Is downgrading necessary?

[Bobbins]

Question: Does the CL having hated relations to NG improve the chances of a successful assassination on NG characters?

I say it doesn't. It's the relations the target character has that is important. Others say it does, having NG at hated gives a bonus (this seems to be the generally held view).

What does the site community think?

[VEO]

I've always been under the same impression as yourself.

The Cloud Lord isn't going to be any more stealthy, or "try harder" as an assassin or even an emissary, if he "hates" you...his job is to kill you. I'd hazard that Hated would work against this kind of work as it'd imply a certain amount of emotion is present to get in the way of doing a professional job... Now, if my pop Hated the Cloud Lord, they'd be ever more vigilant and watchful, and less prone to giving ear to the enemy's emissary spin. If Neutral, the Assassin Wants to kill? Disliked he Really Wants to kill? Hated he Really Really REALLY wants to kill? Doesn't work well for me as a bonus to the Individual, but in regards to groups (the populace, the militia, armies) it's fine.

[DaveHolt]

downgrading helps. it helps defend against agent actions. it helps an aggressor nation execute agent actions. I have lots of anecdotal evidence to support this notion.

My belief (and as we haven't looked at the code, all of us have beliefs on this type of topic) is that agent action success/failure is influenced significantly by the "relations differential".

ever wonder why your A30 fails to stlgold from your ally's unfortified camp? Ans: it's because you're both tolerated vs. each other. if your ally was friendly to you and you were tolerated to your ally, your agent would enjoy a positive relations differential and the agent's actions would have greater chance of success vs. that ally.

Freeps need to downgrade LR & CL. very very important. CL & LR need to downgrade their targets. very very important. the one that doesn't do it suffers. I won't bore you with all of the anecdotes (some from ongoing games that are quite amusing). but I strongly believe in the relations-differential factor in reducing or increasing difficulty of agent action.

cheers,
Dave

[Bobbins]

"ever wonder why your A30 fails to stlgold from your ally's unfortified camp? Ans: it's because you're both tolerated vs. each other. if your ally was friendly to you and you were tolerated to your ally, your agent would enjoy a positive relations differential and the agent's actions would have greater chance of success vs. that ally."

Agree with that Dave. However, it's only because your ally has 'positive' relations to you. Not the other way around.

The instructions say when taliking about agent orders (e.g. 615) "Success is based on..... the relations between the nations". P16 of the rule book is much clearer. It says "Relations effect the chance of success for some agent orders... the better the relations a nation has to your nation, the easier it will be for agents to carry out certain missions".

[VEO]

Because Relations affect groups of people. They'll increase the "defense" of the population centre (group) or increase the effectiveness of the army (group). Popular opinion (relations) have no bearing on the Individual's (character) ability to perform that which he/she is attempting to accomplish, which is based on skill and/or modified skill at that particular (individual) task.

That's simply my spin on the rules, but the rules are pretty clear.

[MichaelK]

The rulebook... well the version that I have... was issued many many years ago when many of the real rules were sort-of-distorted. GSI weren't in the business of giving out the code... MoS helped gather alot of evidence and make some great deductions from all the data collected.

We have had rule changes since the rulebook was introduced, we have had rule clarrifications that the written word sort-of-implied something different, and we have had lots of anecdotal evidence that would suggest that the rulebook isn't a comprehensive set of rules but a guide as to how the program sort-of works.

We know there are limits to certain actions... assassinations... pop centre builds... etc that aren't actually written in the book...

There is a lot of strong anecdotal evidence that suggests that UpStNat and DnStNat are very very important... you ignore this at your peril... and not just for armies. Character actions "seem" to succeed much more regularly when you have the right relations.

Maybe what we need again is another MoS type... ready to gather information and distribute answers... wouldn't work on a list like this though... could be a yahoo group but would probably work as an email distribution group better...

Michael

[Drukarzun]

Don't worry about downgrading with a thought to helping your characters succeed; it's just a wasted capital order. Downgrade enemy agent nations. Downgrade to go from a 10% to 25% modifier in battle.

