Money Does Grow on Trees: How to Convert Forests into Lots and Lots of Gold

I just find it disturbing the AI doesn't know how to win besides culture; it just happens to build a spaceship due to luck rather than planning. :S

@TMIT

I think you hit the nail... the programmers at firaxis and our volunteers aren't that much different in skills.
The difference is that AI programmers here are backed up by strategic articles from other players --> that is the heuristic.
Tough to judge which was going on in firaxis when programming AI ;-).

Don't forget strategies always change, you can see it even yet 2 years after last patch. And when you don't continue with patching the software basically dies (another rule out of school ;-)).

When I was going through some parts of the firaxis code thinking about changing city governor behavior when building wealth I realized that the code in itself isn't that dumb, but it's mostly all some kind of evaluation code with weights for different situations.

The same probably is true for other parts of AI.

i took one conclusion out of it... I don't want to burn there my mid ages though .

I'll take up the challenge explaining why I think the current implementation of failgold is a "failure".

I think failgold was introduced to compensate people for failed wonder attempts, thus reducing the risk in actually trying to build a wonder, and therefore encouraging you to go for it even if you thought you had only a small chance of success. The method used in earlier versions of Civ (transferring the committed hammers to a different wonder) had much issues of its own and was rightly tossed. The basic idea was to give players who failed the wonder the same result as if they had used the solid, always(*) available, never really bad, but usually not great either, option of building wealth instead. In principle I think this is a good idea, but the implementation of failgold is not precisely according to this idea.

ETA: Being a compensation for failing, it should not be the main goal of building the wonder in the first place, thus the compensation should at most be what you could have gotten otherwise.

The problem being of course the treatment of wonder-specific production bonuses. Promoting the current implementation of failgold requires defending the principle that
at seemingly random periods of time you can build wealth more effectively. I fail to see why the game gets more interesting and not just more micro-intensive if you increase the amount of wealth you can build in a limited amount of cities for small periods of times. Besides, even though much about Civ is that more advanced civs get extra bonuses, helping them to increase their lead, having advanced civs get a temporary boost in their wealth production efficiency which backwards civs won't get, even if they reach the same tech-level as the advanced civ had when it got the bonus, seems unnecessary.

This only becomes a significant issue if it is (ab?)used to the max by even more micro involving chops and whip overflow. The AI is not smart enough to do this, and since it does not involve smart thinking (beyond the initial idea) but just lots of micro, I'd prefer it if you got the same benefit if you just build wealth in all your cities instead of having to change the production queue all the time. Thankfully I can decide to refuse to use this tactic and the AI isn't smart enough so the current implementation of failgold does not harm the game too much. (In particular I would agree that there are much worse problems to fix in a patch, but that does not mean that I still think this is a problem).

I thought about how to patch it, but unfortunately I don't think a simple change in the code for the amount of failgold (as in the code for the overflow hammers) will work. The reason you benefit is that the production bonus for building wealth is different from that for building a wonder. You can't however use the instant production bonuses at the time of completion of the wonder to recalculate the amount of wealth you would have built in the same time given the number of hammers invested in the wonder. The problem is that the production bonuses might have changed over the entire period you were building the wonder. Thus you need to know what bonuses were applicable throughout the entire period of building the wonder. This information is not remembered by the computer. (My simplest solution would be to introduce another, hidden, variable which counts the amount of wealth accumulated while building the wonder, but this is certainly more complicated).

To give an example of the issues that can happen using simple patching rules (assuming a non-industrious civ):

Suppose you have 1 base production and build Stonehenge. Suppose you build it without stone for 10 turns and then connect stone and build it for another 5 turns, after which some AI builds it. Then you have 20 hammers in the bag, so you now get 20 failgold. If you patch by using the latest production bonuses you get (20/2=)10 gold. Neither of which is the 15 gold you would have gotten had you build wealth the entire time, which is what I would consider fair.

