The case for Virginia Company in the "Greed" meta
Posted: 07 Mar 2016, 21:14
Fellow Brit players,
First of all there's not enough of you!
This post will attempt to detail the benefits of the "Virginia Company" (VC from here) card which makes British manors cost 88 wood instead of 135. It's an overview of my personal journey from "Haha, that card is terrible" to "Well shit", which underlined for me that confirmation bias is strong in not just the people that I disagree with, but in myself as well
DISCLAIMER: Whenever I state something in this post seemingly as fact but it is not backed up by math, remember it is just my opinion. I will not put "imo" after every sentence.
Let me just state for the record what I will be trying to prove with this post: VC is a viable card in match ups where British do not lose to a rush.
This discussion is not new, and by that I don't mean that I have personally been spamming about this card for a while (which I have). I mean that countless threads have been made on a variety of forums over the years, and I'm sure that they exist here as well, asking about or aguing for VC. Mostly, and this is interesting, by low-ranked players. Not the establishment, if you will, but people who saw the tremendous cost decrease and wondered why no one used it.
I myself have always (and by that I mean 4+ years ago, when I was active) argued against it. I tested the card before, as I'm sure a lot of you have as well, and came to the conclusion that it's just not worth it. “It's inflexible. It's missing 3 villagers early on, which slows your military production. And even with 20 manors, it is still 3 vills behind a player who did not send VC but also maxed manors.”
These are the standard arguments you will hear from people arguing that VC is a bad card, and they are the arguments you would have heard from me if you had asked me 5 years ago.
In my and other high level players' defense: the reason why it has been overlooked for so long, I think, is that it only recently became viable. It simply did not fit into the meta (or on the maps, for that matter) before. This explains why low-ranked players were exploring it while high-ranked players were quick to dismiss it. Of course there's also the part where we ruined Brits by increasing manor cost and pop in AS FP 1.2, which made VC entirely not viable.
Brits, and this is surprising given the civ design, are even now often played relatively aggressively. Forward towers and early military production are a very common sight in Brit games. This is not because they are a "rush civ", but rather because:
1. They often get rushed (or used to anyway) so they need early military presence, or
2. ... they need to prevent the other player from doing an economic semi-FF unharmed.
3. Map control becomes vitally important later in the game as Brits which makes taking the map early on all the more tempting.
4. In a British mirror it is extra important to take map control early because it is easy to hang on to, which makes VC not viable in mirrors. Because most top level games these days are mirrors (although finally in the Winter tournament the rules prevented this), it is not worth it for top players to explore options beyond the auto-pilot standard build.
So that’s my theory about the reasons why VC was not used before. Now, on to the fun stuff.
There's a new kid in town, as trading posts gain popularity and the maps' resource balance is improved. That new kid is Greed, and by greed I mean economic play styles designed to end the game after 10 minutes, not before.
Brit players should be very excited about this. After all, the British boom is the fastest and cheapest in the game, which allows for some interesting new options. And I’ll say even more: I would not be surprised if Brits are currently one of the best civs.
For the past week I have been working on a program which mathematically simulates AoE3 and is able to quickly calculate the difference between build orders. Using this I am able to compare 3v build orders to VC build orders without having to test them ingame, eliminating my own poor execution and the uncertainties of your average AoE3 game.
Now of course I could have simply done the math for this specific occasion and detailed it here, but I am hoping (with enough time commitment) the program will evolve into something which is easily scriptable and works for all civs, and this is a great place to start. For now I am working with a very simple simulation of the British early game. Remember that due to idle vill time (walking from hunt to hunt etc), which my program doesn't take into account (yet), you will be slightly slower than the simulated times in a real game. A benchmark if you are going for a full manor boom: Your final manor should finish around 5:30.
For this simulation I am assuming Brits started with 300 wood and 200 food in crates. In this case Brits will build 2 manors immediately and attempt a 4:10 colonial age up. The program's starting conditions are based on (near) perfect early game vill micro which I tested ingame.
Starting position
Game time: 56 seconds
Vills: 10 (all gathering food)
Manors: 2
Food: 220
Wood: 30
Coin: 0
XP: 187
The script which the program executes (essentially the build order) is as follows. It is always automatically creating villagers unless aging up.
- Send the first shipment (VC or 3 vill).
- Age to colonial with the 500f politician.
- Switch all vills to wood.
- Build manors until colonial age is reached.
- Send 700 wood.
- Build manors until they are maxed. (or until x manors have been built, this is relevant later)
- Build 2 barracks.
- Switch vills to food/wood for longbow production.
- Build longbows.
As you may have noticed, this is the build order which gets the most possible amount of value out of the VC card but, surprisingly, it's also a build order you can easily get away with in many match ups. As for the results of the sim:
3 vills first
- Colonial age at ~4:10 with 6 manors built during transition.
