Optimizing Elmer's Current Capabilities (Full Version)

All Forums >> [Current Games From Matrix.] >> [World War II] >> Norm Koger's The Operational Art Of War III >> Scenario Design



Message


Dabbs -> Optimizing Elmer's Current Capabilities (11/16/2006 7:25:49 PM)

I'm making a list...and checking it twice, to get Elmer to make players pay the price... Okay, well - that doesn't make a very good Christmas Carol...ha...

What are the tricks for making Elmer most competitive?

Thus far, my list is as follows (in no special order)

1. Having more numerous smaller formations vs. fewer larger formations;
2. Provide formations tight "area based objectives" with logical numerical sequences vs. strictly linear objectives.
3. Give formations staggered and overlapping objectives and orders as suits their size and nature.
4. Unless there's a good reason otherwise, assign all formation's loss tolerance to Ignore Losses.
5. Assign one or more small formations to contingency coverage against Player "special ops" and/or intrusion into rear area security.
6. Having rationale and coherence to the entirety of Elmer's objectives and orders - in effect, the scenario designer is laying out an entire battle plan spanning the full course of the scenario vs. going on a turn by turn basis.
7. Make use of the objective tracks for a better defined defensive and offensive disposition; use the third track to augment and enhance the side's primary disposition - offensive or defensive - with some unexpected trajectories.
8. Attempt to make use of natural terrain features in conjunction with defining formation objectives.

Anyone have additional tips?





sstevens06 -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 1:48:26 AM)


quote:

ORIGINAL: Dabbs

I'm making a list...and checking it twice, to get Elmer to make players pay the price... Okay, well - that doesn't make a very good Christmas Carol...ha...

What are the tricks for making Elmer most competitive?

Thus far, my list is as follows (in no special order)

1. Having more numerous smaller formations vs. fewer larger formations;
2. Provide formations tight "area based objectives" with logical numerical sequences vs. strictly linear objectives.
3. Give formations staggered and overlapping objectives and orders as suits their size and nature.
4. Unless there's a good reason otherwise, assign all formation's loss tolerance to Ignore Losses.
5. Assign one or more small formations to contingency coverage against Player "special ops" and/or intrusion into rear area security.
6. Having rationale and coherence to the entirety of Elmer's objectives and orders - in effect, the scenario designer is laying out an entire battle plan spanning the full course of the scenario vs. going on a turn by turn basis.
7. Make use of the objective tracks for a better defined defensive and offensive disposition; use the third track to augment and enhance the side's primary disposition - offensive or defensive - with some unexpected trajectories.
8. Attempt to make use of natural terrain features in conjunction with defining formation objectives.

Anyone have additional tips?





Dabbs,

This is a pretty good list.

One Event Effect which has a significant influence on the PO's (I don't like to anthropomorphize machine 'intelligences' -- like referring to it with a human name...[;)]) behavior is 'Strategic bias 1 or 2'. This effect can be set to the values: 'Very cautious', 'Cautious', 'Neutral', 'Aggressive', and 'Berserk'. There are very clear differences in the PO's behavior depending on how this event effect is set -- you should test it to see.

Having designed several scenarios specifically designed for play against the PO, my goal has not been so much to make the PO as challenging an opponent as possible as it has been to make it follow a more or less historical course of action (with a little variablility thrown in to make things interesting). The 'Strategic bias' Event Effect has been very useful to help me simulate historical changes in force posture, triggered either by turn, or by some other event.




rhinobones -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 6:34:28 AM)


quote:

ORIGINAL: Dabbs
7. Make use of the objective tracks for a better defined defensive and offensive disposition; use the third track to augment and enhance the side's primary disposition - offensive or defensive - with some unexpected trajectories.


Think you are definitely correct about the use of Obj Tracks to randomize the PO’s disposition. Depending on whether the PO is intended to be the aggressor or defender (or switching between the two) there are two basic event strings I use to make the PO unpredictable and difficult to defeat.

