freyj
Sergeant

I have been playing my own kingdom in diplomacy for a while, and I greatly enjoyed it. It is MUCH better than Native, due to all these additional features that are not only convenient but also intuitive. However, lord recruitment remains a challenge. I found a lot of frustration described in this following post very similar to my own experience:
http://forums.taleworlds.com/index.php/topic,116431.105.html#msg3312381
Now, instead of just saying "diplomacy sucks" (which I disagree), I would like to make some suggestion. I hope my suggestions satisfy three requirements:
a) They reduce the frustration in lord recruitment, making it a little bit easier
b) They are intuitive and make diplomacy's simulation more realistic, not less.
c) They will not be of too much work for implementation
Now here are my suggested changes:
1) Give more opportunity for "private" conversations to allow recruitment. More concretely, change the test for lord proximity, so that when I approach a castle and request to meet someone in the court yard, the conversation is considered private.
I believe this suggestion satisfy the three requirements:
2) Make the decision of joining or refusing my kingdom less random. More concretely, once the chance of recruitment is calculated, make the pseudo-random number that it is compared to is generated not from 1 to 100, but from 31 to 70, or something similar.
I again present my arguments following the 3 requirements I put myself:
3) Make the military reality weigh more in the recruitment decision and fix some inconsistencies. More concretely, I have 3 suggestions: 1) the lord should be REALLY easy to convince if he has no fief; 2) the lord should consider the faction-wise power comparison more (i.e. if his faction only has 2 castles while the player faction has 8 towns and 10 castles, a majority of lords should switch side instead of staying with a hopeless faction); 3) if the lord only has village then he should never say "If I switch side I'll be in a sea of enemy" due to location of his fief, because if he switch side he will lose his village anyway.
This completes my proposal, for now. Thoughts and suggestions are welcomed.
http://forums.taleworlds.com/index.php/topic,116431.105.html#msg3312381
Now, instead of just saying "diplomacy sucks" (which I disagree), I would like to make some suggestion. I hope my suggestions satisfy three requirements:
a) They reduce the frustration in lord recruitment, making it a little bit easier
b) They are intuitive and make diplomacy's simulation more realistic, not less.
c) They will not be of too much work for implementation
Now here are my suggested changes:
1) Give more opportunity for "private" conversations to allow recruitment. More concretely, change the test for lord proximity, so that when I approach a castle and request to meet someone in the court yard, the conversation is considered private.
I believe this suggestion satisfy the three requirements:
a) A lot of times I couldn't get a lord because his faction has only 2 castles left, each is populated with 6 lords, and these 6 lords - including the lord I want - never come out. I once did a test: I used the cheat menu to force the lord to "go to another friendly castle" or "raid a village". Then I talk to him, and found that the chance of recruitment is often fairly high if I have the best possible status on myself and have chosen the correct dialog option. But in normal game play I never get such a chance.
b) Now consider this: I am besieging a castle, and asked a lord in the castle to come out and talk to me. That conversation should be assumed to be PRIVATE. All the other lords are still inside the castle and hundreds yards away.
c) I found the code of private conversations in module_dialogs.py and the test for lord proximity in module_scripts.py. However, since I didn't spend time learning all the functions, what I did was simply commenting out the lord proximity test, so that all conversations are always private. Well, that's cheating. I would rather to have Waihti implements a proper change, if he/she is interested in.
b) Now consider this: I am besieging a castle, and asked a lord in the castle to come out and talk to me. That conversation should be assumed to be PRIVATE. All the other lords are still inside the castle and hundreds yards away.
c) I found the code of private conversations in module_dialogs.py and the test for lord proximity in module_scripts.py. However, since I didn't spend time learning all the functions, what I did was simply commenting out the lord proximity test, so that all conversations are always private. Well, that's cheating. I would rather to have Waihti implements a proper change, if he/she is interested in.
2) Make the decision of joining or refusing my kingdom less random. More concretely, once the chance of recruitment is calculated, make the pseudo-random number that it is compared to is generated not from 1 to 100, but from 31 to 70, or something similar.
