necessary things for CMAK

[Apexicus]

I'm not going to comment the issue about tank rotation speeds, but I have to say that I have had no problems with command delays. I must admid though that I haven't played that much with conscript tank crews. I like my tankmen to actually hit their targets.

If the command delays seem too long for two additional waypoints etc. with conscript tanks, you might consider the fact that there is also communication going on inside the tank. The drive has to be told what kind of route they are going to be taking and since the (conscript) driver isn't really used to looking out those litte periscopes and the commander hasn't had any experience in giving orders it can take a long time to explain the route.

For example

Tank Commander: Drive to the left side of those trees and then take a right.

Driver: Which trees? We're in a god damn forest.

So getting the driver to do what you want him to do can be very difficult, believe me, I know. And maybe conscript drivers have difficulties getting the right gear on. I couldn't do it, I have this thing with gearboxes, they hate me!

[Redwolf]

But it wouldn't be difficult to say "drive down that road" or "drive around that house".

A good game fix (but requires some implementation work) would be to only add the delay when the direction of a curve changes. You could have as many waypoints for free as long as they all form a curve in the same direction, e.g. a right curve. The first waypoint to be left of the axis of the previous two would have the delay.

That doesn't solve the "go around the house" problem, but it allows you to do one realistic round curve. It would get rid of that overly edgy movement we need right now.

[Apexicus]

I understand what you mean Redwolf. I could argue that with inexperienced crew even those tasks would take longer than with experienced crew, but I won't since the game doesn't model those kind of things.

And that's a problem since the game isn't really a tactical level game and the player isn't supposed to "micromanage" his troops, but he sometimes has to. This makes the command delay issue more complex since sometimes the waypoints represent orders from platoon/company even batallion level and sometimes when the player has to "micromanage" the waypoints represent orders from the tank commander and given the scope of the game those really should be done by the AI. Since it would be very tough to create an AI that could move the forces in a correct tactical manner and take their experience into account, I really can't be very dissapointed with the current system.

My point is we shouldn't have to be issuing tactical level waypoints/orders, but we have to. The fixes proposed by Redwolf and others are good and I would like them to be implemented, but don't really solve the real problem. But as I said before realistic tactical AI would be very hard to code, so I really can't complain, especially, if the penalty for those tactical level commands are cut down a bit. (Can't say they've bothered me that much though.)

[Moon]

Originally posted by redwolf:

[QB] But it wouldn't be difficult to say "drive down that road" or "drive around that house".

No, but the problem exists on another level. How would a tank commander know where to drive? How would he know that he's going to get that desired flank shot if he drives around the right side of the house? The command delay simulates not only the time it takes to pass a command down the chain (be it from unit to sub-unit or even within units, like the communication between tank commander and driver), but also the time it takes to observe the battlefield, draw conclusions about the best approach, make up a plan, and so on.

The player knows many things that the individual on the battlefield wouldn't. The coordination of forces which players could achieve without a command delay system would yield a simulation with little to no resemblence of WWII combat.

Of course it's possible to spot loopholes in the system we have implemented. There are situations in which a tank commander could instantly decide the right course of action. Driving down a road might be one such example, and there are others without a doubt. The problem is that the TacAI is not human, and that the underlying game system has to account for the majority of situations and provide a credible combat simulation experience - which, in my opinion, it does in it's current form. It is far from perfect and we will continue to work on improving the system and its execution without a doubt. But if you realize what it's been designed to do (the big picture) and what it does (I am not aware of any other wargame with such a flexible TacAI), it looks quite good to me.

I have been shouting at the screen because my tank didn't do this or that when I wanted often enough. I have been frustrated when my road column got stuck at a road bend because one of the waypoints was a little too far to the side, which threw my whole plan out of the window. I have been cursing my AT gun crew because it didn't rotate fast enough to get off a few shots at a flanking armored car. But once again, these are not bugs or design flaws, if anything these incidents are features. Rest assured that as we are perfecting the various game elements, maybe come up with different, more elegant, more detailed command delay systems in the future, the perfect coordination of forces might in fact NOT become easier

Martin

[Redwolf]

Hm, what about this one:

Make the movement path plotted for turn one (and only that one) delay-free.

Background: the most annoying thing about the delay-by-waypoint feature is when you start the battle. You have a bunch of tanks on a road in your setup zone at your map edge, all open, not under fire and 15 meters from each other. Now you have to move them to contact and you just want to follow the road or go around a bunch of houses in a smooth curve.

If it takes 2-3 minutes to get them started (early Russian kids) then this goes right off your attack time. You could as well start the battle later. The alternative is to plot just as many waypoints as the tank can move the next turn and repeat this every turn. I'm not going to do that as it is an excessive waste of player time for no realism gain over giving them same path completely in one step (if anything, the latter is more realistic).

Giving the player a free shot with the initial move to contact sounds realistic. As we know the initial plan falls apart quickly and then you would delete the remaining waypoints and plot new ones (suffering the usual delay). You have no intelligence about the enemy positions at the time you plot the delay-free path, so this cannot be abused for fancy combat maneuvers.