PO on offense, as an example supposes:

Obj Track #1 - Right Flank Attack, Center and Left defend
Obj Track #2 - Center Attack, Right and Left defend
Obj Track #3 - Left Flank Attack, Center and Right defend

Use the events so that 1, 2 and 3 each have a probability of happening and then cascade the events so that later in the scenario there is a probability that the Obj Track may change. 1 changes to 2, 1 stays 1, 2 changes to 3 and then changes to 2 again . . . and so on. Do this two or three times depending on the length of the scenario and the effect you want to produce. The combinations you can make are virtually endless when combined with other events.

PO on defense

Obj Tracks for the front line defensive formations are a bit more limited. You can vary the direction they fall back, or their primary objective that they are assigned to protect.

However, for reserve formations there are more options. I like to set up reserve formations with objectives that are spaced to the left and right of their starting position. In this case the movements of the enemy crossing a trip wire (the formation objective) determines which direction the formation will move and the position it will try to defend. Of course this can backfire if the enemy attacks every where and all of the trip wires are crossed. A good scenario design can guard against this happening.

In addition to the use of Objective Track randomization, events to marry Strategic Bias (as Stevens suggested above) Weather effects and Shock with the appropriate Objective Tracks can create a wide range of realities for any one scenario. I think this is true whether the scenario is hypothetical or historical. Depends on the skill of the designer.

Regards, RhinoBones




Dabbs -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 10:49:46 AM)

Thank you Stevens and RhinoBones - very helpful, Strategic Bias is too important to miss and I was missing that one.  The PO is doing considerably better on the offensive, but enabling it to provide a challenge against veteran players is my goal for the scenario I'm working on.  The PO's main liabilities are very limited local situational awareness and more limited front-wide situational awareness.  And, in some respects that is good - offensively if in one area it is advancing but in another area it is stalled, it doesn't redirect forces to the stalled area (except as directed by its objective path). 

At the same time, units that have fulfilled their orders don't just remain static.  I need to test and watch this further to better understand the behavior, but the units did engage actively in hex conversion.  So, the test should extend to assess the extent to which units under various sets of orders (and a different strategic bias) having already secured their objectives, "will take the initiative" to seek out other enemy units and/or objectives. 

The defensive game is going to be the real trick from what RhinoBones references with respect to the frontline defensive and reserve formations.  For reserves, I think - and there are limits to its effectiveness, the trick is to break the reserves into smaller formations - to the extent it is logical:  reducing the number of tripwires accordingly. 

Defensively, what are the tricks to get units to dig in more frequently?  I'm noticing some units digging in, but it is still a fraction of the units that should.  I'm guessing a combination of orders and strategic bias are the primary factors applying, anything else?

Thanks for the help, information and advice...  Have a good one!

Mark




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 11:44:46 AM)


quote:

ORIGINAL: sstevens06

Having designed several scenarios specifically designed for play against the PO, my goal has not been so much to make the PO as challenging an opponent as possible as it has been to make it follow a more or less historical course of action.


This is a much more attainable goal. The main problem with the PO is predictability.




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 11:54:40 AM)


quote:

ORIGINAL: rhinobones

Obj Track #1 - Right Flank Attack, Center and Left defend
Obj Track #2 - Center Attack, Right and Left defend
Obj Track #3 - Left Flank Attack, Center and Right defend


This leads me in the direction of a possible future feature for the game: to have different parts of the force on different objective tracks. As it is, with three tracks one is still left with a fairly basic pattern of behaviour- for example in the above case the defender can count on only having to face one offensive at a time.

As a further advance, one could allow events to trigger off loss penalty- so if there was a large differential in loss penalties, the event engine would then choose one of the two "passive" sections of the front to switch to offensive mode.

A more speculative approach would involve the PO making its own decisions about which objective tracks to use. But that would involve a lot of work.