On the other hand, in the same situation suppose that you only get stone at the 15th (and last) turn. But instead you have grown and increase your production to 2 base hammers at turn 12. Then you still have (12*1+2*2+1*4=20) hammers in the bag, and are connected to stone. So the final situation is identical to the previous one. However now, building wealth would have netted you 18 gold, which is different from the 15 you got before.

(*) Of course, you need currency to build wealth, but currency comes pretty early and before currency you usually have better things to build than wealth anyway.

TheMeInTeam said:

What I dislike is blanket and nonsensical removals of some things on the sole basis that experts use it to do well in unexpected ways (without testing it or considering costs whatsoever), while leaving things that are stronger in the game and completely ignoring the basics.

Click to expand...

The soapbox aside, I appreciate that in general, there's a problem with developers patching bugs without any concern for the possibility that the bug makes the game better. But I also appreciate there's also a problem with players whining about patches without any concern for the possibility that the bug makes the game worse.

Your comment was phrased as a whine, but I was curious if it was actually a reasoned comment rather than how it appeared. (and to a smaller extent, I was in the mood to call the whining if that's what it was)

Just thinking, there are tons of alternatives, I wonder why they chose to go with gold

1. There is no fail gold, you lose the hammers flat. But when any civ starts building a world wonder, an announcement is made across the world like "The Egyptians have started building a great wonder, The Pyramids". Then if you join the race after that you know what you're getting into and it's your own fault if you lose.
   
2. As above and also the wonder adviser screen shows updating %complete so you can decided to pull out of races you're clearly losing.
   
3. Not fail gold, but the unmultiplied hammers are immediately paid back in as many of your best available unit as they can build, applying any active unit-build multipliers.
   
4. The unmultiplied hammers are paid back in city instead of gold. Arguably most realistic for such cultural things as wonders, and less exploitable.
   
5. Let more than one civ build world wonders, then there's no failure and so no fail gold. For example you would build "some pyramids" and not "The Pyramids", if you felt the benefits were worth the hammers. Or a stone henge, or a statue of liberty.
   
6. Only one civ can build world wonders, but partial attempts are "culturally linked" to the actual wonder for a short time and confer the regular benefits for a few turns. E.g. 100 into the pyramids gets you 10 turns of representation.
   
7. The unmultiplied hammers are paid back over time via a temporary boost to the city tile, say +5 until they're all used up. Probably most realistic - e.g. you can build the hanging gardens a little more quickly because you already have some cut stone lying around the place from your half henge.
   
8. Unmultiplied hammers are paid back in fresh forest growth in random places across your empire.
   
9. No payout but partial wonders become "city ruins" on a random tile in the BFC, conferring random benefits in the future.
   
10. Failure results in a diplo bonus (Either "We sympathise with your cultural loss " or "Your pathetic attempt to build that wonder has entertained us! ") with one or all civs.
    
11. Failing a wonder pops up a screen offering you the choice from several of the above, like the random event screen. Some choices would depend on how many hammers you'd put in - some could require an extra gold investment like random events do.

@Kid

some of solutions you offered operate on the basis that player should get rewarded for failing building the wonder.

I like most 1,2,5 and 9.

5 has btw great logic... if you have the resources/technology (knowledge), why should you care if civ B built already something similar?
Granted some balance check would be needed. But for that we have already national wonders.

Well not so much reward as compensation for the fact that some of the wonders are game-breakingly expensive for their time-period if you fail. Losing 100 on stonehenge is sort of meh but 400 on the pyramids at the same stage of the game you can't really recover from.

There's only a handful of early wonders I'd say it really matters for. From the medieval period onwards the wonders are reasonably cheap and a few hundred hammers here or there is nothing. If the mids, MOM, maybe parthenon and sistine were a bit cheaper or came a bit later the whole idea of failure compensation would maybe never have been created because it wouldn't be needed.

