[WIP] bCombat infantry AI Mod [SP]

[RSF TheCapulet 59]

I saw your .rpt. Indeed there are some problems.

Did you enable the mod within configure / extensions menu?

Enabled through modline. Does it make a difference?

[fabrizio_t 58]

Enabled through modline. Does it make a difference?

Maybe.

Try enabling it from the menu and let's see what happens.

---------- Post added at 09:28 ---------- Previous post was at 09:27 ----------

Yes fabrizio, that will not be a problem.

Ok, adding you to the list.

[Psilocybe 3]

Fabrizio, what parameters would you like us to test in particular, my main tests I have been conducting have been around ambushes in area's with and without cover, checking how the AI utilize it vs snipers and other assaults. Also testing reactions to hearing local small arms fire and grenades etc. I can write up my findings but I want to know if I am testing in vain and should try focusing on other things.

[fabrizio_t 58]

Despite being a bit late on my schedule due to some real-life duties, testers should receive shortly the link to next testing package (v0.12).

New version comes with some optional features ON by default:

DEBUGGING FEATURES:

* Debug Balloons

* bDetect stats pane

Debug balloons are small spheres hovering on top of units head.

Their colour shows morale level:

WHITE = 100%
GREEN = 75%-100%
BLUE = 50%-75%
YELLOW = 50%-75%
RED = 0-25%
PURPLE = Fleeing unit

During a battle WHITE / GREEN / BLUE balloons should be usual.

During pitched combat, when casualties or suppression fire weight up, you'll usually see some YELLOW ones.

RED balloons should show episodically, mostly after some combat.

PURPLE balloons shoud be much rare, meaning units are fleeing, thus ignoring any enemies.

These units may occasionally surrender.

NOTE: These features cause some FPS hit.

You may toggle these features on/off by commenting these lines at the bottom of bcombat.sqf:

call bdetect_fnc_benchmark; // Uncomment this line to activate bDetect live stats panel
...
call bcombat_fnc_debug_balloons; // Uncomment this line to activare bCombat debug balloons (as alternative to bcombat_fnc_debug_text)

GAMEPLAY FEATURES:

* bcombat_allow_fast_rotate = true

* bcombat_allow_fleeing = true

bcombat_allow_fast_rotate

By default units use fast rotation mode.

This means that they rotate faster towards target.

I'm sure we will be able to further fine-tune rotation speed, this is supposed to be some staring point.

I've noticed that the activation of fast rotation leads to a sensibly lower bullets / kill ratio.

bcombat_allow_fleeing

By default units adopt bCombat custom fleeing behaviour.

So units can flee, temporarily and individually.

Fleeing units usually RUN away and are supposed to be completely combat uneffective.

NOTE: You can toggle this stuff by editing relevant entries within the "Custom Configuration" section into bcombat.sqf:

// -----------------------------
// Custom Configuration
// -----------------------------

...

bcombat_allow_fast_move = false;
[b]bcombat_allow_fast_rotate = true[/b];
[b]bcombat_allow_fleeing = true[/b]; 
bcombat_allow_surrender = false;
bcombat_slow_leaders = false;

...

---

The new version comes with many enhancements, here is the final changelog:

Core:

* Morale hits are now applied non-linearly: it should be a bit harder to achieve complete morale break
* Better filtering of BIS danger.fsm events. Distance limit of 500m. from source for events collection. 
* AI units may open fire upon on targets of opportunity from distance, if:

	* target is near unit's planned destination ( < 50m )
	* target is near group's current waypoint position ( < 50m )
	* units are "hidden" and not moving

* Disabled vanilla fleeing behavior when bcombat_allow_fleeing = false (it was creating mayhem)
* "Fast rotate" optional enhancement fixed
* Minor tweaks and bugfixes.

Balancing:

* morale hits lessened about 15-20% in average, due to new non-linear logic.

