(CONFIRMED) Marathon/Espionage Bug

[DrewBledsoe]

If you try to posion a cities water supply on normal speed, It gives 8 for 8 turns. Marathon incorrectly triples both the duration, and the effect, hence:- 24 for 24 turns. This effectively destroys a city in the middle game more completely than if it was nuked. I presume this was overlooked, and it should be 8 for 24 turns. Triple the duration, same strength. The same thing applies to unhappiness.It again produces 24 for 24 turns instead of 8 for 24 turns.

 

[Seidrik_The_Gray]

I agree with you. Is there an xml line somewhere we can change to fix it until a patch?

 

[DrewBledsoe]

I agree with you. Is there an xml line somewhere we can change to fix it until a patch?

I haven't checked yet.

 

[ainwood]

Do you have a save from both of these situations?

 

[DrewBledsoe]

Do you have a save from both of these situations?

Fraid not, but here is the screenshot in another thread http://forums.civfanatics.com/showpost.php?p=5705230&postcount=497

 

[ainwood]

Thanks. The XML doesn't have information for varying game speeds, so it looks like its a scaling problem in the DLL.

 

[Todd Hawks]

I thought a bit about this in order to fix it until the patch: The xml only allows one factor to be entered, i.e. the number of turns and the effect is always the same. In normal, the city is losing 8+7+6+... = 36 food in these 8 turns. Since the marathon city takes thrice the amount of food to grow vs. the normal speed, at first glance it would make sense to lose 3 times as much food, which means that the loss should be 36*3 = 108, which means about 14 turns with an effect of 14/13/12/...

Now, a normal city might switch citizen assignment in order to get more food to counter the effect. The marathon city can do the same but it can't get more food than the normal city (same 21 tiles to work with), so it would be hit harder if using 14 as starting effect.

Therefore I chose 12 as compromise for the time being: Less overall effect (only 78 vs. the 108 calculated above) but the city gets hit harder at the beginning, so it has a harder time countering the effect and the actuall loss of population might be about the same. In order to get this, open the xml and change the corresponding number from 8 to 4 before you play a marathon game.

Foment unrest works a little differently but the reasoning is roughly the same, so I suggest to use the same number.

(PS: I'm not sure where I got the idea that the effect diminishes over time but the numbers also roughly work if it stays constant)

 

[Zoolooman]

Good thinking Hawk. It's a wonderful interim solution until they set it properly, like so:

-8 health/happiness for 24 turns, decreasing in effect by one point every three turns.

Also, has anyone looked into the fiscal costs of espionage in comparison to time? There is a potential imbalance there.

Consider this: when you research a technology at normal speed, you gain or lose a certain amount of cash. In Marathon, since you spend 3 times as many turns researching, that quantity of cash is effectively tripled. So if you're running on a gain, you'll earn 3 times as much money because it takes three times as many turns to finish researching. And likewise, if you're running a loss, it's 3 times the loss. This means that running on a gain in Marathon gives you more money.

Generally, this causes only minor problems: upgrade costs are effectively 1/3rd of normal, great merchants are effectively worth only 1/3rd as much, and the trading value of gold for a technology is miscalculated--the AI thinks a technology is worth 1/3rd as much gold as it ought to cost. But the first problem balances out, since Marathon armies are 2 to 3 times larger; the second problem is significant but not game-breaking; and the third problem isn't really a problem, since the AI and the player never stockpile enough gold to buy and sell technology at Marathon values, and it's best if the technology evaluation stays the same.

But consider espionage. If I can afford to poison the water every 10 turns on normal, the enemy will recieve 2 poison-free turns. But if the costs aren't multiplied for espionage on Marathon, then I can still afford to poison the water every ten turns, except that my last poisoning will have 14 more turns to go! In effect, I have 3 times as much money to put towards active espionage.

 

[Roland Johansen]

It seems to be generally accepted in this thread that the poison water supply should be -8 health for 24 turns, decreasing in effect by one point every three turns at marathon speed. While that sounds logical, it will result in a far higher population loss at marathon than at normal speed. The reason is that it takes exactly the same number of turns to lose a population point when out of food at marathon than at normal speed, namely 1 turn.

