slithy
Chieftain
Ok, first of all let me say that this is already an extremely good mod, one which definitely improves upon an already great game. I really like the "feel" of the different civilizations, which makes playing each one almost a different game. The magic system is already quite good, and is powerful without dominating the game, and the religions are well differentiated so that choosing one is a decision with serious consequences for your strategy. As a whole, it conveys a distinct aura of quality crafting and thought which is very impressive.
I have played 10 or 11 games of 0.13 now, and after due deliberation I thought it would be appropriate to throw some ideas at the creators, as a kind of payment for the enjoyment I've already pulled out of this excellent creation. For the most part, these aren't criticisms but more in the vein of "I really like this idea but it would be even better if it *this* were added" type of suggestions.
First off, I really think that the "early game" feel should be stretched farther than it is now - the idea of capturing animals (and slaves for some civs) is an excellent one, but by the time you can really exploit hunter-type units, most of the wilderness seems to be gone. As the Balseraphs, I was reduced to summoning tigers to put into my cages, which was not nearly as much fun as going out and capturing them would have been... Further, while the barbarians have been given a larger and better role, and Orthus did manage to wipe me out once (ironically, the very first game I played, and there was a definite "wtf? Nooooooooooo!" factor as he mowed down my fragile cities with their one puny warrior defending each), it would, in my opinion, be even better if a good long portion of *every* game involved lots of barbarian/animal combat and subdual, rather than the generic Civ feel of "get those settlers out the door and expand!" To me, this would not only further differentiate the gameply from generic Civ, but is also much more flavorful from the standpoint of being a fantasy-themed game.
I think the best mechanic to delay the mid-game would be to require a tech *and* a building before one can build settlers - this being preferable to simply increasing their cost as it is a hard barrier, rather than a soft one (and I would assume that hard barriers are easier for the AI to understand, for one). Further, I would add greatly increase the cost of the default civics, so that it is difficult to expand past 2 or 3 cities unless you have made some progress into the tech tree. This would ensure that there would remain large tracts of wilderness for a long time, which in turn would make even the later hunter units, which currently are of little use (except as city defenders and raiders for civilizations who have made poor military tech decisions), both interesting and powerful.
What would make this fun is that you have important decisions to make right from the start - is it better to try to rush for the settler tech and see if you can get a lead on the competition, or should you try to make your first city into a science powerhouse and grab a military or religious tech advantage you can use to dominate your bigger but more backward neighbors, or make that city a production powerhouse and try to grow by conquering barbarian cities rather than settling your own, or should you try to hunt and trap your way into an "empire of the animals" and send your hordes of wolves to ravage your rivals' lands and feast upon the flesh of their citizens - it really opens up the early game.
Secondly, while I understand that the tech tree is split into many different areas for a reason, I find there is too much choice - once you start going down the tree, you almost always have at least a dozen different techs you *could* research, each of which leads to other techs which also sound intriguing, with the end effect being you spend a lot of time trying to figure out which of these techs or linked techs actually does something useful for your civilization (because there are more than a few which are more or less only indirectly useful for any particular civ), or else you just give up and go with the cheapest one that sounds cool. I haven't looked at the .14 tech tree design yet, but I'd prefer to see a flatter design that .13, or, which would be even cooler, something like the design they had in the old Civ3 Middle Ages scenario (and which is used to some extent as well already in this mod, I think) where different civilization groups had entirely different tech trees, due to the base tech being either non-researchable or fabulously expensive to civilizations that didn't start with it. It would be too much work to have unique trees for each civilization, but 4 or 5 groups might be possible, especially if part of the tech tree was shared and another, smaller part was not.
