Thoughts/Discussion RE: PO/Elmer Sr.

[USXpat]

On the verge of getting heavily into working with the PO, thought I'd share a few things and try to pick up some additional tips from others.

The PO has improved considerably but still has some inherent weaknesses. The primary weakness, imo, is inability to see the global strategic picture. Instead, with each formation having its own orders, and such, it develops lots of "tactical snapshots". Developing it to respond with the right amount of force to massive breakthroughs (and even small breakthroughs)... or even to respond at all, is our nightmare as designers.

What I've found to offset the tendency to "overstack" - can be resolved in how formations are defined. Obviously, this can vary a lot, suffice that for TOAW it is better to define formations with regards to how the PO handles them, then how they were defined historically. The "system" advocates for a variety of different formation types, in terms of size and function. I'll use some stereotypical WW2 Russian formations as examples.

1. Triggers - independent brigades/divisions - just something to hold specific landmarks. Very small formations.
2. Objective Takers - Shock Armies, Guards Armies and other large formations that concentrate in force.
3. Main Lines - medium-sized formations expected to hold "x" amount of frontage.
4. Gap Holders - small-size formations expected to hold "x" amount of frontage between Main Line formations and/or provide defensive coverage when the Main Line engages offensively.
5. Trigger Response - biggest variable - depends upon what is available to be called upon, akin to a second line Objective Taker Formation.
6. Waves - For situations where there is an abundance of units/formations - can serve as Trigger Response and/or replacements for Main Line formations (and alternate function) managed through objective tracks.

Considering what any given formation can be reasonably expected to accomplish on its own will go a long way to help in "PO stability". A lot of Russian Armies are pretty small - 3-5 divisions; and some can be pretty big, 10-12 divisions. The objectives of each formation have to be considered with what it can be reasonably expected to handle - such as covering 3 hexes of frontage on the offensive, or 3 hexes of frontage with defense in depth. That's decent for a small army, but will lead to bunching up for a large army.

One method that may help is standardizing the size of some of these larger armies to that of the smaller armies - and then branching off its excess units into Operational Groups. Say 64th Army historically had 10 divisions - in TOAW that 64th Army could be represented as an Army with 6 divisions plus 2 operational groups of 2 divisions each (left/right flanks). Of course, that means four divisions will not have a normal HQ - but the only functional drawback to that appears to be they will not be subject to the proficiency loss in the event that the 64th Army HQ is destroyed. They'll still gain a bonus in supply if a friendly, cooperating HQ is adjacent to them.

In my own and with other's in regular play, when conducting an offensive - we all like to have at least something dug in on the front line when we go on the offensive. Having more formations should lend to the PO being able to do the same. Haven't tested this in a while - suffice that in my lasts tests - the PO did more or less what I told it, and if it got it wrong, it was owing more to my issuing the wrong orders or messing up the objective sequence.

Covering the big what-ifs - major breakthroughs and potential encirclements - I think is the biggest problem though. In theory, this comes down to setting up the means to 1) Prevent (or attempt to), 2) If it does happen, enable the PO to (attempt to) determine that it has happened... and then... .... ha... 3) try to respond to it. Not asking for miracles or anything... hahaha...

I think the above mechanisms can "help" do a lot more, in some scenarios, with respect to prevention. The Trigger Points, provided there are enough of them, should be able to help in defining for the PO what has happened. Might require 2 or maybe even 3 layers of Trigger Points, though that can get expensive in unit count.

With just three objective tracks... I think being able to engineer an appropriate response would be pretty difficult. Now that there are five objective tracks - it could be do-able. Leastwise, I can see it being do-able in situations like late-WWII Soviet side situations - where there are endless hordes of reinforcements.

These are my thoughts and observations up to this point, suffice that I welcome other's ideas and observations, too - techniques for simplifying PO components, etc.

I'm looking forward to lots of fun in figuring out how to define the Axis response to the Balkans...

[sPzAbt653]

Good stuff Mark. I'm a solitaire player and like working with Elmer. Having scenarios playtested by different players is a big key in being able to set up a good PO. A few other thoughts:

The primary weakness, imo, is inability to see the global strategic picture.

There is no book on Elmer so I only know what I've seen in playtesting and from looking at a few pologs. I don't think there is a strategic layer built into Elmer, but he does have a few parameters that he checks in order to make decisions as to what to do. Maybe in the future a strategic parameter will be added. Meanwhile, designers can figure out what are the most important strategic aspects of a scenario and can then set up Elmer accordingly.

What I've found to offset the tendency to "overstack" ...

If it is determined that overstacking is never a good thing, then it shouldn't be impossible to tell Elmer not to do it. However, before doing that I think all us human players would have to admit that we actually never overstack ourselves. But I think we do, and when we do we go thru a thought process that may be too much for Elmer to reasonably handle.

... that means four divisions will not have a normal HQ ...

You could at that point arrange the units in Corp Formations, which would give each of them an HQ. Either way, the smaller the formation the better. Human players may consider the overall situation for a formation, but will use each unit individually to gain a result. Having multiple units in a formation for Elmer makes him less efficient. The smaller the scenario, the more this is magnified.

... and if it got it wrong, it was owing more to my issuing the wrong orders or messing up the objective sequence.