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 1:27:09 PM)


quote:

ORIGINAL: Dabbs

At the same time, units that have fulfilled their orders don't just remain static.  I need to test and watch this further to better understand the behavior, but the units did engage actively in hex conversion.


This could be a very useful behaviour in scenarios with guerrilla and other rear-area activity. Some formations could have their objective chains stop short, building up a force of anti-guerrilla troops, dependent on how much territory has been captured.




Dabbs -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 2:30:15 PM)

Just noting some ideas/thoughts for the purpose of ongoing consideration...theoretical

With the PO (and not against a human player) the differential in loss penalties can be closely approximated relative to the loss of objectives with as few as 3 - 5 (more is better)  Player vs PO playtests - best obtained prior to defining the 2nd or 3rd objective track.  Obviously, that information would be desirable prior to defining even the 1st track. 
Additional data specifically for ballpark figures can be acquired through the saved games of like-sized/structured scenarios (following similar objective VP values, similar force sizes, etc.)  Alternatively, the events can key off the destruction of specific units - which, if destroyed, are likely to signify the destruction of one or more full formations. 

You're right that it would be a significant plus to adjust the objective tracks for just portions of a force - especially in relation to "local/regional" formation loss values (if/when & which units and formations have been destroyed or rendered ineffective).   Logically, we can presume that the failure of the PO to hold a specific objective by a certain point in the game is indicative of an absence of sufficient force to obtain that objective.  Problems arise in under or over committing in response to this...

This is fun, I need to test a few things and watch how strategic bias works before speculating further, this is very interesting...




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 5:55:16 PM)


quote:

ORIGINAL: Dabbs

With the PO (and not against a human player) the differential in loss penalties can be closely approximated relative to the loss of objectives with as few as 3 - 5 (more is better)  Player vs PO playtests - best obtained prior to defining the 2nd or 3rd objective track.  Obviously, that information would be desirable prior to defining even the 1st track. 
Additional data specifically for ballpark figures can be acquired through the saved games of like-sized/structured scenarios (following similar objective VP values, similar force sizes, etc.)  Alternatively, the events can key off the destruction of specific units - which, if destroyed, are likely to signify the destruction of one or more full formations.


These approaches have serious shortcomings. The former leaves the PO open to the one of the most popular strategems of all time: the feigned retreat. The latter could be exploited by an unscrupulous player who disbands individual units and tricks the PO into seeing heavy losses- it's much more difficult to fake a high loss penalty. Of course I suppose if players want to cheat against the PO they're not hurting anybody.




Dabbs -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 8:04:08 PM)


quote:

ORIGINAL: golden delicious

These approaches have serious shortcomings. The former leaves the PO open to the one of the most popular strategems of all time: the feigned retreat. The latter could be exploited by an unscrupulous player who disbands individual units and tricks the PO into seeing heavy losses- it's much more difficult to fake a high loss penalty. Of course I suppose if players want to cheat against the PO they're not hurting anybody.


You have a very good point there - the feigned retreat works wonders for breaking up an unpenetrable line. I kind of considered this with respect to the PO of not likely conducting a feigned retreat....and that if the player has to resort to that to turn the tide, then we've "almost" done our job as designers. If players are going to cheat against the PO though...ha...can't help it, but that just sounds funny considering its early incarnation - then probably the effort to make the PO even more difficult probably won't be very welcome....ha...

Now...as designers we will have fulfilled our objective in the truest sense if even after a player conducts a feigned retreat, that the PO is able to deal with the counteroffensive that follows on reasonable terms.

It looks like I'll have some extra time, so will design a more comprehensive test that what I've been working with to test each set of orders in conjunction with each level of strategic bias, probably across 5 identical situations, and see if perhaps the behavior can be defined in more qualitative and quantitative terms. I sort of expect it will be in vain and futile to analyze it this way, but should contribute to a better understanding of this alien entity...

It is becoming easy to appreciate Stevens aversion to anthropomorphisizing Elmer...

