Investigation Arithmetic

[Yudishtira]

The investigation check is calculated from:
1) Full amount of the LE unit with the highest current Investigation value
2) +10% of the cumulative investigation amounts of all remaining units with investigation present
3) +Investigation amount from buildings in the city (including player modifiers)
4) - The city's amount of Insidiousness (including player modifiers)
5) - The criminal who's being investigated's natural Insidiousness

The resulting Investigation amount is either multiplied by .1 or .01 (can't recall offhand) and this becomes the % chance of a successful investigation check. One is made every round against each local criminal, even your own, and the primary LE unit involved in the check gets XP for doing it, as does the criminal for attempting to avoid being investigated.

1) is about 22
2) is about 2 x 6 = 12
3) no idea: question: do LEs still get this if it's not their city?
4, 5) from alt-mouseover of the criminal, I get 130. The crime autobuilds are certainly included in that. Is there an amount of city insidiousness that is not included here, and if so, where do I find it?

Unless the remaining amount from 3, 4 & 5 gives me a big surprise when I calculate it, the result is currently a little below -100. Even if the crime level could be returned to zero, investigations will still have a -ve chance of success. When they ever push their nose above zero, they will be reduced by a further 90 or 99%.

All the criminals, but only one LE, are getting XP from these impossible investigations, so the LEs need to get out of there to avoid going backwards. But of course then they will not be able to reduce the crime property.

Maybe this belongs in the bugs thread after all...

 

[Thunderbrd]

no idea: question: do LEs still get this if it's not their city?

I think it has to be same team but I could be wrong on that.

City insidiousness amounts are shown in the defense help hover. On the old interface you'd hover over the rook symbol... maybe that's still the same on PPIO? Not sure.

All the criminals, but only one LE, are getting XP from these impossible investigations, so the LEs need to get out of there to avoid going backwards. But of course then they will not be able to reduce the crime property.

But that one LE is getting XP for each check. The most important thing is actually to train enough to overwhelm the crime curve and bring the crime down and in the process you're also supporting investigation efforts. Investigation chances may not ever seem like they can get all that high - they can't usually grow all that high over the insidiousness, but if you're on top of the crime, they are inevitable unless you're really out-teched by the criminals. The next trick is to have a unit strong enough to make those arrests without dying. LEs have an edge up on the balance where combat is concerned, usually, but it's not impossible for some criminals to give them a real run for their money, particularly dangerous to differing degrees by era. Some Criminals are stronger than the normal curve and they are very dangerous and require very well trained LE units for taking down criminals.

This is all part of what's intended. There's supposed to be a fighting chance for both sides and not completely decided and assumable. Use of criminals as an act of hostility should be a good strategy in general and if it were any easier to arrest them, it wouldn't be. Use of criminals should not be an overwhelmingly powerful strategy and if it were any harder to control crime and defend against them, it would be.

 

[Yudishtira]

There are two insidiousness promo lines - Conniving and Corruption. Do they both have LE counterparts?

I think this AI city might be too far gone for me to even be able to experiment with. 7-8 criminals, all with Corruption 9 or 10 (Corruption gives less insid than Conniving does, but it also gives bulk additions to the crime property). Crime property at 3000+ (albeit coming down currently at around 100/turn). My LEs can only get 21XP at birth, enough for investigation values of 22ish. I've got 10ish involved in the experiment presently, and none more on the way.

The criminals are mostly many eras behind the LEs, but they've been around/encouraged longer, with far more promos, so that the tech advantage is cancelled out many times over (in this context at least).

You make a good point that one LE is going to gain XP 7-8 times as fast as the crims do. I shall have to see if he ever catches up, and whether that tips the balance.

