[Submod] Improved Spotting Distances for more Tactical AI RELEASED Feb 2

[MagicCuboid]

Hi everybody!

Due to developments on this thread, and the ingenious idea of Krixux, a few of us have been searching for methods to improve the tactical ability of the AI.

What this mod does:
It improves, across the board, units' minimum ability to spot distant targets without affecting the hiding mechanic. Hiding in scrub, trees, etc. have been left at vanilla DeI values.

Vanilla Spotting

Modded Spotting

Why?
During some initial testing, a few of us have noticed an improved ability for the AI to deploy its troops in the initial fight if it's aware of the entire player army. With vanilla values, the AI was clearly nearsighted, and would chase only the units immediately in front of them (usually skirmishers) with WAY too many troops than were necessary. We believe this was actually a design choice by CA for casual players who wanted to "bait" with skirmishers, but it's ruined the AI for even moderately skilled players.

The Effect
AI Armies will demonstrate notably superior flanking tactics than in any other mod! Now that the AI is aware of what it's facing, you will see it keep reserve forces, hold back to regroup, and use its cavalry more intelligently. Now, the AI still is of course no substitute for a human, but at least the player will have a little more management to do in the early stages of battle in order to keep up/counter AI moves.

Installation:
A) Here you go

B) Alternative Version (Increased spotting distance, AND lowered spotting distance of hiding units for more exaggerated guerrilla tactics)

C) Butan's Extreme Version (NO more hiding whatosoever)

Compatibility
Either version is fully 100% compatible with ANY other mod.

Special Thanks to
Krixux, for the original idea
Butan, for helping with testing and exploring the DB.
KAM, for his tireless devotion to Rome 2 combat.
The Rest of the DEI Team... without you guys I would not be playing Rome 2 anymore.

[Aethyr]

, ooh interesting. Downloading when I get home from work.

[rjacko10]

I've been following the conversation intently on the other thread. Well done guys!

Can we see an additional pack for the increased hiding abilities? Might make hiding worthwhile, as long as it isn't OP.

+rep all round

[MagicCuboid]

Alternative version added, rjacko. Reduced spotting of hidden units by 25%. Let me know how it goes!

[rjacko10]

[krunsh]

amazing. Will start using this ASAP (With better hiding!). Hopefully this is good enough that it gets implemented in DeI after some testing!

[Butan]

MagicCuboid, do you mind if I sometimes post reports on completely opposite work on hiding? Like, greatly affecting hiding mechanics and heavily tinkering with spotting distance. I want to know exactly how it "can" work and then I will tell you so that you may use it for an improved AI.

[Krixux]

been in hospital today ( not to serious ,but...) and can hardly stay at my PC(this is hell !)
got this huge needle in my back and the effect will wear of in a day or two ( I hope)
not been able to test and concentration is low (due to meds )
now that i've got your sympathy ( and drown all attention to me : yay )
back to business:

this threat is my dream come true!

MagicCuboid is just that : Magic !
and Butan you just rock no more no less !

keep digging
there is stuff still to be "unearthed" in data (after all this time who would have thought ...)
lol

[Butan]

Dont lift too heavy Krixux!.. and rest!
Also what were you about to say last time on the other thread about hidden data?



Also for the benefit of that thread I come with further data conclusion:


Using those values, I tested the impact of "visibility" columns: I put the minimum visibility to almost the maximum value




Here's the impacts:


* none on hidden units (still hidden at up to 10m, 2499 meters value not used)

Spoiler Alert, click show to read: 

* the only impact is on non-hidden units (still visible through cliffs at very long distance thanks to 2499 meters value)

Spoiler Alert, click show to read: 

Conclusion:

- visibilitymax_spot_range= when Line of Sight isnt a problem (direct visual) + the unit isnt hidden + the distance is equal/below maximum visibility, the unit is seen.
- visibilitymin_spot_range= when the unit isnt hidden + the distance is equal/superior to the minimum visibility, the unit is seen


If you put an extremely high visibility minimum spot range, you will see EVERY UNITS on the map that arent in bush/forests, even if they are behind cliffs.

So the visibility values are really all about non-hidden and LoS, while the spot distance in scrub/forest are what applies when hidden.



If one want to make the AI more responsive to tactical challenge, I believe the Line of Sight AND the hiding spotting distance must be increased by a fair margin.
In the end it means that we have to sacrifice Line of Sight camouflage and forest/scrub camouflage in order to increase AI tactical response capacity.
I dont know of any other values to tinker with.

[MagicCuboid]

Krixux, stay healthy! Take it easy and eat some soup

Butan, nice testing . That would explain why some Gallic cavalry came screaming at me from behind a hill earlier. Definitely gave my skirmishers a bit of a surprise!

