 production) - consider them on an flat open terrain, if you have food to spare. However, Workshops slowly gain production bonuses with subsequent technology, and become very well suited to cities on this map. Remember, you normally can't irrigate fields, while Cottages are replaced by coastal water. With the State Property civic, the -1 food disadvantage is lost, and Workshops become better than Lumber Mills. As discussed later, it is highly likely you will beeline to, and run, State Property for its financial benefits alone. By the time we can build Lumber Mills, we'll have have Workshops which are broadly as good as Lumber Mills, why spare the trees? Forests are certainly useful very early in the game, but don't adopt the AI mindset that trees are too valuable to chop.
4. EARLY WONDERS AND TECHNOLOGY
Fishing, and then Sailing should be early research priorities, but not necessarily the first technologies to be discovered. Also consider technologies that allow you to exploit your starting position, especially to improve resources or exploit a (leader's) trait. Beyond that, initial technology research should be dictated by Wonders.
There are 2-and-a-half wonders that can radically change the outcome of the first half of the game (and consequently, the second half). You should aim for at least one:
1. The Great Lighthouse - +2 trade routes in all cities - on a map where likely every city is a coastal city, available very early. Research up to Sailing (and then build a normal Lighthouse), then research up to Masonry. Obsoletes with Corporation, however Corporation research can be avoided until you want something like Mass Assembly. You will need to beeline straight for The Great Lighthouse, especially on harder difficulties. That will almost certainly slow initial expansion.
2. The Colossus - +1 gold all water tiles - on a map where probably half the workable tiles are water. Research to Metal Casting, which also requires Pottery. Beelining may not be essential, because although the AI prioritises Metal Casting, it rarely goes there immediately due to the cost. Unfortunately The Colossus obsoletes with Astronomy, which cannot be avoided without immediately losing naval supremacy.
2.5 The Oracle - 1 free tech - then build The Colossus. While the free tech may be useful to you, completing The Oracle is often more beneficial because it prevents an AI civ picking Metal Casting quickly, and immediately going on to build The Colossus! For players founding early religions, The Oracle into Metal Casting into The Colossus is a solid strategy.
Both The Great Lighthouse and The Colossus provide very substantial financial bonuses, which should become obvious moving into the medieval period. An extra pair of trade routes can add over 5 (base) commerce to a large city. Same for water tiles. The Great Lighthouse is the clear winner for its potential longevity, and is also less restrictive - all cities trade, but not all cities will be dominated by working water tiles. However, The Colossus is less of a gamble. The ideal is to build both The Great Lighthouse and The Colossus. If you fail to complete both, consider military options to "acquire" them. Especially if they were built by Financial civilizations, who are otherwise very likely to power ahead technologically.
Unless you have a very specific strategy that benefits from them (such as a Cultural or religious approach), other pre-Renaissance Wonders can be ignored. Technology to focus on includes Compass (huge trade and health benefits of Harbors - far more useful than Aqueducts), Currency (more trade, more commerce), and Codes of Law (courthouses). Our aim is to build a robust economy, without which we can't effectively expand or develop. Don't be afraid to ignore technologies you may not be able to immediately use, like Agriculture, Horseback Riding, or some of the religious technologies.
This guide aims to help intermediate players adjust to Tiny Island maps, by highlighting how and why conventional play styles need to change. Arguably the absolute optimum strategy for such maps is primarily Cultural. Culture is discussed as an advanced strategy, towards the end, because Cultural players still need to understand the quirks of Tiny Island maps.
Tiny island gameplay is genuinely different from standard large land masses. Tiny island maps:
- Emphasise careful settlement and financial planning. With so little land, finding locations which can be turned into viable colonies can be challenging. That land tends to be separated from your Palace by sea, so distance greatly raises maintenance costs.
- Emphasise non-standard strategies. Prioritise technology and wonders that can seem almost irrelevant on other maps. New approaches emerge, like powerful one-tile island specialist cities, which we probably wouldn't consider if land wasn't scarce.
