Looking for a real life platoon sized AAR for a CM experiment

[Michael Dorosh]

I'd like to find a real-life account of a battle to use as a "control" with which to compare and contrast various tactical game systems, both board and computer (the idea being to develop small scenarios for various systems using the same topic and see how the different systems treat the material). Given the amount of research some of the designers have done here, I thought I'd throw out the question - perhaps Kingfish or someone can suggest something. I have some specific parameters in mind.

US vs. German - most game systems seem to include these nationalities - Russian (Soviet) vs. German is another option.

Platoon sized - a little smaller than ideal for many game systems, but the idea is to make the compare/contrast a manageable size and is for illustrative purposes rather than competitive play.

Fine weather, no armor, ordnance or vehicles, no fortifications, no special terrain like hedgerows - not all game systems have special rules for these things so to be as scientific as possible, best to leave these out.

I'd like to find a historical action. The best one that comes to mind is the meeting engagement in Tunisia that LIFE Magazine reported on, when fog lifted and a US platoon armed with M-1 Garands shot the crap out of a German platoon a few dozen yards away; however, ideally I'd like to find a situation that is well documented with the units of both sides identified, and perhaps even the unit commanders. Would ideally also have a specific date and location. US vs. German is optimal for this reason also, as high-res Google Earth imagery and period maps are easier to find than Eastern Front ones, at least in my experience, ditto translations of German unit histories regarding units serving in the west vs. those fighting in the east.

Other actions that come to mind are some of the MOH actions - for example Brecourt Manor. That one has fortifications, is not platoon sized (Winters had only a squad) and involves ordnance. But something similar that is infantry only, if anyone could suggest something, either from NW Europe, Italy or Tunisia, would be of interest.

[David I]

Michael,

SLA Marshall did some great AAR's during the Korean Conflict. I think he did some WWII stuff too. Curahee (spelling probably wrong) comes to mind.

Just a thought.

David

[Michael Dorosh]

Hi David - many thanks, I think I found a better idea. I've broadened my scope to look at Singling, as per the thread in the GF. Marshall's work is indeed good, though.

[Kingfish]

Mike,

If platoon size is your cutoff then you will be hard pressed to find something. I'm not sure even the byte battles went that small. The smallest I've done is 2 platoons + a tank, which doesn't meet your requirments.

Now, if you can work with a scenario that is slightly bigger than platoon size then I recommend you take a look at Wild Bill's "A walk in the sun". 1-1/2 American platoons + some odds and ends. German force is about the same size. No vehicles, and nothing bigger than a HMG.

[Michael Dorosh]

Originally posted by Kingfish:

Mike,

If platoon size is your cutoff then you will be hard pressed to find something. I'm not sure even the byte battles went that small. The smallest I've done is 2 platoons + a tank, which doesn't meet your requirments.

Now, if you can work with a scenario that is slightly bigger than platoon size then I recommend you take a look at Wild Bill's "A walk in the sun". 1-1/2 American platoons + some odds and ends. German force is about the same size. No vehicles, and nothing bigger than a HMG.

Yeah, I'm coming to the same conclusion re: my parameters. Is that scen based on the movie or a real life occurence? In any event, I am looking at Singling - a much bigger action, but insanely well documented, and I've not only located a Steel Panthers scen (done by a CMer) and a CMAK scen, but just joined an ASL group that is doing a historical map for this one - they asked me to be a playtester. Bigger is probably better; if I can find a version for Panzer Leader, I'll be laughing...

Thanks for the suggestion - I'll keep it in mind. Singling may be too big; it also has Panthers and stuff the original Squad Leader did not have - I had wanted a situation I could use across many different games but I suppose I can fudge - that was the name of the game as far as the original SL went in any case.

I may decide to do both - a company/bn sized situation and a smaller one like Wild Bill's just as an illustration of how the game works. Much appreciated.