Thirdly, there are far too many military buildings that do nothing other than enable one particular unit or unit type - which is especially annoying when they show up through the majority of the game as a build choice for civilizations which can't even build that unit type anyway. If a building is useless to *any* civilization, almagamate its function with another building, for example, because I can't always build Monks, make any Temple + a Barracks the prerequisite for training Monks, and dump the Monastery. Further, buildings you keep after pruning the list, like (for example) the Archery Range should be National Wonders - build 6 Barracks and you can make one Range (except call it something like the Yew Academy), but make it add +4 or +6 XP to archery units instead of +2. That way, you can specialize your military cities even further, and sometimes you might have to make hard choices about what types of units are most important to you and deserve an Academy. (I'd also make it so you need at least one 10 XP unit of a particular type before you can build the National Wonder for its genre). So you'd have the Way of the Sword; the Yew Academy; the Siege School; the Hunting Preserve; the Horselord's Household; the High University of the Magi; the Admiralty; the Temple School; the Parade Ground (for gunpowder units); and the Guild of Assassins.
I'd like three levels of Barracks as well - Training Yard, Barracks, and then Military College, each adding +2 XP to all units, although at a high cost for the Barracks and an extremely high cost for the College. This would allow the player to specialize their units with all the cool promotions more easily, which adds both to the fun factor and the flexibility, while draining some production which has been freed up by the loss of the other building types. Then drop most (or all) of the automatic promotions some units get on creation, as they shouldn't need them anymore when built in the appropriate environment.
Fourthly, there are too many buildings in general, and (strangely) too few tech-enhancing buildings. (There really needs to be Universities, as otherwise Science Academies become far too powerful a tool for a GP farmer - if the tech costs need be adjusted upwards then so be it). Fewer but more expensive buildings means less time clicking thru build lists (and the same can be said for military units, come to think of it). One way to lessen the perceived number of buildings would be to make some of them inaccessible to different civilizations depending on their tech tree group as previously outlined - for example, Inns could be unavailable to civilizations based around the evil non-humans, as presumably the only travelling they do is to go off pillaging and destroying. Gambling Houses then might not be available to good civilizations, and by increasing both the production cost and the coin benefit of both Gambling Houses and Inns, the effects on commerce remain roughly the same for both groups while adding yet another area of flavourful distinctiveness between the groups.
Three levels of science buildings is a good way to go (as something to put all the freed up production into, for one), Libraries, Monasteries (reclaimed), and Universities, each adding +25 percent research. It would be nice if most of the buildings went along with this same rule of three - build a Herbalist, then Apothecary, then Infirmary; Town Well, Aqueduct, and then Sewers; Carnival, Theatre, then Hippodrome (stripped of its military use but adding culture and maybe a bit of money instead). By having pre-requisites for each type, it makes the build list shorter, forces the player to build some of the less immediately useful buildings to get to better ones, and is much more comprehensible to the casual or beginning player as well.
Lastly - and this is probably extremely premature, but nevertheless is important (and I am assuming something can be done about it, or this is moot) - the default CivilizationAI really sucks at warmongering, and in this mod where war is even more crucial to success, means it really sucks in general. Now, I don't claim to be any kind of expert on AI, but I generally see the problem as one of focus - the computer tries to do everything when it is fighting, and thus ends up doing nothing effectively. For example, it'll send one stupid horseman into your territory to tear up a Farm, but is 70 hammers or whatever it cost to build him ('cause you KNOW he's going to die to any competent player who keeps a reserve to deal with just this type of idiot) REALLY worth slightly inconveniencing a city which is unlikely to lose that Farm for the 35 turns it would take to equal in food what you've just lost in shields? Then, once it builds another "pillage" unit, it'll send that unit in to die again, and again, and again, and again... Either build a honking big stack or two and do the job properly, or don't bother doing it all. The same with attacking cities - sorry, your two Bloodpets and a Catapult are not going to beat my 7 Arquebuses, so why not just stay home and use them for defence? My understanding is that the AI builds units and then assigns them a pool for attack, defence, or pillage, but that is, quite simply, a bad idea. Units, as a rule, are useless, the AI should be thinking in terms of stacks, which are not only more useful, but, because there are obviously less stacks than there are units, make it less complex to think about what to do. What it should do is relatively simple - build STACKS of units, EVALUATE the situation, then ATTACK the enemy (attacking being in the general sense of exploiting his weaknesses and covering your own).
