'Llo Bros,

While I definitely enjoy SB Pro, and have gotten so I actually (in single-player) win more scenarios than I lose, it is rare that I manage to do so without losing anywhere from half to three-fourths of my force. Also, it seems I am more-often-than not compelled to play 'fireman'; i.e. 'F10' all over the battlefield (re-)directing AI controlled units.

OK, this may truly be a lame question but, allow me to ask: is this generally how most scenarios can expect to run, or I am likely missing some fundamental things. I.e., have some of you gotten tactically proficient enough that you are generally able to beat most scenarios with considerably fewer losses?

Robert

(No, I am not asking for broad strategy advice, but any links and/or pointers in re will certainly be appreciated. )

I always route units under my control so that if they're entering an area I suspect they may meet opposition in a way to ensure they don't just sit there and take punishment.

As a general thing give this a try. Set "guard" tactics at a unit's final waypoint. Before the unit gets there set a retreat route with no "embark if..." instructions to a way point with no tactics set, and another route with engage tactics from that to another waypoint with "guard" tactics.

This will result in your unit embarking on the retreat route if it comes under direct or inderect fire, then from the waypoint at the end of the retreat route, it will move to a new fire position.

It sounds fairly longwinded, but give it a try and you'll understand when you see it inpractice.

Also, with the unit under your direct command, never remain stationary for longer than two minutes if you suspect your unit has been spotted by the enemy. If you do, you'll have arty raining down on you. And once in contact, redeploy to fresh positions using whatever cover the landscape presents after taking a few shots. Do not sit there and try to slug it out.

Hope I explained that clearly, and that it helped a bit.

You are correct in your impression that scenarios will frequently be run with massive losses on your side, and still achieve a victory. This isn't realistic, but is more 'fun': After all, in real life you wouldn't launch an attack unless you were pretty sure of winning in the first place.

There are a small number of scenarios out there where it is possible to win without losing a single vehicle, or very few. When correctly created, the scoring system takes this into account so that you can still have a challenge in achieving a 'win' on points even if taking the ground wasn't a problem. Unfortunately, properly creating scoring is the biggest area where the scenario creators are lax. It's a lot of extra work for little perceived gameplay effect.

NTM

Thanks for the perceptive feedback, ladz.

Jester: Not a long-winded post at all--thanks for taking the time to detail out this tactic. It seems you have hit upon a very solid formula for ensuring that AI units will take care of themselves. This (i.e. routing) seems to be a facet of the game where I am sorely lacking...

3Star: Er, how exactly does SB rack up VPs? Is it just simply far easier for a scenario designer to assign VPs for enemy units knocked out, than for conditions met?

Cheers,

Robert

I have to admit, I can't take any credit at all for the tactic of setting waypoints like that.

Ssnake suggested it in a thread (can't remember if it was here on on ESim's own forums), I tried it out and found it very effective.

The one downside to it is when things get very hectic you could find yourself spending longer than you'd like in map view (especially if you're controlling a force larger than a company). I would suggest that if that's the case, make sure you're in the C/O's vehicle and keep yourself out of the frontline unless you absolutely can't avoid it.

It isn't that complicated, actually:

You can give points for force strengthes - either survivors, or kills. You can include all forces, or pick individual platoons. You can have several of these sub-formulas. That way you can have big points for one or a few high value targets, and/or differentiate between vehicle classes (e.g. a truck kill gets more points than a tank if the player's job is to eliminate a supply convoy).

Then you can give points for regions either defending, or taking them. You can give different point values for minimum and maximum time thresholds, e.g. the player is to defend an area for at least 25 minutes (that's when he gets a point score at all), and no more than 60 minutes (because the defense is part of a delaying operation, so he must avoid encirclement by not staying in place for too long). Say, you assign 50 points for 25 minutes, and 225 points at 60 minutes. The player managed to hold it for 45 minutes and 15 seconds. The score is 50pts + 5pts per additional minute = 50 + 5x15 + 1 = 126
(1 point since 15 seconds are more than 12 seconds but less than two fifth of a minute).

