Just a spectator?

[Fred]

Los,

thank you for your posting

---

"But there are plenty of "non-soldier" gamers on the beta team and now with the demo and their opinion is arguably more important since this is a game after all not a west point training tool and their considerations have to be taken into effect."

-----

You are exactly "on target" with your posting, seeing the pros and cons of both sides. Totally agree.

Fred

[Moon]

Los - a short sidenote to your point 1. I do have the feeling, after playing some PBEM games against "fresh meat" that many people in CM try to do too much in one turn, e.g. order suppression fire and an assault at the same time, in the same turn. That's why the TacAI might bother them when it suddenly retargets. When I am about to assault, I usually take at least one full turn of suppressive fire before assault, since I want to see some results first before leaving cover. Maybe that's one of the reasons?

[Fionn]

They're definitely trying to do too much...

I won't embarass anyone by naming names but ANYONE who tries to suppress three squads, fire at two tanks AND move a platoon forward into new positions (all while being outnumbered two to one) needs to reconsider how much his men are capable of

------------------

___________

Fionn Kelly

Manager of Historical Research,

The Gamers Net - Gaming for Gamers

[Guest Mikeman]

After some 12-15 hours of play I would agree that the Tac AI seems a wee bit too anxious to disregard my targetting orders. When I was in the Army we were trained to follow orders without question. We were always told not to think since we never knew what the greater strategic plan was. We were taught that following orders would keep us alive. We were to trust our leaders. In combat there are only a few good reasons for a soldier not to obey TARGETING orders that I can see:

1) The soldier is completely frozen with fear and panic or running away from battle for the same reasons. Unless the man is a coward he will recover from this panic within a few minutes.

2) An unexpected IMMEDIATE threat to the soldier's life pops up. I'm talking close range here! (Five men at 50 meters charging straight at the soldier's position.) In this case I'm sure those giving the orders would change them if time permitted and they were aware of the soldiers personal situation.

3) The soldier has lost all faith in his commanders' abilities and judgement based on previous experience with them in firefight situations. I would rather not see this sad situation simulated anyway.

4) The soldier is heavily pinned down making it suicidal to raise up and concentrate fire on a specific area or target. In this case the soldier would not target anything at all.

5) The soldier misunderstands the targeting order, or does not receive the order in time (command & control).

"Plaster those guys comin' out of the treeline to the west! NOW! You hear me! Very important. Don't let anything get by ya." "But sir, there's millions of 'em crossing the stream over by the church! They're gonna waste 3rd squad!" "Don't you worry about 3rd squad son, I'll take care of 'em. Now do what I tell you before those guys make it to the cover of the wall."

If I were that soldier I'd be shooting up the guys coming out of the treeline, as ordered.

More weight should be given to the player's orders. Maybe the way the game stands now is more realistic, but it would be more fun if I were allowed a bit more latitude with targeting. It seems that I make very few correct targeting choices as far as my men are concerned. I check LOS and range, etc..

I don't feel my targeting orders are obviously stupid, so why do my men? Even if my targeting orders are not the best I would still expect my men to obey them unless one or more of the five above conditions are present.

My impression of reality in a firefight may not be correct since I've never experienced combat. However, my impression of reality (and perceived reality is what I want) makes me think my troops are doing a little bit too much independent thinking. Make the Tac AI treat my decisions with more respect. I don't mind paying for poor decisions. (My real life counterparts certainly had to pay). I'm not asking for automatons, just more disciplined troops who have some faith in their commander.

I find it very interesting that no beta testers have brought this issue up. Could it be, like Moon said, a bug in the demo? Maybe the game just does not handle a poor player very well, and chooses to play itself. By this I mean that perhaps one needs to be very proficient at tactics to derive the most enjoyment from the game because of the way the Tac AI treats player decisions. Fionn says 90% of his targeting orders are obeyed. I would like to experience that too. Do I need practice or a less defiant AI? Probably both. Still a great game.

Mikeman out.