- Last manor starts building at 5:30 - 5:40.
- Vill count at 6 minutes is 39.
- 2 barracks are done by 6:10.
- Longbow count at 7:00 is 13.
- Longbow count at 8:20 is 35.
VC first
- Colonial age at ~4:10 with 7 manors built during transition.
- Last manor starts building at 5:00 - 5:10.
- 2 barracks are done by 5:30.
- Vill count at 6 minutes is 36.
- Longbow count at 6:00 is 8.
- Longbow count at 7:00 is 23.
- Longbow count at 8:20 is 45.
(If anyone would like details on how I implemented the math, as I'm sure some of you will want to poke holes in it, feel free to ask)
Above shows that the VC boom is about 30 seconds faster. Keep in mind I did not implement the second colonial shipment just yet, but due to the faster XP gained from the manor boom the VC build gets this shipment significantly faster as well. Of course, the VC build is still 3 villagers behind and this is a deficit that will stay for the duration of the game.
So if you can get away with a greedy build, which you often can in the current meta, the discussion is now about +3 villagers versus a 30 second faster build order. Given the nature of British and their relative weakness to early fortress timing attacks, I have a very easy time making a case for the faster build here. We shouldn't forget that +3 vills may seem like a lot, but because the British economy is so big this is only a 6-7% increase. It will pay off over time, but it will take long to do so. A 30 second faster build on the other hand can be the difference between losing to a mid colonial or early fortress timing attack or holding it off with an easily winning economy.
But ask yourself, how often do you really max manors before 7 minutes if you sent 3 vills? Not often, you can't really afford it because it slows you down a lot having to chop all that wood, so the +3 vills wouldn't even come into play until way later. You may argue “why not just build 17 manors instead of 20, save 405 wood and have equal vill count compared to the VC build?” The thing is you’re still slower. Slightly, but you’re still slower. Let’s throw in a couple of graphs:
And to reflect relative vill counts instead of manors built (adjusted the VC graphs 3 to the left, meaning the 3v build saved 405 wood):
We mostly need to be looking at the second graph, because this represents the actual vill counts that you end up booming towards. The VC graphs have been adjusted to the left in order to compensate for the -3 vills. Keep one thing in mind though: In the second graph the VC build is always ahead by 81 XP compared to the 3v build due to the XP gained from building the extra 3 manors.
The data tells us that the builds turn out about equal if you build less than 13-14 manors with 3 vills first (equivalent to 16-17 with VC first). However, 3 vills quickly starts falling off if you want to boom to 14+ manors (or 17+ with VC). This is due to the 700w being spent on manors and the player being unable to build their barracks in a timely fashion.
Not only does the VC build produce significantly more units in the given time frames, it also starts building units quicker. The VC manor boom ends sooner so it will have the barracks up before the 3v build, even if the 3v build chooses to build 3 less manors to compensate for the lost time.
And for the record (this will be obvious to most of you I hope), if you build less than 9 manors (chosen arbitrarily) before 6 minutes, which is equivalent to less than 12 manors with VC, this can’t be considered a boom and the 3 vill build is always superior. That’s the reason the graph starts at 9. We are looking at the best boom builds for Brits here, not the best rush.
So with the data out of the way, let’s dig into the strategy a little more. This very fast boom opens up some interesting options. Brits can now think of going 1 rax (build the rax slightly before 700w so you are prepared for cav) 5 pike 700w 5v 700g semi-FF against opposing semi-FFs and not lose to a 9 minute timing attack. Against colonial-focused build orders, they can try to get away with a fast boom and rely on minutemen + longbows and vills to tank in order to hold any 6-7 minute timings. 23 longbows instead of 12 at 7:00 is a big difference, and 5-10 pikes at 6:20 compared to zero can absolutely decide a game against a cav semi-FF. This goes to show how important speed is.
And there’s 1 last argument that I know will be coming which I should get ahead of: “If your opponent scouts VC he will just rush you and you wasted a card.”
But there’s an important factor you may be missing if you were thinking this: Rushing sucks. If you went VC and you scout a rush from your opponent, you are actually happy. Yes, 3 vills would’ve been better, but if you had sent 3 vills they wouldn’t have rushed. Remember that by rushing, your opponent is sacrificing his own economy as well as hurting yours, and with good wall + longbow laming he is not going to break you. Then if the rush is over and you feel comfortable behind your cozy wall, you immediately switch gears and max manors. This way VC wasn’t all that bad in the end, and meanwhile your opponent opted out of his economic semi-FF just to stop you from booming to 40 vills at 6 min. Of course there are some match ups where a rush will actually kill you, and those are the match ups where VC is not viable. Which match ups are those? I’ll leave that up to you.