If it helps agent or emi activities, why is it easier to steal/InfOthr to/from allies? Assuming you're both Tolerated (or both Friendly, whatever), then the relations modifiers cancel out, right? Same argument for stealing/InfOthr'ing an enemy then? Both Dislikes cancel each other out? I don't think so.

Drew

[marinersk]

could be a yahoo group but would probably work as an email distribution group better... Uhm...a YahooGroup is an E-Mail distribution group.

The ability to read and write on it through its web interface is an added feature.

It amazes me how many people don't know this. YahooGroups started as little more than another listserv...

- Steve M.

[MichaelK]

I guess you're really trying to make a point here... :)

You are right up to a point, but yahoo group is no longer just an email distribution group... it has more added features that make it similar in a way to this bulletin board.

What I was suggesting (though not saying in as many words as I will use now) was having a setup similar to MoS... a person/group of people collating information and disseminating it through email to a wider audience... with that wider audience participating by sending in relavent information for each issue.

We learnt a hell of a lot of usefull stuff about the game, made lots of inferences about the game mechanics, shared riddle information, dragon and other encounter information, etc etc... OK we sort of do that here, but with the MoS you had tables of information collated and updated... does that happen with this forum? Does someone take riddle answers and update the riddle file... do people send in dragon info and have the encounter tables updated... do we have people ready to have their strategy guides disseminated and discussed... I don't think so but am willing to be corrected...

Michael

[darrell]

but with the MoS you had tables of information collated and updated... does that happen with this forum? Does someone take riddle answers and update the riddle file... do people send in dragon info and have the encounter tables updated...

The encounter/riddle/dragon/etc. information on this site is just something I input on my own from various other online sources. If someone wanted to volunteer to maintain it actively, I could set up permissions to allow that. I just never figured anyone would care enough to do it, and I'm lazy enough as it is.

[Overworked1977]

Don't worry about downgrading with a thought to helping your characters succeed; it's just a wasted capital order. Downgrade enemy agent nations. Downgrade to go from a 10% to 25% modifier in battle.

If it helps agent or emi activities, why is it easier to steal/InfOthr to/from allies? Assuming you're both Tolerated (or both Friendly, whatever), then the relations modifiers cancel out, right? Same argument for stealing/InfOthr'ing an enemy then? Both Dislikes cancel each other out? I don't think so.

Drew

It has been buzzing around in my head for a couple of days, and I think that I might have it...or might not.

Think about it from a practical standpoint.
1. Hate vs. Hate: Emmy: Have you ever tried to win an arguement? It is next to impossible with the other guy shouting back at you. Convincing someone of something requires the other person to actually be listening. Not likely in a Hate vs. Hate environment. Agent: I believe that someone pointed it out earlier that their hate towards you makes them more watchful. Your hate towards them makes your agents more malicious and disdainful.
2. Neutral vs. Neutral: Emmy: Imagine both sides not really caring one way or the other about their way of thought. This will make convincing the other guy easier, but since neither really care it won't be much of a bonus over a Disliked environment. Agent: It's kinda like having a neighbor in a duplex. You don't really expect him to come through the back door when you're not home, but you're not going to leave it open...Just in case. A skilled burgler will still get in, but that's not why they put locks on doors. The locks are there to keep honest people honest.
3. Friendly vs. Friendly: Emmy: Now imagine that the guys that you are trying to convince actually think the same way that you do. If only life were this easy...:rolleyes: Agent: Now, your neighbor is your brother, and his wife babysits for you all the time, so you give them a key to your house. You don't really expect them to steal anything, but there is no accounting for a bad apple in the bunch.
4. Hated vs. Friendly: Emmy: Those poor saps will believe anything that you say because they think that you are only looking out for their best interests. Agent: Can you believe that those idiots that ran over your favorite dog and your daughter's bicycle have asked you to house-sit? Suuuuuuuuuuuure you will....;)
5. Friendly vs. Hated: Emmy: So what if the neighbor is a cocaine dealer. He's a nice guy, so he won't deal to my kids... Agent: Ever notice that there are always places that you shouldn't park your car unless you are trying to get it stolen? But for some reason you seem to think that it won't happen to you...