[Kraut]

I'm with fred all the way on this one. I want gameplay, i want to make mistakes and see which orders don't work and which ones do. I do not want the AI to take over most of the time. Now, granted, you have no control for 60 seconds, and the AI has to make some decisions for you, but in my view it is making some very poor choices. Poor chaoices in relation to MY overall plan and MY style of play.

Yesterday i played the last defense battle again, using a lot of the ambush commands. I must say it did work a lot better. However, it still feel this is a poor work-around.

MK

[Guest L Tankersley]

Walking the tightrope between gameplay and realism is a tough chore. I fully understand the frustration several people cite when their targetting orders are ignored. But I think CM's gaming model is much closer to "player as company commander" than "player as individual squad leader." If you went all the way down that path, you would issue all your orders and assign objectives to your subordinate units before the battle started, and then just watch the battle unfold, perhaps issuing a few corrective orders to units in your immediate vicinity. Quite realistic and challenging, but not all that much fun.

To leaven the realism with a dose of fun gameplay, I think the CM designers have allowed the player to step down into the shoes of the platoon leaders a bit as well. You can order your squads to move to a certain position, but if they start taking fire they might think better of it. You can direct them to shoot at one target, but they may decide otherwise. I don't think the targeting order is meant to simulate an explicit "order from the top," but rather just a way to allow the player to convey a general intent/objective without having to develop a complete op order/frag order. The men aren't disobeying an explicit order, rather it's that Sarge and the boys weren't briefed on the intricacies of the captain's defense plan; all they know is that the lieutenant sent them out to this building and now that the Germans are advancing, it makes sense to shoot those guys there and then skedadle before the 88s start landing nearby.

That being said, I don't think it would be a bad thing from a gameplay standpoint to increase the persistance of a player's targeting orders for units that are "in command" - not strictly realistic, but it _would_ allow players a bit more influence and permit them to make their own mistakes. I expect that this effect is already in the code somewhere; maybe it should be strengthened a tad.

Leland J. Tankersley

[Guest Big Time Software]

OK, time for us to weigh in. Thanks to various Beta testers holding the fort while we were busy doing other things

There seems to be two concerns here:

1. That the TacAI is screwing up and missing more critical target choices.

2. That the TacAI is too easily shifting fire.

In the answer to #1, although the TacAI is not perfect, we have not seen horrendous errors that can not be explained through a deeper understanding of how the system works. You guys are at a disadvantage here because the documentation you have is minimal and you are just starting to play the game. Our testers that have been playing it for much longer, and with access to more info, have not reported even one such error. Opening up too soon in x situation, yes (and we are constantly tweaking to eliminate these cases). But not "my guy ignored, for no reason, a lethal and obvious threat". So not having seen the exact situations you speak of, I have to firmly believe that you missed something that caused what you saw to happen. If the TacAI was making mistakes that were so easy to find, then we would ALL be noticing these problems. At least that is the much more likely case.

Now, in terms of the second point about the TacAI being a little too eager to switch targets... we have not found this to be the case, BUT, we have just tweaked the threshold values so that the TacAI will keep your orders a weebit more close to heart. This is more along the lines of argument put forward by Fred. It won't make them robots, but might make them a little less likely to switch.

Kraut, your comment about the Bazooka is partly based on your lack of understanding about what the "20%" represents. The % to hit factors in things like the weapon's accuracy, range, the speed of the target, etc. 20% for a Bazooka is actually quite good. You will rarely see better chances in a dynamic situation in CM. If you snuck up on the rear of a tank at 50m while it was sitting still... then you might see something more like 70%. But I very much doubt you will see something more than that. Main reason is that the Bazooka has an inherent inaccuracy that will never allow that number to get to 100%.

Steve

[Guest Mikeman]

Moon and Fionn,

Your latest posts, which went up while I was writing mine, may be the answer. It's possible I've just been trying to accomplish too much at once. I never realized it. This fact may explain everything!

For the sake of sales though, I think mandatory obedience should be able to be turned on or off. Call it a "realism toggle" and explain what it does. A wider cross section of the public would enjoy their first impressions of the game, stick with it, and say good things about it. It appears that many of us have had 'problems' with the Tac AI. How many will decide it's no fun because of this?

Mikeman out.

