Lesson learned: Choose own forces in QB!

[Sgt Bonehead]

Just tried my first QB tonight (usually play pre-built scenarios) and discovered what happens when you let the generator pick your force disposition. Wound up attacking with two companies of mostly Green Pioneers (huh?) and a single Conscript AC across a large expanse of open steppe against T-34's supported by infantry. I *knew* I was doomed as soon as I heard the first main gun round fired in my direction on turn 1, but tried to stick it out ... hopeless situation. I had nothing at all to counter the armor with, and noplace to hide. Ugh.

Anyway, what do you guys leave as Random, and what do you select, when setting up a QB? Just curious how I can cut down on the "you gotta be kidding" factor without taking the spontaneity out.

[Soddball]

I had a fabulous game when the computer picked units - I had a battalion of conscript infantry to take a town. That was a tough battle, but well worth the challenge. However, I inherently distrust the AI's penchant for picking wierd equipment.

[Michael Emrys]

The "spontaneity" thing may present an insurmountable hurdle, unless you can get someone else to pick your forces for you. I always pick my own forces (and sometimes, albeit rarely, the other side's as well), but then I have a very different goal it mind, which is to explore certain kinds of tactical situations. So the spontaneity factor is less important to me.

Sounds like your best bet is to always play designed scenarios. Since there appear to be a wealth of those to be downloaded, you shouldn't run out too quickly. And if you do decide you want to play a QB, and it looks like the computer has chosen a ridiculously inappropriate force mix for you, ditch it immediately and start over.

Michael

[dieseltaylor]

I like to choose my force but to prevent uber-tanking if you leave the terrain at random , and any other factors you like, you need to get a balanced force.

I always choose the area i.e. North, Central etc so I have a rough idea what the genaral terrain and weather ought to be .[Thinks does the random engine look at region?]

I always avoid not knowing if there will be snow because that makes a huge differnce to tactics and therefore force selection. Particularly if you play over very large maps!!!

Best of all chose your own troops and play a human , failing that someone who posts on the bulletin board ....

[Joachim]

I tried another apporach vie scenario editor:

"Random amount of enemy points"

Buy own forces.

Buy most of the enemy infantry and some common tanks "on board".

Buy more inf and more tanks - among them rare beasts able to turn the tide - as reinforcements with very low probabilities per turn. Make sure some of these these beasts can kill any cat I have.

I have my objectives, I have my forces - but I don't know what the enemy has exactly or how much he will get. All I know the longer it takes, the stronger he gets.

The map has some depth

a) forcing me to an approach march so some reinforcements are already there.

leaving space for enemy reinforcements in covered areas (ie not in plain view). Infantry reinforcements on "covered" approach routes already a bit forward.

c) no first turn waste of AI arty ammo

d) I have to tow any big guns brought into play

The availability of reinforcements for the AI forces me too attack a bit faster - reinforcements come from the rear, so if I already reached the objective, it will be more like a meeting engagement instead of a human attack vs AI.

Here is a small table how likely it is that reinforcements at a given percentage arrive on the battlefield depending on game length. Changing the reinforcement turn helps varying it a bit.

I do not use any formula to calculate the actual exact value of the reinforcements, but I thought about weighing the earlier turns a bit more. As any weight is debatable, a guestimate will do, too. Thus we have:

Count any reinforcement approx at the percentage with which it will be there after 60% of the game length.

Thus a 1000 point tank force which will show up with a likelihood of 30% until turn 24 is counted as 300 pts (and I don't count beans, if it is a 979 pts tank force at 29.5 %, it is still 300 points for me.)

Appendix:

Reinforcement probability.

Calcualtes how likely it is for a unit to have shown up n turns after the specified earliest entry turn with a given probability

%\turn 10 15 20 30 35 40

0,01 0,10 0,14 0,18 0,26 0,30 0,33

0,02 0,18 0,26 0,33 0,45 0,51 0,55

0,03 0,26 0,37 0,46 0,60 0,66 0,70

0,04 0,34 0,46 0,56 0,71 0,76 0,80

0,05 0,40 0,54 0,64 0,79 0,83 0,87

0,06 0,46 0,60 0,71 0,84 0,89 0,92

0,07 0,52 0,66 0,77 0,89 0,92 0,95

0,08 0,57 0,71 0,81 0,92 0,95 0,96

0,09 0,61 0,76 0,85 0,94 0,96 0,98

Gruß

Joachim

[ August 07, 2003, 08:09 AM: Message edited by: Scarhead ]

[Michael Emrys]

Interesting approach, Joachim. I've considered something along these lines, but never actually gone to the trouble.

Michael

[Lars]

Originally posted by Sgt Bonehead:

Anyway, what do you guys leave as Random, and what do you select, when setting up a QB? Just curious how I can cut down on the "you gotta be kidding" factor without taking the spontaneity out.