You can, but need not, give bonus points and penalty points for every minute below or above the thresholds; if so, you should set a maximum of bonus/penalty points to avoid excessive penalization or excessive expectations of the player's performance; e.g. if you set the average time to kill a target to 20 seconds (50 points) and a bonus of 10 points per second, and the player managed to kill all targets within 11 seconds (which would be incredibly good in a combat situation with target saturation)... the computer would assume a best possible score of 50 + 20x10 = 250 points. The player would get 50 + 9x10 = 140 - barely above the 50% level of the theoretical maximum (which isn't possible to reach at all since the loader needs 5 seconds and the projectile flight time might add another 2 seconds...).


You can also give points for events and conditions. This is probably rather self-explanatory.


Finally, each of these sub-scores can be activated by the state of events and conditions at the end of the mission. This is useful if the player is supposed to make a choice among mutually exclusive mission goals - e.g. he could try to reach his own troops after being cut off behind enemy lines, or follow that high value target that is passing by near his location to eventually kill it, at the expense of his platoon's survival. Either choice may be valid, but only one goal can be achieved during a 1 hour scenario. Therefore the player should be rated against a formula that only takes into account those sub-scores that are of relevance towards his initial choice of his course of action.


Now, each of these sub-scores adds to the total score. The points that you assign to them are indicators of the relative importance of the various mission goals. If force preservation is your top priority, then about 50% of all points should be given to the survival of blue forces. Give 30% of the points to the secondary mission goal, and the remaining 20% should be distributed to bonus goals (treat this as a rule of thumb).

Steel Beasts will then sum of all sub-scored, divide it by the maximum possible score, and multiply this percentage with the "maximum scaled score" which can be any value. I would recommend using anything from a 100 to 1000 point scale to indicate whether a mission is easy (100) or almost impossible to win (1000).

The other question to ask Robert is what size force is he playing with?

The larger the force the more you're locked into really hammering out detailed routes and actions up front to help from being overwhelmed. If you are in charge of a platoon, you can do more "seat of the pants". For myself I find normally a company level action is about what I can do and still stay on top of things so long as I've done a certain amount of waypoint setting and detailed planning ahead of time. Larger forces lock you into a much more detailed planning and waypoint setting process to keep from just being along for the ride. However even with the detailed planning given how the AI works especially with units like the M901, if you don't give a little personal attention you end up getting them killed easy.

Either way you have to balance the work of fighing your own vehicle with the responsibility of husbanding your entire force. (A problem in no means limited to virtual commanders). All of this is mitigated and facilitated with greater experience with the planning tools.

I like to use the trigger system. After analyzing the mission you can build each likely COA you may embark on as a seperate trigger, which represents a radio call.


Los

Well thought out thread. I've been experimenting myself with many a scenario and have to concur with Los, ssnake. The name of the game in this sim is "planning" as in the planning phase. The more time you spend in this phase before you hit the "start scenario" button, the more control you actually have, depending on size of forces involved.

I've always promoted the use of the planning phase. We added it to the game from the start, and for a reason. It is remarkable how few people learn from their failure to achieve low casualty rates that proper planning prevents piss poor performance.

I hear ya. I'm only now really starting to understand how the different "routes" and "tactics" map commands actually work for the different types of units and situations. I've tended up to now to make sure that each waypoint in the combat area has approximately 3 different routes emanating from it for different situations. The "Embark if" command is quite powerful. Once I start giving the radio commands (triggers), I then move up out of the turret scanning the battle with binoculars, watching the battle unfold...simply amazing. Hopefully in my favor.

Count me among them, but only as far as being ignorant of the subtlies of the planning phase; I do wish to learn more. One of you ladz needs to write the 'strat' article to be posted either here or on Tanksim.com--I'm too lazy to learn the hard way.

Robert

I do have to say this though. Once you have to go off plan, due to unforseen circumstances, then it's easy to get overwhelmed dealing with larger forces. The game gets a little tough when you are trying to change a company or more wrth of units when all you re hearing on the radio is
"2-1 contact"
"this is 2-1 we lost a tank!"
"this is 3-1 we lost another tank!"

That's where the beauty of multiplayer coop comes in.

Of course that's no different than RL too. I always love sitting through a Bde OPORD at NTC and the smug little staff weenies, wrapped in the comfort important of all...powerpoint, rattle off in detail everything that they're going to do to the OPFOR including how many % of the FSE or the FD will be attrited, and then the OPFOR just simply lays a huge smoke screen down across the central corridor, conducts a raid against the BDE headquarers, and all these guys are scattered in all direction and the whole thing falls apart. Great looking plan and presentation, but wargaming COAs and considering what the OPFOR might do "off template" left a little to be desired. That's always ROFL stuff.