Nice ideas Kid R, but I don't think any beat the "unmultiplied hammers turned into gold (as if you had build wealth)". The only problem with the current failgold is that it is about multiplied hammers, thus failgold beats building wealth. Many of your options use "unmultiplied hammers" but knowing that number is the problem. I've not heard many people complaining that building wealth is overpowered so using the hammers to build wealth as default is not a silly choice, in fact gold is probably the most versatile of resources in civ. The benefits of turning hammers into units/culture/forests/diplo bonuses is much more situational.

Kid R said:

Just thinking, there are tons of alternatives, I wonder why they chose to go with gold

Click to expand...

They probably chose gold because it's something you could have done otherwise via building wealth. (neglecting the bug that lets you benefit from wonder bonuses)

On the compensation for failure (if any), there are two significantly different ways to do things:

• Failing the wonder is better than not even trying (and possibly better than succeeding)
• Never trying is better than failing the wonder

that lead to games of very different flavor.

In the latter, aiming for a wonder is a strategic risk, with the potential for great^* reward, with the danger of your effort being wasted. Planning goes into making sure that you will succeed for the wonders you want, and identifying when you're unlikely to succeed. The consolation prize for failure is to tweak how painful the penalty for failure is.

In the former, however, building a wonder is a strictly positive^* investment. Planning is centered around making sure one is available when desired, and influencing which of the possible outcomes you get.

*: these is situational, of course

Just based on the type of mechanic involved, in the case where the best possible outcome is to try and fail the wonder, I expect that one of the two following things are likely:

• The outcome is actually unbalanced and needs to be nerfed.
• The game would be better off if you had access to the outcome directly, rather than indirectly through wonder failure

[*]Failure results in a diplo bonus (Either "We sympathise with your cultural loss " or "Your pathetic attempt to build that wonder has entertained us! ") with one or all civs.

Click to expand...

Ah, the old joke.

• If you build the Stonehenge, you will achieve cosmic joy.
• But what if I fail to build it?
• Then the cosmic joy is on you.

The outcome is actually unbalanced and needs to be nerfed.

Click to expand...

Please don't try to fix something that isn't broke.

We've been through this before when the amateurs were touting their horns, and sadly, the rest of us have to now pay for it (indefinitely).

It's certainly not unbalanced. Most games, failgold is a (slightly earlier, slightly weaker) version of building wealth. 9/52 leaders are IND, and failgold is one of the things keeping IND competitive; take it away, and IND would be a bit on the weak side. Sometimes you'll get marble or stone in a game (1/3 games?); then it is a big boost to your economy. But that's not because failgold is strong; that's because marble and stone are strong. Perhaps too strong - I might have been happier if they gave a 50% boost to production instead of 100% - but given that they're fairly uncommon resources that often take a real sacrifice of getting an early weaker city to secure them, I'm OK with the current balance.

vranasm said:

@Htadus

if you would play plain 3.19 game without any mods, you won't get any gold from OF whipping warriors.

Click to expand...

Thank you. I did not know that. All I play is buffy.

TheMeInTeam said:

Doesn't work with overflow being bugged in 3.19. Used to work.

Click to expand...

Just learned. I wonder why buffy allow that then.

Buffy allows it because it was deemed to be a bug, and so buffy fixed it, but does it without modifiers (I believe)

coanda said:

It's certainly not unbalanced. Most games, failgold is a (slightly earlier, slightly weaker) version of building wealth. 9/52 leaders are IND, and failgold is one of the things keeping IND competitive; take it away, and IND would be a bit on the weak side. Sometimes you'll get marble or stone in a game (1/3 games?); then it is a big boost to your economy. But that's not because failgold is strong; that's because marble and stone are strong. Perhaps too strong - I might have been happier if they gave a 50% boost to production instead of 100% - but given that they're fairly uncommon resources that often take a real sacrifice of getting an early weaker city to secure them, I'm OK with the current balance.

Click to expand...

It's not just Marble and Stone. Ivory for Statue of Zeus. Copper for Colossus.