And...I should have something better to do...




ralphtricky -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 11:21:54 PM)

quote:

ORIGINAL: Dabbs
I'm making a list...and checking it twice, to get Elmer to make players pay the price... Okay, well - that doesn't make a very good Christmas Carol...ha...

What are the tricks for making Elmer most competitive?

Thus far, my list is as follows (in no special order)

4. Unless there's a good reason otherwise, assign all formation's loss tolerance to Ignore Losses.

Anyone have additional tips?

Ignore losses has a definite impact on Elmer's behavior. It means that he will ignore losses, and will also ignore supply.

On the other settings, he has a chance to rest and allow his foramtions to recover if they're below a cretain minimum.. On ignore losses, he assumes that you want him to press on no matter what.

Ralph





ralphtricky -> RE: Optimizing Elmer's Current Capabilities (11/17/2006 11:25:43 PM)

quote:

ORIGINAL: sstevens06
I don't like to anthropomorphize machine 'intelligences' -- like referring to it with a human name...[;)]

I'm going to tell Elmer you said that[sm=00000436.gif]




sstevens06 -> RE: Optimizing Elmer's Current Capabilities (11/18/2006 2:22:02 AM)


quote:

ORIGINAL: ralphtrick

quote:

ORIGINAL: sstevens06
I don't like to anthropomorphize machine 'intelligences' -- like referring to it with a human name...[;)]

I'm going to tell Elmer you said that[sm=00000436.gif]



Oh I'm in trouble now...[sm=Crazy-1271.gif]




larryfulkerson -> RE: Optimizing Elmer's Current Capabilities (11/19/2006 2:28:47 AM)

I'm not 100% certain that it makes a difference whether the intelligence is "meat"-based or "silicon"-based but whether or not it really has intelligence.  Maybe the differentiating quality is the pattern of logic being executed rather than the pathways of the electrons/ions conducting the information.

After all 100% of everything is made of subatomic particles, even meat.




Dabbs -> RE: Optimizing Elmer's Current Capabilities (11/19/2006 1:53:22 PM)

Larry, are you saying that with enough subatomic particles it is possible to achieve concentrations of mass, and if exercised properly would evolve into concentrations of force, which in operational terms could produce a breakthrough on the tactical level leading to broad strategic gains replicating endlessly across the entire multiverse?
[X(]

[:D]




freeboy -> RE: Optimizing Elmer's Current Capabilities (11/19/2006 4:07:15 PM)

now we are in ttrouble.. call goustbusters




ralphtricky -> RE: Optimizing Elmer's Current Capabilities (11/19/2006 7:27:39 PM)


quote:

ORIGINAL: Dabbs

Larry, are you saying that with enough subatomic particles it is possible to achieve concentrations of mass, and if exercised properly would evolve into concentrations of force, which in operational terms could produce a breakthrough on the tactical level leading to broad strategic gains replicating endlessly across the entire multiverse?
[X(]

[:D]

Nope. He's saying that given time, ELmer may evolve enough to be able to kick you from one side of the map to the other <g>




Raindem -> RE: Optimizing Elmer's Current Capabilities (11/19/2006 9:52:50 PM)


quote:

ORIGINAL: golden delicious

This leads me in the direction of a possible future feature for the game: to have different parts of the force on different objective tracks. As it is, with three tracks one is still left with a fairly basic pattern of behaviour- for example in the above case the defender can count on only having to face one offensive at a time.

As a further advance, one could allow events to trigger off loss penalty- so if there was a large differential in loss penalties, the event engine would then choose one of the two "passive" sections of the front to switch to offensive mode.

A more speculative approach would involve the PO making its own decisions about which objective tracks to use. But that would involve a lot of work.

This would be good as long as we retain the ability to set the force track all at once. I shudder to think of having to do it one formation at a time.