Lets say, you have a food storage in your city of 22 at normal speed. After 3 turns, this food is almost gone and the fourth turn, you'll lose population (assuming you can't compensate and are at the health cap). Then you will continue to have problems for four more turns, losing further population points or trying to get more food or health in the city.

At marathon speed a comparable situation would be a city with 66 food in storage. It would lose most of that food in 9 turns and the tenth turn, you'll lose population points. The problem now is that there are 14 turns left during which the city will lose population if it cannot remedy the situation. So the population loss will be much more severe.

The difference is even bigger when the happiness is lowered in the cities. Health problems can be countered quickly as the city shrinks because you'll stop using the low food tiles and thus can use the extra food to combat the temporary loss of health. The same option is not possible for happiness.

So a -8 health/happiness loss for 8 turns decreasing in effect with 1 per turn at normal speed is not equal in effect to a -8 health/happiness loss for 24 turn decreasing in effect with 1 per 3 turns at marathon speed.

Also, I would argue that in general the spy action which lowers happiness is more severe than the spy action which lowers health. If you lose 8 citizens working in a city, then you will in general lose more than 8 food per turn and the city will shrink quicker than with the health problem. It will also not be easy to combat the problem as with the health problem. Health problems can be countered by using only high food tiles while shrinking. You can barely use any tiles with the loss of happiness as the citizens won't work.

 

[Zoolooman]

It seems to be generally accepted in this thread that the poison water supply should be -8 health for 24 turns, decreasing in effect by one point every three turns at marathon speed. While that sounds logical, it will result in a far higher population loss at marathon than at normal speed. The reason is that it takes exactly the same number of turns to lose a population point when out of food at marathon than at normal speed, namely 1 turn.

Lets say, you have a food storage in your city of 22 at normal speed. After 3 turns, this food is almost gone and the fourth turn, you'll lose population (assuming you can't compensate and are at the health cap). Then you will continue to have problems for four more turns, losing further population points or trying to get more food or health in the city.

At marathon speed a comparable situation would be a city with 66 food in storage. It would lose most of that food in 9 turns and the tenth turn, you'll lose population points. The problem now is that there are 14 turns left during which the city will lose population if it cannot remedy the situation. So the population loss will be much more severe.

The difference is even bigger when the happiness is lowered in the cities. Health problems can be countered quickly as the city shrinks because you'll stop using the low food tiles and thus can use the extra food to combat the temporary loss of health. The same option is not possible for happiness.

So a -8 health/happiness loss for 8 turns decreasing in effect with 1 per turn at normal speed is not equal in effect to a -8 health/happiness loss for 24 turn decreasing in effect with 1 per 3 turns at marathon speed.

Also, I would argue that in general the spy action which lowers happiness is more severe than the spy action which lowers health. If you lose 8 citizens working in a city, then you will in general lose more than 8 food per turn and the city will shrink quicker than with the health problem. It will also not be easy to combat the problem as with the health problem. Health problems can be countered by using only high food tiles while shrinking. You can barely use any tiles with the loss of happiness as the citizens won't work.

Excellent points. While I'd thought about how much more damaging unhappiness is than unhealthiness, I completely missed the fact that Marathon speed effectively kills starving population points 3x faster.

The question is--how many of those 14 turns will result in a dead pop point? If the end result is that in most situations, the Marathon city loses only as many pop points as the Normal city, then it doesn't need a heavy rebalancing. However, if the end result is that the Marathon city loses twice as many pop points or something along those lines, then it needs a severe tweaking.

 

[Todd Hawks]

Roland is right here, I didn't think about one important fact: After losing all food, you continually lose population. It's not the reverse of pop growth (i.e. after you lose a pop, your food bar isn't full again).

So, let's assume you have a size seven city and it takes 34 food for normal and 102 food for marsathon to grow. Let's further assume your storage is nearly full and you absolutely can't grow any more food.
On normal, after 6 turns, your storage is about empty (8+...+3=33) and the next two turns will lose you a pop each. To get the same outcome at marathon (2 pop loss), you'd have to start at 14 to lose 102 food with still two turns left.

Now, let's assume the storage is half full:
Normal: 17 food, gone after about two turns, 5 or 6 pop lost.
Marathon: 51 food, gone after about 4 turns, 9 or 10 pop lost.

