Problem with victory condition

[jep]

Yesterday I played a Quick Battle and began to think something is definitely wrong with victory condition mechanics. Basically, I failed to earn vitory points just because single broken soldier. Problem was emphasized because of the sheer size of the target area.

[Michael Emrys]

This kind of thing has come up before. And looking at some of the QB maps it seems that often the area that has to be controlled is huge. Unless the attacker has adequate forces (mostly infantry) at his disposal and enough time to thoroughly sweep the area, he will almost inevitably not get points for controlling the location. Part of the problem is the all or nothing way that the distribution of points is determined. If a player got some points for having the dominant force on the location, then he would not be totally denied by a single soldier.

Michael

[womble]

There are lots of maps with very large VLs. There are further maps with multiple adjacent, indistinguishable large VLs. Neither makes getting the VPs for the location any easier. "Occupy" is what it is, though, and shouldn't be a surprise to anyone.

[jep]

Kind of kills the more competitive plays. Perhaps this will be fixed in future releases.

[Nelson 1812]

Sometimes it is worth reducing this area in the editor, as long as your opponent agrees.

Playing a large map with medium forces is a good example.

[womble]

Kind of kills the more competitive plays. Perhaps this will be fixed in future releases.

If you're being "more competetive", it's probably worth agreeing the field of battle with your opponent beforehand, so you know the challenges you're facing.

[sburke]

If you're being "more competetive", it's probably worth agreeing the field of battle with your opponent beforehand, so you know the challenges you're facing.

true. Decreasing the vl favors the attacker. Increasing it favors the defender. For a QB versus the AI you generally would want to give the AI some leeway. It isn't going to shift forces knowing it's hold on the VL is tenuous.

[womble]

In a QB vs the AI it's generally not a problem since you'll be forcing a surrender anyway.

[MarkEzra]

I don't see this as a V Obj size but as a win/loss calculation. Consider this: Even a very small V Obj will have many places to hide a cowering enemy.

Now Let me discuss my evolving thinking about V Obj.

Since CMSF and all through CMBN I have generally kept V Obj to a minimum number and a maximum size. I used this thinking to help the AI hold and have a chance of reaching the Obj during counter-attack. CMRT has a large number of QB maps designed in this manner. But I started trying something different near the end of Beta Testing which may be going un-noticed. It's the future.

What is coming in the next patch and later modules is something entirely different.

The maps I am doing at this moment are being designed to use smaller, multiple victory locations and are assigned different values. I believe the improved AI allows for more "group" choices. So instead of just 2 or 3 groups for the AI the numbers are more likely to be 4-8 groups. This of course is based on map size. It doesn't mean that all maps will be multiple V Obj. But it does mean fewer single V Obj than ever before.

[Michael Emrys]

The maps I am doing at this moment are being designed to use smaller, multiple victory locations and are assigned different values.

I think I'm gonna like this. It definitely sounds interesting. This would seem to offer a greater latitude for alternate strategies.

Michael

[YankeeDog]

Yes; this has come up before and I think the problem is not so much with how terrain objective are modeled, but in how they are sometimes used. Unless the goal is to represent a "Sweep & Clear" operation, multiple smaller Occupy VLs is usually better than one big one as this eliminates the issue of a small, broken infantry unit being able to hide out somewhere in the VL to deny points, and also allows for more nuanced scoring, as the attacker will get partial points for occupying some, but not all of the objectives.

[MarkEzra]

...Unless the goal is to represent a "Sweep & Clear" operation, ...

The goal of EVERY QB Map is to "Occupy" the V-Obj. That's because of the Games design, not the map design.

[Michael Emrys]

Is it possible to have touch objectives in a QB? If not, I assume there is a good reason but I am presently unaware of it. Would you care to shed a light in this dark corner, Mark?

Michael

[Volksgrenadier]

The problem is that the victory conditions are only extremely primitively determined in a binary manner.

A game with such a high pretension should calculate correctly how good and how strongly a victory location is controlled by either side.

[MarkEzra]

It is not possible (and never has been) Let's say I want to convert a scen map that has only Touch and Exit Obj. So I copy it to my QB Map file, fire it up and voila!.... all magically become "Occupy". The QB Battle generator cannot recognize another type.

In a QB Perfect World the QB Battle generator would allow us to customize the setup well beyond Assault, Attack and Probe.... BUT.... I cannot imagine the coding difficulty to get that to work... oh, and the AI Group plan layers involved to make it so... Now you've gone and given me a headache... I'm gonna go watch re-runs of Veronica Mars and let her ease my pain.

[womble]

The problem is that the victory conditions are only extremely primitively determined in a binary manner.

A game with such a high pretension should calculate correctly how good and how strongly a victory location is controlled by either side.

Or it should state the parameters of the victory condition clearly in the manual. Which it does.

[Volksgrenadier]

How could I forget! The primitive binary mechanism is explained in the manual, so everything is ok.

[zmoney]

How could I forget! The primitive binary mechanism is explained in the manual, so everything is ok.

So, by binary do you mean your troops are on the objective and no enemy troops are on the objective capable of resistance? If so how is this not adequate as a victory condition?

[womble]

How could I forget! The primitive binary mechanism is explained in the manual, so everything is ok.

You think it's primitive. Whoop de doo. It just is. The binary VC has to exist to allow "clearing" a VC to be an absolute goal. Perhaps there should be a different VC mechanism for where you're only concerned with a "relatively clear" VC, but that's an argument you aren't making because it doesn't suit your bashing agenda.

VCs do need some work. The interaction of Exit and Destroy VCs makes "scoring" some types of scenarios difficult. Condition and Casualty VCs being threshold makes for awkward employment. But they are what they are, and it's not because The Brain in a Jar couldn't code them differently so they're sliding scales, or that calculating them would eat all the processor cycles. They're design decisions which you might not agree with. But then you have some funny ideas in general, so that's no surprise.