A bit lengthy, but I figured to share what I thought. Perhaps someone can shoot holes in it, and clear it up for us...*shrug*

[VEO]

You're just trying to justify your Belief. You're not really listening (is that one of the examples above...?) The Agent and the Emissary don't care how they feel, they're professionals doing their jobs. It's the susceptibility of the Populace that matters. YOU have to downgrade the enemy that is attacking you with characters to give your people protection. HE doesn't have to downgrade with you to give his characters an edge.

[DaveHolt]

Here's an additional $0.02:

while it's intellectually interesting to think about what makes "sense", it is not the best way to approach figuring this (or anything else) out about this game. I mean no disrespect to those of you who like to think about how things "should work." It's just that how things "should work" does not necessarily correlate with how things "do work". Not to mention that you might things "should work" a certain way, whereas Bill Fields might have thought things "should work" a different way.

two superior methods I'd suggest are:
a. measure, measure, measure and measure again. fit to a model and see if your model correlates with your measurements.
b. realize the game is a software program. how would you code it if you were writing the software?

My answer to both b. and a. matches for the particular topic of this thread.
I'd build the software by creating a subroutine that computes relations differential. that subroutine would be called as a part of order processing for those orders where relations matter. I would not have a different piece of "relations" code for each order. different code doing the same or similar things is bad coding practice. yuck.

That implementation concept matches with my observations on how relations differential affects agent & emissary actions in practice, observed over a decade of playing the game.

All that said, I fully well admit that I could be completely wrong. Part of the fun is forming your own opinions on these more esoteric areas of game strategy.

cheers,
Dave

p.s. for example, how does stealth fit into agent order calculation???

[Tuormo]

My personnal belief on stealth is that it adds a random 1-x to your agent rank for purposes of your order. Say your a 62 agent with 25 stealth you would act as a 62 +(random 1 to 25) added to your agent rank. The random 1 to 25 would be similar to how in Personnal Challenge you get a random 1 to 100 and your opponent does likewise and then the difference is the 1 to x amount of damage the loser takes.

[VEO]

I'd build the software by creating a subroutine that computes relations differential.

...snip...

That implementation concept matches with my observations on how relations differential affects agent & emissary actions in practice, observed over a decade of playing the game.


So Friendly vs Friendly is no different than Hated vs Hated? I took this from the word "differential"....

Clarity is required here, as you don't appear to be weighing in on either side here.

Thanks,

Brad

[DaveHolt]

Brad,
sorry. didn't mean to be ambiguous. yes. believe it or not, i'm saying that friendly vs. friendly is identical to hated vs. hated. in character actions where relations come into play.

worse case is friendly vs. hated.
best case is hated vs. friendly.
etc.

It's the differential that matters.

it's easy to implement. So, it's probably implemented that way. It'd be more difficult to implement a split-hieararchal approach based upon each nation's seperate relations, so i bet it's not implemented that way.

Also, my experiences seem to coincide with this belief. I admit freely that it is an "educated" belief.

Dave

[DaveHolt]

My personnal belief on stealth is that it adds a random 1-x to your agent rank for purposes of your order. Say your a 62 agent with 25 stealth you would act as a 62 +(random 1 to 25) added to your agent rank. The random 1 to 25 would be similar to how in Personnal Challenge you get a random 1 to 100 and your opponent does likewise and then the difference is the 1 to x amount of damage the loser takes.

Tuormo, thanks!

This is what I have always used as my "model" as well. I also think this calculation applies for Scout orders. Others disagree.

in fact, the full agent success formula I've used is:

Agents explained

Much of this is from Mouth-of-Sauron, mixed with my edits and comments received from lots of players over the years:

Terms:
Net Agent Rank: Agent Rank + Agent Artifact(s) + Stealth Modifier
Stealth Modifer (1): Rand () * (Stealth Rank + Stealth Artifact(s))
Stealth Modifer (2): Agent Rank * (Stealth Rank + Stealth Artifact(s))/100
Net Guard Rank: Rand () * 2 * Net Agent Rank of Guard
Character Rank: Victim's highest unmodified (no artifacts) skill rank
Relations Adjustment: (see table pg 54 of rulebook. take the difference)

Note: I tend to believe Stealth Modifier (1) is the correct one as it more closely matches the way the rest of the software was apparently written.

