Quick Battles are fun

[Rocky Balboa]

I don't know. I'm responsible for the CMBN QB Maps but not the CMFI QB's. We are talking about a pretty large, time sucking project, but CMBN is worth it. I do have a method in mind that might speed up the process to maybe 30minutes per map.

Mark, Great Job as usual. If it's a matter of time and therefore $ then update the CMBN QB maps and release them as a pack. I'll pay an extra $10 to have those maps updated with the new AI triggers ....

[Childress]

BF installed triggers in the RT QB maps? Doesn't seem possible with the near infinite permutations involved.

[weta_nz]

Childress - I'm not sure of the 'technicalities' of the AI triggers as I haven't looked at the QB maps with the editor. My guess is that with the right combination of units (attacker and defender) they can work very well (my experience). However I'm sure they do not work for all situations.

But that is the great thing about single player QBs - if the AI get's it wrong you can just request a cease fire or stop the game. If the AI gets it right you really get a unique battle with some tough tactical problems to work through - even if the end result for me is total victory - I still have learnt some valuable lessons. All with minimal effort on my part.

On a slighty offf topic note - when I'm playing QBs against the AI I have a couple of 'house' rules so I don't have too much of an unfair advantage.

1) I only use an HQ to arty a known enemy position if the HQ actually 'knows' there is an enemy there as well. e.g at least a question mark via the radio network

2) I will only area fire with units who have at least some idea there is enemy in those trees. Though at very close ranges I will break this rule.

I would be interested to hear if other people use any 'house' rules against the AI?

[Macisle]

I would be interested to hear if other people use any 'house' rules against the AI?

My only consistent house rules are that I give myself Typical infantry or lower in quality (I almost always choose Typical regular infantry for my troops) and give the AI tougher tank models than I choose, and often better quality infantry (like SS or Paratroops in CMBN).

I can't remember the last time I gave myself Tigers and Panthers as the Germans in anything other than a quick test when I bought CMRT. I fight against them a lot though, when playing SP QBs.

[marioa]

> I have found a probable bug in Red Thunder qb, i selected probe tiny battle, the AI purchase either forces now the german have several at team on board of halftrack with 88 antitank cannon's, when unload i see only the at crew but the weapon are not loaded and not limbered, thei are simply missed, i have a save to upload but no clue how to upload. Any help will be appreciated.

> Thanks in advance

> marioa

[Andrew H.]

I would be interested to hear if other people use any 'house' rules against the AI?

I play "probe" against the AI, with me being the attacker. I usually think this gives the AI the best chance. (I will have to revisit this now that triggers, etc. are in the game).

I usually play "mix" and have automatic selection on for both sides; that way, I don't know what I'm facing. (And if I end up with HTs vs heavier armor...well, I have to deal with it).

ATM I've been playing a bunch of QBs where I pick my equipment...but that's mostly because I want to experiment with newer units.

I also find that I learn interesting things from what happens with some of the AI's picks. For example, I don't think I would never buy an ATR Coy. in a 1500 point game. But when the AI did, they ended up really kind of freaking out my tanks. Even Panthers - 3-4 hits in a turn from units that they couldn't spot and the Panthers started to lose morale and wanted to reverse away.

I didn't know that there was a such thing as an AT Rifle Co. until I saw the AI's. 390 points, 27 AT rifles. You know you want one...

[MarkEzra]

send the save game to me... markDOTezra3591ATgmailDOTcom

[Reanimator]

I appreciate the hard work you do with the quick battles because they are generally what I play as well. But I have to say my experience seems different from post people here. I generally get maybe 1 good battle out of 5. Now I don't know if it is maybe I don't wait long enough in them to see the AI do something or I have the wrong settings (usually an attack in a medium/large map) but most of the time I'll get an AI force that is scattered around the map in places that are not in los of the objective or approaches to it. I will easily take the objective and then just quit. Too often I have not even seen the enemy and had to actually try and hunt it down on the map to see any action at all.

Poking around the QB maps in the editor I notice there are a lot of plans that are on the "not used" setting but have actual orders. Are these going to be used in the future when you guys get some more time or are they just sorta failed ideas?

[womble]

...an AI force that is scattered around the map in places that are not in los of the objective or approaches to it...

Poking around the QB maps in the editor I notice there are a lot of plans that are on the "not used" setting but have actual orders. Are these going to be used in the future when you guys get some more time or are they just sorta failed ideas?