[Kineas]

Originally posted by Michael Dorosh:

I'd like to find a real-life account of a battle to use as a "control" with which to compare and contrast various tactical game systems, both board and computer (the idea being to develop small scenarios for various systems using the same topic and see how the different systems treat the material).

That sounds very interesting. Using a well documented event as a testbed for game mechanics comparison. Could you tell us more about the purpose? Do you want to compare different game genres, or to assess how much they approximate "reality"?

[JasonC]

Singling is so well documented because it was one of the army historian studies in "Small Unit Actions". Another that is more nearly pure infantry is in Italy, a US infantry night attack that is the last word in accurately described fearsome f-ups. Details here -

http://www.army.mil/cmh-pg/books/wwii/smallunit/smallunit-smi.htm

[Michael Dorosh]

Originally posted by Kineas:

</font><blockquote>quote:</font><hr />Originally posted by Michael Dorosh:

I'd like to find a real-life account of a battle to use as a "control" with which to compare and contrast various tactical game systems, both board and computer (the idea being to develop small scenarios for various systems using the same topic and see how the different systems treat the material).

That sounds very interesting. Using a well documented event as a testbed for game mechanics comparison. Could you tell us more about the purpose? Do you want to compare different game genres, or to assess how much they approximate "reality"? </font>

[Kineas]

Thanks. This topic is one of my favourites. So let me add a couple of comments quickly, though I'm definitely not an expert.

Launch a game which uses the realtime-pausable engine of "Airborne Assault", e.g. Highway to the Reich, set it up to maximum speed and do not interact. You'll see the bigger events unfold. Armies moves on the ground like amoeba, the forward tentacles being continuously pounded by artillery. If you imagine a complete scenario fastforwarded this way in another games, you get more hindsight about the realism factor of the genres.

IGOUGO: the alternating-oscillating nature is so strong, there is no point in searching realism after a certain level. E.g. the guys on the ASL forum, who love their game, try to explain why skulking can be realistic, because it means the defender hides better etc. But it can't be explained. Nor can the opfire-depleting tricks be explained what you can use in the Steel Panthers series. Beyond this point the only common thing with realistic warfare is that both the virtual and the real commander tries to get the best out of his system, competing with not nature but another human being.

WEGO: though CM don't has the feature, but imagine again a 30 turn game replayed in 2 minutes. You would still get a highly oscillating picture, because the intelligence is injected into the system at every 60 seconds, from that point on the tacai manages the things, and slowly deteriorates till the next heartbeat.

REALTIME(pausable): like the HTTR mentioned above. It's (in my opinion) the most realistic, but still has the borg-spotting effect. I don't mean it in the CMx1 sense. You, the player are the borg. If you see a lone scout in the flank of your army, you change your plan immediately, even if you can't communicate it to your subordinate units. With a CMx2 example, your forward scout sees an ATGM launch. It won't be hard to figure out that the target is your only tank 3000 meters away in the city. So in the next moment you will quick-reverse the tank behind a building. Realistic - not really. Can this be modeled? Yes,

but complicated, evasive actions should require some observations by the crew, and still, a non-evasive reversing could be enough. I guess this is where chasing realism ends...

I recently came across ASL, and realized that this is the mother of everything, really everything. All the odd things I experienced with the Steel Panthers series suddenly were grasped and understood. I think its effect was tremendous to every tactical wargame I've seen, CM very much included.

If you take a fire action in CM, basically it's resolved against an internal IFT. That's why we end up a so suppression-oriented system. The next breakthrough towards realism will be the 1:1 infantry representation and real bullet trajectories by CMx2. I'm interested how much the fire effects will differ in the new model. You can try it in Armed Assault (realistic FPS). If you are an LMG crew, and you catch someone in the open, 200m in front of you, then he is dead within 15 secs, he doesn't really have the time to suppress himself.

[JasonC]

The FPS is right that an LMG will kill a man in the open in 15 seconds. But no system that tries to start from that fact and scale it up will generate correct results even at platoon level, let alone at company or battalion.

