The Caravel Mod

[Durango]

Gollum, what is your stance on restricting units to specific eras only? I'm seriously considering it now, since I'm
tired of fighting outdated units. The Medmod, IIRC, had a similar setup. But the danger is of course, that the
AI (thanks to its stupidity) will just lack the tech to build up to date units...

Also, I'm just generally sick of the AI having larger stacks than necessary, due to the cheap trash units.

[Stazi]

I'm certainly not gollum ;) but I think I can share some of my ideas/possible choices:

1. Restrict units to the specific eras and risk the lack of units for the AI to build. Eventually make some basic units available for all eras.
2. Give the AI some time (HIGH era) to build required buildings - some units EARLY, HIGH and others HIGH, LATE. I'm not sure if the AI can understand and follow this idea.
3. Make new units have the same or very similar requirements as older units.
4. Make 3 versions of the same unit - restrict each version to one era only. Raise the cost and upkeep for the unit in later eras. This should force the AI to develop and build newer and cheaper units.
5. Try to modify 15th column "Unit choices (AI)". btw does anyone know what the number in the brackets exactly means (e.g. POVERTY_STRICKEN(75.75)). It's 75 and three quarters or 75 and 75 (whatever it means)?

They are only the ideas. I don't know which of them will be really useful but you have something to think and test.

[Durango]

I'm certainly not gollum ;) but I think I can share some of my ideas/possible choices:

1. Restrict units to the specific eras and risk the lack of units for the AI to build. Eventually make some basic units available for all eras.
2. Give the AI some time (HIGH era) to build required buildings - some units EARLY, HIGH and others HIGH, LATE. I'm not sure if the AI can understand and follow this idea.
3. Make new units have the same or very similar requirements as older units.
4. Make 3 versions of the same unit - restrict each version to one era only. Raise the cost and upkeep for the unit in later eras. This should force the AI to develop and build newer and cheaper units.
5. Try to modify 15th column "Unit choices (AI)". btw does anyone know what the number in the brackets exactly means (e.g. POVERTY_STRICKEN(75.75)). It's 75 and three quarters or 75 and 75 (whatever it means)?

They are only the ideas. I don't know which of them will be really useful but you have something to think and test.

Thanks for your input, Stazi

1. Making basic units available will, I'm afraid, only create a situation where the AI builds just those anyway.

2. Doubtful that the AI will follow this scheme, though. At least from my experiences.

3. This is a good idea, but a large part of the campaign game itself (teching up to get new units) will be lost.

4. Could work, nice idea!

5. Numbers in the Unit Choices simply mean the probability that the AI will train that unit given the condition it is in.


Generally speaking, the basic problem stems from the fact that the AI, even if it's rich, still prefers to spam cheap units.
This is why the Castilians for ex. loads it armies with Jinetes even though they own all of Iberia and rakes in the cash.
Sometimes the AI builds an "advanced", modern unit, but always woefully few of them.

Modifying the unit choices likely holds the key to solving this issue, but I think it's a damn chore to achieve and
would require A LOT of playtesting. I don't want to be back at square one instead of actually playing MTW...

[PershsNhpios]

They are all good ideas, but I think most of those ideas stray too far from vanilla gameplay - which therefore stray too far from the aim of the Caravel mod.

Forgive me for speaking on your behalf Gollum, but I thought you would most likely see this as such.

I do like your ideas, but I wonder if either of you have played the mod through to the Late era or for a good length into High? (I haven't)

How do you know the problems which you mentioned have not already been fixed by Gollum's attempts?

[gollum]

Hello guys,
the approach in the mod is to take out the armour upgrades, as they are the ones that skew unit stats more than others, and also reduce somewhat the morale bonuses, that again were too many in vanilla, and were skewing the unit match ups too much. It also makes the base stats of units from fort, keep and castle level of the same class (say spearmen) a little bit more close to each other than what they were in the vanilla game. The units however, are all still fully recognisable in role and performance for anyone that has played the game in terms of capability and dependencies.

In this way, the majority of the units of the same class that are recruited from higher levels are not very much better than the basic ones from fort level, but still better. Their prices to be trained and maintained have been tweaked to represent more faithfully their difference in stats - now what you pay for is what you get and you have to maintain accordingly.