- De-emphasise huge militaries. Production is poor, so armies are hard to build and must be used more strategically. Unfortunately, the computer AI uses ships poorly and struggles to mount offensive actions over sea, which can make warfare feel somewhat trivial.
Veterans may find these maps too easy, because the AI isn't a strong opponent. Newer players, especially those inclined to follow "standard" strategies, may be frustrated by the extra economic and naval management. However, for intermediate players, I think tiny islands can be a great place to learn: A chance to understand aspects of Civ IV that don't seem important to the beginner, but are important to understand on harder difficulty settings.
Guide structure:
- Map Concepts
- Setup and Traits
- Early Game
- Early Wonder and Technology
- Medieval Game
- Late Game
- Advanced Strategies
- Learn More
1. MAP CONCEPTS
Before you play, consider what makes tiny island maps unusual, likely challenges, and optimum starting traits.
Most larger-land mass maps are production rich, and initially food and commerce poor. Tiny island maps are production poor, but food and commerce rich in the early-to-mid game. A biologically-farmed or fully-cottaged (towned) conventional map city will ultimately produce more food or commerce, but won't be doing that until quite late in the game. Tiny island maps allow cities to develop strong food and (especially) commerce earlier, with less need for worker improvements.
You'll need that financial advantage to offset increased maintenance costs: Your second or third cities may not even be on the same island as your Palace. Oceans may force expansion in one direction away from your Palace, rather in an area geographically around your Palace. Early-to-mid game expansion will be far more expensive than on a larger-land mass map.
Hills are almost as important in settlement location as resources, but still need to be carefully balanced against food, so that a city has enough excess food to be able to work those hills. Every settlement will be on the coast, with likely half of its productive tiles being water. With a Lighthouse, every one of those tiles will produce 2 food. Working those tiles will sustain the citizen, but won't ever grow the city. Fresh water is rare, so almost the only tiles to produce more than 2 food will be improved food resources. With 1 improved food resource (and especially with 2), excess food will be available, which is essential for serious production: Initially either by working mines, running a Specialist production economy, or using Slavery to complete production. Later Workshops or Lumber Mills can also contribute.
Since likely every city is a port, trade does not need roads (except to connect land-based resources to the nearest port city). This makes it much easier to share resources between your cities, and trade with AI civs. Religions also seem to use easier travel to spread more virulently across the map. While profitable for the religion's founder, rapid spread can make it harder to control religions for diplomacy.
Land-based resources are naturally confined by the lack of land. This creates unexpected quirks in geography. Normally common resources, like Corn, may now be rare. Other resources may be clustered on only 1 or 2 islands - it is not uncommon to find 3 or 4 identical plantation or camp resources (Sugar, Furs, etc) on a single tiny island. Copper tends to appear on the ice-caps, which often means that civs do not have access to it until after they discover Iron. All these quirks open up interesting diplomatic and miltary strategies. And start to demonstrate the importance of boats:
Boats are your scouts, your best defenders, your settler and invasion transports, maybe your workers. On other maps boats are almost irrelevant. Here, the ability to manage boats well is likely to be what wins the game. Naval technology has an especially high priority, because of the very dramatic improvements in the strength (and movement) of subsequent naval units: Frigates devastate Galleys, Destroyers sink Frigates. Promotions, strategy and sheer volume of ships are irrelevant. Even a defensive player will need ships to protect their economy.
Unfortunately for your enemies, the AI doesn't handle boats well. It settles land much slower than in comparative difficulty games. On Monarch or Emperor difficulty, perfectly good city locations can remain untaken until late in the late. The AI also has no effective strategy for defending its empire: Remote colonies are defended by small garrisons, often allowing a well-planned war to be executed and peace sought before the AI has managed to find a boat to transport a counter-attacking army. Indeed, even the most aggressive AIs have a habit of launching half-hearted wars that finish with the sinking of their first Galley! While such combat may sound easy, it is only easy because you are managing your boats better.