Opinions?

[rune]

Wouldn't work. I see a player plotting moves across a 6 km map on turn one so he doesn't take a hit on waypoints later in the battle. Only if the unit was fired upon and orders over-written would it get to the command delays.

also the curve waypoints wouldn't work, as it would require each unit to have a "way point memory". The game would have to know that this unit is on a road, curving one way or another.

Still haven't found if the Panther had neutral steer. However, have been busy with other research...

Rune

[Redwolf]

Originally posted by rune:

Wouldn't work. I see a player plotting moves across a 6 km map on turn one so he doesn't take a hit on waypoints later in the battle. Only if the unit was fired upon and orders over-written would it get to the command delays.

Actually no, TacAI-given order suffer no delay.

So he can plot a path all over the map and he can stick to it as long as he doesn't cange his mind. Why is that a problem given that he will change his mind on contact?

also the curve waypoints wouldn't work, as it would require each unit to have a "way point memory". The game would have to know that this unit is on a road, curving one way or another.

Are you mixing up road move and the curves concept? In any case, the computation of dlays would be done when the path is plotted, so all the waypoints are right for for inspection when they are given, not when they are executed.

Of course it's fairly complex implementation work, I made it clear that this is probably not patchable into the CM1 engine.

[Bastables]

Originally posted by rune:

Since we are on the topic, let's talk neutral steering.

Panzer I, II, III, IV did not have it.

Panzer VI did have it

Panzer V, anyone know?

Rune

All AFV on the Panther chassie and the Tiger/Tiger II chassie are capable of "neutral" steering and therfore rotating on the spot.

Only other WWII AFV with the ability to neutral steer were the British ones Churchill, Comet and Cromwell.

Nope my mistake the French Char B had this capability as well.

I would also like to remind every one Of J.D. Salts point that all the German, British and Russian tanks that did not have "neutral" steering could through Brake-clutch transmissions merely lock up one track and pivoting on that. The Sherman had a glaring weakness in this area in that it could not do even this never mind "neutral" steer with its idiosyncratic transmission system.

[ June 19, 2003, 11:15 PM: Message edited by: Bastables ]

[flamingknives]

The 'free' waypoints on turn one would suit Soviet operation style, especially early and mid-war.

Commentaries describe the units rigidly following a plan rather than adapt to local situations.

[xerxes]

In my mind, the best idea in this thread is the lower command delay for road movement. That would be relatively simple to code and gamey abuse possibilities are pretty minimal.

I would think initial delay would be standard but additional road waypoints should only have somewhere in the range of 10%-25% of the normal "additional waypoint delay."

This would simulate an order of "Follow this road" quite well. Using a similar scheme for 1st turn movement would also be reasonable because it would simulate having an initial plan going into battle.

[ June 19, 2003, 10:04 AM: Message edited by: xerxes ]

[SFJaykey]

Well, on the subject of "Follow Road" commands, everyone seems to be focusing on the command delays. Yes that is a problem I would like to see resolved, but the worst part of it for me is simply the micro-management of placing all those waypoints on a winding road. Especially when dealing with columns of more than a couple of vehicles, it is a real PITA. Not to mention the aggravation when the timing of a column is slightly off and the dummies start running into each other...

Reducing the delay for road waypoints would help with the command delay issue, and might be a small enough tweak to make it into CMAK. But it wouldn't address the micro-management burden, nor the traffic jams. For those, "Follow Road" and/or "Convoy" commands still seem necessary. And realistic: I mean, IRL, which order would you more likely hear: "Jones: drive down that road and stop just short of the ridge," or: "Jones: drive to that bend in the road, then turn and drive to the next bend, then turn and drive to the next bend, then turn and drive to the next bend, then turn and drive to the next bend, then drive forward and stop just short of the ridge?"

[Dschugaschwili]

I think the main reason for traffic jams is that vehicles take evasive actions too often. A condition like

"If you're bumping into another vehicle and the other vehicle is moving in roughly the same direction as you, don't plot new waypoints to get around it, just slow down/stop for a moment and continue with the current movement path."

could probably improve the current situation.

Dschugaschwili

[xerxes]

The other reason for traffic jams is that people use the "fast" command while moving in convoy. Move works a lot better.

And yes, I still try and get away with using fast instead of move.

A "follow road" and/or a "follow the leader" command would be a great addition. But then again, in the desert are we really going to have that many roads? We may see a lot less of this issue simple due to the terrain of North Africa. Sicily and Italy would still have the same issues natually.

[Redwolf]

Originally posted by Dschugaschwili:

I think the main reason for traffic jams is that vehicles take evasive actions too often. A condition like

"If you're bumping into another vehicle and the other vehicle is moving in roughly the same direction as you, don't plot new waypoints to get around it, just slow down/stop for a moment and continue with the current movement path."

could probably improve the current situation.

Actually that works good in CMBB 1.03.

If you have several vehicles moving, the front one is slower and the second one bumps into it, then the 1.03 TacAI will just stop the second one for a second. And not go off-board head-over-heels.