Obviously, this cannot be the case with elite units, like knights, varangians, janissary heavies etc: these units cost a lot in buildings to get and so they should be statistically head and shoulders above the others, and they are kept as in vanilla. In most cases, however, they cost as much as their due to maintain. The vanilla training costs weren't bad at all and in the vast majority of cases of elites the mod sticks with them or thereabouts. Their maintenance though, was judged to be in many cases too low and it was tweaked accordingly. For example, Varangians of late era (that's the varangian unit of all periods in vanilla) are practically speaking a late era unit in terms of armor and capabiliity (huge ap attack, lots of armor and shield when nopt meleeing, very good defence etc), so they cost a bit more to maintain now.

Hence, with this approach, the mod maintains the variety of the vanilla game without flatening the differences between units of the same class (essentially turning them into clones of one another), but also without making easy for the player to have an army of elites that would sweep less developed AI factions on the battlefield.

In the medmod, the other potential approach was used, ie have 1 spear unit per era per faction, one bow unit per era per faction etc, but this approach was taking away the variety of the original game. My goal for the mod was to maintain that variety but make it more balanced and hence less hurtful in the long term for battles.

In the vanilla game, low end units cost little to train but a lot to maintain, while high end units cost a lot to train but too little (less than the low end units in many many cases; for example Nizaris cost less than Desert archers in maint. costs), to maintain. This approach clearly favors the player, that can tech up quickly and efficiently and then have an army of elites at lower cost than an army of low end units.

In the mod, the accessibility of elite units was increased somewhat, and most importantly, buildings that were getting nothing to AI factions, were removed from that faction's building potential. A classic example is the Townwatch line of buildings that for the Eggs and Byzantines gets nothing other than from the first building of teh line. In the mod this is corrected - teh higher levels than level1 of the watch tower are not available for the Byz and the Eggs. This means that the AI won't be wasting many turns and money and the benefit of what he could get with these many turns and money to the higher levels of teh town watch that get him absolutely nothing.

These are the methods and rationale employed and, i can say that stack composition works really really well; you won't see any faction with stacks that are worthless at any stage of the campaign.

The spamming of the AI of cheap units has to do, as i posted earlier, with the ratio of teh (training cost)/(maintenance cost). Once the AI finds a unit that has high such ratio, with small training cost, he spams it (eg siege engines/artillery/javelins/slav warriors etc). This is all the more for units that are very easily available - like the Slav Warriors that are avalable from the fort in vanilla.

This spamming really hurts the AI because, in the long run, many AI factions may end up fighting with units they've build in the first 30 years of the campaign. If they are fodder units, the AI will have to live with them for much of the duration of the campaign, as the AI can't disband to make financial room to get new units, or burn them in battle too efficiently in order to build new ones. In any case, even if he could done so, the chances are taht he would still build...Slav wariors all over again :)

The mod does a great job with this, or so i think. It is its strongest point, imo.

[gollum]

I also quote this post of mine from the Main Hall (the deus nolt! thread, Stazi refered to), as its relevant here.

The mod plays slightly slower than vanilla because faith propagates slower and because the happiness bonuses are given more gradually.

In vanilla you can rush very very quickly because the watchtoewers line of buildings gives +50 happiness (when both build) with 300flrs and in 2 turns - that's very cheap and easy cowering of new provinces

In the mod, only the border forts give +20 happiness and in 3 turns (watchtowers=1t, borderfort=2ts). The other extra +30 is given by buildings that come from keep level (basically teh church/mosque), hence you need to be a bit more careful and also will make a bit less tax money from newly conquered territories.

Also, the faith propagation being slower it takes more time to subdue different religion territories because of that, hence happiness is affected and there is reason for be careful for loyalist revolts.

The mod plays slower for the AI, especially if you put it @180 loyalty with -ian (as reccomended). If you want teh AI taking more initiatives and pester you more and take opportunities, you can drop that to loyalty:150 i'd say. if you do not use the loyalty feature at all, the AI may be hurt by the slower happiness bonuses, especially since many factions are set to expansionist that makes them aggressive and prone to burning out by overstretching and revolts that ensue. But who knows, it may play ok in that too - just i've never tried it.

Having said that, i like the occasional rushing campaign, and i tried to rush in Caravel and it does work without a doubt. Just the opponents are a bit more ready especially ones that are not immediately next to the player and i was able to make less quick cash flow - those were the main differences.

The mod is at this stage finished and there is no further plan to update it for 2011.