Agent Action Success %:
Assass success: Net Agent Rank - Net Guard Rank - Character Rank*0.5 + Relations adjustment
Kidnap success (1): 5-10% easier than assass
Kidnap success (2): same success as assass success, but consequences of failure are less severe

Note: from my own experience I believe kidnap to be easier than assass, so I’d go with (1), but I usually choose K/A based upon whether the target char has agent skill or not. A hostage with no agent skill doesn't escape.

[Alain]

Dave,

Recently I seem to find I have less success in kidnapping then before, I believe some time ago there were some adjustments made and kidnapping is now harder then assassination. I got a lot of data on kidnapping when I played a special scenario where only kidnapping was allowed and I am 90% sure that assassination is easier now.

Cheers,
Alain

[Ed Mills]

Considering the huffy/offended denials that have come out of Cardiff, one would hope that there have been no source code changes these last several years. It might reflect on someone's integrety and judgement.

[DaveHolt]

Alain,
thanks. I'll look into that the next time I get to play a K/A power...
maybe in a team game coming up in Jan... (if we get DS)

Dave

[Clint]

"It might reflect on someone's integrety and judgement."

No changes have been made by us, we've not been able to! Polite request: please don't do this.

If we were to change things we'd discuss it with the players and see what they felt about the changes.

Clint (GM)

[VEO]

A respected player once advised that he takes 2/3 the stealth value and adds it to the agent value straight up. I respect this particular player enought to say "That's good enough for me", although I shall continue to read and consider. I haven't heard anything new on the subject since, certainly nothing with any kind of MoS style data analysis to "prove" a formula, so 2/3 it is in my mind.

[VEO]

Dave,

Recently I seem to find I have less success in kidnapping then before, I believe some time ago there were some adjustments made and kidnapping is now harder then assassination. I got a lot of data on kidnapping when I played a special scenario where only kidnapping was allowed and I am 90% sure that assassination is easier now.

Cheers,
Alain

How can you use data from a scenario where Assassination wasn't allowed to compare K to A..?

Mind you, the common belief that kidnapping is easier than assassinating kind of flies in the face of "common sense" anyway, doesn't it? Just kill the guy or tie him up and carry him off...? Seems easier to get away without carrying an extra body around. Common sense flying in the face of common knowledge? Ah, lovely game, this!

[Celebion]

"It might reflect on someone's integrety and judgement."

No changes have been made by us, we've not been able to! Polite request: please don't do this.

If we were to change things we'd discuss it with the players and see what they felt about the changes.

Clint (GM)

On a side note, does anyone have the article from WotW regarding the changes made to the code making army commanders harder to kill?

[FrightfulTales]

To the best of my estimation, it appears that stealth is a percentage modifier on base agent rank. By this, I mean that if you have 30 stealth, your base agent rank is increased by 30%. This is reflected in the lack of any appreciable increase in agent effectiveness with 10 stealth, and also in the fact that a 20A with 70Stealth will never succeed in assassination anyone. If it were a random roll of 1-70, more some percentage of the time, your agent skill would be 70+ and you would have a chance of assassinating a character, but 70% of 20A skill is 14, which would make you a 34A and very unlikely to kill anyone.

Yet another theory in a long line of many.

[VEO]

...the fact that a 20A with 70Stealth will never succeed in assassination anyone.

"Fact"...? Beware of such flippant use of the language. It's been done on more than one occasion. Careful readers, always insert various synonyms of "opinion" when people speak of behind the fog.

[Tuormo]

I recently had a 32 challenge rank character kill a 124 rank challenge character with but a single blow in a personal challenge. I ran the challenge calculator quite a few times afterwards and it comes out at around 1 in 40,000 to win this challenge.