Try medium woods, hills, village, summer time frame, day, good weather, etc. and then let the computer pick the forces . You should get at least a halfway decent shot at a good game that way. Play around with the random settings a bit till you find what works best for you.

Remember, the computer can make some pretty screwy picks for itself too.

[Andrew H.]

I always let the computer pick my forces in QBs - if you pick your own forces, you will always do a better job than the computer, since it's quite easy to choose the optimum force, and the computer rarely does so. Sometimes the picks will be unplayable, though - but it's not hard to recognize this and just start again. I.e., if its March 1942 and you get two platoons of Pz IIFs in an "Armor" type game, you should just surrender and let the computer pick again, since the Allied player is guaranteed to have T-34s or KVs or both. Pz IIs in June '41 can be playable, though, sometimes you're just facing BTs, T-26s, and armored cars.

Here are my settings (as much as I can remember) - I choose random weather (I think clear and dry favors the Germans and is unrealistic for Russia), random terrain, Armor type battle, whatever kind of battle I feel like (usually ME, attack, or assault), fit troops, no casualties. I usually make the Russian quality Low and the German quality High - this often leads to very interesting battles, especially if the Germans end up crack/elite - because they will have so few vehicles, whereas if the Russians end up conscript, they'll have tons of vehicles. And a conscript KV is *still* a KV.

An elite PzII has to be the biggest waste of trained crew...

If I'm playing the Allies, sometimes I'll set them to "regular" and the Germans to "High" - just because I don't like the large number of troops I have to push if Low turns out to be a bunch of conscripts.

In many cases, though, I think that the Allied players is best served by Green troops - at least in armor heavy battles. This is probably not surprising. Green troops behave okay, and, more importantly, you get a good number of them. Often having a few more vehicles makes all the difference. And if the StuG gets the drop on you at 1000 meters, it doesn't matter how good your crew is.

[Jack Carr]

I find the following parameters make a very enjoyable QB:

1250 to 1500 points - Soviet player +25 to +50%

Me - German defender (Armor)

Opponent - Soviet attacker (Armor)

October 44 - Give or take 2 months.

Central front

Weather conditions - Random

Terrain - Small hills or gentle slopes, village or farmland, moderate foliage

Lottsa nail biting fun!

[Ex Bellator]

If you are playing smaller or medium QBs (1500 pts or less) then the Combined Arms setting definitely causes the Ai problems purchasing tanks, so you will rarely see any worthwhile armour to pound with or against. ISTM that the Ai usually tries to purchase a whole platoon and if it can't gives up on even a semi-decent tank arm and goes for the odd halftrack instead.

A sizeable tank force will be chosen if you select 'Unrestricted'. Of course if you know this it can spoil the fun somewhat. If you leave the force type to 'Random' then from my experience you will almost always get a very infantry heavy force. Also if you set the map type to 'Random' you will usually get a flat and open type map, which will often lead to the pointless situation you found yourself in.

When I do play the Ai I'm always attacking it, for me after all these CM games defending against it is way too easy. So I'd pick something like 1500 Pt Probe/Attack/Assault, Unrestricted Force, Farmland/Lt Trees/Lt Hills (I prefer open maps in Russia). You can randomise everything else including date and you will have a good but winnable challenge on your hands.

[PeterX]

However, I inherently distrust the AI's penchant for picking wierd equipment.

This, IMO, is a myth. Though the computer can deliver up some quirky units, the kampfgruppes you get are more realistic than than ones seen with Human Selection. With the latter you get force repetition due to predjudice, continually refined price optimization, and the 'armor race' fear factor. Thus, ultra common WW2 units as, for example, AT Rifle teams or Russian tankettes, are seldom glimpsed in HS battles between price-savvy players. Much of the depth and richness of CMBB is forfeited.

BFC, Steve and Charles, made it possible with Random Variable selection to come up with a varied yet representative forces for the respectitve fronts and time periods. Give them credit!

I, and another forum member, considered creating a ladder website built around Random battles. I came up with an Excel-based parameter generator that spits out the battle conditions, game ID#, and passwords for each player. We excluded 'Open Trees Coverage' and 'Blizzard' as conditions. Every other parameter is set to random, except Fitness.

Each battle is saved and passworded using the Hotseat procedure then mailed to the two opponents. It's an airtight process; no cheating is possible. There are a few conditions that will entail discarding a Random Battle from consideration during the creation (which takes a couple of minutes per battle).

1- No LOS into the setup zones.

2- The attacking force will not be saddled with Conscripts. (Challenging, but arduous due to the micro-management required)

3- If the Force Type selected is 'Armor' and the Tree Coverage is 'Heavy' the battle is re-generated.