Cheers, and commence the naysaying!
First of all there's not enough of you!
This post will attempt to detail the benefits of the "Virginia Company" (VC from here) card which makes British manors cost 88 wood instead of 135. It's an overview of my personal journey from "Haha, that card is terrible" to "Well shit", which underlined for me that confirmation bias is strong in not just the people that I disagree with, but in myself as well
DISCLAIMER: Whenever I state something in this post seemingly as fact but it is not backed up by math, remember it is just my opinion. I will not put "imo" after every sentence.
Let me just state for the record what I will be trying to prove with this post: VC is a viable card in match ups where British do not lose to a rush.
This discussion is not new, and by that I don't mean that I have personally been spamming about this card for a while (which I have). I mean that countless threads have been made on a variety of forums over the years, and I'm sure that they exist here as well, asking about or aguing for VC. Mostly, and this is interesting, by low-ranked players. Not the establishment, if you will, but people who saw the tremendous cost decrease and wondered why no one used it.
I myself have always (and by that I mean 4+ years ago, when I was active) argued against it. I tested the card before, as I'm sure a lot of you have as well, and came to the conclusion that it's just not worth it. “It's inflexible. It's missing 3 villagers early on, which slows your military production. And even with 20 manors, it is still 3 vills behind a player who did not send VC but also maxed manors.”
These are the standard arguments you will hear from people arguing that VC is a bad card, and they are the arguments you would have heard from me if you had asked me 5 years ago.
In my and other high level players' defense: the reason why it has been overlooked for so long, I think, is that it only recently became viable. It simply did not fit into the meta (or on the maps, for that matter) before. This explains why low-ranked players were exploring it while high-ranked players were quick to dismiss it. Of course there's also the part where we ruined Brits by increasing manor cost and pop in AS FP 1.2, which made VC entirely not viable.
Brits, and this is surprising given the civ design, are even now often played relatively aggressively. Forward towers and early military production are a very common sight in Brit games. This is not because they are a "rush civ", but rather because:
1. They often get rushed (or used to anyway) so they need early military presence, or
2. ... they need to prevent the other player from doing an economic semi-FF unharmed.
3. Map control becomes vitally important later in the game as Brits which makes taking the map early on all the more tempting.
4. In a British mirror it is extra important to take map control early because it is easy to hang on to, which makes VC not viable in mirrors. Because most top level games these days are mirrors (although finally in the Winter tournament the rules prevented this), it is not worth it for top players to explore options beyond the auto-pilot standard build.
So that’s my theory about the reasons why VC was not used before. Now, on to the fun stuff.
There's a new kid in town, as trading posts gain popularity and the maps' resource balance is improved. That new kid is Greed, and by greed I mean economic play styles designed to end the game after 10 minutes, not before.
Brit players should be very excited about this. After all, the British boom is the fastest and cheapest in the game, which allows for some interesting new options. And I’ll say even more: I would not be surprised if Brits are currently one of the best civs.
For the past week I have been working on a program which mathematically simulates AoE3 and is able to quickly calculate the difference between build orders. Using this I am able to compare 3v build orders to VC build orders without having to test them ingame, eliminating my own poor execution and the uncertainties of your average AoE3 game.
Now of course I could have simply done the math for this specific occasion and detailed it here, but I am hoping (with enough time commitment) the program will evolve into something which is easily scriptable and works for all civs, and this is a great place to start. For now I am working with a very simple simulation of the British early game. Remember that due to idle vill time (walking from hunt to hunt etc), which my program doesn't take into account (yet), you will be slightly slower than the simulated times in a real game. A benchmark if you are going for a full manor boom: Your final manor should finish around 5:30.
For this simulation I am assuming Brits started with 300 wood and 200 food in crates. In this case Brits will build 2 manors immediately and attempt a 4:10 colonial age up. The program's starting conditions are based on (near) perfect early game vill micro which I tested ingame.
Starting position
Game time: 56 seconds
Vills: 10 (all gathering food)
Manors: 2
Food: 220
Wood: 30
Coin: 0
XP: 187
The script which the program executes (essentially the build order) is as follows. It is always automatically creating villagers unless aging up.
- Send the first shipment (VC or 3 vill).
- Age to colonial with the 500f politician.
- Switch all vills to wood.
- Build manors until colonial age is reached.
- Send 700 wood.
- Build manors until they are maxed. (or until x manors have been built, this is relevant later)
- Build 2 barracks.
- Switch vills to food/wood for longbow production.
- Build longbows.
As you may have noticed, this is the build order which gets the most possible amount of value out of the VC card but, surprisingly, it's also a build order you can easily get away with in many match ups. As for the results of the sim:
3 vills first
- Colonial age at ~4:10 with 6 manors built during transition.
