Where to build markets - only in fertile lands?

[MCM]

So here's the actual math on upgrading Market->Rice Exchange

Costs 1600 gold
Adds +300 wealth (beyond Market)
Takes 5 turns to build

Also:

Adds +5 town growth/turn (beyond Market)
Subtracts 1 town growth/turn from every Province

So the overall town growth is completely even if you have 5 Provinces. 5 provinces is really small; basically the minimum you need to start fighting your real battles (Shikoku, the Chosokabe starting island, is 5 provinces.) Also, at 5 provinces you probably don't even have the tech to build Rice Exchanges yet, or 1600 spare gold to build a Rice Exchange. Let's go with 10 Provinces. Still not much, you have a long way to go to win the game, even on Short campaign. But at least at 10 provinces, you can definitely be deciding, "Should I invest more in econ, or build more troops?"

So we have 10 provinces. Next, let's just look at the costs of the Rice Exchange itself, forgetting all that complicated town growth stuff.

+300 wealth * .25 tax rate = 75 gold per turn
1600 gold to build / 75 gold per turn = 21.3 turns to pay off
21.3 + 5 to build = profit 27 turns later (Rounded up, since after 26 turns, you have still lost money. After 27 turns, you have turned a profit!)

Wow, that's... pretty bad. And that's with no town growth penalty at all! (Or, basically, having 5 provinces. If you have 5 provinces for 20+ turns... you are probably not going to win the game. :))

So let's look at it with town growth on 10 provinces.

10 provinces mean you are losing 5 town wealth per turn, since +5 growth in the Rice Exchange province is counteracted by -1 growth in every province. So even on the very first turn of our Rice Exchange's life, we don't really have +300 town wealth. We are at +295 town wealth, decreasing 5 per turn, and we paid 1600 for this privilege.

So what does that mean for our Return On Investment (ROI) on the Rice Exchange?

It means that not only does it take more than 27 turns to make a profit, it means that building a Rice Exchange LOSES money, and DOESN'T gain you money. Every town will be just a little bit smaller, and that smallness will add up to much more than the +300 gained.

I made a little spreadsheet, since you have to calculate how much town growth you have at any given time, and how much you have lost, and how much the previous town growth gets you since that changes every turn.

To sum it up, if you have 10 provinces, and you build a Rice Exchange, you do not get your money back until 34 turns later. Your investment peaks at turns 64 and 65, where you have profited 612 gold overall. After that, the overall money you have gained from the Rice Exchange decreases. The Rice Exchange turns negative around turn 95 or so. Yes, that means that building it has overall COST you money. With 10 provinces only, on turn 95, ~24 years later, you would have more money if you had never built it in the first place.

Does anyone else realize how incredibly bad that is? It only accelerates as you conquer more provinces. Let's say I conquer kind of slow, but conquer 1 more province every 4 turns (ie, 1 province per year), going to 11 provinces at turn 6, going to 12 provinces on turn 10, etc. This is probably too slow to win on Short, and definitely too slow to win on Domination, since we're talking about already having 10 provinces, and then only getting 1 more per year. It might be enough time to win on Normal speed. But we know that more provinces is worse for Rice Exchange, so conquering slower is actually better (which should be a huge blinking warning sign that THIS IS A BROKEN BUILDING.)

Anyway, what happens if I am actually conquering provinces over time?

Well, for every additional province, I am now losing 1 more town growth per turn that I would have had. This continues to snowball over time.

It means that a Rice Exchange never pays for its initial 1600 gold investment.

Never.

Ever.

It means that even on a Short Campaign, a Rice Exchange is a total black hole for money. The longer your campaign goes on, the worse it is. The faster you conquer, the worse it is. This is why I say: never build a Rice Exchange, ever.

[xploring]

http://forums.totalwar.com/showthrea...e-Market-Chain

There is a very detailed analysis of the economics of the market chain in the link above. The author made a simple error early on in his calculation, but he has said he will correct it soon, so maybe check back later on if interested.

(I haven't gone through it yet though.)

[MCM]

http://forums.totalwar.com/showthrea...e-Market-Chain

There is a very detailed analysis of the economics of the market chain in the link above. The author made a simple error early on in his calculation, but he has said he will correct it soon, so maybe check back later on if interested.

(I haven't gone through it yet though.)

His math is totally wrong because of the "simple error" he made. He calculated +wealth as +money per turn, not +wealth*tax rate.

[quadalpha]

His math is totally wrong because of the "simple error" he made. He calculated +wealth as +money per turn, not +wealth*tax rate.

Wow, MCM, good work! Now perhaps we can take into account if the need for metsuke changes the equation? Something like: if you need another metsuke, you have to either upgrade a market or take up a building slot in another province, where the use of one building slot ~= one castle upgrade (not really, but you see what I mean). This is necessarily going to be a bit more social-sciency, estimating the value of a metsuke.

[Rothe]

Someone else correct me if I am wrong, but I think you only get one Metsuke per market chain, not per market upgrade.
I think the same goes for ninja and monks. The later buildings only give XP to the agent.