Oo. This chimes with something Proambulator mentions in his "30 minutes with RT" video: he says that he's not sure the "plan use frequency" setting makes any difference. If that's true, and "Not used" gets the same chance of a plan being used as "Used Frequently", then gash plans might get used when they shouldn't...

[Macisle]

Reanimator,

Not all SP QBs are great. I think we are mostly saying that we notice a definite improvement over those in CMBN, particularly with regards to the micro-level performance of infantry.

What kind of force settings are you using? I know it's more fun to have the AI's forces be unknown, but you can usually guarantee more of a challenge by picking the enemy force yourself and making sure that it is of sufficient size and mix to provide adequate opposition. My last "Mix" auto-pick by the AI yielded an opposing force of 2 PIVs and 4 AA vehicles with one team of supporting infantry. So, that problem is still there--especially with smaller force-size parameters.

As for the AI setting up in ineffective locations, using larger maps will likely make this happen more often. So far in CMRT, I'm using medium maps and giving the AI something like 1-2 companies of infantry, a platoon of AT guns and 1-3 platoons of tanks. Doing something like that on a medium/small map can help improve things a lot. Choosing maps with obvious choke points, like having to cross a bridge as the attacker, can help, too.

[Reanimator]

Reanimator,

Not all SP QBs are great. I think we are mostly saying that we notice a definite improvement over those in CMBN, particularly with regards to the micro-level performance of infantry.

What kind of force settings are you using? I know it's more fun to have the AI's forces be unknown, but you can usually guarantee more of a challenge by picking the enemy force yourself and making sure that it is of sufficient size and mix to provide adequate opposition. My last "Mix" auto-pick by the AI yielded an opposing force of 2 PIVs and 4 AA vehicles with one team of supporting infantry. So, that problem is still there--especially with smaller force-size parameters.

As for the AI setting up in ineffective locations, using larger maps will likely make this happen more often. So far in CMRT, I'm using medium maps and giving the AI something like 1-2 companies of infantry, a platoon of AT guns and 1-3 platoons of tanks. Doing something like that on a medium/small map can help improve things a lot. Choosing maps with obvious choke points, like having to cross a bridge as the attacker, can help, too.

I definitely agree they have gotten better, it is why I said I am glad for their hard work in the matter. I appreciate them taking the time to flesh out this part of the game.

I have moved to do exactly as you do, smaller battles and pick the enemy force myself. Though in the latter I have seen some improvement as well, perhaps it is just due to the TOE but I have run into fewer instances of ridiculous force compositions. So now like you I generally do medium battles with an AI mix that I pick myself.

It just seems like the obvious thing to do would be to have the ai either in or or in los of the objective. That way even if there is an 88 spawned it can have a chance of being part of the battle and in this way you could ensure that whatever force happened along it would at least be visible for your average game. Then after you work on the QB system and refine get fancy. A lot of maps I can see that whoever designed it wanted the AI to hide for a bit and then jump out, which is a great idea when it works, but it doesn't really seem to work enough times for it to justify it due to the random nature of placing static (or near static) guns, etc. in random groups.

I think if they could add a function to the editor where a type of unit such as infantry, gun, tank, could be assigned a group this would be a great thing. Or, even if that is not possible, perhaps define a group by speed? Then it seems you could reliably set up AI ambushes.

[womble]

One way I think you can, if you're picking AI forces yourself, mitigate the "putting units in inappropriate groups" problem is by making sure there are enough formations in the force to have "some of everything" in each group. Rather than picking one Battalion "head", then chopping out everything but the AT platoon (4 guns) and a company (3 rifle platoons and a heavy platoon) of infantry, pick four Battalion headers and chop out everything but one AT gun, some Rifle squads and an MG or two from the Hvy Platoon for each Battalion. AFAICT, the AI assigns alternating Formations to the default 2 groups, and will leave one group empty if you just pick the one Formation. Or if you pick a Formation for infantry and another for Armour, you'll have all the Infantry in one setup zone, following one set of orders, and all the Armour elsewhere, operating according to a different timetable which probably won't support your infantry's route.

[MarkEzra]

[quote=Reanimator;1526419

Poking around the QB maps in the editor I notice there are a lot of plans that are on the "not used" setting but have actual orders. Are these going to be used in the future when you guys get some more time or are they just sorta failed ideas?