[Ben Galanti]

Leland brought up an interesting point, which I don't know if it's modeled (or should be modeled for that matter) That troops that are out of command and control will be more likely to break your targeting orders then those inside command and control. If it's already in, that's quite cool If not it makes sense to me at least. If I have some lt. shouting in my ear to target a particular bunch of guys, I'd be pretty likely to do it. On the other hand, if I was an MG team sitting off by myself, I'd probably be much mroe likely to target what I thought was important....

[Los]

"many people in CM try to do too much in one turn, e.g. order suppression fire and an assault at the same time, in the same turn. That's why the TacAI might bother them when it suddenly retargets. When I am about to

assault, I usually take at least one full turn of suppressive fire before assault, since I want to see some results first before leaving cover. Maybe that's one of the reasons?"

It could very well be I suppose. I fall into the camp of not trying to do too much at once, Simplicity and The Objective are two prinicples that serve the gamer well in CM. In a sample attack scenarios I find normally these general phases: (Note YMMV)

1. Make a tentative plan and set my movement orders as such. If everything (by some freak of nature) goes according to plan, I won't have to do very much clicking for a number of turns. Unless the scenario specifies elsewise it's best not to spread everyone out all over the place to turn over every possible rock. This dissapates effort and resources.

2. Movement to contact and trying to ascertain where's the enemy, (weaknesses and strengths). Confirming or denying assumptions made during the planning phase. But again refer to One, you don't need to recon the whole map ala Red Alert or some RTS game looking for resources. Support your plan with your recon that's all.

3. Manouever forces into their support/ overwatch positions while the assault elements move to their best point to "cross that last 100 yards" if there's isn't concealment all the way to the target. (VERY IMPORTANT NOTE: The assault elements should be kept out of harm's way as much as possible during the movement into positions. They will get hammered soon enough when the time comes so the fresher, more motivated and more intact they are when the time comes, the better.)

4. Suppress: Determine which targets are the ones that merit the most attention. I'm not going to try and blast every available enemy target on the board. That dissapates the effectiveness of your massed fires. Lets say there are four enemy units (that I know of) which can possibley see my assault force as I cross open ground, but one or two are 6-700 meters away. Well I'll concentrate most of my fire at those which can bring effective fire. Note also if you have arty or bn mortars then use the smoke that's what it's there for. If you use those valuable fires to wearly they won't be availabe when you need them the most.(Side note: Ammo consumption and concerns about it are one of the many things I love about CM. Keep a close eye on critical weapons such as mortars and bazookas etc. which have only a few rounds. Wasting fire early in the game (trying to do too much) only bites you in the ass later. To this end choose manuevering around enemy fire early instead of suppressing it (If possible).

5. Assault: AFter a few turns or at least one full turn of fire RUN their asses up to whatever cover is available near the OBJ (Or right at it if it feels right). provided that you are supressing the enemy properly they will make it in a good enough state to get the job done.! BTW save those 50-60mm compnay mortars for that last whirlwind of firepower on the OBJ before they jump off. (NOTE: From a defender's point of view you cans ee why it is important to keep some key weapons/forces hidden if possible, even if it means out of LOS until late in the game AND to keep a reserve to counterattack the Obj if necessary.)

6. Consolidate: Once at the OBJ (if you take it) consolidate and reorg your D. Be ready for counterattack. Bring up other forces if you can.

Point is, try to do less, have a single objective (this doesn't mean a single victory location, but one goal) and conserve the strength and ammunition for when you will need it the most.

And of course the standard caveat: these are general guidelines which can be what-iffed in any given situation, they should be food for thought.

Cheers...

Los

[Kraut]

Steve, i don't know, i always feel we are talking past each other ... no side seems to get their poit accross here. It's crazy ... lol

I'm sure that 20% for a zook is good, i believe you. BUT, that's not the point i was trying to make. I was trying to explain that *I*, the commander, set up and ambush to be set at a certain time. Specifically at the time when the stuG was *at least* under 100 meters away (it was coming directly at the zooka's position). I would have liked to wait until about 60-80 meters.

The tacAI decided to disregard this order which could have been very bad for my total plan, which again annoys me A LOT.

btw. ALL my men were in range of a platoon commander.

