In my part of the world, roe deer are the undisputed contour kings. They know their territories intimately, and when surprised in open country invariably dash for the nearest hollow, or retreat behind the closest LoS-breaking rise or hillock. To prevail in my last Armored Brigade 2 ‘random skirmish’ I had to think like a nervous Capreolus capreolus.
This week I’ve been testing the abilities of AB2’s AI with the help of the game’s well-appointed skirmish generator. The scenario I rustled up yesterday was designed to reveal offensive strengths and weaknesses, but ended up testing me just as stringently as my synthetic adversary.
Utilizing an uncluttered region of the Lubeck master map, it was a ‘meeting engagement’ pitting a platoon of M60A3s against a much larger Soviet mechanised force. I went in expecting the computer controlled/picked red team to make short work of my four tanks, two unarmed Kiowa choppers, two Stinger teams, and two infantry scout squads.
Placing my earthbound intel gatherers in two copses atop the transverse ridge that screened one end of the battlefield from the other, my intention was to immediately move my M60A3s to positions behind the ridge and clobber the Soviet AFVs as they skylined. How would Vera (Not the AI’s official moniker) cope with a classic unoriginal reverse slope defence?
Pretty well, it turned out. While my ambushers made fairly short work of the PT-76s that tipped the Soviet spear, and gave the infantry-laden BTR-60s that came along next a good kicking, things started to go awry when T-62s began appearing on the horizon. Struggling to penetrate the frontal armour of the dome-turreted Eastern Bloc MBTs, the M60A3s eventually found themselves outflanked and out-of-order.
Could I do better I wondered? Studying the rolling battlefield from various low viewpoints, and checking hunches with the help of the LoS tool, a new plan began crystalising. Instead of parking my tanks behind the ridge, I’d hurry them to a village on the western edge of the map. Firing from this position, they’d have, I hoped, more opportunities for flank shots and more cover should duels with the T-64s develop.
Playthrough # 2 started promisingly. As before, Vera led with PT-76s closely followed by BTR-60s, and beelined for the objective on the southern slope of the ridge. My scouts in the woods watched as the red team’s AFV’s rolled up the road then, one after another, succumbed to 105mm killshots fired from Havighorst. When the first T-64 was slain by a sabot round a few seconds after trundling into view, an unlikely coup seemed possible.
Hope can curdle into despair awfully quickly on a hot AB2 battlefield. A minute after killing the lead T-64, I was pondering the hulks of three dead American AFVs and a downed Kiowa. Once again, my tanks’ inability to bodkin the armour of their adversaries had cost them dear. I restarted aware that I’d need to engage in sly, proactive flanking, to come out on top in this leathery challenge
By chance the rectangle of Schleswig-Holstein I’d picked as a venue seemed sympathetic to my plight. The LoS tool revealed a patch of low ground just north of Havighorst invisible from almost everywhere on the map. If two of my angry houses rushed to this spot rather than lingering in the village, perhaps they could get the jump on their distracted nemeses, the T-64s.
The plan might have worked had the distractors lived long enough to play their part. Because the tanks in the villages perished before the flankers got stuck in, playthrough #3 ended in defeat #3.
At this point in proceedings I can’t say I was taking much interest in AB2’s solid if unimaginative offensive AI. I simply wanted to prove I could pass the taxing tactics exam I’d accidentally created for myself.
The first ten minutes of attempt #4 followed a, by now, familiar pattern. The M60A3s mauled the Soviet vanguard from Havighorst. The difference this time was that after this opening of accounts, I ordered all four trundlers to the patch of dead ground near the map’s western edge.
From there they launched an attack, that, though it resulted in the loss of one of their number, eventually left the red team tankless and hopeless. At last I had the result I craved!
If Vera had purchased the odd Su-25 or Mi-24 with her points allowance, I suspect no amount of cunning contour exploitation on my part would have given me victory. One of the first things AB2 drums into you is the value of air assets and air superiority. I might still be analysing and experimenting if she’d used her truck mounted mortars more intelligently too. These barely contributed and ended up KOed close to the southern objective. Obviously, a human player would have kept them in the rear.
All-in-all though I feel she handles her assets in situations such as these, pretty sagely. Give her the firepower advantage most military theorists recommend when attacking, and she generally won’t embarrass herself.
What experiments like this also prove is that when a game possesses a skirmish generator as flexible as AB2’s, a folder bulging with authored scenarios isn’t really necessary. ‘Tim05a’ took five minutes to set-up and entertained me for an entire evening. Potentially, there are tens of thousands of battles just as interesting, just as singular, waiting to be generated.
And, of course, the same corker of a battle generator powers the game’s dynamic campaigns. I’ve not tried said campaigns yet, but having dipped into the latest three videos created by Corian33, the Cornerite behind this extremely watchable YouTube channel, I’m itching to.