Tweak to spying needed?

[ArticFire]

ORIGINAL: WanderingHead

Just to let everyone know where this stands now. I've changed Japanese starting security to 15. I could be convinced it should be smaller, but this seems like a number most people were happy to see. I don't expect to make any other changes.

I still don't like this as the only side it hurts is Axis. If the Axis player wanted to increase Japans counter intel they could. Allies will still have the ability to overcome and rip off Japan left and right just like now the only difference is you hurt an option for Germany.

I don't see why a change is needed that only hurts one side. If you don't like Germany stealing from japan cut off the ability for any ally to steal from one another.

imo this does nothing to help the intel gameyness only makes life easier on the Allied side.

[WanderingHead]

I'm perfectly happy to do whatever most people want. Honestly, I must admit that I don't have a strong opinion on the subject myself. I'm happy with 8 or 15.

Personally, when I play I load up the Japanese security and I don't bother trying to have my Germans steal Japanese tech. So 15 would greatly assist me when I play Axis.

[GKar]

I wouldn't mind if it were impossible to steal techs from an ally.
 
One other thing: How about limiting the number of spies you can buy each turn? As of now you can hire and install two dozen spies in a glimpse, given enough supplies. If it was harder to accumulate them, the whole tech stealing business (which is often abused in my opinion) would be slowed down a bit.

[Marshall Art]

I'd challenge the point that only the Allies profit. Yes, Germany would be kept from spying on Japan. But that only directs Germany's spies towards their true enemies: WA and/or Russia. This is both more historical and makes more sense since you rather spy on your opponents and not your "friends". Russia is a special case since until 1941 it could not be exactly called an Ally to Britain. In fact it was closer to Germany than to any Allied power until June 22, 1941. Not to mention the cold war after 1945.

As Japan and Germany (with more SIGINT and/or more spies) can challenge Allied spies in the early years the Allies really need to invest into SIGINT and/or spies. As it is now they get Japanese production, research data almost for free. That is a great advantage since they know what Japan most likely plans to do. In the early years the WA and Russia have little production compared to later and thus would be at an disadvantage rather than profit from the change.

One should keep in mind that these changes are of minute impact since at best you get a couple of tech points worth from a spying action over the course of the game. You still need to put in your own resources if you need the tech level FAST (which is always the case[:D]).

I agree that "creating" 100s of spies in a single turn is not realistic, the number of spies could be reduced to a level that would force the players to continuosly invest in spies much like they continuosly invest into production, supply etc.

How about allowing only one spy per turn and per target nation in 1939, 2 in 1940, 3 in 1941 etc?
That would disallow moves like "let's purchase 20 spies against Russia now" but still would give the WA plenty of new spies after 1941-42, close to when they finally geared up their intel services? Security and SIGINT could also be levelled the same way.

If the spying from Allied powers were to be reduced, I'd suggest to change the spies to the following:

GERMANY
Security - 10; SIGINT - 6; vs Japan - 0; vs Russia - 4; vs China - 0; vs W. Allies - 6;
JAPAN
Security - 15 (increased from 8); SIGINT - 4; vs Germany - 0; vs Russia - 3; vs China - 6; vs W. Allies - 6;
RUSSIA
Security - 15; SIGINT - 5; vs Germany - 9; vs Japan - 9; vs China - 9 (useless anyway); vs W. Allies - 3 (changed from 9);
W. Allies
Security - 8; SIGINT - 10; vs Germany - 3; vs Japan - 2; vs Russia - 0 (changed from 3); vs China - 2 (useless anyway);

[ArticFire]

ORIGINAL: Marshall Art

I'd challenge the point that only the Allies profit. Yes, Germany would be kept from spying on Japan. But that only directs Germany's spies towards their true enemies: WA and/or Russia. This is both more historical and makes more sense since you rather spy on your opponents and not your "friends". Russia is a special case since until 1941 it could not be exactly called an Ally to Britain. In fact it was closer to Germany than to any Allied power until June 22, 1941. Not to mention the cold war after 1945.

I don't see how you see forcing Germany to do things a certain way a benefit. It may scratch some 'realism' itch but it has 0 bearing on the practical effects of such a change. The whole spying system has 0 realism as is so you cannot apply realism arguments to a system fundamentally in opposite to it. As I have mentioned one of the primary problems with this sequel is how it flat forces the Axis to do things in a certain way even moreso then the first incarnation of the game. This change just yet again ties the Axis players hands , meanwhile of course SU and WA continue to steal from each other with impunity.