Gameplay:

* AI Group leaders now usually prioritize fire vs. movement. This was done to avoid them rushing. No changes to sub-formation leaders.
* AI Units positioned over terrain level (e.g. guard posts, some buildings) should prefer crouching instead of going prone, when fired upon.
* "Fast move" enabled by default for subordinates, on Player led groups. It sould make units under player command move faster while in combat mode.

Performance:

* Code optimization for functions (only)
* Simplified, lighter (=faster) line-of-sight checking. Checking accuracy made degrading on distance.

Last minute changes were done to morale logics, making hits applied non-linearly.

Keep in mind that all the morale hit coefficients applied should be carefully tweaked, i'd expect some feedback on this.

Further changes were made also to group leaders behaviour, as per-feedback.

---

Finally, some slight optimization was done on .sqf functions.

Substantial streamlining was applied to line-of-sight checking function: it should now cause less overhead in object crowded areas.

My pseudo-profiling tool tells v0.12 functions set performance index is about 2.1 points, where as v0.11 was scoring 2.8.

The lower the better, so new function set is 25% faster.

This does not mean you'll see a 25% FPS gain, the core scripts part is still unoptimized and it's much more complex.

Still some FPS improvements may be noticeable.

By the way, i've done extensive testing (default settings) on Assault on Charkia mission and measured -2 to -3fps avg. difference comparing vanilla vs. bCombat.

That means about 12-15% overhead in my case. I would call this acceptable.

---

SO, WHAT NOW?

As usual, any feedback is welcome.

That said i would like to prioritize efforts onto some key areas, hence i'd like to know testers' opinions about the following:

1) Are group leaders behaviour any better (e.g. less rushing) ?

2) Are you fine with units crouch instead of going prone when on elevated terrain (including buildings / guard posts) ?

3) Do you find morale driving appreciable effect on AI combat effectiveness ?

4) Do you find morale effects almost fine, lacking or overdone ?

5) What do you think about units combat stance handling ? Does it improve units survivability ?

6) What do you think about actual balancing between movement & fire ?

7) Did you notice further examples of unexplicable activation of combat mode on units ?

8) What do you think about AI firing accuracy / precision ? Good or low ?

9) What's your experience on playing as group leader ? Are micromanaging needs any lessened ?

10) What do you think about units "fast rotation" feature, especially in CQB ? Useful ? Overdone ?

11) Did you notice any FPS performance gain / loss ?

---------- Post added at 16:27 ---------- Previous post was at 15:51 ----------

Package links sent.

Have a nice WE.

Edited November 9, 2013 by fabrizio_T

[stun 5]

It looks like v0.12 is disabling player fatigue when used with dev branch. I haven't tested with the stable branch.

[bravo409 13]

Ok first off thank you fabrizio_T for letting test your mod out.....................................

ok here is what I found out testing far away sniping......

1] I used bcombat and tmr mod and saw when firing from 500m out and it hit a enemy unit they didn't fire back nor did they try and hide much.

2] I used Tmr,wwai,bcombat pretty much same result except more movement and more crawling.

3] I used Tmr,wwai<tpw,bcombat This time they really started moving and really taking cover....

Now also notice they have a hard time seeing you until you get up and move to new spot so that is a plus in my book that shows realism cause you cant see someone that far away and know where there shooting from...........

4]I also did a regular seek destroy movement for both opfor and blufor on another test and the movement and hiding was excellent. Not sure if this mod uses houses or not but didn't see any of that. Did see a lot off covering movement though which is nice....in those test I used both tmr,wwai,bcombat,tpw and found with both mods together it ran nice.... So those are my results so far... I would like to see more moral changing didnt see much specially when I shot a guy and his buddy saw him go down and he did nothing in the sort to run or hide... but when I used all mods together it changed some of that.....

[mr_centipede 31]

1) Are group leaders behaviour any better (e.g. less rushing) ?

2) Are you fine with units crouch instead of going prone when on elevated terrain (including buildings / guard posts) ?

3) Do you find morale driving appreciable effect on AI combat effectiveness ?

4) Do you find morale effects almost fine, lacking or overdone ?