Please condiser tweaking the hide command a bit more, especially with AT teams.

Ohhh, also, what do you think about the case where my Panzerschreck fired at a sherman at 200m ? Damnit ! I didn't want him to fire at that range ... grrrrr ...

MK

[Archer]

All right - here's what I think about the targetting deal. First off, let me point out that I'm probably the world's worst wargamer. I'm horrible. I'm really, really horrible. I jump the gun, I have bad planning skills, I really and truly suck. I always have, and I don't expect it will ever change.

When I first started playing the CM Demo, the AI retargeted pretty much every single thing I did. There wasn't a single command I'd give that wouldn't be countermanded somewhere along the way, usually less than 20 seconds into a phase. It was horrible.

But the thing is, I actually started to WIN. Or at least, come close to it. In all my previous computer wargaming experience (mostly Close Combat), I never learned. I'm an anxious player, and I *do* try to do too many things at once, every time. I give stupid orders. Really, really stupid orders. If the AI were more heavily weighted to paying attention to my targets, I'd never win, or even be able to walk away from the field without bleeding from a thousand holes. Now, though, I just don't bother targeting at all unless I'm REALLY REALLY interested and SURE that what I want to target is a good and valuable target. Otherwise, I pretty much let the AI do what it wants - because it's SO MUCH BETTER THAN I AM at this sort of thing.

I would be in favor of a kind of sliding scale, perhaps, where on one end you have something close to mandatory targeting (or at least VERY heavily weighted toward the player) and on the other hand you have the AI as it stands now - where it makes up for my boneheaded targeting mistakes.

I will admit to not usually understanding why the AI retargets the way it does, but I'm sure that's just because my "situational awareness" is poor. If I understood what was going on, maybe I'd do a better job and the AI wouldn't retarget so much. In fact, I'd like to see how thrashed I'd get if the AI *did* listen to me...

[StuNZ]

Just to throw my 2 cents in on this, I've had no problems with the AI retargetting - admittedly I've only played the last defence so I don't know if the other scenario is worse for it retargetting, but as the americans in last defence, I've had no problems with it - if I told my MG's to fire on the german armour to button up, they did ...

Who knows, maybe I'm just lucky

[Guest Big Time Software]

Hi Kraut,

Yup, understand you here. The problem is obviously one of degrees. The "Hide" command is general and therefore won't work the way you want it to in every situation. That doesn't mean it is perfect as is, but short of some terrible micromanagement system it never will be. There *might* need to be a tweak or two made here and there, but everybody has to be VERY careful of drawing too many conclusions from one particular instance. We need to tweak for general trends, not specific cases. Schrecks firing at 200m is a concern, but if it happens once (in a situation I can not see in detail) that isn't enough for us to tweak a value. This could be robbing Peter to pay Paul. You say it is too far, we tweak, then someone says it is too short, we tweak, then you say... aw, you get the picture Again, I am not saying that tweaks aren't needed here, but please understand that we can't go jumping in and twisting numbers based on one instance. Just to be clear too, there is no über value to tweak. So what an MMG or Bazooka does is not related.

Also note that Hiding units are less likely to stay hiding when they are fired upon.

Los, thanks for that post! Couldn't agree more! Archer, eventually you will be able to learn more about why the guys are retargeting. Takes a bit of time, and perhaps the full manual, but you will!

Steve

[Fionn]

MK,

Some of your last posts have shed a lot of light on this.

1. "Yesterday i played the last defense battle again, using a lot of the ambush commands. I must say it did work a lot better. However, it still feel this is a poor work-around. "

Work-around? MK you simply aren't getting the orders IMO. The ambush command is NOT a work-around it is a vitally important order.

Units ordered to "hide" will basically hide until they get a nice target (20% or so lets say) and then pop up and fire..

What you SHOULD have done is ordered the zook to ambush by setting an ambush market at 50metres from the zook and had the zook target the ambush point.

It's not that the AI is disobeying your orders its that you are giving the "wrong" orders.