I still say no stealing from an ally at all if your going to make a change. But if you insist on jacking up Japans counter intel , I believe that none of the allied countries should get to start with spy's already on their ally's just like the axis side. I also believe that Russia's starting spy counts against Germ/Jap should be drasticly lowered. SU in WW2 was NOT the secret police , cold war Russia of the 1980's. If the Allied player chooses to play SU in that way they should have to pay for the privlige just like everyone else.

[WanderingHead]

ORIGINAL: ArticFire
SU in WW2 was NOT the secret police , cold war Russia of the 1980's.

Currently reading the Gulag Archipelago ... hard to argue there weren't secret police! True, they were mostly pretty inward focused. However, the existence of active Communist cells around the world does seem to make some Russian leg-up reasonable.

The only place that I intentionally steal from Allies is actually China stealing from WA, and it seems perfectly reasonable there. Really, the idea came from the question "What the hell can China do with all these supplies?" when the WA are able to consistently deliver them and China doesn't need them for factory repairs or maneuvering.

But back to the issue at hand ... you have an excellent point. If Japan needs an extra 7 points in security, why not just give them 35 extra supply and let the Japanese player decide what to do with it? In general, more choices make the game better. The only reasons to give the Japanese more security instead of supplies are if the issue is the mechanic and not game balance, i.e. in this case, if we are more worried about Germany stealing from Japan than the Allies stealing from Japan.

At this point, I believe that the best thing would be an adjustment of the mechanic, although this is unlikely to occur. As it is, the only reason I ever bother to spy is to steal technology. It would be nice if actual useful INTELLIGENCE could have equal roughly weight as a benefit of spying. Somehow.

All that is well and good, but now I am even less sure what to actually do. I was happy to make the Japanese security change when it was unanimous. Now I'm inclined to leave security at 8 or increase it only to 12 or so.

[ArticFire]

ORIGINAL: WanderingHead

ORIGINAL: ArticFire
SU in WW2 was NOT the secret police , cold war Russia of the 1980's.

Currently reading the Gulag Archipelago ... hard to argue there weren't secret police! True, they were mostly pretty inward focused. However, the existence of active Communist cells around the world does seem to make some Russian leg-up reasonable.

The only place that I intentionally steal from Allies is actually China stealing from WA, and it seems perfectly reasonable there. Really, the idea came from the question "What the hell can China do with all these supplies?" when the WA are able to consistently deliver them and China doesn't need them for factory repairs or maneuvering.

But back to the issue at hand ... you have an excellent point. If Japan needs an extra 7 points in security, why not just give them 35 extra supply and let the Japanese player decide what to do with it? In general, more choices make the game better. The only reasons to give the Japanese more security instead of supplies are if the issue is the mechanic and not game balance, i.e. in this case, if we are more worried about Germany stealing from Japan than the Allies stealing from Japan.

At this point, I believe that the best thing would be an adjustment of the mechanic, although this is unlikely to occur. As it is, the only reason I ever bother to spy is to steal technology. It would be nice if actual useful INTELLIGENCE could have equal roughly weight as a benefit of spying. Somehow.

All that is well and good, but now I am even less sure what to actually do. I was happy to make the Japanese security change when it was unanimous. Now I'm inclined to leave security at 8 or increase it only to 12 or so.

I still don't see why everyone is so worried about Germany stealing from Japan , there are some pretty good examples of Germany / Japan attempting to share technology historically.

It would be nice if you could tie the spy system into the tactical side of the game. For instance force a country to have a supply link to a territory of the country you wish to steal from that is in turn linked to that countries capital. In this way for instance if Germany / Japan want to share tech Germany has to make a concerted effort to open up the middle east. Same for SU/WA , they would need to maintain the India or northern ocean connections to be able to share tech.

In that way you could also up the effectivness of stealing tech since there would be a tactical component and it would not always be available. SU/GE would always be able to steal from each other. WA from GE as long as WA maintains a transport connection to mainland Europe. GE from WA as long as they gain control of the med and can link to WA territories in India or Africa. That would also have the effect of making the Sub vs WA Transport far more interesting. It would actually be worth it for GE to risk the WR increase to drop WA transports in hopes of breaking the connection and force WA to interact with GE more directly to maintain the direct link threw the Med or coast of Europe.