- Last manor starts building at 5:30 - 5:40.
- Vill count at 6 minutes is 39.
- 2 barracks are done by 6:10.
- Longbow count at 7:00 is 13.
- Longbow count at 8:20 is 35.
VC first
- Colonial age at ~4:10 with 7 manors built during transition.
- Last manor starts building at 5:00 - 5:10.
- 2 barracks are done by 5:30.
- Vill count at 6 minutes is 36.
- Longbow count at 6:00 is 8.
- Longbow count at 7:00 is 23.
- Longbow count at 8:20 is 45.
(If anyone would like details on how I implemented the math, as I'm sure some of you will want to poke holes in it, feel free to ask)
Above shows that the VC boom is about 30 seconds faster. Keep in mind I did not implement the second colonial shipment just yet, but due to the faster XP gained from the manor boom the VC build gets this shipment significantly faster as well. Of course, the VC build is still 3 villagers behind and this is a deficit that will stay for the duration of the game.
So if you can get away with a greedy build, which you often can in the current meta, the discussion is now about +3 villagers versus a 30 second faster build order. Given the nature of British and their relative weakness to early fortress timing attacks, I have a very easy time making a case for the faster build here. We shouldn't forget that +3 vills may seem like a lot, but because the British economy is so big this is only a 6-7% increase. It will pay off over time, but it will take long to do so. A 30 second faster build on the other hand can be the difference between losing to a mid colonial or early fortress timing attack or holding it off with an easily winning economy.
But ask yourself, how often do you really max manors before 7 minutes if you sent 3 vills? Not often, you can't really afford it because it slows you down a lot having to chop all that wood, so the +3 vills wouldn't even come into play until way later. You may argue “why not just build 17 manors instead of 20, save 405 wood and have equal vill count compared to the VC build?” The thing is you’re still slower. Slightly, but you’re still slower. Let’s throw in a couple of graphs:
And to reflect relative vill counts instead of manors built (adjusted the VC graphs 3 to the left, meaning the 3v build saved 405 wood):
We mostly need to be looking at the second graph, because this represents the actual vill counts that you end up booming towards. The VC graphs have been adjusted to the left in order to compensate for the -3 vills. Keep one thing in mind though: In the second graph the VC build is always ahead by 81 XP compared to the 3v build due to the XP gained from building the extra 3 manors.
The data tells us that the builds turn out about equal if you build less than 13-14 manors with 3 vills first (equivalent to 16-17 with VC first). However, 3 vills quickly starts falling off if you want to boom to 14+ manors (or 17+ with VC). This is due to the 700w being spent on manors and the player being unable to build their barracks in a timely fashion.
Not only does the VC build produce significantly more units in the given time frames, it also starts building units quicker. The VC manor boom ends sooner so it will have the barracks up before the 3v build, even if the 3v build chooses to build 3 less manors to compensate for the lost time.
And for the record (this will be obvious to most of you I hope), if you build less than 9 manors (chosen arbitrarily) before 6 minutes, which is equivalent to less than 12 manors with VC, this can’t be considered a boom and the 3 vill build is always superior. That’s the reason the graph starts at 9. We are looking at the best boom builds for Brits here, not the best rush.
So with the data out of the way, let’s dig into the strategy a little more. This very fast boom opens up some interesting options. Brits can now think of going 1 rax (build the rax slightly before 700w so you are prepared for cav) 5 pike 700w 5v 700g semi-FF against opposing semi-FFs and not lose to a 9 minute timing attack. Against colonial-focused build orders, they can try to get away with a fast boom and rely on minutemen + longbows and vills to tank in order to hold any 6-7 minute timings. 23 longbows instead of 12 at 7:00 is a big difference, and 5-10 pikes at 6:20 compared to zero can absolutely decide a game against a cav semi-FF. This goes to show how important speed is.
And there’s 1 last argument that I know will be coming which I should get ahead of: “If your opponent scouts VC he will just rush you and you wasted a card.”
But there’s an important factor you may be missing if you were thinking this: Rushing sucks. If you went VC and you scout a rush from your opponent, you are actually happy. Yes, 3 vills would’ve been better, but if you had sent 3 vills they wouldn’t have rushed. Remember that by rushing, your opponent is sacrificing his own economy as well as hurting yours, and with good wall + longbow laming he is not going to break you. Then if the rush is over and you feel comfortable behind your cozy wall, you immediately switch gears and max manors. This way VC wasn’t all that bad in the end, and meanwhile your opponent opted out of his economic semi-FF just to stop you from booming to 40 vills at 6 min. Of course there are some match ups where a rush will actually kill you, and those are the match ups where VC is not viable. Which match ups are those? I’ll leave that up to you.
Cheers, and commence the naysaying!