2. SETUP AND TRAITS
I dislike rigid recommendations on starting Civs - a good player should be able to make the most of any start, adjusting their strategy accordingly. That said, some starts are better than others. And in the original game, Washington was almost perfectly suited to the core strategy outlined in this guide: The 2 best traits, decent starting technology, and an ideal special unit...
Technologies are unlikely to be critical. Fishing is ideal because of the immediate Work Boats, guaranteed coastal start, and immediate Sailing research. Starting technology is most likely to influence early Wonder choice (discussed later).
Special units are not important. Pre-Medieval special units are difficult to use effectively on island maps, because you can't rush without boats. Gunpower units may be slightly more useful, especially the American Navy Seal (improved Marine). Even India's (non-military special) Fast Worker isn't as useful as it might seem, because there isn't much land to be improved.
The 2 easiest traits to play (based both on personal play, and from watching the AI play) seem to be:
- Financial: On other maps, Financial civs often start slow, and become powerful in the later-game. Here they start fairly powerful, and so often remain powerful. Especially if they build a key wonder (such as The Great Lighthouse or The Colossus). Tiny island maps have the potential to generate more income in the early-game, giving a winning technological lead by the middle of the game.
- Organised: Everywhere will build a Lighthouse and likely everywhere will build a Courthouse. The bonus to civic costs doesn't hurt either. Slow expansion or use of State Property (no distance maintenance) in the later-game may make other traits more attractive.
Other traits are potentially advantageous, but less universally useful than the first 2:
- Expansive's health bonus is less useful on tiny island maps because Harbors (with seafood) help in the mid-game, while city size is generally capped by food production in the late-game. Almost everywhere will eventually want a Granary and a Harbor, but perhaps not as much as they will want a Lighthouse and Courthouse.
- Industrious' Forges are clearly helpful on maps with so little native production. The Wonder production bonus opens up possibilities for building several key wonders (notably both The Great Lighthouse and The Colossus). But overall, Wonders aren't so important on a tiny island map: Lower overall map production means longer production time, means increased risk of getting beaten. Wonders with continental benefits are almost always worthless.
- Creative's instant culture bonus is useful when establishing colonies on isolated islands, especially in the early-game, where key resources are 2 tiles out and Culture cannot be "built". The building bonuses tend to be quite situational - and in the mid-game it is often easier to focus on gaining happiness resources, which are automatically distributed across your empire.
- Aggressive is hard to recommend because the military bonuses are partly negated by the need to ship units overseas. Early rushes are slow! However, if you can get a small army moving, the free unit promotion makes that army fairly deadly, especially in the Medieval period.
Philosophical and Spiritual are interesting to play, but only fit advanced strategies (described towards the end):
- Philosophical's extra Great Person bonus is valuable if running a Specialist Economy. There are plenty of locations suitable for "Great Person farms" (endless water, plus a few improved food resources). Clearly this requires very focused strategy, as discussed later.
- Spiritual is similarly interesting because it allows rapid shifts between key production civics: Notably Slavery and Caste System, but also Nationhood or Bureaucracy. For example, a research city might normally run with lots of researchers. But when a new building, like a University, needs to be finished, switch to Slavery briefly and sacrifice. Again, your strategy needs to be very focused, but avoiding Anarchy makes this possible.
I don't see that the expansions would change this analysis dramatically (although they clearly nerf Washington, and therefore make tiny island maps more balanced).
Finally, longer length games will allow more time to move boats around the map. Settlement and warfare generally take longer on island maps, however ships always move at the same speed.
3. EARLY GAME
This section describes your early moves and settlement, roughly until conflict breaks out - typically near the start of the medieval period. The fourth section below describes Wonders and technology during the early game.
Archipelago maps guarantee a coastal start. Indeed, with tiny islands it is often physically impossible to build an inland city! You will almost always start on a separate continent (islands) from other civilizations, but this is not guaranteed. This guide assumes separation - the early land grab and military would become considerably more important if there was no separation. Similarly, isolationist starts are theoretically possible, but unlikely. This guide assumes early contact is possible with most other civilizations via a combination of coastal tiles and territorial borders.