Having seen the situation for couple of years after a realm divide in my game, I see the problem of rice exchange+ level markets. They really hurt me since I could also use the buffer of food surplus to conquer provinces with large castles but small farms level. Now I have to avoid them and try to build max farms everywhere - land consolidation takes so long to build that my advances in the campaign are limited by it. Having the extra food from my 5-6 market chains would help a lot.

I really wish CA will repair it by making the market chain buildings better (perhaps dependent on the number of provinces) or the food surplus less useful.

I would not mind if a few fully upgraded chains would be the optimal solution for the economy. Still, it seems a bit odd that you have to leave markets at level 1 just because of the food issue. Sake den's are actually much better buildings than markets in the sense that you can get happiness and wealth from them without sacrificing anything really.

[Dead Guy]

Yeah I think you're right about the agents. I think it's one per chain, regardless of level so to speak.

Might be worth it to build one of the top level markets to get +2 xp Metsuke though? If you bother to get that far with the Chi, that is.

Though I consider Metsuke to be the least useful agent to be honest. I rarely bribe, Monks can boost happiness in a region (right?) and ninjas can "apprehend" ;)

[MCM]

Top-level market for a Metsuke isn't such a great idea. The marginal tax increase might not even off-set the amount lost by investing in the buildings in the first place, plus the Chi research issue, as you mention.

You can get +2 xp Metsuke by building a Market in a province with Philosophical Tradition, too. Though I suppose Philosophical Tradition+top level Market upgrade would mean +4 xp Metsuke, so I guess it's the same question either way - how much money per turn does a good metsuke in a good province get you? Not just % tax difference, but real, effective difference. I don't have any clue how administrative costs work, but those seem like they decrease the actual value of a good metsuke by quite a bit.

What I would really love to know more about is how tax rate affects growth. The growth malus from taxes seems to change a lot. It might be that having a good metsuke is overall not that valuable because the increased taxes slow growth more. In other words, taxing 150 at 30% is really not better than taxing it at 25%, if that 150 is going to grow 1 wealth less under 30%.

[MCM]

Yeah I think you're right about the agents. I think it's one per chain, regardless of level so to speak.

This is correct but there is also a max limit of 5, regardless of how many chains you have. You can never have more than 5 Metsuke, 5 Ninjas, 5 Monks, etc.

[Agent Miles]

Hypothetically, if you have +5 food in 10 provinces from the example:

Make RE (Rice Exchanges) in the top 5 income provinces that have a tax boost from a Metsuke:
(The actual tax rate with Metsuke varies, but I've used 50% with the Metsuke and 25% without to keep the math simple. I also know this to be an attainable rate from experience.)

Each single RE costs 1600 koku
This adds +300 per turn at 50% Metsuke tax rate = 150 per turn
1600/150 per turn = 10.67 or 11 turns
150/turn x 11 turns = 1650
It takes 5 turns to build the RE and an additional 11 turns for a ROI of 50 at one RE.

50 x 5 RE provinces with Metsuke = 250 total ROI after 16 turns.

Unfortunately, for the 16 turns you lose +5 growth for 10 provinces.
5 in ten provinces = 50 per turn lost growth
So after 16 turns, that's 800 total lost growth.
Breaking this down:
400 total lost growth for the 5 Metsuke provinces at a tax rate of 50% = 200 lost income
400 total lost growth for the other 5 provinces at a tax rate of 25% = 100 lost income
So, 300 total lost income after 16 turns.

250 (total ROI) - 300 (total lost income) = -50 net loss after 16 turns.

However, you would continue to make (5 x 150 = 750/turn) total RE income from the 5 provinces with Metsuke.
You would now also get +5 growth from the RE over the Market in each of the 5 Metsuke provinces. The +5 replaces the lost +5 growth.
(-5 x 25% = -1.25 at each 5 remaining provinces) = -6.25 rounded down to 6 continuing lost income

750 (total RE income) - 6 (continuing lost income) = 744 koku/turn added income

Thus, you get (744-50 = 694) koku profit after 17 turns and 744 koku profit every turn thereafter. Further, if you conquered the whole map, you would continue to make a profit.
So, build Rice Exchanges as described.

P.S. Here's the argument presented in the extreme cases.

A. 10 provinces with a Stronghold, Land Consolidation, and a Market
or
B. 10 provinces with a Stronghold, Land Consolidation, and a Merchant Guild

With A, you get +20 Food and each province gets +5 growth from the Market for +25 overall growth.

With B, you get zero Food and each province gets +20 growth from the Merchant Guild for +20 growth overall. However, you also get an extra +800 commerce.

At a delta of +5 growth, the total wealth in A would catch up to B after (800/5 = 160 turns). Of course, the wealth in B would have been generating much more in taxes from the additional commerce for those turns. Let's start at A and spend 5,000 koku and 11 turns for the Rice Exchange and Merchant Guild to build up each province to B. Even at 10% tax, that would then get us 80 x 149 remaining turns = 11,920 koku per province. So 11920-5000 for the build up still gives 6920 profit. At the same tax rate, A would need over a hundred turns to earn this sum through the continued difference in growth and would thus seldom ever catch up before the end of the game.