Such a connection should make spying / intel much more effective then it is right now so that cutting it a worthwhile tactical goal.

I expect that this would take large code changes though which are just not possible. But I think it's something worthwhile to think about for the next incarnation of the game.

[GKar]

ORIGINAL: ArticFire

For instance force a country to have a supply link to a territory of the country you wish to steal from that is in turn linked to that countries capital.

Not a bad idea. I guess it'd benefit the Allies again though, as there's probably little point in still sharing/stealing techs for Germany and Japan when they have a land or transport connection.

I still think that Japanese security should be increased, as this will make the occasional theft by the WAllies or Russia harder. And I'd really like to see a limit of investment in espionage per turn like the one Marshall Art proposed - if it is feasible at all. Being able to buy 20 spies at once if you have the supplies takes away some of the strategic thinking.

[WanderingHead]

One idea is to make it it easier to catch spies when there are more of them.

From: tm.asp?m=1302037

Code: Select all

1. If Random (Security) > Random (50) then opposing player loses a spy. 
 2. If Random (Security) > Random (Defending SIGINT *8) then opposing player loses 1 SIGINT. 
 3. If Random (Attacking Spies) > Random (Defending Security *3) then research progress is exposed. 
 4. If Random (Attacking Spies) > Random (Defending Security *3) then production units are exposed. 
 5. If Random (Attacking Spies) > Random (Defending Security *3) then SIGINT screen is exposed. 
 6. If Random (Attacking Spies) > Random (Defending Security *4) then that region is exposed (check for all regions). 
 7. If Random (Attacking Spies) > Random (Defending Security *3) then lower WR range bracket increased by 1 (unless minimum of range already at real value). 
 8. If Random (Attacking Spies) > Random (Defending Security *3) then higher WR range bracket decreased by 1 (unless maximum of range already at real value). 
 9. If Random (Attacking Spies *2) > Random (Defending Security) then technology can be pillaged. 
 10. If Random (Attacking Spies) > Random (Defending Security + 5) then gain 1 research point in technology you are behind in. 
 11. If Random (SIGINT) > Random (Defending Security) then gain +1 to attack for that unit (repeated for all units in battle. 

One thing I notice here is that there is no attrition of security. They never disappear. There should be a running cost to this.

Also, I believe that tech can be stolen even if the tech levels are the same but the target player has just RESEARCH which is ahead. Perhaps a full level difference should be a requirement, or the probability should reduce for having nearly the same level (the enhancements to a weapon are more obvious when there is a big difference observed).

But anyway, I was thinking of something like this, say

Code: Select all

1. If Random (Security)+ Random(Attacking Spies) > Random (50) then opposing player loses a spy. 
 12. If Random(Security) > Random(50) then lose one security.

I'm reluctant to limit spy builds because of UI changes (how to indicate, etc).

Also, I don't know that I even want to make the above changes. I should be conservative on changing anything, and while something like the above seems preferable, it doesn't seem like a no brainer to me.

[MrQuiet]

To me the issue is that the WA can easily afford to spend the supplies to steal a lot of Japans teck and keep easy tabs on there production and research.
When I say a lot I mean one point of reasearch in at least 3-5 fields every single turn.
Thats about 80-100 points over the course of the game.
I do this every game.
Forwarn45 is currently doing this to me in a tourney game.
It just doesnt feel right to me that the WA can steal so efficiently against Japan.
Japan can not afford to stop this if they want to build a effective fighting force.
I think starting Japan at 15 security is the right move.
 
Wether or not allied nations get some teck research from each other really does not bother me either way.
 
-MrQ
 
 

[WanderingHead]

another possible tweak is to change the criteria for which tech fields can be stolen. Right now, you have to have at least one point invested towards that field for it to be eligible. This is fairly meaningless, I routinely invest exactly one point just to enable theft.

Instead, the requirement could be that you have invested in that tech THIS TURN, i.e. you are actively researching it.

[Forwarn45]

ORIGINAL: WanderingHead

another possible tweak is to change the criteria for which tech fields can be stolen. Right now, you have to have at least one point invested towards that field for it to be eligible. This is fairly meaningless, I routinely invest exactly one point just to enable theft.

Instead, the requirement could be that you have invested in that tech THIS TURN, i.e. you are actively researching it.