I again present my arguments following the 3 requirements I put myself:
a) If you activate the debug message, you often see something like this: "chance of success = 80; random = 98" (a very good attempt is failed, due to random chance), or "chance of success = 25; random = 10" (a very bad attempt is successful, due to random chance). This is frustrating, because in the game the decision is final - you never get a second chance.
b) Many people have complained about the lord saying "I like everything about you, BUT I refuse", or "You suck a lot, BUT I pledge", and "this word is final". When I first played without the debug message, I thought the dialog was simply fake. I was then surprised to see that Waihti has the decision simulated exactly as the dialog goes, it is the high randomness that is causing the discrepancy. Therefore I think reducing the randomness would be beneficial.
c) Implementing this change is simple. Just change that time-based mod formula would be enough. However, you might dislike that, because then the "chance of success" is no longer a REAL chance of success. So a better fix would be changing parameters in the chance of success calculation, so that when an attempt is really bad, the chance is smaller than 0, and when an attempt is really good, the chance is greater than 100. I have seen chance of success to be -10 or 112 with the current parameters, but I think it is too rare.
b) Many people have complained about the lord saying "I like everything about you, BUT I refuse", or "You suck a lot, BUT I pledge", and "this word is final". When I first played without the debug message, I thought the dialog was simply fake. I was then surprised to see that Waihti has the decision simulated exactly as the dialog goes, it is the high randomness that is causing the discrepancy. Therefore I think reducing the randomness would be beneficial.
c) Implementing this change is simple. Just change that time-based mod formula would be enough. However, you might dislike that, because then the "chance of success" is no longer a REAL chance of success. So a better fix would be changing parameters in the chance of success calculation, so that when an attempt is really bad, the chance is smaller than 0, and when an attempt is really good, the chance is greater than 100. I have seen chance of success to be -10 or 112 with the current parameters, but I think it is too rare.
3) Make the military reality weigh more in the recruitment decision and fix some inconsistencies. More concretely, I have 3 suggestions: 1) the lord should be REALLY easy to convince if he has no fief; 2) the lord should consider the faction-wise power comparison more (i.e. if his faction only has 2 castles while the player faction has 8 towns and 10 castles, a majority of lords should switch side instead of staying with a hopeless faction); 3) if the lord only has village then he should never say "If I switch side I'll be in a sea of enemy" due to location of his fief, because if he switch side he will lose his village anyway.
Again the 3 requirements:
a) Again, the frustration comes from the situations when all these 12 lords stay in the last castle of a faction. And if you capture that last castle, they will join other factions (who are often also responsible for destroying the original faction), but never join you. I don't know if diplomacy handles the destruction of factions itself or leaves it to native, so I didn't propose to change that. However, if the change I am proposing could be made, my life as a conqueror would be much easier.
b) Now I do believe my suggestions are intuitive. We are talking about feudal world, and all lords are fairly practical when making all their decisions. I would imagine not having a fief by itself would be a very legitimate reason to switch side for most lords. The faction is failing by itself is also a very legitimate reason. The suggestion about village-only lord is by itself intuitive.
c) Again, I have found the code responsible for all these calculations in the code, but I don't know enough syntax or function to change them. I might be wrong, but I imagine none of these should be too hard to implement.
a) Again, the frustration comes from the situations when all these 12 lords stay in the last castle of a faction. And if you capture that last castle, they will join other factions (who are often also responsible for destroying the original faction), but never join you. I don't know if diplomacy handles the destruction of factions itself or leaves it to native, so I didn't propose to change that. However, if the change I am proposing could be made, my life as a conqueror would be much easier.
b) Now I do believe my suggestions are intuitive. We are talking about feudal world, and all lords are fairly practical when making all their decisions. I would imagine not having a fief by itself would be a very legitimate reason to switch side for most lords. The faction is failing by itself is also a very legitimate reason. The suggestion about village-only lord is by itself intuitive.
c) Again, I have found the code responsible for all these calculations in the code, but I don't know enough syntax or function to change them. I might be wrong, but I imagine none of these should be too hard to implement.
This completes my proposal, for now. Thoughts and suggestions are welcomed.