I'm working on putting up some AARs and movies which will show how to do this correctly via some PBEM games I'm playing and I hope that this should really help clear up the mistakes you and others are making with ordering.

------------------

___________

Fionn Kelly

Manager of Historical Research,

The Gamers Net - Gaming for Gamers

[Guest Mikeman]

I've come to the conclusion after reading all these posts that this targeting thing is mostly a learning curve issue as are other recent topics posted to the board. Like good flight sims, this game will take time to learn, and be more enjoyable with experience. Chess is the same way. The better you get the more you enjoy the game. I do look forward to a slightly more agreeable Tac AI though. (realistic or not) Like the new intermediate Fog of War setting Fionn just posted about, it will make the game more enjoyable for newbies like me. Any realism feature which can be turned on or off is good for newbies and yet doesn't detract from the hard core realistic game experience desired by the super grognards. Since the tweak BTS is doing concerning targeting is not an on/off deal I say don't tweak it too hard! Just a little bit. I'll be wanting max realism soon enough.

Mikeman out.

[Guest L Tankersley]

One other thought on the subject of targetting orders, inspired by some of the discussion of suppressing targets before an assault. First, note that instead of utilizing one or more full turns of preparatory fire before launching your assault on a turn boundary, once you have a feel for the effectiveness of different concentrations of fire you can time assaults to begin _during_ the turn by judicious use of the 'pause' command. In fact, you can have a unit fire for the first half of the turn and then start moving forward by assigning it a target, pausing twice and then issuing a move order. (Necessary pause times vary according to in-command status, of course.)

Doing this can let you turn inside your opponent's decision cycle, a Good Thing . But consider how much more (unrealistically) effective it would be if units would not ignore the player's targeting orders when vulnerable targets appear. Say you order your MG to fire on a squad in the tree line in front of you to suppress it. All well and good, but then another squad next to it starts crossing the open ground. Wouldn't you WANT your MG to switch targets? I sure would!

The effect of this is to really reduce the impact of "turns" as much as possible, which I think is a good thing since they are an entirely artificial construct used to regulate the gameplay.

Leland J. Tankersley

[Guest Mikeman]

The 15 second order pause key! I forgot all about that. Thanks Leland!

Mikeman out.

[Dale H]

Pardon me, but am I missing something: what about an infantry squad 40 meters from my structure doesn't make them THE imminent threat (considering there are no others closer or there is a farther threat that could blow the structure up within the next minute)? I'm starting to get that little crinkly feeling on the back of my neck even thinking about this kind of situation as I write. Seems to me the only thing to do is to open up on the advancing US squad. I happen to think Fred was right to raise this question & you must explain to me how the program intelligence would ignore this threat in terms a simple VolksGrenadier would understand. And don't talk to me about work-arounds. Either the programmed intelligence can handle this situation - or it can't and I think we need to know the answer, plain & simple. Sorry to be blunt about it, but I ain't no grognard and all these wordy explanations leave me scratching my head.

Dale

[Fred]

Steve,

----

"...we have just tweaked the threshold values so that the TacAI will keep your orders a weebit more close to heart. This is more along the lines of argument put forward by Fred. It won't make them robots, but might make them a little less likely to switch."

-----

once again impressed by your responsiveness! I really think that tweaking this will make the game even better in gameplay value and will rise the ratio of player input to game output(and I'am sure a lot of others have the same opinion).

Thank you

Fred

[Dreghorn2]

L Tankersley, interesting point regarding the pause button.

In the situation described (targeting.. two 15 second pauses.. then an advance), does the unit continue to fire while under the pause order? Does the pause order only extend to movement, or firing, or both? Important to know, because you are right this can be a very potent tactical tool.

------------------

'A mans greatest pleasure is to crush his enemies, sweep them before him, to take from them that which they possess, to see their women and children in tears'

[Guest L Tankersley]

<BLOCKQUOTE>quote:</font><HR>In the situation described (targeting.. two 15 second pauses.. then an advance), does the unit continue to fire while under the pause order? Does the pause order only extend to movement, or firing, or both?<HR></BLOCKQUOTE>

I am virtually certain you continue to fire while "paused." I know that vehicles do. After all, when you issue movement orders to guys and they don't react for a while, they _do_ shoot in the interim. The slightly gamey aspect of this is that you can look at the unit's command delay after you issue an order and use that to tweak the number of pauses you need to issue, taking advantage of the C&C delay to actually improve your coordination somewhat.