Even worse, empty storage:
Normal: 8 pop lost
Marathon: 14 pop lost
(it's only a size 7 city, though, so it's down to 1 either way)

Now, using 12 as basis for marathon, we get:
Full storage: No pop lost
Half storage: 5 or 6 pop lost
Empty: 12 pop lost

So basically with 12 as basis, you lose more when your storage ist empty and less if your storage is full and about the same if your storage is half full (marathon vs. normal).

Next assumption: You can get 4 more food by reallocating your workers, reducing the poisonous effect to 8-4=4 on normal and 12-4=8 on marathon.
Full storage: no losses
half storage: no losses
empty: 4 pop at normal vs. 8 pop on marathon

I'm interested to see how Firaxis will fix this mess. I'll probably stick to 12 for poisonous effect and 9 for unrest for the time being.

 

[Zoolooman]

Todd - There is a lot more work to be done. Two factors to consider!

1. Don't forget that a Marathon empire essentially acquires cash and (maybe?) espionage points at 3 times the normal rate. Unless espionage missions cost 3 times as much in Marathon, they will occur three times as often.

2. As pop points die and turns pass, unhealthiness decreases. The city will rapidly reach a health/disease equilibrium.

Consider this:

Size 7 Normal city, half storage capacity (assuming granary), healthiness is equal to unhealthiness (7/7). It has 17 food to burn. The player is working five 2 food tiles and two 1 food tiles, and the city is producing 2 food. As you said, the normal player loses most of his storage in 2 turns, leaving him with a measly 2 spare food. A pop point dies next turn, taking out a worker on a 1 food tile. He ends up with -11 unhealthiness (-6 from the pop, -5 from 3 turns of poison) over his 7 healthiness. Only 2 more pop points will die before his unhealthiness and healthiness reach equilibrium with 3 pop points killed after five turns of poison.

Size 7 Marathon city, half storage capacity (assuming granary), healthiness is equal to unhealthiness (7/7). It has 51 food to burn. Let's first play out the so-called standard scenario, where the city receives -8 for 3 turns before dropping down to the next level of poisoning, and so on and so forth for 24 turns. The city reaches exactly zero points of food after seven turns--one turn into the -6 poison turns. At this point, the city's unhealthiness is -13 compared to a healthiness of seven. It loses one pop point per turn for two more turns, reaching -11 healthiness. But it's also dropped to -5 poisoning, for a total of -10 unhealthiness. If you follow this pattern to its end, 3 more pop points will die before equilibrium is reached. Overall, the city loses 5 out of 7 pop points. Ouch.

Also, I calculated it out. It doesn't matter in which order you apply the turns of poisoning--as long as Marathon delivers 3 turns of poisoning, the same number of population points will die. In fact, one Marathon poisoning would be equivalent to three sequential Normal poisonings, and it'd occur at an effective 3 times the normal speed for potentially 1/3rd the cost. Triple ouch.

 

[Todd Hawks]

Well, I only wanted a rough estimate in order to figure out what value to use in the xml.

Your points are interesting, though. I really didn't think about the pop <-> unhealthiness issue.

Not sure about there being thrice as many espionage missions in marathon, though. Did you test it? I'd assume that all costs are tripled, considering the points you gain per turn stay the same. Everything else is more expensive as well, after all. But of course it's possible that they forgot about that as well when they implemented it.

To be honest, I'm starting to extremely dislike those two spy missions, no matter the game speed. It smells like lot of cheese. Want to have the greatest effect? Just conduct the mission right after the city grew. First of all I'm pretty sure the AI doesn't factor that in and secondly it's completely unjustified. Why should my town/city suffer way more from poisoning or unrest just because it just grew? It's illogical. What is needed is some way that hurts the city independently of how much food you currently stored.

Maybe let it work like this:
- all excess food generated from the moment the mission starts is accumulated
- if it reaches a certain amount after x turns, nothing happens (so you can combat the effect)
- if it reaches less, one or two or three pop points are lost, depending on how far away from the needed amount you are.
- scale this with city size

Kind of like an event. Since the excess food of a city is independent of the speed setting, it would work the same way for all speeds.