Dabbs -> RE: Optimizing Elmer's Current Capabilities (11/20/2006 1:06:23 AM)


quote:

ORIGINAL: ralphtrick

Nope. He's saying that given time, ELmer may evolve enough to be able to kick you from one side of the map to the other <g>



Which map? [:D]

My testing of Elmer only reiterates what is stated in the manual - on the Attack/Advance and Secure/Screen orders.

1. Units move in the most direct path toward their objectives per strict sequential order, sometimes taking advantage of roads even if doing so is slighter longer but costs fewer MP's.
2. Units generally will not go around enemy units en route to their objectives, but will ignore enemies not blocking the route to their objective.
3. Upon taking their objectives, they respond to enemy units in their vicinity according to their Strategic Bias:
a. Very Cautious - units will tend toward hex conversions in an arc/circle around their last defined objective.
b. Neutral - the formation will divide its efforts between hex conversions and responding to the nearest enemy.
c. Beserk - the formation will seek out the nearest enemy units typically starting the turn after securing its last objective.
4. Units assigned to Secure do break down/divide after securing their last objective - and begin to screen - convert ownership of hexes. If one of their objectives is taken, units will sometimes recombine in pressing an attack.
5. Units assigned to Attack may break down a small portion of their formation, but tend to keep units whole.
6. Formations and units appear 100% oblvious to the presence of other objectives not defined in their list.

Nothing really new - but the tests did tend to demonstrate that if you want formations to continue to be active even after its taken possession of the last objective defined in its list, then the Secure and Screen orders appear best; whereas if you want them to be inactive then the Attack and Advance Orders are best. Of these, the Attack Order reacts to threats in rear areas while the Advance Order does not.

In total though, it is very much a matter of having formations with overlapping objectives and different sets of orders, relative to terrain and all other battlefield conditions. The Screen Order is good for formations probing along the extreme flanks; while Secure is better for formations following behind the heavy/primary formations dedicated to Attacking. Advance order sets might be best given to recon elements, cavalry, partisans and the like.

These are issues good to keep in mind even when developing the OOB - The PO may look at the combat values of its individual units, it does not appear able to distinguish which of its units is a tank unit or a headquarters....so it has a distinct tendency to rush its HQ's into exposed positions when it has the chance - without protection. That tends to advocate reducing the movement allowance of the HQ grouping it with a few other appropriate units (artillery, air defense and some kind of infantry for protection or local reserve), and assigning that formation objectives 1-2 hexes behind the true front line units. Or, it may mean assigning the HQ icon to a front line unit.

Basically, we have to do the majority of the thinking for the PO making use of its logic. Leastwise, understanding its logic enables us to get it to behave closer to our designs. That starts with giving it units that it can work with, assigned to formations with orders matching their purpose.

Just as long as Elmer doesn't become Hal...we should be safe. Right Dave?





golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/20/2006 11:33:23 AM)


quote:

ORIGINAL: Dabbs

Now...as designers we will have fulfilled our objective in the truest sense if even after a player conducts a feigned retreat, that the PO is able to deal with the counteroffensive that follows on reasonable terms.


Well, not if the feigned retreat becomes the strategem for beating the PO. You've removed an 'interesting decision'.




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/20/2006 11:37:40 AM)


quote:

ORIGINAL: Raindem

This would be good as long as we retain the ability to set the force track all at once. I shudder to think of having to do it one formation at a time.


Yeah. Also, one could get a lot of extra functionality by adding an OOB layer higher than the formation. This could be used for objective tracks, shock, force proficiency, replacements, strategic transport, co-operation levels etc., etc.




freeboy -> RE: Optimizing Elmer's Current Capabilities (11/20/2006 12:04:24 PM)

hello GD.. since I have not used the editr, forgive if this is an ignorant off the wall ?.. could you use an event trigger to simulate some of the above.. ie shouck etc for larger groups?




golden delicious -> RE: Optimizing Elmer's Current Capabilities (11/20/2006 12:16:17 PM)


quote:

ORIGINAL: freeboy

hello GD.. since I have not used the editr, forgive if this is an ignorant off the wall ?.. could you use an event trigger to simulate some of the above.. ie shouck etc for larger groups?