I think I'll go verify that right now, actually!

Hmmmmm....just had an idea - BTS, are targetting orders affected by the C&C delay? I wonder if this might be causing some of the targeting woes some people are having - they tell their guys to target one unit, but it takes a while for the order to percolate through and in the interim they're still shooting up their preferred target. Just a thought...

Leland J. Tankersley

[Kraut]

Steve said, while scratching his head and wondering if this will ever end:

"Schrecks firing at 200m is a concern, but if it happens once (in a situation I can not see in detail) that isn't enough for us to tweak a value. This could be robbing Peter to pay Paul. You say it is too far, we tweak, then someone says it is too short, we tweak, then you say... aw, you get the picture Again, I am not saying that tweaks aren't needed here, but please understand that we can't go jumping in and twisting numbers based on one instance. Just to be clear too, there is no über value to tweak. So what an MMG or Bazooka does is not related."

Well, then i would say you'll have to hand over a few more scenarios for us to "test". lol

Just a serious side-note to all this "value-tweaking" business. When we were testing cc3 for MS, there wasn't enough time for us to experience the game, to say "hey, the infantry suck ... needs tweaking". Sure, there were voices raised about this issue and other issues, but we didn't have enough game-time and enough variety to say with definity: "THE INFANTRY STINK !"

I hope that your "testers" are enough to eliminate balance-problem. Was the retargeting issue raised by them, or did it only appear after the public demo ?

I think from this group, which has DLed the beta-demo, you're getting good feedback on issues, so releasing a few more scenarios wouldn't be too bad for ya Preferably some with JT's and other heavy guns, please *cough* ... just to see how they *cough* stack up.

Fionn said:

"It's not that the AI is disobeying your orders its that you are giving the "wrong" orders"

Great, blame it on the customers ... now where did i hear this type of talk before ? lol

Man, reading your responses is always painful.

Remember, you have a genuine cc-player here, who has only played SP and EF and a few demos of other hex-based wargames before. This player wants control ! This player is GOD of the battlefield ... he needs that fix or he'll die of ego-spatistic-hyper-spasm.

I gotta check out the ambush some more. Maybe after that, i'll come back with more ammunition to fire at you, or i'll be converted and join you in the holy crusade, to enlighten the world with CM. ROFL

A few new scenarios could aid me in my quest for wisdom you know ...

MK

[Renaud]

Interesting debate!

My .02 cents says that, although the retargeting has irked me at times, it contributes to the feeling of being in a real-life unpredictable combat situation. That's what I like most about this game, the feeling of witnessing a WWII battle without having to really be there. I can hear the rebuttals already, 'well go and rent a movie then!'

A small suggestion: Why not tweak the target-switching tendency down for units within LOS and C&C of a command unit. This would be a great simulation of a commanders ability to more closely direct the fight when he is within screaming distance of his troops. Sure he might get them killed with his annoying butterbar micromanagement but then 90-day-wonders were a reality of WWII on all sides. This tweak would also stress the importance of C&C which is always a good thing in my book.

-Ren

[Guest JonS]

Going back up the thread a bit someone mentioned that the turns are 60 seconds long (I know, I know, it's obvious - but wait there's more). And the TacAI takes over when it thinks it sees something beter and will retarget. Well, at the start of the turn the Human has a very good idea of what he wants, but can't predict what the situation will be in 60s, so how about using a sliding scale of 'stickyness' for orders, so that at the start of a turn a unit will obey the order most (~85%?) of the time, at 30 seconds it will be down to some of the time (~60?), and near the turn end the unit will givce it a shot but don't put any money on it (~45%). I guess this might be hard to code with out getting a resolution that was too chunky, but having a 'stickyness' variable with 60 possible values, tied to the current second, would I think do the trick.

Now I'll probably get both sides flamming me. Hehehe

JonS

------------------

Quo Fas et Vino du Femme