Considering how screwed up this all is, I'm thinking about deactivating the mission (if possible) or reducing its effect to (close to) zero. But that probably won't go down well with the AI. Maybe increasing the cost extremely would prevent the AI from using it?

 

[Zoolooman]

Considering how screwed up this all is, I'm thinking about deactivating the mission (if possible) or reducing its effect to (close to) zero. But that probably won't go down well with the AI. Maybe increasing the cost extremely would prevent the AI from using it?

I plain dislike both of the missions. If I had a choice, poisoning the water would kill one population point, and the mission couldn't be repeated in that city for x number of turns. Fomenting unhappiness would add some reasonable amount of unhappiness (like +3) for x turns, and it also couldn't be repeated for x number of turns. These restrictions on repetition would prevent someone from simply locking down an enemy city in an absurd string of spy missions. This would also encourage players to choose the best time to activate the mission.

Not only would these effects be vastly easier to balance, but they would also scale across the speeds evenly. Furthermore, they would be easier to understand at a glance. The player wouldn't have to work through long calculations to determine the effect of the mission that was used against him.

Hell, imagine how easy they'd be to notice at a glance!

"Your city's water supply was poisoned! A population point is lost in ______! Security will be stepped up for the next x turns." And. "Some of your citizens are in a plot to disturb the peace! +3 unhappiness for the next _____ turns in _______! Security will be stepped up for the next x turns."

Edit: A better foment unhappiness mission might simply assign one population point as an "angry citizen" for the next x turns.

 

[Todd Hawks]

I concur.

I went ahead and removed those two missions (simply delete the entries in the xml) and I like it.

Not only due to the reasons discussed so far but also because to me a spy is mainly responsibly for gathering intelligence and maybe small sabotage missions. It's not responsible for bringing down whole cities. Of course it might theoretically happen that (in real life) a spy plants a nuke or something but that's a one time event and nothing that should happen constantly.

So by removing these missions, I can concentrate on using espionage as intelligence gathering mechanism with sabotage acts when necessary and advantageous. And waging war stays a matter of using my army instead of nuking a city with poisoning spies. Not to mention the decreased micromanagement.

I also slightly increased the cost of the other missions to make up for the two missing ones.
Edit: No, I didn't. I wanted to, but since the costs are scaled to different values, I decided to not screw up the balancing of the other missions.

A different approach might be to make the missions expensive enough that you can only conduct them very rarely and that they will lead to a huge hit to your espioinage points (and therefore your ability to conduct further missions). Since that's hard to balance, though, and I don't know how the AI will react, I won't do that right now. Maybe after many games and after checking out what Firaxis does in order to fix it.

 

[PieceOfMind]

If you try to posion a cities water supply on normal speed, It gives 8 for 8 turns. Marathon incorrectly triples both the duration, and the effect, hence:- 24 for 24 turns. This effectively destroys a city in the middle game more completely than if it was nuked. I presume this was overlooked, and it should be 8 for 24 turns. Triple the duration, same strength. The same thing applies to unhappiness.It again produces 24 for 24 turns instead of 8 for 24 turns.

May I just say...

This particular bug needs to be addressed rapidly. Fortunately for myself I play on normal speed but I understand this bug is absolutely ruining the game for some players. I will be quite disappointed if Firaxis leave this bug for a couple of months - the next patch with this fix should be released ASAP even if it means it will be a relatively small patch.

 

[DrewBledsoe]

May I just say...

This particular bug needs to be addressed rapidly. Fortunately for myself I play on normal speed but I understand this bug is absolutely ruining the game for some players. I will be quite disappointed if Firaxis leave this bug for a couple of months - the next patch with this fix should be released ASAP even if it means it will be a relatively small patch.

Concur with just about everyone in this thread

Its very easy to do your own xml "bandage" fix, but I messed around so much with warlords, I couldn't stop, and have vowed to leave everything alone with BTS, so that would break rule one

Nice analysis Roland (as normal) its not so simple when you actually do the math behind it. The point is also, that if you're city is just about to grow, and it gets poison nuked, the effect is MUCH, less than a city that's near to starving anyway. As you said, it only takes one turn to lose a pop point, but 3x the food to grow of normal. It's complicated anyway.