5) What do you think about units combat stance handling ? Does it improve units survivability ?

6) What do you think about actual balancing between movement & fire ?

7) Did you notice further examples of unexplicable activation of combat mode on units ?

8) What do you think about AI firing accuracy / precision ? Good or low ?

9) What's your experience on playing as group leader ? Are micromanaging needs any lessened ?

10) What do you think about units "fast rotation" feature, especially in CQB ? Useful ? Overdone ?

11) Did you notice any FPS performance gain / loss ?

As for my own feedback,

it seems AI squad leader no longer rushing like in v011, so I guess that's good

About morale, I once saw a unit turned purple and turn tail, so I guess that's working as intended. Then I saw a unit turned red, but didn't noticed any noticeably different behavior. They still charge the enemy if ordered too, (but in prone only)

The rest haven't properly tested yet, those above are my early observation. But good feeling overall

[bravo409 13]

Commanding is so much nicer now they Actually move to the area right away and not think about it.... Also I said to fire on a curtain unit and they did without pause. I have had a steady Fps of 15-20 and I'm running 2 full rifle squads.

Edited November 10, 2013 by bravo409

[lordprimate 159]

Ok first off thank you fabrizio_T for letting test your mod out.....................................

ok here is what I found out testing far away sniping......

1] I used bcombat and tmr mod and saw when firing from 500m out and it hit a enemy unit they didn't fire back nor did they try and hide much.

2] I used Tmr,wwai,bcombat pretty much same result except more movement and more crawling.

3] I used Tmr,wwai<tpw,bcombat This time they really started moving and really taking cover....

Now also notice they have a hard time seeing you until you get up and move to new spot so that is a plus in my book that shows realism cause you cant see someone that far away and know where there shooting from...........

4]I also did a regular seek destroy movement for both opfor and blufor on another test and the movement and hiding was excellent. Not sure if this mod uses houses or not but didn't see any of that. Did see a lot off covering movement though which is nice....in those test I used both tmr,wwai,bcombat,tpw and found with both mods together it ran nice.... So those are my results so far... I would like to see more moral changing didnt see much specially when I shot a guy and his buddy saw him go down and he did nothing in the sort to run or hide... but when I used all mods together it changed some of that.....

probably worth wild to point out that this is to be tested WITH OUT other mods ..... feedback with other mods running doesnt really help at this stage of DEV... whats to say that its not the other mod causing problems.....

Thanks Fab, for the patch. as soon as i have time ill test it (monday)..

[fabrizio_t 58]

probably worth wild to point out that this is to be tested WITH OUT other mods ..... feedback with other mods running doesnt really help at this stage of DEV... whats to say that its not the other mod causing problems.....

Thanks Fab, for the patch. as soon as i have time ill test it (monday)..

Yes, while feedback is welcome, having other mods running will hinder possibility to evaluate bCombat problems and feaures.

By the way, by running bCombat AND TPWCAS you may end up having 2 different instances of bDetect running at once, which is really bad for performance.

---------- Post added at 12:42 ---------- Previous post was at 12:36 ----------

As for my own feedback,

it seems AI squad leader no longer rushing like in v011, so I guess that's good

About morale, I once saw a unit turned purple and turn tail, so I guess that's working as intended. Then I saw a unit turned red, but didn't noticed any noticeably different behavior. They still charge the enemy if ordered too, (but in prone only)

The rest haven't properly tested yet, those above are my early observation. But good feeling overall

Good to hear.

I find this solution to the rush a bit overdone, since now group leaders tend opening fire at will, immediately and from unsafe positions.

I'm curious about general testers' opinion on this.

"RED" units do obey commands, but they prioritize movement to fire to the extents of being close to useless in combat.

Also their accuracy is very low. This is supposed to mimic fear to the edge of panic ( "PURPLE" / fleeing ).

They should still be able to stand / crouch however, can you confirm the issue with them staying prone all the way?