Instead you will see casualty counts in time frames that do not occur in practice, except at the most extreme outliers of the war.

The electronic units will be mashed together far more forcefully than they actually collide in real life, and then your FPS atomic calculation will take over, and the net result will be a prediction that WW II should have lasted a week and ended in the death of half the combatants, for something like 1/1000th of the ammo expenditure actually used.

Tobruk tried to use a loss focused, firepower count combat system. It was fine for tank fights - many of its elements recur in CM even -but completely falsified the role of infantry in the desert. Which became, in that system, to sit at the bottom of holes or die in minutes if it tries to move above ground.

Why? Because neither differential acquistion nor real suppression can be modeled as fully as the imaginary ideal type of the final 15 seconds of trigger pulling. (Implicitly, the thinness of a man's shirt was being accurately modeled, but not the man's stealth).

Thinking that the purpose of a combat system is to model what happens between fully seen target and end of trigger pull is simply a mistake, it is not the determining factor in real combat. And no, modeling the flight of each bullet does not help - it hardly matters whether a probability is used instead, because the mean result will rapidly build to "he's dead, Jim".

Because the actual rate determiner is, does the man on the MG see that particular moving man, and that in turn turns on, is the man on the MG up and aggressively scanning an assigned fire sector or is he heads down behind his parapet clenching his teeth?

And we simply know that every system ever devised that does not start from the other end, the known outcomes, has and will get the frequency of those items completely wrong and will model automata not men in combat.

One has to start instead from the typical known outcomes, and tailor the model to fit. A typical outcome of platoons in contact is not, one side shot down to the man in 60 seconds. But a perfectly ordinary one is, half a dozen men hit on one side and a couple on the other, the former also featuring a dozen running away out of control of their commanders, and the rest hugging the ground.

None of which has anything seriously to do with ballistics, and instead has everything to do with fear and the patchy broken LOS of an empty battlefield populated by scared men avoiding each other.

Engineering realism attempts from atomic level up are hopeless - they can't even see the rate determining step in the reaction. If you can see that step and model it even approximately, you get an excellent system that fits historical tactics and combine arms relationships rather well - hence the (relative) success of ASL and CM.

What is the greatest remaining inaccuracy in CM? It is the high ammo expenditures, the way units run dry before hurting each other appreciably. It is true that even low levels of fire sustained over CM time scales will run everyone dry. But it is not true that men in combat fire that much. The reason is again the scattered patchy LOS of an empty battlefield - most men are not firing 6 bursts a minute because they can't see anything.

The load we run through inside 10 minutes of active infantry firefight, in practice lasted a typical US infantryman up to a week.

This could be handled by simply letting ammo use decline with target cover. If e.g. each shot had a chance of using an ammo point equal to the square root of the target exposure percentage.

It wouldn't be perfect (it would still put ammo use too high) but it would result in more accurate tactics and effects from cover.

Cover usually prevents shots entirely and only secondarily makes shots both occur and miss. Occur and miss is what drains ammo away faster than losses accumulate, prevented completely does not.

If occasional lethality is to reach realistic LMG catching 'em in the open levels while average lethality is to remain as bounded as it needs to be, fire effect variance needs to rise dramatically.

Though that would result in a more frustrating system for players, who would no longer be able to "learn by doing", what typically works and what typically doesn't. (I mean, a platoon advancing on an LMG would have a different outcome for the same command actions, nearly every time, 10 times running).

More generally, one cannot start with the bottom end and make a fetish of it, and assume that getting that right will make the whole compounded thing work. It won't. Nearly all the important effects occur higher up, and accurate modeling needs to work from multiple measures on both ends. Of the two, the upper end (average overall outcome) is far more important to get approximately right, than the atomic individual shot.

[Michael Dorosh]

All good points from both of you. I think the fidelity of simulation in all platforms extends beyond ammo usage. I think wisely, games like ASL don't even attempt to simulate ammo usage; it may have been a mistake for CM to include it.