And I understand the need to avoid investigation %ages getting too high. I still think that this "straight line" approach means that the %age is negative far too much of the time. I suggest the %age should be allowed to get positive "sooner", but then being reduced by a modifier that gets higher as the percentage does. Eg. 0.1% is reduced by 1% (to 0.099%), 1% is reduced by 10% (to 0.9%), 5% is reduced by 50% (to 2.5%), 10% is reduced by 70% (3%), 20% reduced by 80% (to 4%), 30%+ reduced by 83.3% (to 5+)

Something like that anyway...

I will continue the experiment. It is perhaps too extreme an example to base suggestions on.

 

[Thunderbrd]

There are two insidiousness promo lines - Conniving and Corruption. Do they both have LE counterparts?

Yes. Conniving is primarily for insidiousness while Corruption is primarily for Crime enhancement. Policing is the opposite of Corruption for LE units and Sleuth is the opposite of Conniving for LE units.

My LEs can only get 21XP at birth, enough for investigation values of 22ish.

The criminals are getting a lot of XP from the crime autobuilds too. However, the intended counterbalance to this is the +1 xp you can give to LE units by specialist assignments, which in a city at this stage of the game would be a huge amount. If you don't have that option on, then simply training even more LE units is always possible. It should always be easier to overwhelm the amount of criminals than it actually is to overwhelm an enemy city with your crime units, particularly since the autospawned criminals get harder and harder to spawn with each criminal present in the city (there's an ultimate limit of locally spawned criminals due to that factor) yet there's really no limit to how many LE units can be trained (other than gold to support them.)

Sure, criminals can sink a nation by overwhelming the economy - so can military units and spies by their own more direct and indirect means.

I still think that this "straight line" approach means that the %age is negative far too much of the time.

It just means you have to first focus on addressing the crime levels first and foremost. THEN the criminals will be something you can tackle. Once you have, you're left with a veritable army of LE units that hopefully you can use effectively before they need to be called to another huge problem city. It is much easier to maintain good control on crime than it is to gain control once it's been lost. A player cannot let it get too far out of hand without recognition of the compounding consequences. There are still a few weaknesses in the AI that allows this to happen at times and we still need to figure out exactly where those issues are located.

I will continue the experiment. It is perhaps too extreme an example to base suggestions on.

It might be. It would certainly be a good test case save for us to evaluate how the AI is failing to address the problem for themselves. Sure, crime can spiral out of control and be very tough to manage, but I'm curious what the key flaw for them is in this case.

 

[Yudishtira]

It might be. It would certainly be a good test case save for us to evaluate how the AI is failing to address the problem for themselves. Sure, crime can spiral out of control and be very tough to manage, but I'm curious what the key flaw for them is in this case.

You have my save. The city is the Maya capital, Yax Mutal, in the south-east of their original territory ie. on the 'middle' continent next to my main colony.

 

[Thunderbrd]

You have my save. The city is the Maya capital, Yax Mutal, in the south-east of their original territory ie. on the 'middle' continent next to my main colony.

I have hundreds of bug saves. You'll have to at least give me the name of it.

 

[Yudishtira]

The latest I've uploaded is here:
https://forums.civfanatics.com/thre...0th-of-july-2019.647928/page-55#post-15539315

 

[Thunderbrd]

You have my save. The city is the Maya capital, Yax Mutal, in the south-east of their original territory ie. on the 'middle' continent next to my main colony.

Thanks. I don't have too much time to look into this but initial observations suggest he's dramatically understaffed throughout his whole nation. Not sure how he ended up like this but it looks like maybe there's some reason that he cannot respond properly. I can't determine that from a cursory look but it seems like it would have to be a combination of a lot of factors. Maybe they aren't building the prereq buildings like they should? I'm not sure. It's a very advanced game and difficult to analyze in depth but we'd have to check on a lot of factors here. It's possible he's just been plagued so hard for so long on so many things that he simply has been unable to keep up.

Interesting game. I'm not sure I've seen one this advanced and varied in a while - it's certainly past what I'm familiar with in terms of much of the design at the moment - but I make no secret of putting my primary focus on the first eras until they are complete.