Before you engage autopilot and build a Warrior or Worker or Scout or Settler or whatever you normally do, stop and take a look around: Chances are you starting island is almost explored already. Surrounded by water, with no (or almost no) fog-of-war tiles on your island, you are perfectly safe until the Barbarians start spawning in boats. There may be no tiles a Worker can improve until you've research a few techs, so why prioritise the Worker? Don't be afraid to start building something that doesn't get completed until much later (like a Barracks) while you let the population grow and wait for the first technology to research.
On this map scouting means either Work Boats or Galleys used to explore. You will normally be able to reach most areas of the map while remaining in coastal (or territorial) waters, but the route won't be direct. Work Boats can be used to scout, but can't defend themselves. Galleys take slightly longer to build, but can often survive a Barbarian Galley encounter. Barbarian Galleys will not appear immediately, so Work Boats make excellent Scouts at the very start, when you need them the most.
The most important scouting objective is to be the first to circumnavigate the globe, which gains the permanent +1 naval movement speed. This is a huge boast in the early game, when otherwise ships only move 2 tiles per turn. It remains useful later - from faster troop travel to chasing down enemy shipping. Another civilization will beat you to this bonus if you delay. Ideally, build 2 Work Boat scouts as soon as you have researched Fishing, sending one east and one west. With luck, each will navigate half of the globe, which is good enough (the 2 boats don't even have to meet, merely reach the same longitude). Theoretically circumnavigation might not be possible until Caravels, but at least you'll discover that fact quickly, and know to prioritise Optics research.
While circumnavigating, you are likely to contact most of the other civilizations. Such early contact allows you to plan diplomatic alliances before the AI civs have contacted one another. This helps survive a potentially hostile map. At worst it gives you an advantage in any conflict: Often you will know where the enemy's cities are on the map, but the enemy will not have discovered your cities. Early contact and scouting (of their city locations) also helps establish trade routes. Trade routes are especially profitable on these maps, because almost every city is on another "continent".
Depending on the map, you may need to prioritise Writing research, for Open Borders agreements, allowing easier exploration and mapping trade routes. Of course, negotiating Open Border can be difficult (even impossible) if you have already converted to a "heretic" religion. Note that unless you have a relatively isolated starting position, it is rarely possible to stop other civilizations exploring and settling land by "blocking off" territory and denying Open Borders agreements. Also be aware of locations that can only be reached by moving Galleys through ocean tiles within territorial waters: Suddenly losing an Open Borders agreement can cut off supply lines in ways you probably weren't expecting.
Ideal settlement sites generally have a double-food resource, or a food resource and hills, or a food resource and another useful early resource (such as Iron). Food resources are important: Without at least one, you're likely construct a city that can do nothing except feed itself - a poor choice when so many thriving locations are available to be settled.
However, the distance maintenance penalty can become crippling. 90%, 80%, maybe even 50% research might be viable if it means you are expanding. But it's also easy to get into such terrible financial difficulty that the first genuine financial shock (like a war) might kill you. Colonising land close to the AI and then back-filling later settlements (in the gaps around your Palace) doesn't work well here: Water means that even cities which are on islands close to your Palace gain significant maintenance burdens. And if you leave those sites empty, the AI will often sail right round and settle them: Even well-developed cultural borders tend to leave gaps on a tiny island map.
Build order will vary by location. But Lighthouse (to quickly maximise food output) and Courthouse (once available) are recommended. A Granary is especially important in cities where production will be based on Slavery, since it speeds population growth. Harbors also have great health and trade benefits (once researched). Beyond those 4 will probably depend on what role the city is intended to perform.
In a larger-land mass games you might aim to build 1 or 2 Workers to improve every city. Here, 1 Worker will comfortably serve 2 cities. Indeed, if you wage war early you'll probably capture all the Workers you need. Instead of Workers, you should prioritise Work Boats. Ideally these can be sent out with new Settlers, and used to improve seafood tiles the same turn the city is founded. Remember, each Work Boat is a single-use improvement, but they cost half the price of a Worker and don't stop population growth while being produced.