Yes, 'Elmer Sucks' is not a true statement. 'Elmer needs some design assistance' would be more accurate.

... major breakthroughs and potential encirclements ...

With IGO/UGO this is quite an issue to tackle. Using the additional objective tracks is quite helpful, but still leads us into predetermined reactions. Having reserve formations is a must. Set up properly, these formations can give a good kick to a human player. if research for a scenario didn't provide enough reserve formations to help Elmer, then we have to go outside of the scenario parameters in order to provide some. An alternate 1944 Balkans Invasion is a good example. As it never happened there will be no historical references (unless someone comes across a document outlining a possible Axis reaction). So the designer would be left on his own in programming an objective track and in pulling units from all the other fronts in order to defend the area. Sounds like fun [:D]

[ColinWright]

ORIGINAL: sPzAbt653



If it is determined that overstacking is never a good thing, then it shouldn't be impossible to tell Elmer not to do it. However, before doing that I think all us human players would have to admit that we actually never overstack ourselves. But I think we do, and when we do we go thru a thought process that may be too much for Elmer to reasonably handle.

For what it's worth, I can't think of any situation where I have risked exposing stacks with more than a yellow light to enemy fire.

I don't even like that, and only resort to it under certain conditions. If the enemy gives me an orange stack, I generally find it profitable to bombard it, and if the stack is red, I'll stop everything to wreak relatively painless carnage.

Then on the other hand of course, while traffic penalties can be an issue in some scenarios, there's little reason otherwise not to go to red out of artillery range and/or observation unless the enemy has impressive air power.

However, to simplify matters, Elmer could be told just to never go past yellow no matter what. As usual, it would be best if this could be a designer controlled setting, not 'bundled' with other settings, etc.

[USXpat]

Good points, thanks! Just re-read Ralph's thread on improving the AI with regards to distance between objectives and the "What's New" where there's a point that the AI performs better relative to "slow advances".

There seems to be plenty of "givens" - small formations being more efficient, objectives 5+ hexes apart, probably something with respect to units of the formation having similar movement values. The little bit of strategic layer is given to defining hex ownership within a radius around an objective, and where it has a VP value - ownership of the objective hex itself.

It sounds like artillery is working effectively, but that other non-front line units are frequently being used by the AI as "front line units". Really think this is a matter of "designing the oob for how the PO handles things" vs. "historical".

The one thing I was disappointed with initially was that "objective tracks" are determined globally vs. by formation. Still thinking about this, but five objective tracks should be sufficient. So, then it becomes a matter of defining the events that lead to changing objective tracks. With a large scenario, this would seem to require ownership of a series of hexes.

I can see that is going to be a bit more complicated for multiple fronts.

Am thinking, that say switching from Track 1 to Track 2 would be % dependent relative to ownership of 4 - 5 objectives; each objective adding to the EEV and then when it fires, the EEV resets to zero. Then does the same thing relative to Track 3, 4, and 5. I'm inferring from the manual that this is possible - any other ways to handle it without using the EEV?

Another method of developing the PO that I've considered over the past couple years is using an excel spreadsheet "as a map" - which might be getting far too involved, and requires a lot of work setting up in its own right. Leastwise, once set up it can be copied easily multiple times within the same file. Thinking that this could chart the hexes relative to whether they are passable/impassible/objectives. Then make it easy to define "Formation Groups" and color code the "excel map" relative to their operational area. The only thing this needs that I haven't come up with is a method of reflecting "overlapping" operational areas.

Leastwise, in defining each formation's objectives - remembering what all of the other formations are doing can get more than fuzzy. Especially with a starting OOB and reinforcements arriving over time, each within their own formation - can help with making sure each area of operations is adequately covered.

Now... what would be ULTRA COOL... for a geek like me, would be the ability to import a map directly into Excel... but we'd need to change Excel to use Hexes instead of Rectangles...ha.

Time for more map work... really homing in on some of the last stuff to do before setting up the Events.

[sPzAbt653]

any other ways to handle it without using the EEV?

The EEV sounds like the best way to accomplish what you want. In doing a large scenario with multiple fronts, you may consider only using 3 or 4 tracks initially, tied to what you feel might be the most important front line changes. Then you will have 1 or 2 tracks available for when playtesting has revealed what else is really important.

remembering what all of the other formations are doing can get more than fuzzy.

I wouldn't waste time on it, it's not that critical. I suggest laying down objective tracks one at a time and uninterrupted by other work. This way things stay in your mind. Then you can get to playtesting, which will reveal areas of concern. Track adjustments can be made at that time.

the ability to import a map directly into Excel

I don't use Excel but I have pulled the map from a scenario and put it in a graphics program to make notes on. A quick easy reference.

[USXpat]

Thanks! I'll try to constrain it to 3-4 then and leave that last one open for later use.  Good idea with the graphic too - will give that a shot as looks easier than excel... but I'm always using Excel and rarely get into graphics. 

Have another pretty good week to work on this, maybe part of next week.  Work's ramping up heavily and fast, so - anything that saves time is great.  About the only thing that really saved me on this project was taking notes, charts and keeping every print-out.  Surging on!  Thanks for the input!