Surviving Urban Ops #1- long post

[Chelco]

The problem with urban settings is that the distance between good cover and a kill zone is just a couple of meters. I find the maneuvering of whole squads in urban terrain too unwieldy because how unpredictable is the final destination of each squad member. For the time being, I'm very happy with the results I get spliting my squad in three teams and applying the "hunt" command as explained in this thread.

[Pandur]

i think the real challenge will be found in RED vs BLUE or RED vs RED battles.

i for myself have no real problems implement tactics into my game wich work things out "more or less" like i want(i play wego only) and next to the inherent "sideeffects" of the hunt command and the action spot system i find no unbeatable challenge in a red opponent, myself playing as blue, in most scenarios.

that said, i have mostly played red vs red and have had some RED vs BLUE scenarios.

currently i play "rebell hell" and i really have a hard time advancing the 300m into a build up area!

the big problem here is that its dawn and my guys dont see any ****

i took 5-6 casualties so far, wich i had to leave behind for now as the groups cant stop their assault for a single man in open terrain in that case.

i have enough dismounted personell, BMP2 overwatch directly behind them, enemy tracers comming from 100 to 200 meters from different buildings in town. everyone looks but noone spotts

this is the first scenario i had to "force" myself to use "area fire" on a position where i just saw tracers comming from(i hardly used area fire so far).

i even cut down a "suspected" shack with 30mm autocannon fire becouse i used intel i would NOT really have since my foot troops didnt spotted anything so far, and i really mean nothing.

i destroyed 2 ATGMs, these where spotted @150m by my BMP´s and also destroyed by these.

now if you go back playing blue a bit than, you just breeze through the scenarios "more or less" when you faced such problems befor...

[Guest]

The_Capt and metalbrew -- awesome. Thanks for your hard work.

I feel like I'm going to be "gaming" the system now in using these tactics, though. It seems to me, The_Capt, as though these are *applications* of real-world tactics to the game's constraints, not real-world tactics per se (which I would be able to use, say, in CCM, for want of a better example). I'm really looking forward to a stronger system with better TacAI.

I should be able to stack up on a corner, have my corner guy check the corner (even abstractly), and get a Spot based on whether or not he can see the enemy. I should not have to quick-cancel my Hunt order or mess around.

Since the Hunt command currently takes control out of the hands of the player (if you let it run its course) I'd really like it if it did the right thing. Even at a company level it will get prohibitively expensive time-wise to "game" every corner.

Anyhow, thanks again. I've got some more comments about acceptable casualties, but I'll come back later for that, work is pressing.

Cheers.

[The_Capt]

Phillipe,

I hear you but what are you going to do..it is a game. Maybe a "peek" command or have hun do it automatically but we are not there yet.

It is time consuming to "game every corner" so in later write ups we will talk about cutting the corner, getting eyes on etc. But sometimes you will run into a situation when you have to and we have seen a couple ways to do it.

[Guest]

Well, what can I do? Offer to help work on the TacAI I suppose.

You guys think BFC would go for it? I've been considering it from day one but just haven't raised it with them because I wasn't sure I wanted to do it. But... if it's going to take a while to get the TacAI right, I could at least help it take *less* long perhaps.

Hmm.

Edit: Oh, also because I understand what a massive undertaking and commitment it would be and I'd want to give it 110%, and I still haven't figured out how that would work (let alone if they'd need or want the help). That's a factor in why I haven't brought it up as well.

[KNac]

I always though that one single programmer for a computer gaming company is just innapropiate, I think nthat since some years they could afford more programmers at least to work partial time or help in some issues.

I'm sure Charles is a very good programmer, but one man can do so much, not more.

Ah well, is heir company so they must know how to run it hehe.

[Guest]

Yeah, that's what I've been thinking. They know how to run their company, so they would've found help if they wanted it. Just a thought I suppose.

Edited because I'm a word-choice Nazi when it comes to my own writing.

[ September 10, 2007, 03:54 PM: Message edited by: Phillip Culliton ]

[Guest]

Sorry for thread-jacking there, The_Capt, as well as for the double-post. Any further info on those tests?

Also, if you guys need additional reps done let me know and I'll do what I can to help.

[The_Capt]

Hey Phillipe,

All you can do is send a resume to Steve. I know they have other help but in what capacity I am not sure.

I am going to do up a write up on the assault next but it will be a couple of weeks cause I am away with work.

Then I was going to get into moving a platoon. Then vehicle integration.

[Guest]

Sounds good.

