The tax collection definitely needs an overhaul. Deleting it completely would be a detrimental nerf and it also doesn’t fit into the game. Almost every civ has some way to gain gold even if there are no more gold mines on the map. The problem is just that Tax collection is so unreliable and its effectivness relies entirely on the player. Meanwhile, French, English, HRE, Rus all have extremely easy yet gold collection. In my opinion the Chinese gold collection should be more effective than what Rus and English get because its also much more complex. HRE needs to get 3 relics for their 900 gold/minute atleast, so there is some skill involved and you can somewhat deny it.
Not sure I’d agree with a hard and fast collecting tax is never worth it.
Late game one IO running between say 4 buildings every 30 seconds, collecting taxes that have built up earlier in the game, works out at 320 gold per minute. I don’t think that’s an implausible set up - especially if you’ve not been collecting tax and so its built up in various pools.
To collect that from overseeing you’d need to mine over 1500 gold in a minute. This is do-able - but it entails getting around 25~ villagers with with the level 3 mining upgrade on a single mine. In practice its a “why not both” scenario. By contrast farms are basically impossible to optimise this way - and wood lines can get spread out unless you are diligent in putting down a new lumber camp every other minute (which I guess you probably should - but you may have other concerns on your mind.)
My main complaints are
A) The frankly needless mini-game micro early on of supervising say that starting mill, and having to click back at 20 to tell him to gather the tax, walk to the TC, then click supervise again. I think auto-collecting tax on a supervised building (either instantly, or just 20/40 every 30 seconds) would be a nice quality of life change which I feel would be a relatively modest buff.
B) While I think it stops if you turn off auto-tax collection (which I realise is just one keystroke, so probably something you just have to learn), if an IO has 20g or 40g he will effectively instantly cancel any order to go supervise something. I don’t really like my units going “I don’t think you really want me to do that”. This doesn’t feel like a feature and it could just be changed.
3 Likes
320 gold per minute would assume absolute efficiency with zero walk time. That absolute peak is going to drastically decline the further out you have to walk and even with a modestly placed mill on the nearest berries to the TC you’re looking at 10+ seconds of walk time from the TC to collect the tax and then walk it back, unless you get really lucky with resource placement. The video below shows this walk time first hand with a fairly modest berry bush placement. The Officer spawns at 3:31 and does not make it back to the TC to deposit until 3:42. Factor the 30 cooldown and you’re looking at 40~ seconds for 20g (40 if you have the upgrade). This is for one collection point relatively close to the TC. Obviously, that walk time is going to significantly increase, as the game progresses.
Now, compare that to 3 villagers on gold (which is pretty standard in Dark Age). Villagers gather somewhere between 0.70 and 0.75 gold per second, as a base gather rate with no upgrades. Even using the lower tier of that, each villager is generating +5.6 gold in that same time frame (40~ seconds) you are generating 16.8 additional gold with all three villagers. Once you have 8 villagers on a gold mine, you are generating more gold from supervision than the Officer would from tax collection, even with the double carry capacity (44.8). Again, this is the base gather rate with no upgrades.
To be fair, I suppose there could be an argument made on using an Officer for tax collection in the very early game, but it could also be argued that the 150 food you spent to create the Officer could have been spent to get 3 more Villagers (which can generate 84 gold in the same amount of time it takes an Officer to generate 20). So, we have to really consider whether that kind of investment is worth it in the early game. Either way, once your resources spread out and/or you have 8 or more villagers mining gold, then Supervision is going to outpace tax collection every time.
2 Likes
To avoid order cancelling, you have to disable auto collection, that way it will obey =)
I really like this idea.
1 Like
As you say the Officer Spawns at 3.31, goes to the mill and returns at 3.42. We are sort of splitting seconds here - but that’s 20/40g in 11 seconds. The IO shouldn’t then just sit idle until the cooldown is over to go collect again. Its hard to see due to the graphic - but I think he can get from the TC to the Lumber camp later in the video and in about 10 seconds. Now if the IO then has to walk out to say a mining camp at the other end of the base this might be quite inefficient. But its quite plausible - to imagine 2 barracks (or whatever) just north of the Imperial Academy, effectively next to the TC. It isn’t going to take 10 seconds to gather and return from these. Therefore in around 30 seconds the IO can circle between the mill, a lumber camp and two barracks gathering 80 or 160 gold - which becomes approximately 160/320 gold a minute, providing these is tax to collect.
And this brings up point 2. As I understand - and I could be wrong - you only get the 20% bonus resources when they are delivered to a supervised mill etc. The rest of the time IO is effectively just sitting there watching. Therefore in terms of optimal resource collection (especially early game), assuming you value resources on a 1:1 basis, you want to use that idle time to ferry gold back to the TC. In practice this is basically impossible to get exactly right - but collecting in 20 (or 40) gold is worth missing out on say 4 extra food/wood/gold from say 2 villagers handing in before you can get back. This is especially true if you get Wheelbarrow - as the workers spend more time collecting up 15 resources - although this also effectively nerfs tax collection to 1 per 15/18 resources, rather than 1 per 10/12. (This raises theoretical questions of whether its worth getting for China - and the question of whether this is intended.)