At a high level (which, of course, is easy for me to say), this shouldn't be all that complicated - if the AI has 400 points of units, then maybe a stack is evaluated as 10% of that number - 40 points. Have it decide to make one of several preprogrammed types of stack, which can be at random as long as the stack types are combined arms - example, 1 unit of 5 is siege, 2 of 5 are melee, 1 of 5 is arcane, 1 of 5 is bow/gunpowder. Pick a rally point, preferably one that is deep inside your own borders. Build the stack, adjusting build queues as necessary in your cities. If you have overflow units, start a second or even third stack. Once the stack is ready, look at the situation - do you need to attack, defend, or pillage? Are you losing cities - move the stack to a threatened city and defend! Is the enemy stronger than you, or are there dozens of defenders in his nearby cities - pillage! Is the enemy weak or overextended with only a few units defending a juicy target - attack! Every 5-10 turns (make it random particularly to make it harder to manipulate the AI ingame) have the stack reevaluate its mission, and also decide if it needs to "re-stack" and get reinforcements, whether by merging with another stack or retreating and rallying new troops to its location. In that way, the AI can react to situations as they arise - if it's cities are not being threatened, then it makes sense to counter-attack rather than wait for the enemy to restore its strength and return; if the enemy is in your heartland, pillaging time is over.
The key concept is exemplified by the famous Stalin quote "Quantity has a quality all its own". Any decent player can get kill ratios in the neighborhood of 3 or 4 to 1 over the AI in normal play, but if the AI resists the temptation to send inadequate forces anywhere at all and keeps its quantities up, it can try the old Soviet tactic of burying you with numbers alone. All it has to to is bring the kill ratio down just to 2 to 1; at higher difficulty levels where it can outproduce you by more than that ratio, you are going to lose in the long run, and even at lower levels if the human gets dogpiled, the AIs are going to come out on top. In general, the troops it loses running around on their own doing dumb-ass things like suicide pillaging, finishing off wounded or weak "bait" units, attacking fortified units on hills, and other individual acts of idiocy, are a large proportion of its losses, so forcing it to forgo these pleasures by fighting in stacks should give it a chance to be merely clueless instead of spectacularly incompetent.
I have played 10 or 11 games of 0.13 now, and after due deliberation I thought it would be appropriate to throw some ideas at the creators, as a kind of payment for the enjoyment I've already pulled out of this excellent creation. For the most part, these aren't criticisms but more in the vein of "I really like this idea but it would be even better if it *this* were added" type of suggestions.
First off, I really think that the "early game" feel should be stretched farther than it is now - the idea of capturing animals (and slaves for some civs) is an excellent one, but by the time you can really exploit hunter-type units, most of the wilderness seems to be gone. As the Balseraphs, I was reduced to summoning tigers to put into my cages, which was not nearly as much fun as going out and capturing them would have been... Further, while the barbarians have been given a larger and better role, and Orthus did manage to wipe me out once (ironically, the very first game I played, and there was a definite "wtf? Nooooooooooo!" factor as he mowed down my fragile cities with their one puny warrior defending each), it would, in my opinion, be even better if a good long portion of *every* game involved lots of barbarian/animal combat and subdual, rather than the generic Civ feel of "get those settlers out the door and expand!" To me, this would not only further differentiate the gameply from generic Civ, but is also much more flavorful from the standpoint of being a fantasy-themed game.
I think the best mechanic to delay the mid-game would be to require a tech *and* a building before one can build settlers - this being preferable to simply increasing their cost as it is a hard barrier, rather than a soft one (and I would assume that hard barriers are easier for the AI to understand, for one). Further, I would add greatly increase the cost of the default civics, so that it is difficult to expand past 2 or 3 cities unless you have made some progress into the tech tree. This would ensure that there would remain large tracts of wilderness for a long time, which in turn would make even the later hunter units, which currently are of little use (except as city defenders and raiders for civilizations who have made poor military tech decisions), both interesting and powerful.