What i will/would update next, would be the rebel garrisons. Some of them are a bit too big and some would make for better gameplay between factions if they were considerably less strong without hurting the mod concept and intended gameplay. This should/would help people who like different unit settings to play the mod better, and also improve teh AI performance of some factions without upsetting the intended difficulty.

The fixes i would have done would be to:
-Make Navarre completely empty- it stands historically as teh Navarrese dynasty was merged with the Aragonese and Castilan one at the time the early era starts and so there is no reason to have a strong Kingdom there.

This would impact gameplay positively as it would increase teh chances of competition between Aragon/Castile/England/France and also, in case the Ibearians are the first to get it, it would aid them to power status sooner than they now do in the mod.

-Make the Prussia, Volhynia rebel stacks smaller. This is relatively historically accurate, and it would also help the Polish AI to develop a bit better than it currently does.

-take out a few units from the rebels in Lithuania/Livonia. As they are now they are overflowing and even attack the Poles and teh Russ occasionaly which is ok, but not as much as they are now.

An additional thing i would try to do would be to fix the strong shot/slowreload missile units in terms of battlefield role, if there is a way.

Other than these things, which won't be done for a while, the mod is done.

[Durango]

Gollum, I seem to recall that you wanted no river crossings in your mod, no? I found some oversights in the startpos.
These are the lines:

SetBorderInfo:: ID_BLACK_SEA ID_MOLDAVIA
SetBorderInfo:: ID_POLAND ID_CARPATHIA

SetAttributes:: ID_CONSTANTINOPLE
SetAttributes:: ID_EGYPT
SetAttributes:: ID_HUNGARY
SetAttributes:: ID_ILE_DE_FRANCE
SetAttributes:: ID_PEREYASLAVL
SetAttributes:: ID_ROME
SetAttributes:: ID_SCOTLAND
SetAttributes:: ID_SILESIA
SetAttributes:: ID_VALENCIA
SetAttributes:: ID_WESSEX

I might be mistaken in that you have decided to keep rivers in some places, and I apologize if it is indeed so

[Stazi]

5. Try to modify 15th column "Unit choices (AI)". btw does anyone know what the number in the brackets exactly means (e.g. POVERTY_STRICKEN(75.75)). It's 75 and three quarters or 75 and 75 (whatever it means)?

5. Numbers in the Unit Choices simply mean the probability that the AI will train that unit given the condition it is in.

I know it's a probability but I don't fully get it. Is value 75.75 a fraction or two different values?
1. If they are two different values what are they referring to? What's the difference between them?
2. If it's a fraction - What's a point of making a fraction value when you have @0-300 range at your disposal? It's strange.
The probability - how to convert these values to the easily understandable and comparable percentage values?
What's the min-max range for this probability?

[Durango]

@Stazi: It's most likely a fraction value, even though it may seem redundant.

@Gollum: Making duplicate units is probably the most foolproof way of restricting mercs - albeit at the cost of variety. The problem with
regular units as mercs is that while you can change the magnetism value of the Inn, you can't stop lots of advanced units flooding
the borders when there is a war.

Also, you should consider making the Inn unique (as I have) and having it cost more and take longer to build. The AI, fortunately,
can be left out of the equation when it comes to Inns, so feel free to raise costs/build time however you want.

[gollum]

Hello Durango,
making the Inn unique is something that i have done in my home modding efforts in the past considered and rejected for this mod for the following reason that detracts from the vanilla gameplay: once the borderline moves and the Inn is no longer in it, the magnetism of mercs does not work, which means the player will have to manually destroy the Inn and rebuilt it on a frontline province to get the mercs, which is confusing if someone doesn't know how thge mercs are attracted (as many players i bet don't). On the other hand, leaving the Inn in the game can give a cheap/quick building that aids teh AI faction leaders (and teh player) to get the "builder" line of traits easier that makes them more stable (as it gives province happiness and general loyalty eventually).

In vanilla the Inn takes 2 turns to build and 400flrns. Since the AI is not using it, the cost to build it was downed to 200flrns, which is a reasonable price to pay for the AI leader's improved trait.

I was thinking to make available some of the extra units that come from the VI as mercs only, which won't impact on variety in any way, as most are not used in the mod and are sparsely used in vanilla.

[Durango]

I'm not sure I follow you here. You are saying that razing and rebuilding the Inn makes it an exploit? Well, in that case, every building
in the game can be abused in that way. And what is the point of making the Inn cheap, when its disuse by the AI leaves you carte
blanche to make it considerably more expensive for the player?