Let's hope for a quick fix.

 

[Roland Johansen]

Nice analysis Roland (as normal) its not so simple when you actually do the math behind it. The point is also, that if you're city is just about to grow, and it gets poison nuked, the effect is MUCH, less than a city that's near to starving anyway. As you said, it only takes one turn to lose a pop point, but 3x the food to grow of normal. It's complicated anyway.

Let's hope for a quick fix.

Thanks. I also like the discussion by Todd Hawks and Zoolooman. These missions need some changes, that much is clear. It's probably pretty difficult to balance them across difficulty levels so you will basically have to redesign them for every difficulty level. The effects are pretty big, even at normal level (which is probably the standard version of the mission). But I can't comment on that as I haven't played the game enough at the moment and don't know the costs of the missions. Personally, I also don't like that a single mission can seriously decrease the population of a city. 2-3 population points lost seems ok, but more seems very much for just a spying mission. Nukes do less on cities with bunkers.
And of course, the effect of the amount of food in the storage on the strength of the mission is rather weird.

Maybe we should discuss some alternatives here. It could very well be that Firaxis takes a look at this thread as it is in the bug report forum which is designed to be helpful for Firaxis.

2 very simple missions:

Poisson water supply: -2 population. Cannot be performed again for 10 turns (scaled for gamespeed).

Ferment unhappiness: 3 population points don't work (unhappy) for 10 turns (scaled for gamespeed), but also don't cost food maintenance. (edit after DrewBledsoe's post) This unhappiness effect cannot be reduced in any way until the mission period has finished. (end edit) Cannot be performed again while the mission (effect) is active.

Of course, both mission need an adjustment of costs because these missions are seriously weaker than the original ones. Please criticize these missions as I haven't thought long about the ramifications. The missions are very similar to the suggestions by Zoolooman. They could be used to slow down wonder production or space ship part construction. Or maybe, they could be used to harm important cities from the enemy. But they won't be able to really 'destroy' cities.

 

[ChrTh]

@Todd Hawks:

What was the name of the XML file you modified to remove the missions? Also, if a save is passed to another player (say in a Succession Game), do they need to have removed the missions from the XML file, or is that a change that remains within the save?

 

[DrewBledsoe]

Thanks. I also like the discussion by Todd Hawks and Zoolooman. These missions need some changes, that much is clear. It's probably pretty difficult to balance them across difficulty levels so you will basically have to redesign them for every difficulty level. The effects are pretty big, even at normal level (which is probably the standard version of the mission). But I can't comment on that as I haven't played the game enough at the moment and don't know the costs of the missions. Personally, I also don't like that a single mission can seriously decrease the population of a city. 2-3 population points lost seems ok, but more seems very much for just a spying mission. Nukes do less on cities with bunkers.
And of course, the effect of the amount of food in the storage on the strength of the mission is rather weird.

Maybe we should discuss some alternatives here. It could very well be that Firaxis takes a look at this thread as it is in the bug report forum which is designed to be helpful for Firaxis.

2 very simple missions:

Poisson water supply: -2 population. Cannot be performed again for 10 turns (scaled for gamespeed).

Ferment unhappiness: 3 population points don't work (unhappy) for 10 turns (scaled for gamespeed), but also don't cost food maintenance. Cannot be performed again while the mission (effect) is active.

Of course, both mission need an adjustment of costs because these missions are seriously weaker than the original ones. Please criticize these missions as I haven't thought long about the ramifications. The missions are very similar to the suggestions by Zoolooman. They could be used to slow down wonder production or space ship part construction. Or maybe, they could be used to harm important cities from the enemy. But they won't be able to really 'destroy' cities.

Not sure about the poison water mission, maybe something like that but not quite exactly, can't really put my finger on it at the moment.

The second I like, and would also refine it, so that its flat 3 pop WON'T work whatever. Even if there are 40 troops in the city under Hered Rule. (part of the reason the original 24 although wrong, isn't quite as nasty as the poisoning,,if you have 16 troops, and say therefore a hapiness cap of around 30 for a pop 12 city, then the unhappiness isn't so nasty as the unhealthiness)