For your last paragraph, I don’t think its fair to compare 150 food for an IO with 3 villagers. This is because while this is true in terms of resources spent, its not true in terms of efficient use of time. I.E. (before Song Dynasty) it takes 60 seconds to get out 3 villagers. The IO player wouldn’t (hopefully) buy an IO and then leave his TC idle for 40 seconds - so in practice you don’t get the equivalent of 3 extra villagers gathering. You’d get 1.
Its a fair point that you are down 100 extra food to get say 1 IO and 2 villagers, versus 3 villagers - but whether this matters depends on what that extra food would have got you, and how quickly supervising say a Mill will get you the resources back (plus the tax you can collect as per the above). I guess its an interesting question whether there’s a build order that initially sends fewer workers to food (because you don’t need to catch up on the additional food outlay of an early IO) - and say devote them to wood or stone for a fast TC. But I suspect even then, you are going to need to collect more than 500 food to sustain villager production and get to feudal in a viable time (5ish minutes?), so the IO will pay for himself by supervising a mill and catch up the extra resources that 1 villager would have gathered instead.
But its a complicated question because how quickly you recover that 100 food deficit depends on how many workers are effectively supervised at any one time.
This kills the base planning part of the tax mechanic. If you are collecting from a mine in the middle of the map there is a trade off of not getting tax revenue and that’s OK.
Yes but the tax mechanic doesn’t need tradeoffs right now, because it’s not a very good mechanic. Most people use their officials to supervise production because it’s blatantly superior to collecting tax. If the supervise mechanic added a bit of tax then you’d actually have tax getting collected.
The only benefit of the tax collection system is that after both parties run out of gold resources, there will still be a steady stream of gold available to Chinese players. But, hey, how many times will you encounter this situation in 1v1?
The tax collection system is a hassle. If the tax collection conditions are too loose, the tax collection system will be OP, but the current tax collection system does not seem to be cost-effective, at least not more cost-effective than supervision. Or the developer design empire officials. The main purpose is to monitor? Taxation is just an additional value added? I’m not sure.
Alrighty, folks. So, rather than sitting around and talking speculation, I actually did a series of isolated tests. Mostly for my own understanding and making sure I had my facts straight, but hopefully others find it useful.
So, in the test, I set up five mills around the TC, each one exactly one tile away (because this exact distance was brought up earlier). I found some…interesting things. Here are the numbers:
Control (3 Villagers, no upgrades): 110-120 gold per minute.
3 Villagers + Supervision: 132-144
3 Villagers + Supervision + Wheelbarrow: 144-162
3 Villagers + Supervision + Wheelbarrow + Specialized Pick: 162-180
Tax Collection with no upgrades:
High: 160-200
Low: 58-100
Imperial Academy Aura (3 Mills):
High: 160-200
Low: 100-160
Imperial Academy Aura (3 Mills) + Imperial Examinations:
High: 320-400
Low: 78-117
So. There are the numbers. During this test, I found that tax generation generally does not keep up with collection, when pickup points are that close to the TC. The exception to this is if the pickup points are within the Imperial Academy’s influence AND you do NOT have Imperial Examinations. Once you get that upgrade, then the collection rate outpaces the generation again, which is why there is such a disparity between the low numbers of Imperial Academy Aura with and without Imperial Examinations. This identifies yet another bottle neck in the tax design, as it drastically decreases the income per minute, as collection starts draining the pickup points.
It’s also critical to note that this is with the absolute optimal placement with the least amount of walking time. Obviously, in an actual game, you will not have the benefit of setting them down that close to your TC until very late in the game, once your natural resources have been exhausted. In the early game, you are going to be dealing with a considerable walking time between the different points, which may offset the tax collection > generation bottleneck, but then you will have to factor in the walking time.
So, the main take away that I found is that there is a small window in the Feudal Age, right after you research Imperial Examination, that the tax generation can be very beneficial, while you are building up your gold mine (providing that you have given your pickup points enough time build up a tax pool). Once you start upgrading your mining and have more saturated gold mines, however, the mechanic becomes severely diminished, especially once your pickup points have been depleted or you factor in walking time for more distant points, as resources become more spread out.
Given this, I can see cause to cycle through and supervise, while you let your taxes build up, before sending them back to supervising, but to determine how viable that actually is, I would need to do more testing with more distant pickup points to factor the impact has on income with a more standardized set up that you’d see in a typical game and compare that to more saturated gold mines.
1 Like
The opportunity cost of an official is 1 vill + flat 100 extra food.