---------- Post added at 12:44 ---------- Previous post was at 12:42 ----------

Commanding is so much nicer now they Actually move to the area right away and not think about it.... Also I said to fire on a curtain unit and they did without pause. I have had a steady Fps of 15-20 and I'm running 2 full rifle squads.

Units under player command are indeed supposed to move to destination a lot faster, while in combat mode.

Also they're supposed to keep firing when given a target ATTACK order.

However as you were running other mods we can't be sure this is caused by bCombat.

---------- Post added at 12:47 ---------- Previous post was at 12:44 ----------

It looks like v0.12 is disabling player fatigue when used with dev branch. I haven't tested with the stable branch.

Yes, overlooked that, it's disabled.

I imagine this is not optimal, i'd like to make this optional for player.

On the contrary i think that it would be better to keep fatigue off for AI units, because:

* there are still issues with fatigue , sometimes animations pay in slomo;

* since units do crouch / go prone more a lot, they tire too quickly;

Edited November 10, 2013 by fabrizio_T

[stun 5]

Yes, overlooked that, it's disabled.

I imagine this is not optimal, i'd like to make this optional for player.

On the contrary i think that it would be better to keep fatigue off for AI units, because:

* there are still issues with fatigue , sometimes animations pay in slomo;

* since units do crouch / go prone more a lot, they tire too quickly;

Thanks for confirming this is deliberately disabled. I agree, definitely better to disable fatigue while testing. I'm also worried that the changes the BI devs are making to fatigue will cause some heavily loaded members of AI squads to have issues keeping formation, however that is a completely different topic.

[bravo409 13]

Here are some concerns of mine... When I only used Tmr,and bcombat mod only when I shot at the ai they didnt what to move or take cover like when I had tpw and wwai and bcombat together...So question is does TMR make this mod run different too and if does why? So that said in my testing of bcombat and tmr I only saw green dots never saw purple or red like others have seen. I will do another test today and just use bcombat without any mods even tmr which I dont think makes a problem but I will still give it a go. Also worth noting I'm using Dev build of Arma 3...........................................................................................................................................................................

---------- Post added at 09:53 AM ---------- Previous post was at 09:38 AM ----------

I would like to record my findings what is a good recording software I know about fraps but you have to buy it?

Edited November 10, 2013 by bravo409

[fabrizio_t 58]

Here are some concerns of mine... When I only used Tmr,and bcombat mod only when I shot at the ai they didnt what to move or take cover like when I had tpw and wwai and bcombat together...So question is does TMR make this mod run different too and if does why? So that said in my testing of bcombat and tmr I only say green dots never saw purple or red like others have seen. I will do another test today and just use bcombat without any mods even tmr which I dont think makes a problem but I will still give it a go. Also worth noting I'm using Dev build of Arma 3...........................................................................................................................................................................

here's some explanation about cover usage.

bCombat does not handle movement to cover the way other mods usually do.

Usually some "take cover" script is triggered and units are ordered to move somewhere.

There are many shortcomings for this:

* cover object checking is approximate, cumbersome and causing overhead;

* sometimes units never get to the planned cover position, due to combat mode constraints;

* if cover is scarse units may just bunch into the same cover position (addressable, but with further overhead);

* once in position further logic should be deployed, in order to avoid units moving away / around;

* moving units into some arbitrary position breaks the planned formation, meaning groups need time to regroup before moving further;

All in all it's a quite inefficient all-around solution, in my opinion.

I coded this stuff myself about 2 years ago and i was never 100% satisfied.

That said, some mods are still using parts of my original code, further enhanced, apparently with good results.

Believe it or not, while not immune by problems, vanilla AI does make a decent work on putting units into cover.

The problem is that under some circumstances units are blocked from doing that, and being ordered to stop in-place.

bCombat is circumventing the issue the easy way: by removing / reducing the causes of unneeded blocking, units may "naturally" flow into planned cover.

As somebody noticed, this usually works quite well. It's very effective in urban CQB, for instance (except units don't use buildings interior).