I asked in another thread about the "uncanny valley" and CM may have approached that even with CM:BO. I won't discuss CM:SF which includes 1:1 coverage, so keeping in mind these comments are all CMX1 related. But the more detail one throws in, the more one expects reality and the more one is disappointed at the result. At least with ASL you could understand that the artillery rules were a simple fudge and that all nationality's artillery procedures were treated as if they were Americans (no self-respecting British FOO would ever "request" a simple battery fire mission, he would order it). In CM, with every round depicted, the critics can come out of the woodwork and talk about sheafs and dispersal patterns. Design for effect, as John Hill saw it, is out the window. This applies to what JasonC is saying directly. Gamers want to see every bullet strike, every expended cartridge casing.

Trouble is, guys in battle didn't shoot much. Even if you take Marshall with healthy doses of salt, Strome Galloway of the British and Canadians said that the rifles his men carried in Tunisia, Sicily, Italy, and the Netherlands were mostly security blankets and could have been replaced by pitchforks. Gamers don't want to see that. They want to blaze away like Audie Murphy, not herd cats. They want to take Singling in a single gaming session - 15 minutes or so of game time, not four days' worth of fighting. They want the fighting for the Barrikady to take place in bite size chunks, not play it out for 5 weeks in real time, where most of the time you're hiding in the cellar.

Command and control will always be a huge issue. Player omniscience, and the need for artificial controls. There is nothing wrong with artificial rules to simulate real world constraints - if they work. Simultaneous movement was a pain in the early days of tactical board war gaming, say 1969 to 1974 or so. WEGO in CM works very well; real time massive multi-player seems to be the best way to simulate it but the hardest to co-ordinate and really kind of inconvenient.

[ June 24, 2007, 09:51 AM: Message edited by: Michael Dorosh ]

[Kineas]

Jason, you are right, a game can't be designed bottom up from the damage system, unless it's a damage system simulator. Nevertheless special engagements like that LMG example can be used for later validation tests.

I think it's possible to come up with a ruleset which covers everything. Abstract systems will perform poorly at the micro, 1:1 simulators will perform poorly at the macro level. Simulators can't start from the desired outcome, they have to produce it. In this case you have to simulate humans with automatons. Besides the normal weapon physics you need reflexes, skills etc. but I don't go into that because it's not ontopic.

I think getting the correct results on every scale can be achieved in an 'abstract'(board) game. If you have the data, you only have to cover it with some dice produced probability distribution. You just need enough branches. I guess 30 years ago this was not important, there were no FPS games around to compare, and they had to keep the rules under control.

I don't know what is a typical course of a platoon sized engagement, I ought to read some AARs now. But I don't think they were ducking eachother for hours long, I guess the attacker had to perform, and the defenders tried to keep them away.

But realistic combat is not a game, more like a job I guess, whereas gamers always pursue fun and action. FPS players always rush into the biggest action (you hardly see any defenders on public servers), and wargamers too want to simulate action heavy firefights, not logistic details of a supply system. This is fact, and it's a good question how much it is reflected in the present rules/game models.

[JonS]

Michael,

I've got a couple of Cdn-vs-Fallschirm AARs heading your way soon. One is Cdn inf pn + 3 x 75mm Shermans. The other is Cdn inf coy with 2-in and 3-in mtrs in spt. Both require what would probably be considered 'special rules', so may not be useful for this, but you'll be getting 'em anyway.

Jon

[ June 24, 2007, 06:03 PM: Message edited by: JonS ]

[pavlov]

I would be rather interested in the final outcome of this experiment . . . but Michael Dorosh banned? Guess I missed something.

[Bannon DC]

I guess there is truth to the rumor... once you post 13,966 times you get banned.

Now, he finally has time to figure out how many licks it really takes to get to the center of a Tootsie Pop.