If you hit them at angles or the front one is not moving things are less rosy, but the simple convoy thing works pretty good now.

[xerxes]

That's very good news redwolf.

[markl]

From a beginner,

How about incorporating some of the good threads into the tutorial. Most games come with good tutorials these days. Some purpose build scenarios taking you through e.g. JasonC's crossing 200m in the open, scouting and a real combined arms tutorials. This would really help the beginners. Some of the more experienced players should be contracted to do this as they have demonstrated often enough they have the skills.

And how about posting the pictures in the manual on the web for download. The pictures printed in the manual I hate to say are a waste of ink!

The future of any game lies in getting and keeping new people in the ranks. This is one area that could be improved on with little effort.

Small things like the ability to tab back to the desktop should be maintained. Something you can not do with most games these days. Makes it easy to work and play at the same time.

As well as game improvement which I can not comment on as yet, the game should be made a little easier to get into, and some people do not have a lot of time to check out the forum for all their answers.

Cheers

[SFJaykey]

Originally posted by xerxes:

.....A "follow road" and/or a "follow the leader" command would be a great addition. But then again, in the desert are we really going to have that many roads? We may see a lot less of this issue simple due to the terrain of North Africa. Sicily and Italy would still have the same issues natually.

Given the mountainous terrain involved in much of the Italian campaign, I think the issue is even more pressing for CMAK than for BB.

And additional Commands on the menu aren't really necessary.

As JasonC has suggested "Follow Road" could be applied automatically anytime two consecutive waypoints are placed on a road, or a vehicle already on a road has its next waypoint placed on the road. (The game should probably test the length of the on-road path vs the direct path from point to point, and only apply Follow Road when the difference is under a certain %, say 30-40%. That should eliminate most weird applications. Also, as has been said the computer-generated waypoints should be visible to the player during the orders phase, so inadvertent applications of the order can be corrected.)

"Convoy" could likewise be triggered automatically, when a following vehicle has its waypoint placed on the vehicle ahead of it. Long convoys could be linked up quite easily this way, with no additional commands cluttering the menu.

[JasonC]

Many points.

First the basic one about rotating. The issue is that CMBB allows a "rotate" command at the end of the movement for free, while every waypoint gives additional command delay. But rotations in place are glacially slow, while rotations on the move, faster yes, are still relatively slow.

This present a strange trade off of "delay before moving" or "delay after moving", when there will be a change of direction (which is practically always, if not moving straight at something). But delay is unavoidable.

Yes changes of direction on the move are faster already. But when I rode AFVs, they made right and left turns (90 degrees, as at intersections e.g.) in a handful of seconds. I don't mean 30, I mean more like 3. I'm not talking armored cars, but ungainly SP howitzers.

If the added delay from the extra waypoint to make the turn in motion were on the order of 5 seconds, there would be only a modest problem. But to simulate command difficulties, radio advantages, etc, CMBB quite properly imposes stiff command delays on many vehicles.

On gun movements (rather than set up), the rate does seem slow to me, particularly for the lighter guns. I'd like to see it tied to the weapon's transport class. If the lightest stuff moved almost as fast as a 81mm mortar, and only heavy stuff (e.g. transport class 8 say) moved as slow as now, that would work. Prorate the decline in speed between 3 and 8, perhaps?

Also, road movement should help more with guns than with many other things - they are on wheels, and are heavy enough that avoiding obstructions is a major speed issue. Along a road the lightest ones should manage something close to the "move" rate. This would be a strong incentive to make major moves along a road, which is realistic.

On gun set up, several issues and perceptions. First, many artillery tasks are routinely done much faster than their manual specs. The US field manuals establish training task times of 2 minutes to "march order" a particular battery, e.g., when even reserve troops I served with can do it in 30 seconds - trying, and pushing for speed, to be sure. But not "crack" gunners.

However, when moving a gun there are definitely things involved that take more than 5 seconds. Including getting the ammo operation up and running again at the new location.

I typically see 2 minute set up times for field guns in CMBB today. For the lighter ones it could easily be only half that. That figure is about right for a 105mm howitzer - with all the trail laying, sighting, and even ammo op up and running bells and whistles. A good crew could do even that faster, so 2 minutes is not unreasonable (only surveying for indirect fire reference points would take that long).

It can't take fully that long for a light direct fire piece like a 75mm infantry gun or a 37mm PAK. Again, one might set the standard at 2 minutes for high transport class (class 9 guns can be longer, as an outlier exception), 1 minute at the low end, and prorate. One might also allow ~15 second improvements or penalties for qualities better or worse than regular.

On road following algorhythms, I think it is absurd to ask the CM AI to solve any, however simple. Instead, if a follow road vehicle reaches *any* fork, it should halt (in confusion!). An additional waypoint explicitly ordered *should* be necessary for actual road forks.

Follow road should only work for stretches with only 1 route. It eliminates the need for extra waypoint delays to follow a curvy road, nothing else. The CM AI should not be a GPS. But it also shouldn't take a buttoned green T-34 2 minutes to figure out how to travel 500m down Route 324, just because it bends 2-3 times.