Sure there are some edge cases to be handled, i'll name two:

* Vanilla AI group leaders are very inconsistent in moving into cover. They sometimes stick to some waypoint center (maybe in the middle of a road) and don't care about cover at all.

* When there's no suitable cover very near, units may stop in place.

While i'm considering how to address these problems, my current priority is to try getting the most out of vanilla covering mechanics.

Edited November 10, 2013 by fabrizio_T

[bravo409 13]

fabrizio Can you tell me why a least most the time I only see green dots and no red or blue specially when that unit or group is pinned down? Seems like there morale isn't low enough when under fire... Example I used to play a old game call operation flashpoint dragon rising and one of ai I was testing for another ai maker had a ai morale system where when the ai got scared they would be running for cover and you could here in there voice of being scared. This mod reminds me of it but missing the that low moral when scared or being shot at effect. on the positive side I see some nice flanking maneuvers. would like to see snipers try for high ground and not get caught in cqb. Although I did see the sniper once in cqb pull his pistol out. Also worth noting firefights last longer another positive that shows thinking abilities and planning very nice

Edited November 10, 2013 by bravo409

[mr_centipede 31]

Another thing I noticed when playing as a squad leader, whenever I ordered a group of units (colured team for example) to move somewhere, they always bunched up at the destination, unless they come under fire, then they dispersed and look for cover. This is when in contact.

[bravo409 13]

This the morale code from that ai mod I just mentioned .................................................................

..........................................................................<morale>

<!--regeneration rate - morale recovery points per second-->

<property name="regeneration_rate" value="15" />

<!--hold time - after a negative morale won't recover for x seconds-->

<property name="hold_time" value="15" />

<!--experience multipliers - percentage modifiers to absolute morale values, 0 is no effect-->

<!---0.25 changes a +10 modifier to +7.5 and a -10 modifier to -12.5 -->

<!--+0.25 changes a +10 modifier to +12.5 and a -10 modifier to -7.5 -->

<property name="green_modifier" value="-5.0" />

<property name="regular_modifier" value="0" />

<property name="veteran_modifier" value="2.5" />

<property name="elite_modifier" value="5.0" />

<!--Morale state bands - minimum value for each state-->

<!--State_Max is the absolute limit and not a state in its own right-->

<property name="State_None" value="0" />

<property name="State_Broken" value="5" />

<property name="State_Breaking" value="40" />

<property name="State_Failing" value="60" />

<property name="State_Lowering" value="85" />

<property name="State_Full" value="95" />

<property name="State_Max" value="100" />

<!--state values - values applied per second -->

<property name="InCover_Mod" value="5.0" />

<property name="Pinned" value="-21" />

<property name="Suppressed" value="-19" />

<!--event values - combat/self - values applied once on event-->

<property name="Light_Wound" value="-15" />

<property name="Serious_Wound" value="-25" />

<property name="Low_Ammo" value="-40" />

<property name="No_Ammo" value="-60" />

<!--event values - enemy-->

<property name="target_Light_wound" value="10" />

<property name="target_Serious_Wound" value="20" />

<property name="target_Killed" value="40" />

<!--event values - Fire Team state modifiers are adjusted based on the number of fire team members in each state-->

<!--and then applied across all team members. The values listed are for when all members of the fire team are in the-->

<!-- given state. When 50% of the fire team memebers are in a state then 50% of the modifier is applied to the fire team-->

<!--The fire team modifier is applied to the individuals morale level-->

<!--when calculating their morale state but does not modify their morale directly-->

<property name="FT_Ally_Light_Wound" value="-15" />

<property name="FT_Ally_Serious_Wound" value="-25" />

<property name="FT_Ally_Killed" value="-60" />

<!--fire team events, applied to individuals of other fire teams in range when a fire team triggers an event-->

<property name="Allied_FT_Fallback" value="-25" />

<property name="Enemy_FT_Fallback" value="20" />

<property name="FT_Detection_Range" value="32" />

</morale>

---------- Post added at 12:02 PM ---------- Previous post was at 12:01 PM ----------

this part in the code is the important part>>><property name="State_None" value="0" />

