Freyberg

Thanks very much The last few months I've been practising with AI plans, trying to get a bit more variety without having the AI do anything too reckless. Incidentally - I'd be thrilled if you had any screenshots

Agreed. I've been using Macs for decades, installed and played every Combat Mission title since CMBO without a hitch. Their excellent Mac games alone make Battlefront a cut above nearly every other gaming company. ...And the whole Mac vs. whatever argument is so very very passé...

Wow - that's an amazing and really informative series of tests! Thank you

.50 cal is an amazing weapon, but it is a long-range weapon. Stay at least 500m away, preferably further, and your gunner will be very effective - a wonderful weapon against buildings and light armour. It's a very good indirect-fire weapon. The bullets will go straight through buildings, trees and brick walls. The only problem is most vehicles don't carry much ammo.

Both of those features would make a big difference - agreed!

I just ran a very simple test 5 times. You're right. Enemy trigger objectives are ignored, and the touch objective disappeared. Such disappointed...

I'm not calling you a liar, but are you sure about that...? I don't use them very often (sheer laziness), but I could swear that the times I have tried they have worked. I wonder if the terrain objectives that get converted to 'occupy' are just the 'touch' objectives (even then, I'm sure these have worked in some of the QB maps that come with the games, but I may be remembering Scenarios - I've never played with touch objectives myself). The way I've done it is to set a very wide time range for the event and make use of the trigger, so that both the trigger and the timing are activated. I didn't do a careful test, but it seemed to work. And this was in a QB map. Has anyone else tried this and could shed some light on it...?

There are lots of little random oddities, dice rolls so to speak, that make the game what it is, that make the pixel-troops more unpredictable - it makes the game more exciting, realistic and fun.

I used to do that, but I don't find it necessary with the current version of CM

That's an odd little bug - it happens in the WWII titles too

In fairness, close assault on an unsupported tank shouldn't be hard. I don't believe it was in RL.

Yeah, those 57mm AP shells really fly - great penetration. They don't always do enough damage for a kill, but they're lethal often enough.

Certainly it is less likely to deflect than a high velocity projectile, but the jet still has to penetrate a certain width of steel, which can vary depending on the angle it strikes at. Plus, as a general observation, shaped charges are dependent on the energy and dimensions of the explosive - and I think the Tiger I was right at the upper limit against which the 60mm bazooka round could be expected to have any effect.

An extremely minor feature I would love, would be a menu item for supply trucks, to alter the mix of what they have on board. For example with Commonwealth, I often don't need 9mm (many nations don't use it); and I would love to able to buy more 2" mortar rounds (I often buy carrier platoons just for the ammo); 3" mortar rounds would be amazing (though obviously, you'd need to make it expensive); in urban maps, the option for extra .45 cal would come in handy; .50 cal wouldn't hurt (although plenty of vehicles store it); and I agree with those who've asked in the past for the option to be able to acquire grenades, rifle grenades and demo charges.

I do something similar. I'll set up a single zone, with no orders just 'setup', covering the whole defender setup zone, set to 'ambush 1000m', so the AI can put AT guns wherever it wants (clever placement of AT guns is what it does best). Most of the time it works - but sometimes the AT guns (or some of them) end up 'limbered', trundling along slowly like sitting ducks.

Agreed

One thing that annoys me a little with AI plans in QBs is that sometimes (not always) the AI seems to choose inappropriate groups for a weapon type - the most annoying one being a group with movement otrders for AT guns. When I set up an AI plan, I usually have one (or two, on a larger map with more groups) completely static AI group for things like AT guns, but the AI doesn't always select it.

It's just a game thing...

I think the editor is a great tool, I'm slowly becoming more skilled at using the AI, which can work very well on defence and in limited counterattacks, and I would also like to see the AI and its interface improved. However, while I don't object to the suggestions made, my wishes are completely different. A lot of these suggestions seem to be asking for more micromanaging, when I would like to see less. It's probably some time away, and I'm not at all displeased with the AI as it currently works, but I would like to see the day when you could just define one or two broad attacks (like: feint left, attack right), and rely on the AI to use recon, seek cover, use combined arms, respond to observed enemy and pace its attack, defence or fallback based on the actions of the human opponent's forces. It actually does not a bad job of this now. I generally get better results with fairly simple AI plans: 3-6 groups, painting broad areas of the map, and long overlapping time periods for each action. Micromanaging requires you to guess what the human player will do, and I just can't get that to work, even against myself!