[FlashHeart07]

Trying to wrap my head around this so bear with me guys.

So where we are now is.. You have found out what tables control LoS and what controls whether a unit is hidden in scrub, forest or not?
And to get the effect of a more reactive BAI we need to lower the ability to remain hidden in scrub, forest?

Should this be the case I would say that it is something that we (the users) can be more than happy about.
Then we (I say we because I hope to be able to help testing, unless you get it done) will just have to figure out the precise numbers that will give the best results.
And just to be clear, Butan. We dont need to change values the each individual unit right? One can just change the values overall?
or would it perhaps in the future make more sense to spend the proper time to balance every unit?

[Butan]

What I would advise if one wanted to balance every unit, would be :

- to modify only which scouting level is attributed to every units into the "land_units" table, according to preference
- balance each scouting level spotting/visibility values, according to preference
- in the end only modify "hiding_spoting_values" because it supercedes every individual "land_units" spotting values so the "land_units" are useless except for attributing which scouting skill values are used.


I see only 1 scenario where individual balancing might be needed: special units (like how unit cost for chariots/elephants/wardogs were customised) that we would like to give special scouting roles.

Its also definitely possible to use all the 9 different scouting levels, to custom them so that we could have (for example) 1 "general unit scouting level" and "8 custom tailored scouting levels" that we could give to only 8 different kind of units.
A bit like how we have limited army stances slots, so we can only custom those slots and not add/remove anything.




IMO since scouting levels are not displayed in the stats, the players doesnt knows which units can spot to what distance, so I believe every units should have the same scouting potential on the battle map. When a battle is engaged there is no real reconnaissance units, every units on the battlefields are simply going to fight and detecting the real-time enemy positions on the go, thus I discuss the need for 9 different scouting levels. Its all about camouflage and direct visual - human eye precision. Thats why in my own tests I put every scouting levels to the same values (that and also for.. testing itself!).

Maybe give a bit better bonus to mounted units, because they simply have a taller observation situation to see things infantry wouldnt at the same eye precision quality.

[FlashHeart07]

Okay. Think Im a bit tired because I have a hard time getting it.
But it wouldnt be necessary to edit the values for each unit to overall get the same results as you are getting atm.
Cannot wait to get the time to proper try and test this awesome idea.

True. If we somehow was able to add a tooltip that told people "This unit is specialized in detecting enemy ers hiding in the grass" it would make sense to create 8 custom scouting lvls. Ofc it should be discussed which units should have this and if it is necessary at all if the chance of getting ambushed in battles are so low anyway.
So yeah, lets just have it as it is atm and one can always play around with it afterwards. Right now further testing and if all goes well a possible implementation of this into the mod or as an official submod is number one priority.

[Zonac]

OK I have done some tests end it really seemd that the enemy flanked and so on! they double used their cav and countered my cav and skirmishers, and after some time where a gap opened the enemy even pulled some infantry through it (from its reserves!!! ) also the general unit was used to counter immediate threats like my skirmishers shooting in the back of his main line. however the only "bad" thing was, that the enemy general even countered my cav and therfore his general got down pretty fast.. :/ ^^
However, there is still a need for some more tests , I really think that this could be something!!!!

[Grizwald714]

Just downloaded the pack. I'll report back after some testing.

Krixux I hope u get better and best of luck

[rezoacken]

So you guys would recommend this + Kam's experimental ? Where does Battle Dynamic Reworked fits into this ? :p Too many submod at once for me to follow

[Butan]

KAM's experimental and Alexander's Battle Dynamic Reworkd cant fit together because they modified the same things but differently.

[MagicCuboid]

So you guys would recommend this + Kam's experimental ? Where does Battle Dynamic Reworked fits into this ? :p Too many submod at once for me to follow

Yeah, I haven't tried Battle Dynamics Reworked but they are both going for the same goal. I think KAM's strategy is to increase damage and decrease morale, while BDR actually increases unit speed and decreases defense (I think). Anyway just pick the one you like and use it with this mod

however the only "bad" thing was, that the enemy general even countered my cav and therfore his general got down pretty fast.. :/ ^^

Yeah I wondered if suicide generals would become a problem. A workaround would be to go into land_units and divide the general units' spotting by 2.5x, but then this mod would be way more invasive than it currently is.

[Alexandre Lange]

Following you guys. Pretty good ideas, I must say. I am just waiting for you to test further and find the sweet spot. I think we will have something solid here.

[balparmak]

What I see here is the future of DeI -and of course, Rome 2. Amazing discovery, keep up the good work guys!