[Buzz]

What Mark said.... Good insights.

I am often in the "Just play it for the fun of blowing things up." time to relax from the day and what better way to .."affect your overall enjoyment" than... "blowing things up!" and no one gets hurt

[z1812]

I find very small battles work very well in QB's.

I usually choose a small map and use a mixed force. Generally an attack or probe.

I choose the forces myself. Usually a few Squads. Sometimes with a vehicle or 2.

The tactical situations that arise are quite interesting and challenging.

For instance 2 German Straggler squads, with a light mortar, a HMG, 1 Puma and a half track will defend against 3 Soviet squads, a HMG, an ATR unit, a Tank Hunter unit and a 82 mm mortar.

That particular combination resulted in a tie the first game and a Soviet Minor during the second. Lots of tense moments. Great if you like micromanagement.

[Collingwood]

In CMRT I usually get quite satisfactory QBs by:

Don't let the AI select the forces - I choose forces for the AI that I would take if it were me - and as per Womble's excellent advice: I make up kampfgruppen/battlegroups.

I give the AI up to 50% more points than me, and select meeting engagement.

I don't take Arty - and medium calibre mortars at most. I don't take Air support.

I also sometimes try to avoid micromanaging squads and instead issue orders that will cover several turns, on a platoon or even company level. Unless there is something going obviously pear shaped I don't step in for a few turns - I let my plan play out and enjoy the chaos and carnage that ensues.

Mark makes a good point: if you play with the mindset of using all available tricks to win (as you would against a human) then the AI will come off worst always. Finally, against the AI try not to thoroughly research before each turn all the LOS and the perfect action spot to move your men to: instead make reasonable orders with a general view of the terrain.

[MikeyD]

The only thing QB can't give you (something I miss) is a 'backstory' in the orders section.

But you can always provide one yourself. As the opening screen is slowly processing just start making up stuff - "We're now 4km from Minsk. Oleg can just see the rooftop of his family home on the far side of the valley, a home that he hasn't seen in four years. Once when he was a boy a Roma fortune teller had cursed him, told him he would die on his doorstep on his birthday. And today, 14 July, is his birthday." ...and then the battle starts!

[Macisle]

I thought I'd post some pics of my current SP QB as fodder for conversation. I'm using map Village-Forest Attack 007. All units are "Typical" quality and I started with an hour on the clock in Wego.

I chose the following German units for my side:

1 X Company of Panzergrenadiers (comes with 1 module of 81mm off-map)

2 X PIV Platoons

I chose the following for the AI:

2 X Companies of Regular 44 Rifles

1 X ATG platoon

1 X T34/85 Platoon

1 X T34/76 Platoon

I decided to experiment again with a TIB ("Teutonic Iron Broom") maneuver (tanks in a line with infantry just behind them), going through the woods along the East map edge to put me in a jump-off position to flank the village defenses.

I lost three men (1 to friendly ordnance splatter) and one tank (AT grenade from last two hiding stragglers just last turn) along the way to my jump-off point. I think I destroyed most of a platoon of defenders, with the survivors escaping towards the village as shown by the contact icons in front of my advancing forces below:

About 5 turns ago, I noticed enemy infantry beginning to reposition from the center woods near the Northern friendly map edge, towards the village. I tried to interdict with HMGs from my Eastern woods line, but my MG teams could not get good LOS.

Normally, I would have units scouting and probing other parts of the map, but for this, I'm going all-in in my chosen direction.

Around the same time that the enemy infantry began to reposition from the center-forward woods, a T34/76 (dubbed "The Wild Rover") also repositioned from the right map edge woods towards the village.

My infantry has just spotted 3 more tanks, two T34/85s and another T34/76 ("A"), which took a shot at my advancing infantry (a green team soiled their pants, but took no casualties as yet). My woods-TIB took a whopping 40 minutes, and I've just finished turn 1 of my actual attack. Now, I've only got 19 minutes left to take the town and have to hurry. Things are going to get heavy fast!

There are still two unspotted tanks (where the Wild Rover came from perhaps?) and three ATGs lurking out there somewhere. And, I've only just stomped on the Soviet infantry's foot, as it were. So there are lots more of them out there, too.

As you can see, the AI has covered the likely approaches with something and has put all of its tanks in concealment terrain, except for the wandering Wild Rover (a juicy target, indeed!).