There's a pretty high chance of getting Marble OR Stone somewhere near your capital.

And Obsolete, the forest-fail-gold economy is basically a hammer economy, that's TWICE AS EFFECTIVE AND IT CAN USE FORESTS FOR COMMERCE. So any strategies that use hammer economies can also use the forest-fail-gold economy.

To be honest, I would never waste my precious forests in such a way. As others have mentioned, I want/need those hammers for more important things than commerce.

The larger the map, and the slower the speed, the more you need commerce instead of production.

Even on standard maps, the 1 for 2.5 trade is pretty damn good.

Something to consider.

To get more production, just whip more. You can afford it with a big stack of money.

The problem being of course the treatment of wonder-specific production bonuses. Promoting the current implementation of failgold requires defending the principle that
at seemingly random periods of time you can build wealth more effectively.

Click to expand...

You assert this is a problem. Let's see some evidence. These multipliers are part of trait balance...and "seemingly random"? The wonders are on the same part of the tech tree with the same hammer cost and same tech cost to reach them every single game. What's random about fail gold? Simply when you're beaten to the wonder...IF you are beaten to the wonder...you don't have to be.

having advanced civs get a temporary boost in their wealth production efficiency which backwards civs won't get, even if they reach the same tech-level as the advanced civ had when it got the bonus, seems unnecessary.

Click to expand...

So what, you're saying we should ban oxford? Sorry but your logic doesn't follow.

This only becomes a significant issue if it is (ab?)used to the max by even more micro involving chops and whip overflow. The AI is not smart enough to do this

Click to expand...

Derrrp . AI incompetency is not a valid basis for making the feature an issue. Once again, you might as well ask to ban things like heroic epic, efficient 2 pop whips, oxford, and drafting. BAN DRAFTING! The AI is not smart enough to use it, amirite? Let's please keep discussion of the mechanic to specifically the mechanic, and not start using "players that don't play the game or try" as a basis for whether the mechanic works.

What you have (rather profoundly) failed to do in the entire post I quoted parts from is to provide a single piece of valid argument that shows evidence that wonder fail gold provides an overpowering advantage. If you use AI incompetency with a mechanic as the basis, you'd have to rip out the majority of basic human activity (stop people from chopping, drafting, chain-whips). After all, the AI DOES build wonders and fail them, it just executes that poorly. There's no fundamental difference between that and an axe rush. Lacking this basis, your argument has no supporting examples or logical evidence that fail gold is overpowered. In other words, you're calling for the industrious trait to be nerfed without basis. Unfortunately, there are people at failaxis that will listen to such nonsense. HAVE listened to it. They've hurt the game.

I appreciate that in general, there's a problem with developers patching bugs without any concern for the possibility that the bug makes the game better.

Click to expand...

First of all, you have 0 evidence it was a bug. I can call the oxford bonus a bug and drop it to 50%. Would that make you happy? Would it make the game better? No, it would be nonsense, precisely like the overflow issue. Failaxis INTRODUCED A BUG there, they did NOT get rid of one. Overflow doesn't work in any capacity now...they couldn't even patch something as effectively as the unofficial patch. Painful much?

And whining? Seriously? I'm calling out incompetence. Introducing a bug in overflow while the game controls are *actually* bugged (control, shift, alt click do not work consistently) is flat-out incompetent. Leaving a victory condition in a broken state is incompetent. Introducing a bug in spread culture espionage mission as a panic reaction is incompetent.

Remember, kids! Using a governor glitch to get 2x as many as you could normally get is FINE. We will keep that in the game along with infinite oracle techs alllllllll the way to 3.19. But having protective be a somewhat competitive trait?! Oh no, we can't have THAT ........

Tell you what, rather than calling me out for "whining" because failaxis has seriously screwed up their last few civ IV BTS patches, why don't you provide some evidence that something was/is bugged, or present an argument of some kind?

They probably chose gold because it's something you could have done otherwise via building wealth.