Heh, a resume. Wow, I've been in the same industry for so long (despite taking different jobs) that I haven't needed a resume in about seven years. "Oh, you ran development at X competitor? Come on down!"

[19Kyle72]

First time post from a long time lurker of a fantastic online community.

I feel that I should state up front that I've yet to even purchase Shock Force, as money is extremely tight for us right now. That said, I am very interested in purchasing this game anyway, and combing through these forums is going to eventually persuade me one way or the other.

This thread has been one of the better reads. Its depth is appreciated, that's for sure.

The_Capt, is there any way that you could post your tutorials on YouTube? I feel that using such a method would go a long way in clearly demonstrating your points. For that matter, anybody willing to take on such a feat would be certain to win a lot of applause for supplying the community with such a detailed and clear resource.

Perhaps it'd be valuable to to name such tutorials with a common naming structure, so that they're easier to find later, for example "CMSF Tutorial #, Beating a Blind Corner"...

I'm really surprised to discover that the A.I. is incapable of peeking around corners to assess the environment that one's in. While I agree that no game is free from having to rely on some level of abstraction in regards to translating The Real World into a virtual reality, but peeking around a corner has got to be one of the most fundamental means of survival yet invented by man, and it's not even simulated in the title?

That is utterly bizarre to me. Is there a way to implement such a vital behavior into the game, or, at the very least, remedy how the squad can't comprehend what it sees till the "center" of the unit's icon is close enough to the region of confict?

I'm really looking forward to more of your tutorials, many of which I'm sure will prove valuable in even the first Combat Mission series. Keep up the good and hard work--I'm appreciating it!

Yours,

Kyle

Sept. 11, 2007

[Thomm]

Originally posted by 19Kyle72:

I'm really surprised to discover that the A.I. is incapable of peeking around corners to assess the environment that one's in. While I agree that no game is free from having to rely on some level of abstraction in regards to translating The Real World into a virtual reality, but peeking around a corner has got to be one of the most fundamental means of survival yet invented by man, and it's not even simulated in the title?

Well put.

This is how I perceived the 'action spot' concept when it was first mentioned. Something like the 'Brothers in Arms' system, where troops know how to interact with their environment, where to take cover, etc.

I guess the issue is that CM:SF tries to do huge, wide open areas and, at the same time, densely packed urban areas, thus requiring trade-offs.

Anyway, we shall see what adjustments and improvements will be made ...

Best regards

Thomm

[The_Capt]

Thanks to everyone for the encouraging feedback. I will be doing up more of these in the near future. Unfortunately real-life has taken over for a couple of weeks and I will be away from a computer terminal.

If anyone out there finds something that works don't be shy to post it

[General_solomon]

Originally posted by Rollstoy:

</font><blockquote>quote:</font><hr />Originally posted by 19Kyle72:

I'm really surprised to discover that the A.I. is incapable of peeking around corners to assess the environment that one's in. While I agree that no game is free from having to rely on some level of abstraction in regards to translating The Real World into a virtual reality, but peeking around a corner has got to be one of the most fundamental means of survival yet invented by man, and it's not even simulated in the title?

Well put.

This is how I perceived the 'action spot' concept when it was first mentioned. Something like the 'Brothers in Arms' system, where troops know how to interact with their environment, where to take cover, etc.

Thomm </font>

[Knaust1]

No help by popping smoke up?

[thewood]

I still think the problem is that some squad members move way beyond the squad's center and thus move past the corner, or whatever cover they were supposed to stop at. When they do this, they are not allowed to spot for the rest of the squad, but can get killed. This borders on being just unfair, let alone unrealistic. Once again, it seems the action spots are the cause of this behavior.

It also should be noted that when members move beyond the squad, the entire squad starts doing some kind of reorganization dance that will leave some exposed.

[metalbrew]

Originally posted by thewood:

I still think the problem is that some squad members move way beyond the squad's center and thus move past the corner, or whatever cover they were supposed to stop at. When they do this, they are not allowed to spot for the rest of the squad, but can get killed.

I believe this is slightly wrong. When squad members move ahead of the unit icon, it's true that they cannot spot for the squad unless they are in the same action spot as the unit icon. The enemy however also has no chance to spot the advanced squad members (assuming they're way in front of their unit icon) because the enemy has no LoS to the unit icon.

The problem arises that when the unit icon finally gains LoS to the enemy and now the advance squad members are very far away cover and artificially exposed. This often creates the situation that the squad didn't notice the enemy laying 10 feet away, in the open, right around the corner, until they've fatally moved too far past the corner.