<property name="State_Broken" value="5" />

<property name="State_Breaking" value="40" />

<property name="State_Failing" value="60" />

<property name="State_Lowering" value="85" />

<property name="State_Full" value="95" />

<property name="State_Max" value="100" />

[fabrizio_t 58]

fabrizio Can you tell me why a least most the time I only see green dots and no red or blue specially when that unit or group is pinned down? Seems like there morale isn't low enough when under fire... Example I used to play a old game call operation flashpoint dragon rising and one of ai I was testing for another ai maker had a ai morale system where when the ai got scared they would be running for cover and you could here in there voice of being scared. This mod reminds me of it but missing the that low moral when scared or being shot at effect. on the positive side I see some nice flanking maneuvers. would like to see snipers try for high ground and not get caught in cqb. Although I did see the sniper once in cqb pull his pistol out. Also worth noting firefights last longer another positive that shows thinking abilities and planning very nice

Are you running the mod alone?

WHITE / GREEN is common, as well as BLUE.

YELLOW / RED on heavy fire and with casualties.

PURPLE is rare.

---------- Post added at 18:10 ---------- Previous post was at 18:06 ----------

Another thing I noticed when playing as a squad leader, whenever I ordered a group of units (colured team for example) to move somewhere, they always bunched up at the destination, unless they come under fire, then they dispersed and look for cover. This is when in contact.

Thank you for the report, i've verified this.

fast movement was applied also in NON-COMBAT mode, it's a bug.

Fixed in v0.13.

If you want to fix it before next release, please change line 495 in common.sqf, from:

&& ( behaviour _unit == "COMBAT" || _mode == "LEADER PLANNED")

to

&& ( behaviour _unit == "COMBAT" )

Thanks for confirming this is deliberately disabled. I agree, definitely better to disable fatigue while testing. I'm also worried that the changes the BI devs are making to fatigue will cause some heavily loaded members of AI squads to have issues keeping formation, however that is a completely different topic.

I agree that fatigue is breaking more stuff than it fixes ...

I would like to record my findings what is a good recording software I know about fraps but you have to buy it?

I think you can run fraps for free, with some limitations.

I know also bandicam: http://www.bandicam.com/

---------- Post added at 18:31 ---------- Previous post was at 18:23 ----------

It's sunday, time for some OT.

Combat photography with bCombat (and balloons).

A tribute to the digital soldiers sacrified during testing :)

Killed by mortar fire.

Assaulting under artillery fire

Crawling under artillery fire

Careful!

Go, i'll cover (true, this time)

Broken, but kicking

At night.

Edited November 10, 2013 by fabrizio_T

[Psilocybe 3]

Time for some more sniper testing :D

[fabrizio_t 58]

Testing v0.13, with custom grenade throwing.

It's a bit clunky on animations and trajectory check is far from perfect, but still units are able to throw grenades on your foot from 10m. to about 50m.

So i had some truely brutal CQB, with grenades and suppressive fire ON: hiding behind walls is not safe anymore and i was sniped by a grenade lobbed over a roof :)

---------- Post added at 00:04 ---------- Previous post was at 00:00 ----------

Time for some more sniper testing :D

Beware, bDetect is set up for 500m. max distance and 3 sec. bullets lifetime.

Tweak accordingly and activate bCombat hearing.

Give to any "sniped" groups GUARD waypoints, so they look for you once detected.

Edited November 11, 2013 by fabrizio_T

[bravo409 13]

RainF I tried sniping with just bcombat mod and they still dont react good. I even shot some standing in front of another unit and they didn't even hide They just duck and 2 sec later back up. when reaction that should have happen is hide or run behind something. Also I tested having another sniper and directing his fire to curtain enemys and that was hard because half the time he would say engaging and not even fire or I would say fire and he wouldn't so it still has some work...But there where few times when I had him on holdfire and then told him to engage atsolider for example and he said ready to fire and then i said fire and he hit him first shot so I know it can happen.