I don't favor this change. Increasing Japanese security to about 15 is a sufficient adjustment. I think the spy system otherwise works pretty well as is.

[WanderingHead]

OK, bearing in mind that I am in a kind of brainstorming mode, just throwing out ideas.

One issue I see is that if you are behind in one tech, you have some probability of stealing it. If you are behind in N techs, you have the same probability for EACH tech, so now you will actually steal N times as many tech points (on average) just because you are further behind in so many. For the same investment.

Perhaps the incremental probability of stealing techs should be reduced the more techs you are actually behind in. For example, change item #10 (#9 shown just for reference)

Code: Select all

9. If Random (Attacking Spies *2) > Random (Defending Security) then technology can be pillaged. 
 10. If Random (Attacking Spies) > Random ((Defending Security) * (1 + (number of eligible techs behind)/2)) then 
 gain 1 research point in technology you are behind in.

If we set N=(number of eligible techs behind) and look at the probability of stealing a tech, for attack<defend (this restraint just for simplicity of analysis, and with this formula it is valid for large N anyway) we have

Code: Select all

probability(#10 satisfied) = att/def/(2+N)
    = att/def/3  (N=1)
    = att/def/4  (N=2)
    = att/def/5  (N=3)
    = att/def/6  (N=4)
   ~= att/def/N  (N large)

Which means that the average number of stolen techs passing #10 (assuming #9 already passed for the time being), given that the above is applied to all N techs, is

Code: Select all

ave #techs passing #10 = N * probability(#10 satisfied) 
    = 1/3 * att/def  (N=1)
    = 2/4 * att/def  (N=2)
    = 3/5 * att/def  (N=3)
    = 4/6 * att/def  (N=4)
    = 5/7 * att/def  (N=5)
   ~= att/def  (N large)

So you always get a little more when you are behind in more techs, but not linearly. In other words, if you are behind in 1 tech you can have a reasonable chance to steal it, but if you were behind in 10 techs you'd only have a reasonable chance of stealing about 3 or 4 instead of stealing all 10.

I know the math may look too much for the non-mathematically inclined, but I think it just plain makes sense, frankly. You shouldn't be able to steal so much more for the same cost just because you are behind by more.

This would also discourage throwing a point of research here and there just to be able to steal the techs, because you'd now rather focus on the techs you care more about. So you'd only throw a point into something that you were really prioritizing.

[ArticFire]

What about the idea of making it so you can only steal 1 tech per turn per country yous teal from. That way instead of someone ripping off every single field you are ahead of them at all in almost every turn they get randomly just 1 notch in one field. In this way spying may boost something your working on but you cannot rely on it to do half your reasearch for you while you happily research other things.

This would also encourage people to spread their spying around so that they can benifit from a few advances per turn rather then just 1 or 2 from the main antagonists. This would also provide a law of diminishing returns. Rather then "buy 30 spys and steal everything they are doing" it would make it so going way past what your opponent is prepared to defend against is largely a waste. Thin turn would give more focus to security / signit as spys would be less bought by bakers dozens.

This would also have the added benefit of needing no other modifications to the system. Granted I'm biased to my own ideas but I think it provides a good solution to the abuse of spying without needing much in the way of code change and no need for any UI modifications.

[Marshall Art]

Not sure about limiting just 1 steal per turn, seems to go to the other extreme to me (too little bang for the buck). The formulae above which limit the actual steals to 3-4 out of 10 - do they treat each tech identically (i.e. same probability) or are the first that are calculated by the game engine preffered? There should be an equal chance of each tech being stolen, I am not sure just from looking at the formulae.

[Marshall Art]

If you cannot (or are unwilling) to limit expenditures per turn for spies, maybe "rolling formulae" that reduce spying efficiency each turn would be an easier solution which would force players to keep investing into spies? E.g. what 10 spies can achieve against 10 security in 1940, it would take 15 in 1941, 20 in 1942, just as an idea? This would go along the lines that the spying business became more and more complex and the obvious and easy steals became more difficult over time as the counter-intel got more sophisticated as well?