What would make this fun is that you have important decisions to make right from the start - is it better to try to rush for the settler tech and see if you can get a lead on the competition, or should you try to make your first city into a science powerhouse and grab a military or religious tech advantage you can use to dominate your bigger but more backward neighbors, or make that city a production powerhouse and try to grow by conquering barbarian cities rather than settling your own, or should you try to hunt and trap your way into an "empire of the animals" and send your hordes of wolves to ravage your rivals' lands and feast upon the flesh of their citizens - it really opens up the early game.
Secondly, while I understand that the tech tree is split into many different areas for a reason, I find there is too much choice - once you start going down the tree, you almost always have at least a dozen different techs you *could* research, each of which leads to other techs which also sound intriguing, with the end effect being you spend a lot of time trying to figure out which of these techs or linked techs actually does something useful for your civilization (because there are more than a few which are more or less only indirectly useful for any particular civ), or else you just give up and go with the cheapest one that sounds cool. I haven't looked at the .14 tech tree design yet, but I'd prefer to see a flatter design that .13, or, which would be even cooler, something like the design they had in the old Civ3 Middle Ages scenario (and which is used to some extent as well already in this mod, I think) where different civilization groups had entirely different tech trees, due to the base tech being either non-researchable or fabulously expensive to civilizations that didn't start with it. It would be too much work to have unique trees for each civilization, but 4 or 5 groups might be possible, especially if part of the tech tree was shared and another, smaller part was not.
Thirdly, there are far too many military buildings that do nothing other than enable one particular unit or unit type - which is especially annoying when they show up through the majority of the game as a build choice for civilizations which can't even build that unit type anyway. If a building is useless to *any* civilization, almagamate its function with another building, for example, because I can't always build Monks, make any Temple + a Barracks the prerequisite for training Monks, and dump the Monastery. Further, buildings you keep after pruning the list, like (for example) the Archery Range should be National Wonders - build 6 Barracks and you can make one Range (except call it something like the Yew Academy), but make it add +4 or +6 XP to archery units instead of +2. That way, you can specialize your military cities even further, and sometimes you might have to make hard choices about what types of units are most important to you and deserve an Academy. (I'd also make it so you need at least one 10 XP unit of a particular type before you can build the National Wonder for its genre). So you'd have the Way of the Sword; the Yew Academy; the Siege School; the Hunting Preserve; the Horselord's Household; the High University of the Magi; the Admiralty; the Temple School; the Parade Ground (for gunpowder units); and the Guild of Assassins.
I'd like three levels of Barracks as well - Training Yard, Barracks, and then Military College, each adding +2 XP to all units, although at a high cost for the Barracks and an extremely high cost for the College. This would allow the player to specialize their units with all the cool promotions more easily, which adds both to the fun factor and the flexibility, while draining some production which has been freed up by the loss of the other building types. Then drop most (or all) of the automatic promotions some units get on creation, as they shouldn't need them anymore when built in the appropriate environment.
Fourthly, there are too many buildings in general, and (strangely) too few tech-enhancing buildings. (There really needs to be Universities, as otherwise Science Academies become far too powerful a tool for a GP farmer - if the tech costs need be adjusted upwards then so be it). Fewer but more expensive buildings means less time clicking thru build lists (and the same can be said for military units, come to think of it). One way to lessen the perceived number of buildings would be to make some of them inaccessible to different civilizations depending on their tech tree group as previously outlined - for example, Inns could be unavailable to civilizations based around the evil non-humans, as presumably the only travelling they do is to go off pillaging and destroying. Gambling Houses then might not be available to good civilizations, and by increasing both the production cost and the coin benefit of both Gambling Houses and Inns, the effects on commerce remain roughly the same for both groups while adding yet another area of flavourful distinctiveness between the groups.