Thoughts?

[Macisle]

One way I think you can, if you're picking AI forces yourself, mitigate the "putting units in inappropriate groups" problem is by making sure there are enough formations in the force to have "some of everything" in each group. Rather than picking one Battalion "head", then chopping out everything but the AT platoon (4 guns) and a company (3 rifle platoons and a heavy platoon) of infantry, pick four Battalion headers and chop out everything but one AT gun, some Rifle squads and an MG or two from the Hvy Platoon for each Battalion. AFAICT, the AI assigns alternating Formations to the default 2 groups, and will leave one group empty if you just pick the one Formation. Or if you pick a Formation for infantry and another for Armour, you'll have all the Infantry in one setup zone, following one set of orders, and all the Armour elsewhere, operating according to a different timetable which probably won't support your infantry's route.

Very interesting. Thanks for the info, womble. I'll start putting this into practice with my next QB.

[Reanimator]

So can we accurately know what will go in each group, as in if there are 3 groups then if we pick 3 battalions in the QB unit selection screen we will end up with the first batt in the first group and so on?

If so then it seems to me the obvious solution would be to standardize the QB map pool into something like 2 groups in tiny/small where 1 group is static and 1 is mobile, 3 in medium 2 static 1 mobile, 5 in large/huge where 3 are static and 2 mobile. Obviously this is just the first idea that popped into my head, but with such a system a player would be able to use a random map and be assured when he picks the AIs forces those troops can be put into the plan where they will assuredly be used.

It just seems like if we CAN do QBs where we can reproduce a result in the AI always being able to use its forces to full effect then we ought to do so and in such a way that it can easily be explained to new players easily.

[womble]

So can we accurately know what will go in each group, as in if there are 3 groups then if we pick 3 battalions in the QB unit selection screen we will end up with the first batt in the first group and so on?

I believe to be the case, but I haven't really done enough tests to eliminate the possibility that I saw a fluke repetition.

If so then it seems to me the obvious solution would be to standardize the QB map pool...

They already are standardised to a large degree, because MarkEzra did them all and used a common approach. Things are changing a bit with RT/v3 because Triggers make the environment different, but in BN most maps have 2 groups.

...2 groups in tiny/small where 1 group is static and 1 is mobile...

I'm assuming you're mostly referring to the defender plan here

It just seems like if we CAN do QBs where we can reproduce a result in the AI always being able to use its forces to full effect then we ought to do so and in such a way that it can easily be explained to new players easily.

It seems to me that if such a "rule" can be developed, it will be just as easy to "explain" it to the autopicker, too...

[Reanimator]

I just see it as a temporary solution until BF decide to actually spend some time on the QB portion of the game.

Look at the old cm games. When you picked a QB and put it on map generator and random forces, you pretty much knew what you were going to get. The AI, regardless of what it had would line up across the map so even if it was easy it was still a game. With the current system unless I know what to expect from the map plan I cannot even pick a force that I'm sure will even show up anywhere near where I'll send my troops.

I understand MarkEzra is the man and I appreciate all his hard work. But I just feel that "standardized to a large degree" is just not good enough. Give us a framework that we can use on every map of a type and suddenly everything opens up. I teach a lot of new guys this game and most of the time I find them intimidated by scenarios. If I could say "just run a few quick battles and do x y and z" to get them comfortable with the GUI and the way the tacAI responds it would be extremely helpful.

Plus I would enjoy the hell out of it too.

[Jock Tamson]

There are still two unspotted tanks

The QBs (and the improved AI behaviour therein) have garnered quite a few plaudits since release. I would hope that would provide some momentum to keep polishing the single player experience - which is still the experience for the vast majority:

Force composition by xml or similar. Would enable players to easily exchange forces to play against in QBs. Would allow QBs to be set up extremely rapidly - pick a map, browse to the XML.

Force composition chance modifiers - let me choose to face 2 to 6 AI T34s, rather than knowing the exact composition. Totally changes the way you play the QB if you don't know what you are facing.

[poesel]

Force composition by xml or similar. Would enable players to easily exchange forces to play against in QBs. Would allow QBs to be set up extremely rapidly - pick a map, browse to the XML.

+1 - that would be very useful for QBs and campaigns.

[Lets_All_Fight]

I'm loving the Qbs. They are an element of the game I haven't played very much since CM1 but they are beginning to come into their own.