Absolutely not 3 vills.
Those numbers you calculated are not very meaningful.
1 Vill costs 50 food. 1 Imperial costs 150 food. 50 * 3 = 150. And, by all means, if you think my calculations are “not very meaningful”, then run your own tests and get your own numbers. Let’s compare notes, instead of coming in with a lazy, condescending retort that lacks both substance and merit.
2 Likes
Yeah agree 3-1 cause your spending the 150 and not getting a villager
An IO takes 20 seconds to build at a TC which is the same as a villager. Your TC constantly builds vills for 20+ minutes straight therefore the cost of an IO is 1 villager + 100 food. You either make 1 villager, or you make 1 imperial official. You do not choose to make 3 villagers or 1 IO and idle your TC for 40 seconds–this is not the decision anybody ever makes.
Very early game IO are amazing since you can micro them to get 100% supervise uptime on a mill while also close to 40 GPM (Gold per minute) just off your mill. This increases if you mill next to lumber camp and get closer to 80GPM (gold mining rate of 2 vills) on top of being able to supervise a mill constantly. This would be perfect micro, so you probably get about 60 GPM early game from tax collection once your woodline is saturated if you micro it.
If you setup a production area around the imperial Academy near a TC, you generate about 240 GPM while pumping MAA and Xbows. More if you pump trash units or have more production buildings since a single IO with the upgrade can support quite a few (10+) production buildings depending how close to a TC they are. Units built generate 8 tax and upgrades generate 64 in the range of the Imperial Academy.
VIlls on mills will generate roughly 3 gold per minute each to be collected by an IO. If you have 5 mills and 40 on food, that’s 120 GPM from a less efficient IO walking by mills. A single IO easily collects this can could manage a few more mills and 160 GPM.
This example is 360 GPM from 2 IO collecting taxes.
Relative to 2 villagers, this is a pretty significant economic gain.
The mechanics are clunky and you need to turn off auto-collection early game to micro your IOs, but mid-game+ if you have good building placement just letting them roam and collect tax is very effective.
2 Likes
The test you did with production and upgrade buildings is very interesting. I also found it interesting that your example of 2 IOs generated about the same amount of gold as a single IO (which makes sense, since tax collection is on a universal cooldown).
The problem that I found with the mills test, is that their tax generation did not keep up with the IO’s tax collection. So, you end up with fairly high numbers, when you let their taxes build up, but at some point the tax collection outpaces the tax generation and the IO is not getting the full 20/40 per visit and it drastically reduces the income per minute. This is why there is a disparity between the high numbers and low numbers in the tests I run. The high numbers were generated when every building had a surplus of taxes, while I gathered the low numbers once the taxes in every building had dried up and collection wasn’t at its peak.
I agree with your findings, though, which I stated in the last paragraph, that if you want the best optimization, then you need to micro your IO and cycle them between supervising and tax collection. Supervise, while your taxes build up, send them to collect the surplus and then put them back on supervision. I think your tests with the production and upgrade buildings was very interesting, because it SEEMS like the generation in those buildings would be able to keep up with the collection rate, as long as you maintain constant production. Particularly so with Spearmen and Archers, since they produce in only 15 seconds, which means you can generate 32 tax per minute. Add the 30 second GCD to that and they would generate more than enough taxes to keep an IO going without a drop off (theoretically).
I’m still not sure if getting an IO in the Dark Age is worth it, though, because you are spending almost 40% of the food you need to age up on one unit and the IO isn’t going to be able to pay for itself, before you reach Feudal, to make up that deficit. No matter when you get it in the Dark Age, you’re going to be putting yourself at a slower age up time (assuming you are playing a 50/50 game). Although…I could speculate that the Chinese faster build time COULD make up for that. Would need to run some tests to confirm.
To me, it seems better to build up a tax surplus through the Dark Age and then getting an IO and Imperial Examinations as soon as you hit Feudal to go around and collect those built up taxes to give you a big gold boost for upgrades in early Feudal. At that point, you’ll have both Imperial Examinations and the Academy working for you, as well as sizeable stockpile of taxes to work through, netting that 320-400 GPM right out of the gate (which is the same generation as 8 Villagers).
This whole analysis has been interesting. When I was first looking at Chinese, I was convinced that the IOs were completely useless on tax collecting, but I’ve found that they can be quite beneficial, when used in specific conditions and with a reasonable amount of micro. I still think that they could use some tweaks to reduce the amount of restrictions they have with this mechanic, but now, really, I think that’s just a matter of removing the GCD on tax collection, so that their gold collection is a bit more consistent and doesn’t have as much disparity between their lowest possible yield per minute and their highest.
IO in dark are absolutely worth getting. Early game with micro they pay for themselves very fast. Standard Chinese opening right now uses IO to generate 100 gold ~3:30 game time to start aging up while supervising the mill and never needing to mine any gold.
1 Like