Three levels of science buildings is a good way to go (as something to put all the freed up production into, for one), Libraries, Monasteries (reclaimed), and Universities, each adding +25 percent research. It would be nice if most of the buildings went along with this same rule of three - build a Herbalist, then Apothecary, then Infirmary; Town Well, Aqueduct, and then Sewers; Carnival, Theatre, then Hippodrome (stripped of its military use but adding culture and maybe a bit of money instead). By having pre-requisites for each type, it makes the build list shorter, forces the player to build some of the less immediately useful buildings to get to better ones, and is much more comprehensible to the casual or beginning player as well.
Lastly - and this is probably extremely premature, but nevertheless is important (and I am assuming something can be done about it, or this is moot) - the default CivilizationAI really sucks at warmongering, and in this mod where war is even more crucial to success, means it really sucks in general. Now, I don't claim to be any kind of expert on AI, but I generally see the problem as one of focus - the computer tries to do everything when it is fighting, and thus ends up doing nothing effectively. For example, it'll send one stupid horseman into your territory to tear up a Farm, but is 70 hammers or whatever it cost to build him ('cause you KNOW he's going to die to any competent player who keeps a reserve to deal with just this type of idiot) REALLY worth slightly inconveniencing a city which is unlikely to lose that Farm for the 35 turns it would take to equal in food what you've just lost in shields? Then, once it builds another "pillage" unit, it'll send that unit in to die again, and again, and again, and again... Either build a honking big stack or two and do the job properly, or don't bother doing it all. The same with attacking cities - sorry, your two Bloodpets and a Catapult are not going to beat my 7 Arquebuses, so why not just stay home and use them for defence? My understanding is that the AI builds units and then assigns them a pool for attack, defence, or pillage, but that is, quite simply, a bad idea. Units, as a rule, are useless, the AI should be thinking in terms of stacks, which are not only more useful, but, because there are obviously less stacks than there are units, make it less complex to think about what to do. What it should do is relatively simple - build STACKS of units, EVALUATE the situation, then ATTACK the enemy (attacking being in the general sense of exploiting his weaknesses and covering your own).
At a high level (which, of course, is easy for me to say), this shouldn't be all that complicated - if the AI has 400 points of units, then maybe a stack is evaluated as 10% of that number - 40 points. Have it decide to make one of several preprogrammed types of stack, which can be at random as long as the stack types are combined arms - example, 1 unit of 5 is siege, 2 of 5 are melee, 1 of 5 is arcane, 1 of 5 is bow/gunpowder. Pick a rally point, preferably one that is deep inside your own borders. Build the stack, adjusting build queues as necessary in your cities. If you have overflow units, start a second or even third stack. Once the stack is ready, look at the situation - do you need to attack, defend, or pillage? Are you losing cities - move the stack to a threatened city and defend! Is the enemy stronger than you, or are there dozens of defenders in his nearby cities - pillage! Is the enemy weak or overextended with only a few units defending a juicy target - attack! Every 5-10 turns (make it random particularly to make it harder to manipulate the AI ingame) have the stack reevaluate its mission, and also decide if it needs to "re-stack" and get reinforcements, whether by merging with another stack or retreating and rallying new troops to its location. In that way, the AI can react to situations as they arise - if it's cities are not being threatened, then it makes sense to counter-attack rather than wait for the enemy to restore its strength and return; if the enemy is in your heartland, pillaging time is over.
The key concept is exemplified by the famous Stalin quote "Quantity has a quality all its own". Any decent player can get kill ratios in the neighborhood of 3 or 4 to 1 over the AI in normal play, but if the AI resists the temptation to send inadequate forces anywhere at all and keeps its quantities up, it can try the old Soviet tactic of burying you with numbers alone. All it has to to is bring the kill ratio down just to 2 to 1; at higher difficulty levels where it can outproduce you by more than that ratio, you are going to lose in the long run, and even at lower levels if the human gets dogpiled, the AIs are going to come out on top. In general, the troops it loses running around on their own doing dumb-ass things like suicide pillaging, finishing off wounded or weak "bait" units, attacking fortified units on hills, and other individual acts of idiocy, are a large proportion of its losses, so forcing it to forgo these pleasures by fighting in stacks should give it a chance to be merely clueless instead of spectacularly incompetent.