Legendary ...bloody potato mashers...

I admire your commitment to good statistical methodology and very much look forward to your results

Another very unscientific test. 1943, same as before (M4 early v. PzIVG) 10min shootout. No orders given beyond setup. This time with limited cover (CMFI tree-type D (small, low tree), one tree per action square, tanks positioned just inside the treeline, line staggered at 1-2 action squares). Damp ground (last time was 'dry' so there was a lot of dust) ~ 200m LIGHT COVER, starting buttoned PzIV: 4 kills Sherman 1 kill (winner: PzIV) LIGHT COVER, starting unbuttoned PzIVs 2 kills Sherman 2 kills + 2 bad crew kills (winner: Sherman) ~ 500m LIGHT COVER, starting buttoned PzIV: 1 kill Sherman: nil (winner: PzIV) LIGHT COVER, starting unbuttoned PzIV: 2 kills Sherman: 1 kills + 1 dismount (draw) ~ 750m LIGHT COVER, starting buttoned PzIV: nil Sherman: 1 kill (winner: Sherman - some crew casualties each side) LIGHT COVER, starting unbuttoned PzIV: 2 kills Sherman: 1 kill, 1 dismount (draw: some crew casualties each side) OPEN TERRAIN, starting unbuttoned PzIV: 6 kills (in 3 minutes) Shermans: 1 kill, 2 dismounts (clear winner: PzIV) ~1000m LIGHT COVER, starting buttoned PzIV: 0 kills Sherman: 1 kill (winner: Sherman) LIGHT COVER, starting unbuttoned PzIV: 4 kills Sherman: 1 kill (clear winner: PzIV) OPEN, unbuttoned PzIV: 5 kills Shermans 2 kills, 1 mostly killed crew (stil manned), 2 dismounts (draw...? 1 functioning tank each at end - most of the action in the first 2 minutes - then they buttoned up and couldn't spot each other) ~1500m LIGHT COVER, starting buttoned PzIV: 2 kills Sherman 1 kill, 1 dismount (draw: a few crew kills each side) LIGHT COVER, starting unbuttoned PzIV: 3 kills Sherman: 1 kill (winner: PzIV both teams spotted very quickly, Pz got the range sooner, 3 Shermans killed to one PzIV in 2 minutes, no more kills once buttoned up) OPEN TERRAIN, unbuttoned PzIV: 3 kills, 1 dismount, 1 severe damage Shermans 3 kills, 2 dismounts, 1 gun damage (Axis team out of action) (winner: Sherman Within 2 minutes, each down to one functioning tank each PzIV: 6 wins (and more decisive wins) Sherman: 4 wins 4 draws What did I learn? 1) Buttoning tanks and using them at close range is a mug's game. 2) The two tanks are very closely matched, but the PzIV has an edge overall (especially unbuttoned). Is it enough of an edge to account for the point differential? Who knows.... 3) Running little tests is fun, but running statistically significant, large tests would be like work, so this probably proves nothing

A totally unscientific test (1943, M4 early, PzIVG) Straight shootout, Italy, flat map, no orders given, six tanks per side (Regular, 0, 0) 150m starting buttoned: win PzIVs starting unbuttoned: win Sherman M4s 500m starting buttoned: win Shermans starting unbuttoned: easy win PzIVs 1000m starting buttoned: PzIVs starting unbuttoned: PzIVs

I used to enjoy PBEM, although I got beaten a lot, but I couldn't maintain a respectable turn rate - always too much going on in RL. But the AI is cool too. I would probably enjoy multiplayer in RT if I had the time.

In QBs, I give the AI an observer and otherwise let it control its own barrages - the result is usually the very efficient use of short, accurate surprise barrages on any troop concentration I am careless enough to leave for more than a few minutes. The AI seems very good at using artillery...