On a tiny island map you can get away with almost no city defenses at the start. A single (ideally City Defense-promoted) Archer is normally adequate in the early game, mostly to avoid the unhappiness created by having no garrison! This minimalist approach to defense works because the AI doesn't scout water effectively (so is unlikely to discover many of your cities in the early game), nor does it launch war effectively over water (early wars tend to involve a single Galley), nor does it have the economy to produce enough units to contemplate serious warfare. Ultimately you will need slightly stronger defenses, and don't completely neglect garrison upgrades. But ideally use boats to intercept invaders before they land.
Lumber or Workshops? Unimproved forests provide an extra hammer when worked. Instead of chopping, it is tempting to work these tiles unimproved, and much later add Lumber Mills to them. Unfortunately, Lumber Mills appear relatively late in the game. When first researched (Metal casting) Workshops are only for the desperate (-1 food, +1 production) - consider them on an flat open terrain, if you have food to spare. However, Workshops slowly gain production bonuses with subsequent technology, and become very well suited to cities on this map. Remember, you normally can't irrigate fields, while Cottages are replaced by coastal water. With the State Property civic, the -1 food disadvantage is lost, and Workshops become better than Lumber Mills. As discussed later, it is highly likely you will beeline to, and run, State Property for its financial benefits alone. By the time we can build Lumber Mills, we'll have have Workshops which are broadly as good as Lumber Mills, why spare the trees? Forests are certainly useful very early in the game, but don't adopt the AI mindset that trees are too valuable to chop.
4. EARLY WONDERS AND TECHNOLOGY
Fishing, and then Sailing should be early research priorities, but not necessarily the first technologies to be discovered. Also consider technologies that allow you to exploit your starting position, especially to improve resources or exploit a (leader's) trait. Beyond that, initial technology research should be dictated by Wonders.
There are 2-and-a-half wonders that can radically change the outcome of the first half of the game (and consequently, the second half). You should aim for at least one:
1. The Great Lighthouse - +2 trade routes in all cities - on a map where likely every city is a coastal city, available very early. Research up to Sailing (and then build a normal Lighthouse), then research up to Masonry. Obsoletes with Corporation, however Corporation research can be avoided until you want something like Mass Assembly. You will need to beeline straight for The Great Lighthouse, especially on harder difficulties. That will almost certainly slow initial expansion.
2. The Colossus - +1 gold all water tiles - on a map where probably half the workable tiles are water. Research to Metal Casting, which also requires Pottery. Beelining may not be essential, because although the AI prioritises Metal Casting, it rarely goes there immediately due to the cost. Unfortunately The Colossus obsoletes with Astronomy, which cannot be avoided without immediately losing naval supremacy.
2.5 The Oracle - 1 free tech - then build The Colossus. While the free tech may be useful to you, completing The Oracle is often more beneficial because it prevents an AI civ picking Metal Casting quickly, and immediately going on to build The Colossus! For players founding early religions, The Oracle into Metal Casting into The Colossus is a solid strategy.
Both The Great Lighthouse and The Colossus provide very substantial financial bonuses, which should become obvious moving into the medieval period. An extra pair of trade routes can add over 5 (base) commerce to a large city. Same for water tiles. The Great Lighthouse is the clear winner for its potential longevity, and is also less restrictive - all cities trade, but not all cities will be dominated by working water tiles. However, The Colossus is less of a gamble. The ideal is to build both The Great Lighthouse and The Colossus. If you fail to complete both, consider military options to "acquire" them. Especially if they were built by Financial civilizations, who are otherwise very likely to power ahead technologically.
Unless you have a very specific strategy that benefits from them (such as a Cultural or religious approach), other pre-Renaissance Wonders can be ignored. Technology to focus on includes Compass (huge trade and health benefits of Harbors - far more useful than Aqueducts), Currency (more trade, more commerce), and Codes of Law (courthouses). Our aim is to build a robust economy, without which we can't effectively expand or develop. Don't be afraid to ignore technologies you may not be able to immediately use, like Agriculture, Horseback Riding, or some of the religious technologies.