Total War: WARHAMMER

Total War: WARHAMMER

Less OP Agents - (NO MORE CHAIN SUCCESSES) - 'On Guard' Fixes and Edits
Biggus Thiccus  [developer] 1 Jun, 2016 @ 9:09pm
The New Agent Debuff, Why I changed all the numbers and some basic Math
Hiya,

So I just wanted to outline / explain some of the decisions I made regarding the changes to numbers following the introduction of the new debuff. I know in theory it looks like a buff to agent actions, and the end result is that yes, it is. I didn't actually want to make an army completely immune following a successful assault, I just wanted it to be very difficult for it to happen - if it happens now you will know that you (or the AI) just got very lucky!

I'm just gonna run through examples, using only the base probabilities of success as there's too many skills/traits for me to realistically balance around - I think the skills are balanced around countering each other so it shouldn't be too significant.

Previously, whenever an agent succeeded the army would become practically immune - the highest base chance in the game is 60% and the +65% defence stopped anything happening. This was regardless of whether it was the same agent attacking you repeatedly or a new agent trying his luck.

Now, there are two different avenues of chance. Take a champion assaulting an army's units. The base chance is 50%, the base chance of defence is 0% so the total chance is 50%. If the agent succeeds he now receives a -15 penalty and the army receives a +25% boost. If that same agent wants to assault again, he now faces (50-15) - (0 + 25) odds, which is only a 10% chance. If however, a new agent appears to assault, he faces a 50 - (+25) odds, for 25%. I think this is an acceptable amount, because if we were to roleplay it, different agents would have different approaches so an army defending against the actions of a previous agent wouldn't fully know what to expect, but may be on high alert.

For many this 25% for a second agent to succeed may be too high, and I can completely understand that. I am willing to adjust the numbers according to popular feedback, but for now I would appreciate anyone who is willing to give the current version a go and let me know how it plays out!

I should also mention that due to the difference between 'Blocked' and 'On Guard' it is possible for these defence buffs to stack (for all I know just On Guard could stack, I don't have a save where I succeed 2x in a row to test it on...) - this means if a Spy successfully hinders you and the next turn a champion successfully assaults you (bearing in mind that's only 25% chance) you will then have +50% defence during the overlap, which is about as close to immunity as you can come!

Sorry for such a long and rambling post, I might have stayed up till 5am getting this done...

Thanks for reading and I look forward to your feedback!
< >
Showing 1-8 of 8 comments
Xeryx 8 Jun, 2016 @ 8:29pm 
I am currently using another mod, that removes the assasination from all but one agent. It has majorly changed the game! The assination chance is also lowered, too much I feel. Maybe a happy medium of your mod and the mod I am using would make a winning combination?

I love the idea, about lowering chain atttacks with the on guard! and 3 turns is fair.

The main function that is OP is assasination. Hindering and sabotage are all good with me as far as successes. As long as it si for the first attempt and not able to keep doing it turn after turn.

With the mod I am using, I have seen the AI use agents much better to help or hinder provinces, instead of as agents. There must be a threshold of success that the AI sees as it is ok to attempt a block or assisnation.
Last edited by Xeryx; 8 Jun, 2016 @ 8:37pm
Quvano 9 Jun, 2016 @ 9:05pm 
@Xeryx. What is the other mod you are using?
Xeryx 9 Jun, 2016 @ 10:53pm 
I am testing yours right now. The other mod I was originally using was called. Agents tweak mod.
Xeryx 9 Jun, 2016 @ 11:09pm 
Does this effect, affect any other skill attempts, such as blocking after an assasination by another agent?
Biggus Thiccus  [developer] 10 Jun, 2016 @ 1:31pm 
Originally posted by Xeryx:
Does this effect, affect any other skill attempts, such as blocking after an assasination by another agent?

Yup! It's a general debuff, any action can trigger it (except attacking settlement walls, which I don't think I can fix because the walls seem detached from the settlement itself, database wise.

Xeryx 10 Jun, 2016 @ 3:13pm 
Ok, I have been playing around and testing the mod. I have decided that I like the stacking, as it will primarily deter a human player from constantly trying multliple attack on armies. My three wraiths were able to decimate the the Orcs specials and lower generals overall because of my ranks being 15. It did however take me much longer to do so.

I do feel the vanilla percentage to assassinate is a bit too high, what is your opinion on that?
Biggus Thiccus  [developer] 10 Jun, 2016 @ 3:37pm 
When I was making the mod I actually dug through and looked at the base values - most of them seem reasonable to me 25%-30% base. Spy spec ones are the biggest issue - their base chance is 35% which is high but not too OP, especially as it's their primary function. The issue comes from the 'Assassin' skill point, which after 3 points puts it up 50% base chance, which is probably too high.

I'm floating a couple of ideas of how I could address it:
#1 I just nerf the % chance gianed from assassin - easy but I don't like it at all; spies should be better at killing, but the current skill-tree eco system really doesn't support defending against assasination / hinderance as much as I feel it should.
#2 Rework the entire skill trees so that they make some sense - I'm not super satisfied with my other mod that just removes % gains from skillpoints, especially paired with this mod it feels a little bit too easy and it removes any sense of innate veteranship. I think I can make it better by making % gains tie more to the feel of skill point (defensive skills give defence, offensive skills offence, etc) but I need to find the time to properly research *every* skill tree, make the thing and analyise for balance, which will definitely not be a quick process.
Last edited by Biggus Thiccus; 10 Jun, 2016 @ 3:37pm
Xeryx 10 Jun, 2016 @ 3:50pm 
Here is my recommendation, for now lower the base success rate by maybe 10-15%. I think for now, it would still give the feeling of accomplishment, without frustration that I was feeling with the other mod I played.

I am pretty dissatisfied with most of the skill trees as they were not really well thought out. There are many duplicate skills and not enough unique skills for the different classes. I would like to see them more like the research tree, where the are branched more, not where you pick and choose each one, but have to take a path for a career. That is a major undertaking though and will take a lot of time, Maybe others out there have the time for that.

Cheers
< >
Showing 1-8 of 8 comments
Per page: 1530 50