[YankeeDog]

The goal of EVERY QB Map is to "Occupy" the V-Obj. That's because of the Games design, not the map design.

Yes; my point was that large Occupy objectives obligate the player to clear every single action spot of even a token enemy presence, which is more representative with a "Sweep and Clear" operation. If a whole town is "Painted" as an objective, then the player has to clear every single building of even a token enemy presence to get the Occupy Objective points. This would imply that the attacker's goal is to gain *total* control of the town, to the point where rear-area units could be brought up and the town could e.g., be used as a staging area.

In contrast, if only the 4 largest, heaviest buildings in the down are painted as separate small Occupy Objectives, this compels the attacker to gain substantial control of the town in order to put men into these buildings, but does not obligate him to clear out every sniper and vehicle crew hiding in the rubble. This also allows the attacker to gain points for achieving some of the objectives, but not all.

I would argue that the latter is usually more consistent with the kind of combat CMBN tries to represent. Not always, but usually.

[Volksgrenadier]

Perhaps there should be a different VC mechanism for where you're only concerned with a "relatively clear" VC, but that's an argument you aren't making because it doesn't suit your bashing agenda.

And I thought I cricisized that it was only a binary decision algo and no calculation taking place how well a VC is possessed and how well it is tactically controlled.

So much about your agenda...

[MarkEzra]

...In contrast, if only the 4 largest, heaviest buildings in the down are painted as separate small Occupy Objectives, this compels the attacker to gain substantial control of the town in order to put men into these buildings, but does not obligate him to clear out every sniper and vehicle crew hiding in the rubble. This also allows the attacker to gain points for achieving some of the objectives, but not all.

I would argue that the latter is usually more consistent with the kind of combat CMBN tries to represent. Not always, but usually.

This is an excellent idea! For a scen game.

I am compelled to make QB Maps that account for EVERY conceivable Unit choice players wish. The single platoon of King Tigers the player bought may have difficulty clearing out that top floor of your map. Look, it's not a total disagreement I have with your idea regarding the size of "Occupy" space. You will see QB Town and village maps in CMRT that use only the town edges as V Obj. So what you want is Already there.

My Concern in map design is how well the AI can function on it, not how easy it should be for the player to win. That is always my driving vision. I do not think my emphasis is mis-placed.

Speaking of mis-placed.

QB Battle Generator is the Sand Box for CM play. It's fast, simple, and fun.

Scenario's and Campaign offer the more precise "binary whatever you all are going on at" code. They provide a riveting storyline, finally honed TOE's and the maximum use of the AI pathing tricks (many of which just Cannot be applied successfully to a QB style map).

Attempting to make one into the other is a common wish for us all. And we all wish for more (and I think we have and will get more). I just need to keep my eye on the ball, and never forget that I design for the "game" end of a CM wargame.

[Pelican Pal]

You think it's primitive. Whoop de doo. It just is. The binary VC has to exist to allow "clearing" a VC to be an absolute goal. Perhaps there should be a different VC mechanism for where you're only concerned with a "relatively clear" VC, but that's an argument you aren't making because it doesn't suit your bashing agenda.

Womble, what are you on about? A bashing agenda?

The occupy VL has long been a problem. The current setup is poorly designed and rarely reflective of the mission the player is actually tasked with. It really is a primitive system and an especially poor design choice given the context of the game and the times in which they are used.

Now it isn't game breaking by any means. I mean we've lived with it since CM:SF, but it definitely isn't the occupy that this game really needs.

I mean look at the campaigns. A number of missions in the Soviet campaign consist of a battalion supported by attached armor attacking occupy objectives. I think it is pretty clear that a broken crewmen with a pistol hiding in a house wouldn't deny that objective to those men.

And the German campaign has missions where you are commanding 250+ men and 20+ tanks.

These are not situations where having a single man, regardless of his morale status, in the VL should deny the objective to the attacking force. These also happen to be the vast majority of the missions in the CM games.

Now at platoon sized mission or counter-insurgency Op like those we saw in CM:SF might justify a clear objective but most CM missions do not.

[YankeeDog]

@MarkEzra

Good points. I must admit I don't play QBs much except as "test drives" to try out units and formations I'm not very familiar with, but I've played them enough to understand that since QB VLs & AI plans must be "one size fits all", the construction requirements are very different than for a designed scenario.

In general, I still think more smaller objectives is a better way to go than fewer large objectives, but obviously there are limits to what can be done here when it has to work for a wide range of possible force sizes and dispositions.

[Lethaface]

Womble, what are you on about? A bashing agenda?

The occupy VL has long been a problem. The current setup is poorly designed and rarely reflective of the mission the player is actually tasked with. It really is a primitive system and an especially poor design choice given the context of the game and the times in which they are used.

Now it isn't game breaking by any means. I mean we've lived with it since CM:SF, but it definitely isn't the occupy that this game really needs.

I mean look at the campaigns. A number of missions in the Soviet campaign consist of a battalion supported by attached armor attacking occupy objectives. I think it is pretty clear that a broken crewmen with a pistol hiding in a house wouldn't deny that objective to those men.

And the German campaign has missions where you are commanding 250+ men and 20+ tanks.

These are not situations where having a single man, regardless of his morale status, in the VL should deny the objective to the attacking force. These also happen to be the vast majority of the missions in the CM games.

Now at platoon sized mission or counter-insurgency Op like those we saw in CM:SF might justify a clear objective but most CM missions do not.

This is not something Mark can change I guess, but I agree on the functionality issue regarding 1 broken reserve crew man denying victory to the battalion outside the shed he is hiding. However, I'm not stricken by it often as the 'Surrender' event usually kicks in in those type of battles.