Click to expand...

No. For the wonders where doing this provides the largest advantage and is most readily available, you couldn't have built wealth actually.

In the latter, aiming for a wonder is a strategic risk, with the potential for great* reward, with the danger of your effort being wasted. Planning goes into making sure that you will succeed for the wonders you want, and identifying when you're unlikely to succeed. The consolation prize for failure is to tweak how painful the penalty for failure is.

In the former, however, building a wonder is a strictly positive* investment. Planning is centered around making sure one is available when desired, and influencing which of the possible outcomes you get.

*: these is situational, of course

Click to expand...

This is better. We're on the right track now. The thing is that going for the wonder is *not* strictly beneficial, even when failing the wonder provides bonus gold.

Basically...

The outcome is actually unbalanced and needs to be nerfed.

Click to expand...

Is horsecrap. If the mechanic were something you would always prefer in every situation no matter what, or inordinately frequently, it would be different. If there were no research investment costs to reach wonders over alternative things (like actually getting currency), an imba argument would hold more water. If industrious consistently outperformed other traits on a routine basis, maybe you could put for this argument.

But none of those things are available. This is precisely the same scenario that led to the idiot nerfing of protective, the weakest trait. Pre 3.19, do you think protective was the strongest trait? Did anybody claim/demonstrate it was so? Did MP human games all pick protective? Did players routinely win outright or even get incomparable advantages over other players with superior traits? No, no, and no. In order to make a change, a COMPETENT designer/programmer needs some reasonable basis for making it. Maybe it is WHINING to suggest that, but good designers have that basis when they patch things. They do not patch based on jerk reactions to community discoveries, as failaxis has repeatedly demonstrated it does.

We had 0 evidence that protective was too strong in 3.17 and we have 0 evidence industrious is too strong now. Just because SOMETIMES it is beneficial to fail a wonder does not mean that it's a broken mechanic. You need to demonstrate some in-game comparison that shows that it's too strong compared to alternatives. Nobody has done it. I have a feeling nobody can do it. On some maps, attempting fail gold will make you lose or set you back, even with IND and/or stone/marble. But NOPE, MUST PATCH IT! It's UNORTHODOX trolololol .

And Obsolete, the forest-fail-gold economy is basically a hammer economy, that's TWICE AS EFFECTIVE AND IT CAN USE FORESTS FOR COMMERCE. So any strategies that use hammer economies can also use the forest-fail-gold economy.

Click to expand...

That is, of course, assuming you want the forests for the money and not for units, infra, or actually completing the wonders . On any given map, these opportunity costs might be too high.

Of course, spawning near jungle belts/with limited forest can cause grief for this strategy too .

Why yes, but if you DO have the forests, then it makes it pretty worthwhile, wouldn't it?

Obviously using praets is not possible without Iron. Similarly, forest fail-gold is not possible if there is no forest. But does that make Praets a bad UU? No!

And I'm starting to think that besides the Pyramids, Oracle, the Colossus for Archipelago maps, the GW for spy economies and huge maps, and the GLH, the other Wonders are over-rated.

^ I was more pointing out the necessity and alternative uses of forest to demonstrate valid opportunity cost/usage potential concerns, and thus weaken any "overpowered" arguments.

In the old day before Patch 3.19, someone (crusher1 I believe) already advocated the same idea of translating forests into cash, in a more general and easier way by utilizing the leader's trait. Such as protective wall, aggressive barrack, and creative library. I was not sold on such idea. Forests are precious and limited resources, there are better ways to use them.

If you go warmonger route, ~5 cities should be enough, then chop those forests to grab the decent sites as soon as possible is far more important, and the remain forests will also greatly speedup your military buildup when you have your expected military tech, which is crucial in high level.

If you go peacemonger route, then again chopping the 1st few settlers is also more important, moreover, wonders are far better to support your economy in the long term. Oracle, GLH, TGL, there are just too many.