Of course being taken down a notch out there is normally the first step to real learning.


It would be nice someday to be able to "train" formations in SB so battledrills could be executed with a minimal of clicks to facilitate the single player.

While I'm at it, does anyone ever give any serious consdideration to infantry in this game? They seem to defend somewhat OK but the articulation needed to pull off decent infantry attacks seems to be something still in the future still. They definately still reflect the tankers bent that infantry are just something to run over or leave overwatching some area or soemthing as opposed to what they really are. To bad we didn'thave some CM type options for our infantry here.


Los

In open terrain infantry cannot conduct assaults against a mechanized force with much hope for success unless they have a massive artillery support. Which would work in SB as well, if you're willing to throw wave after wave of your men into the fray.

Infantry assaulting infantry ... different matter entirely, especially in restrictive terrain, but then again that's not what Steel Beasts attempts to simulate. SB is a vehicle-centric simulation, and like all simulations there's a limit to what you can model in it with hope for realistic results. It does work for some encounters of vehicles and infantry, but not in all of them, last but not least because the current engine doesn't offer dismounts as much concealment as reality often would, and because the freedom degrees of action are severely restricted to get them act somewhat controllable. E.g. if we'd allow them to climb trees you'd for sure find absurd situation when suddenly entire battalions of infantry would occupy the canopy of a forest - sooner or later. Even sophisticated shooters like Half-life 2 don't have computer-controlled forces use successfully the same options that the player has, e.g. I've never even seen AI monsters attempting to stack crates in order to climb to a higher place, let alone grab a gravity gun and then use it in a similarly creative way as players do.

Having said all this, the coming upgrade will let them occupy buildings, boosting their defensive strength in urban terrain massively. Again, it's defense where they get better, but that's really where infantry is strongest in the combined arms complement.

I'll be looking forward to the upgrade, hopefully you'll be able to send infantry into those buildings to close combat and clear thenm as well.

I've been spending some time on some scenarions which entail combat in towns and as far as the attacking side goes, infantry can be made to work somehwat more realistically in that environment. However a company of infantry (combined arms) clearing a Shugart-Gordon sized village can turn into a bit of a click fest.

Los

I really prefer the smaller scenarios with about 10 vehicles or so, the ones where it's possible to complete without any losses and stay in the same vehicle.
Some of the epic scenarios like fulda gap or airport assault that use triggers, while they are fun and spectacular always mean big losses and hopping to about half a dozen tanks to complete, you know the ones where you have about 5mm of turret peeping over a ridge hull down and take about 6 hits in 2 seconds. I also learned to appreciate the power of crunchies, for a while I kind of ignored them as they pile out of vehicles and let the AI take care of it, but actually deploying them particularly in any urban aspect, these guys take out T-80's like there's no tomorrow.

One of the best crunchie defence scenarios IMO is hold on boys where you defend a town against superior odds. Apart from your armour you have 45 troops in the town itself, you can effectively destroy everything entering the streets by deployment and movement of troops and a couple of Bradleys.

Some more notes on infantry...

I think the two keys to use in infantry defense scenario-designing are spawning and of course bunkers.

Spawning:
Set spawning to react to enemies arriving in certain regions or waypoints. This gives you an ambush type event. Or at least lets them ride out any prep bombardment by simulating that they have been in cellars or what not.

Bunkers:
Even if you have guys defending an urban area put them in bunkers since as of now they can't go in buildings. Set the bunkers so that the main area of threat is covered by a building or something so that they will get very close flank shots on enemies. I like to use a 2:1 ratio of squads in bunkers vs squads out of bunkers (in a built up area the squad does a bit of roaming out of course and this works well too.

Los

It's a matter of a few days before we'll release it. I dare a prediction: I'll be this week. We're almost done with the work on the tutorials, and the testers haven't found a critical bug yet.



It does make sense to have bunkers in towns nevertheless, and be it just to present multiple possible locations for infantry even if they aren't occupied. It's also a visual cue that the town might actually be defended.

Wow and I wasn't even trolling for a release date. Good news...

LOs