All the effects I listed, except for co-operation levels, are currently set for the entire force only, most of them by events. I was suggesting making it possible to define a new level of organisation between formation and force which could be used for the application of these effects- again mostly by events.




ralphtricky -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 2:13:38 AM)

quote:

ORIGINAL: Dabbs
quote:

ORIGINAL: ralphtrick
Nope. He's saying that given time, ELmer may evolve enough to be able to kick you from one side of the map to the other <g>

These are issues good to keep in mind even when developing the OOB - The PO may look at the combat values of its individual units, it does not appear able to distinguish which of its units is a tank unit or a headquarters....so it has a distinct tendency to rush its HQ's into exposed positions when it has the chance - without protection. That tends to advocate reducing the movement allowance of the HQ grouping it with a few other appropriate units (artillery, air defense and some kind of infantry for protection or local reserve), and assigning that formation objectives 1-2 hexes behind the true front line units. Or, it may mean assigning the HQ icon to a front line unit.

Elmer should keep HQs and other support units away from the front. It also knows the difference between Infantry, Recon and Armor. Is this still happening?
quote:


Just as long as Elmer doesn't become Hal...we should be safe. Right Dave?

Right [;)]




sstevens06 -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 3:13:47 AM)

quote:

ORIGINAL: ralphtrick
...
Elmer should keep HQs and other support units away from the front. It also knows the difference between Infantry, Recon and Armor. Is this still happening?
...



Interesting you should mention that: I have an Egyptian divisional HQ unit going rogue on me in a playtest of Suez Canal 1973 with the PO playing the Egyptian side. Not only does the PO place it in a rather exposed position - not on front lines but dangerously close - that position is NOWHERE NEAR its assigned objectives! I'll send you the pertinent files and a detailed report via email.




Adam Rinkleff -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 4:06:28 AM)

While we are on the topic, how can I get the AI's artillery to actually take a position close to the front, rather than far in the rear? 




JAMiAM -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 4:25:38 AM)


quote:

ORIGINAL: sstevens06

quote:

ORIGINAL: ralphtrick
...
Elmer should keep HQs and other support units away from the front. It also knows the difference between Infantry, Recon and Armor. Is this still happening?
...



Interesting you should mention that: I have an Egyptian divisional HQ unit going rogue on me in a playtest of Suez Canal 1973 with the PO playing the Egyptian side. Not only does the PO place it in a rather exposed position - not on front lines but dangerously close - that position is NOWHERE NEAR its assigned objectives! I'll send you the pertinent files and a detailed report via email.


Has the formation been given "Independent" orders?




ralphtricky -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 4:32:11 AM)


quote:

ORIGINAL: AdamRinkleff

While we are on the topic, how can I get the AI's artillery to actually take a position close to the front, rather than far in the rear? 

How close? It should come within range of the enemy units and wait there.




Adam Rinkleff -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 5:00:26 AM)

quote:

ORIGINAL: ralphtrickHow close? It should come within range of the enemy units and wait there.

Right, and that's exactly what it does -- except, artillery firepower apparently decreases with range, especially with extended range units, and so the result is that AI artillery sits out at such a far range that it minimizes its potential.




ralphtricky -> RE: Optimizing Elmer's Current Capabilities (7/24/2007 5:58:58 AM)


quote:

ORIGINAL: AdamRinkleff

quote:

ORIGINAL: ralphtrickHow close? It should come within range of the enemy units and wait there.

Right, and that's exactly what it does -- except, artillery firepower apparently decreases with range, especially with extended range units, and so the result is that AI artillery sits out at such a far range that it minimizes its potential.

I need to investigate the whole ranged equipment thing more. Between extended range and the mixture of equipment with different range, it's cofusing.




Page: [1] 2   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
1.154297