Just to give you some ideas that spying and secret police were a part of the Soviet system from very early on. Lately I watched a program about Soviet spies on TV (German TV that is). The Soviets "hired" their spies starting in the 1930s all over the western world. Most of them were sympathetic to the communist ideas and young poeple like students. Over the decades many of them rose to positions in various fields of professions, just a few examples I can still remember:
One, being a physics scientist, became the spy who stole the plans of the atomic bomb from the US around 1945, which game the bomb to the Soviets many years earlier (in 1949) than if they had not received this information.
Another made his carreer in the British counter-intelligence and became their chief of counter-intel during the 1950's. You can imagine what damage that brought to Britain's secret services.
A third was sent as a reporter all over the world to "report" from the countires that were believed to be potential enemies of the Soviet Union, among them was Japan in the late 1930's. He reported lots of the Japanese intentions to Moscow but Stalin (as he mostly did) did not believe him! Of course as he was caught in Japan the Soviets denied everthing so he got hanged in 1944(?). The first two I mentioned spent their elder years as heroes in the Soviet Union.

[Marshall Art]

If it is more of an agreeable solution, how about increasing Japan's Sec. to 12 + give them 15 supplies more. That way the stealing business in Tokio is somewhat limited and Japan's player can decide if he wants to further increase sec. or buy some RPs or units from the 3 points worth of supplies he does not have to produce actively.

[WanderingHead]

ORIGINAL: Marshall Art
The formulae above which limit the actual steals to 3-4 out of 10 - do they treat each tech identically (i.e. same probability) or are the first that are calculated by the game engine preffered? There should be an equal chance of each tech being stolen, I am not sure just from looking at the formulae.

The forumula I proposed would be applied equally to all techs eligible for theft. It wouldn't actually limit the number of possible thefts, it would just make it less probable that you steal a huge number while still allowing you a reasonable probability to steal a few. At the same time, buying more spies would still allow you to steal more, because it is not a hard limit, just a probability adjustment, and the probability still gets better with more spies.

I was trying to solve the problem "How do you make it more difficult to steal 5 techs if I'm behind in 5 techs, without making it inordinately difficult to steal 1 tech if I'm behind in only 1 tech?" This technique solves that problem.

Limiting it to one tech per turn is actually slightly harder to accomplish (we have to bear in mind that the SW can't just use the first tech to pass, that would bias towards certaint techs always being stolen and other techs never being stolen), and I think that being a hard limit it is less desirable (with high numbers of attacking spies you'd always steal exactly one tech, removing the uncertainty entirely).

ORIGINAL: Marshall Art
If you cannot (or are unwilling) to limit expenditures per turn for spies, maybe "rolling formulae" that reduce spying efficiency each turn would be an easier solution which would force players to keep investing into spies?

This is more easily and straightforwardly attained by removing attacking spies, increasing their attrition rate. I think the most sensible way to do this is to make it more likely that you lose spies the more of them you have. You could even say that you round(die(number_attacking_spies)/10) (which on average is 5%) spies every turn, regardless of security, just to represent the ungoing cost of spying operations.



Let me suggest 3 concepts, generically (details for my specific proposals are above)

1) the probability of theft reduces inversely to the number of techs you can steal, so it can be hard to steal all 10 if behind in 10 without being too hard to steal 1 if you are behind in 1.

2) attrit the number of spies in a way that you tend to lose more if you have more (two different ways suggested in the thread, modified security roll or straight proportional attrition independent of security).

3) you only benefit from spies this turn that you had at the beginning of the turn. Spies acquired this turn don't help until next turn.

[Forwarn45]

I appreciate the effort you are making to consider tweaks to the game, but I just don't think this is an area that needs much change other than increasing Japanese security a bit.

I think that the more areas the enemy is ahead, the more opportunity for advancement by stealing secrets and/or reverse-engineering captured equipment from the battlefield. I don't know if I like the idea of "focused" spying. Although I might be persuaded it was a good idea to consider some type of change if the game was still in development, it just seems that the spy system has worked fairly well through a lot of play! The way the game plays, spying can be a long-term investment with somewhat mixed rewards as you can miss some turns due to the security roll, and then returns diminish as you begin to catch up to the enemy.

Other than modifying Japanese security, the only type of suggested change I think I could get behind (which wouldn't affect balance, just the dynamics of investig in spies) - would be some type of limit on the per-turn investment in spies, security, sigint, etc.

Just my 2 cents.

[ArticFire]

Ya know , its a shame it woudl require UI tweaks to make it so that the more spys you have the more they start to cost to make more. There have been many limiting steps to prevent wonder weapons , spies really should be held to the same standard.