[froggyluv 2136]

Just tested 200AI battle and the FPS drop was relatively minor on Standard settings with only a 3-5fps drop after 10 tests.

[mr_centipede 31]

Leading AI is so much fun now. You can travel in column formation, when in contact, order some unit(s) to go somewhere and they take cover once at destination and generally facing the enemy, instead of staying in column formation. Thus, reducing micromanagement. Coloured teams make sense now, to act as fireteam.

In my test patrol mission, I will always first go to safe mode, and ordered 1 fire team to go to some location and I will follow them, thus mimicking a travel overwatch pattern. If the lead fire team comes under fire, I will order all units to stop, and free my self to move around and check up with the lead fire team. If I found them not in a favourable position I will order them to move to a general location that seems to have more cover, and they will follow and take cover again as necessary. Watch command also helps. Overall, leading AI is so much fun now. Less micro, but you can micro when needed.

Good job fabrizio_T

[bravo409 13]

Question someone said to activate hearing for bcombat where is that located at in the script? sorry if dumb question....Also How do you get rid of the black box that say detected,fps,kills ext...?

Edited November 11, 2013 by bravo409

[fabrizio_t 58]

Leading AI is so much fun now. You can travel in column formation, when in contact, order some unit(s) to go somewhere and they take cover once at destination and generally facing the enemy, instead of staying in column formation. Thus, reducing micromanagement. Coloured teams make sense now, to act as fireteam.

In my test patrol mission, I will always first go to safe mode, and ordered 1 fire team to go to some location and I will follow them, thus mimicking a travel overwatch pattern. If the lead fire team comes under fire, I will order all units to stop, and free my self to move around and check up with the lead fire team. If I found them not in a favourable position I will order them to move to a general location that seems to have more cover, and they will follow and take cover again as necessary. Watch command also helps. Overall, leading AI is so much fun now. Less micro, but you can micro when needed.

Good job fabrizio_T

Good to hear.

As leader, i play quite the same way as you.

I don't micromanage units at all, except for special situations (ambushing).

While advancing, i usually keep sending couples of units to advanced spots near cover, doing some sort of commanded bounding overwatch.

These units act as scouts and take some risk, but they will switch to combat on their own if needed.

It is important not to send units much far way and sending at least couples, so they can cover each other.

I keep with me AT, autoriflemen, medics.

These are supposed to be deployed depending on situation.

When things heat up i usually escort autoriflemen to some spot suitable for suppressive fire, this allow me to relieve scouts from danger and further move on.

---------- Post added at 10:34 ---------- Previous post was at 10:31 ----------

Question someone said to activate hearing for bcombat where is that located at in the script? sorry if dumb question....Also How do you get rid of the black box that say detected,fps,kills ext...?

see my post:

http://forums.bistudio.com/showthread.php?167288-bCombat-test-candidates&p=2553802&viewfull=1#post2553802

For hearing add into custom configuration:

bcombat_allow_hearing = true;

---------- Post added at 10:40 ---------- Previous post was at 10:34 ----------

Just tested 200AI battle and the FPS drop was relatively minor on Standard settings with only a 3-5fps drop after 10 tests.

Good. What were your average FPS?

Edited November 11, 2013 by fabrizio_T

[lordprimate 159]

Ideal candidates should ideally have:

* Willingness to playtest the mod alone (eg. no other mods running) and to give feedback;

* Good experience with BIS games gameplay (being myself a player since OFP, i would appreciate long-timers);

* Prominent interest into infantry combat;

* Spare time to playtest mod features;

* A decent rig, since mod currently lacks optimization;

Just in case anyone missed it.. just wanted to make sure were all on the same page... Cough Cough

PS. FAB, i did a quick test with my 50v50 ai mission and i had horrible frames (11fps) i had to delete 20+ AI from each faction to get back to 20 fps.. that was not the case before.. That is the only thing i had time to notice, it was a quick test... Ill be doing more testing this week, as i now have time.

Edited November 11, 2013 by Lordprimate