I wouldnt say fact when saying a 20a with 70 stealth cant kill someone, it could happen but it would be extremely rare as in my little guy killing off the big guy.

[VEO]

Depends on the formula. If the 70 stealth adds "70% of 20" then it's highly unlikely. If the 70 stealth adds Random(1-70) to 20, well, it's entirely feasible.

[Alain]

I recently had a 32 challenge rank character kill a 124 rank challenge character with but a single blow in a personal challenge. I ran the challenge calculator quite a few times afterwards and it comes out at around 1 in 40,000 to win this challenge.

I wouldnt say fact when saying a 20a with 70 stealth cant kill someone, it could happen but it would be extremely rare as in my little guy killing off the big guy.

Well it is definitely NOT as rare as that. I have had Carrog, agent rank between 30 to 40 with 30(70) stealth rank, kill more then one Eothraim army commander, though their command rank would only have been between 30-45. My experience leads me to think (same as some of you have already mentioned) that a random number between 1 and the stealth rank is added to the agent skill and then another roll is made for the success of the agent action.

Alain

[Alain]

How can you use data from a scenario where Assassination wasn't allowed to compare K to A..?


Well, assuming of course that the formula for assassination and kidnapping is the same for every scenario, I compared that data with other games where I was going for assassinations more often then kidnappings.:)

[BaaBaaRox]

We are getting a little off topic here, but....

I am v happy for people to think what they like on downgrading for agent actions. Hopefully the people I play against dont think it is important :-)

In BOFA, on T2 Bilbo, a 20 agent, killed the Necromancer on his city castle. Necromancer was in charge of an army and was guarded. This is a pretty tough assignment.... Bilbo did have 80 stealth to go with the agent rank though....

[Clint]

Bilbo - I had the exact opposite with Stealth - had Bilbo die with similar stats on a Goblin commander so my take it that's it more likely to be a random modifier of 1-X on stealth but that's just me "feel". I've also had Carrog do some early kills so...

As for downgrades - I'm a big fan of them. Often seems to be the difference in combat and more so in agent/ems v enemy action. I always try to attack nations that won't have downgraded (stealing gold seems to be a good example - normally 5k max for a MT, 7500 with a downgrade - from memory but you get the idea).

Clint (very much player here).

[MichaelK]

From memory, the random Stealth modifer *was* the way to add to Agent rank before the last (many years ago) change by GSI.

I was, and still am under the impression from data stats that Stealth is very much more important than just a 1-x random number... I would even go as far as suggesting that it is 1+y - x+y where y is the number GSI added to the Stealth modifier to make it a more usefull stat, while at the same time making ordinary Agent characters have some sort of -ve modifier.

Haven't any of you ever wondered why a A40S10 succeeds more often at identical orders than an A50S0 character? Or do you guys have different data collections?

Michael

[MichaelK]

Oh... and that's my opinion... not Fact... now give me the code for a few days and I'll give you Fact :)

What do you reckon Clint... you don't even have to look at it :)

Michael

[DaveHolt]

Michael,
I wonder if every character might have a unique "Y" for each game that is persistent throughout the game? Some characters never seem to fail. Others seem to fail all the time. The vast majority fall in the middle and seem to conform to "expected" results.

Thus, I think at game-in, a character gets a luck factor, or your "Y", which is persistent for them for the entire game. I have different emissaries right now that fall into either the "never succeed" or "always succeed" categories.

And I do think your point is interesting and has merit now that you bring it up. Stealth is a VGT (Very Good Thing).

[Ed Mills]

David, I think you are right. In one game my two Dun mages never got more than one skill point whenever they prenticed. The game went around 40 turns. This is streaching the random factor beyond breaking.

[MichaelK]

Dave,

The reason I pointed towards stealth was the update that GAD sent out from GSI re Agent order changes... long long time ago... can't place my hands on that piece of paper at the moment... but all talk from then on and data collected seemed to point towards Stealth being made more important.

It's an interesting concept... a Luck factor allocated to each character at character creation.

Michael