This is consistent with the assault command. If you watch this command closely, you'll see the unit icon floats roughly between the two split teams. I've done some testing and seems that the advanced team does no spotting until the rear team begins to move forward (and this makes the unit icon also advance). When the unit icon finally reaches the advanced team's action spot, then they have a chance to see an enemy 10 feet away.

In one comical example, the advance team moved into a building occupied by an enemy squad. The rear team had some weird problems and refused to move (I think it was a bug, they just stayed permanently split from the advanced team). The advanced team stayed in the building for 5 minutes on the same floor right next to the enemy and never spotted them since the unit icon was floating halfway between the teams roughly 20m away.

I believe that until a unit's 'location' is comprised of every action spot it's members occupy (instead of the single action spot the unit icon occupies), this will continue to happen indefinitely.

[metalbrew]

Double post

[thewood]

I have seen squad members move well past a wall openning, get shot, but not even spot whats shoooting at them.

[Guest lmgroth]

Has anyone tried incorporating the "Evade" order to The_Capt's blind corner example instead of "Halt"?

[redbear]

nice article but my question is in the case your playing with a human and he goes to the buildings how to go there

[RT North Dakota]

Originally posted by The_Capt:

.

Tactics are actually quite simple..they have to be. Executing them in real combat is the hardest thing one can ever do...well I've never given birth but..well whatever. In combat your brain pretty much starts to shut down. The stress of the noise and confusion means that higher cognitive functions start to seriously degrade. Here is where heart of tactics come into play..drills.

Drills are something you can do in your sleep...I can remember dreaming about getting an order and being half-way out of my hooch to execute before I realized it was the middle of the night and I was suppose to be sleeping.

.

While in no detracting from the usefulness of this post, (and only if you are interested) if you were trained in the Australian, British or Israeli Army, Drills are not Tactics. They are tasks applied to enable a tactical advantage.

The UK 1942 Battle Drill is some of the best infantry doctrine ever written, and the basis for all 50-90% of modern teaching, but "Drills" are not tactics. They are taught differently. - which is why German infantry was so good in WW2 - They were taught tactics, not drills. - where as UK infantry generally under performed.

[Cpl Steiner]

I don't know how I missed this thread - probably spending too much time in the tech support forum - but it has been a joy to read.

The description of how the squad icon seems to be used by the game engine for spotting purposes is a particular eye opener. When you start to get a picture of the inner workings of the game a lot of unusual behaviour starts to finally make sense.

On a positive note, I think the game could be very easily tweaked to provide a lot more intelligent behaviour. Assuming each square on the map is an action spot (is this true I wonder), and that the game does spotting from action spot to action spot, then every second or so it must go through a list of all the units that can spot and compare their action spots. As has been noted, it must only take the "average" action spot for each unit, i.e. the one the unit icon is floating over.

However, the graphics engine does know where every squad member is, as it has to position the 3D model of each and every one of them on the map. It should therefore not be too much of a leap to adjust the unit list to have multiple action spots for each squad. Some sort of action spot array could be held for each unit and adjusted as a unit moves around so the spotting system knows how many action spots the unit occupies and how many pairs of eyes (human or enhanced) are in each spot.

From a computational point of view, it means a longer list to check, but this is no different to adding units to a scenario. Maybe the upper limit of units the game can handle will be lowered but that is not a big issue compared to the potential improvements in spotting behaviour.

This is all guesswork I know, but I have a gut feeling we will see many improvements in the AI based on refinements of the action spot system as CMx2 matures.

[SlapHappy]

I feel that infantry in a game should work one of two ways...either you have a myriad of choices and options to control your micro soldiers and they obey you implicitly. They live or die by your competence.

Or

Each army represented in the game should behave as competently in their AI tactics as they did historically. If you have a crack Wehrmacht infantry group they should behave and perform as such when facing off against a green US infantry unit. I think these damn RTS games have given twitchers the idea that a wargame has to have "equal" sides or they start whining. It was hardly ever that way in most of the board type tactical simulations I played long ago. In many cases, you were either at some advantage or disadvantage depending on the scenario. The point is that the scenario was trying to simulate something historical, not necessarily be "fair".....

I thought the whole point was trying to give the gamer a visceral experience of a historical event or battle....

Or did I miss something?

[metalbrew]

Slappy,

I don't understand your point about fairness and parity. It doesn't seem to fit into this thread, how does your point relate to infantry control and behavior?