4- Severe Armor mis-match (Perhaps. The player can overcome this, IMO, more often than recognized).

Also, the attacker receives a 10% bonus in Snow terrain and, if Armored, in conditions conducive to frequent bogging like Damp or Rain. The force size varies, in the generator, between 800 and 1500 pts. In my experience, only 1 or 2 battles in ten had to be re-generated.

Most of the battles, we've played have been interesting, reasonably balanced, varied, and competitve. Due to unexpected force compositions, weather, and troop morale one has to strategize on the fly. There's a cool suspense element in getting your first file due to the unknown factor. Our current battle, a Green Soviet battalion (me) attacking a force of Regular Hungarians, is going down to the wire.

We didn't pursue this project due to a disinclination to acquire web page creation skills, but if anyone's curious, I can email him the Excel file.

-Peter

[ August 07, 2003, 08:17 PM: Message edited by: PeterX ]

[Ex Bellator]

BFC, Steve and Charles, made it possible with Random Variable selection to come up with a varied yet representative forces for the respectitve fronts and time periods. Give them credit!

Sure it's possible, but it isn't probable. I've started more all random generated battles than I can remember, and given up on seemingly most because the majority of the time you will get an open type map and an infantry type force.

If you look at the parameters and assuming each choice has an equal chance of occurring it is easy to see why. Hills, 3 out of 5 chance of being virtually flat. Trees 2 out of 4 of being almost none. Force type 3 out of 6 chance of being massively infantry heavy (check out a few C/A AI selections!).

I think the force selection problem may occur because either A) The AI attempts to buy whole platoons of tanks and gives up when it can't, or The AI buys the biggest single infantry unit it can immediately (eg. a full regiment or company) and then is often left with nothing to spend on armour.

I've been interested in how the AI buys it's forces for some time, and if I'm way of the mark as usual then I look forward to being corrected by those in the know.

[ August 07, 2003, 09:28 PM: Message edited by: Rex Bellator ]

[von Lucke]

Actually, I find letting the AI choose forces setting works rather well with points +/- 800. Above that... Well, I hope you like Engineers and half-tracks...

[lcm1947]

Just a thought but you could get someone from the forum to pick the AI forces. You'd know the exact force of course but I would imagine that in real life that was the case a lot of the time anyway. So while it wouldn't be a surprise in what you were going up against at least you wouldn't get a weird ass mix against you. Also if you would pick your force first before finding out what the other force was it would even be fairer. Just a thought as I never allow the AI to pick either force because of the things it sometimes picks Although it normally is Ok if you fight an all infantry game or such.

[IntelWeenie]

I had a reply all written up last night, but my internet connection was cut. :mad: I'll try to reconstruct what I was going to post.

Originally posted by Rex Bellator:

[QBIf you look at the parameters and assuming each choice has an equal chance of occurring it is easy to see why. Hills, 3 out of 5 chance of being virtually flat. Trees 2 out of 4 of being almost none. Force type 3 out of 6 chance of being massively infantry heavy (check out a few C/A AI selections!).[/QB]

These results sound pretty close to reality to me. Assuming each region has an equal chance of being randomly picked, 50% of the time you will get Central or Southern. The majority of the terrain in these areas was wide open and fairly flat. The manual (at work, forgot exact page number. I think it's 178?) says that terrain generation is affected by date and region, so a goodly proportion of 41' - '44 battles should end up with cetral-southern wide open plain maps.

I also think the tendency of the QB generator to pick infantry is actually closer to reality than some may realise (or want). Even at the end of the war, the great majority of untis on both sides in the east were non-motorized infantry. I understand that playing with tanks and TDs is often more fun, especially on open maps, but that wasn't always the historical reality. If I want to guarantee less infantry, I just choose "armor" or "pure armor" force selections instead of leaving it on random.

[Joachim]

Originally posted by IntelWeenie:

I had a reply all written up last night, but my internet connection was cut. :mad: I'll try to reconstruct what I was going to post.

</font><blockquote>quote:</font><hr />Originally posted by Rex Bellator:

[QBIf you look at the parameters and assuming each choice has an equal chance of occurring it is easy to see why. Hills, 3 out of 5 chance of being virtually flat. Trees 2 out of 4 of being almost none. Force type 3 out of 6 chance of being massively infantry heavy (check out a few C/A AI selections!).

These results sound pretty close to reality to me. Assuming each region has an equal chance of being randomly picked, 50% of the time you will get Central or Southern. The majority of the terrain in these areas was wide open and fairly flat. The manual (at work, forgot exact page number. I think it's 178?) says that terrain generation is affected by date and region, so a goodly proportion of 41' - '44 battles should end up with cetral-southern wide open plain maps.

[/QB]</font>