bazooka10165

radio check

I just played a scenario called "Son" in which it is very tough to win as the Americans against the AI. This was true even after I had played it as the Germans and knew exactly what I would be facing. I'm not sure if Son ships with the CM or if you have to D/L it from somewhere. [ 04-20-2001: Message edited by: bazooka10165 ]

I must respectfully disagree with the statement that "losses don't mean squat." It has been my experience that, even if I take all or most of the victory flags, if my losses have been very significant, the best I will get is a draw. This seems to be the case even if I have inflicted equally high losses on the enemy. If any one is interested, I can e-mail the AAR from my last game. I took two out of three victory locations (the third one was unclaimed) but had 9 out of 16 tanks KOed. Despite the fact that the other side actually had a higher loss ratio, the game was a declared a draw.

Does this seem odd to anyone. Playing Singling Shootout as US against the AI. Captured all three vitory flags, inflicted 2 to one casualties on Germans. Final score 55 to 45 - a draw. The only thing that I can think of is that I lost about half of my Shermans. Do these KOed vehicles count so heavily in determining victory points? What about the fact that the AI lost all but two of his AFVs?

I'm sure that this has been covered before (then again after 13,000+ posts, what hasn't been), but I too have been puzzled about why I am able to spot eliminated units and abandoned mortars in areas to which I never had LOS. I am playing a PBEM now in which I called in artillery on a wooded area in which I suspect the enemy was hiding. It was night so I could only get LOS for about 10 meters into the woods. Many of the shells however landed further back in the woods. After the barrage, I was able to "spot" an abandoned mortar about 50 meters back in the woods. No way any of my units ever had LOS to that area. Seems a bit quirky. I am sure that there is some compelling programming reason for this given Steve's high standards for realism.

If everyone doesn't already know about it, there is a review of CM in today's NY Times at page G11. I hope that Steve and Charles have plenty of CDs ready to ship.

Live tanks do not affect line of sight. I believe that wrecks (burning or otherwise) do.

I also get transparent smoke and buildings with my Voodoo2 card (Diamond Monster II 12 Megs i think). However, the game tends to lock up f I have run any other program between boot-up and launching CM.

[This message has been edited by bazooka10165 (edited 07-14-2000).]

I second the motion for unit info on the purchase screen. Beleive it or not, there are some people out there who do not have innate knowledge of the differences between the 47 Sherman varients. It is currently unduly difficult to make intelligent unit purchase decisions. Also, there are times when it would be nice to have access to esoteric info during the purchase phase. For instance, in a scenario with rain or snow, it makes sense to purchase AVFs with low ground pressure ratings to minimize the chance of bogging down. I doubt that even the most hardened grog knows those figures by heart. [This message has been edited by bazooka10165 (edited 07-14-2000).]

Given the limits of artificial intelligence, it is safe to assume that no matter how much the TAC AI is tweaked, it will still not approach a human level of situational awareness. We are just going to have to accept this fact. No knock intended to BTS, I think that CM's AI is first rate - its just that given the current state of the art, artificial intelligence is going to be far more artificial than intelligent. Nevertheless, I believe that something can be done to address a major gripe. In my own experience, and based upon the posts in this forum (i.e. www.battlefront.com/discuss/Forum1/HTML/007263.html ) it seems to me that one of the most complained of problems created by the TAC AI's limited situational awareness is its targeting selection for our AVFs and guns in situations where we have intentionally not designated a target (or have designated a target with no LOS). The problem manifests itself most dramatically in tanks with slow turrets, turretless AVFs and for guns, since the penalty imposed upon these units for the TAC AI's poor targeting choices is (on the average) significantly higher than it is for infantry or fast turreted tanks (which can quickly engaged new threats appearing from a different direction). To be even more specific, our main gripe seems to arise when the TAC AI orders these “slow traversers” to target infantry units (which we either purposely did not target or which appear during the execution phase) rather than have them remain on the look out for enemy AVF's that we suspect are lurking on the battlefield. Although we will never be completely happy with the way the TAC AI handles our units, it should be possible to correct this particular grievance. I do not think that it would be unrealistic to be able to order tanks and guns to essentially ignore all infantry targets unless they are immediate threats to survival (i.e. AT teams within effective range and close assaulting infantry). Isn't it reasonable for the crew of a Tiger, knowing that it has a very slow turret and that there is enemy armor nearby, to face the direction of the suspected enemy armor and to make a conscious decision to forgo targeting the Bren carrier crews and rifle teams scurrying about hither and yonder. A "target armor" command for AFVs and guns, which eliminated the unit’s interest in all but the most threatening infantry targets would be useful and realistic. This command would be more akin to the "hide" command than to the target command since it would not actually tell the unit to target any particular unit, but would rather place the unit in a state which influenced it targeting decisions in a desirable fashion. The targeting decisions of a unit in this state would be similar to that already made by infantry AT teams (target only enemy armor, shoot at infantry only if absolutely necessary for survival). If the addition of a new command telling guns and AFV to aquire targets in the same manner as a bazooka team is an unrealistic coding request (as I appreciate it likely is), could the TAC AI be tweaked so that slow traversers are much less likely to target non-threats appearing at anything other than narrow angles to their weapon facing. This way, our Tigers and IG 150 guns will tend to keep the business end of their weapons pointed in the general direction we tell them unless a bona fide threat pops up elsewhere. This will of course have its own draw backs (damn, I wish my StuG had rotated 90 degrees to kill that Platoon HQ which appeared on the hill 500 meters away instead of just sitting there facing the ridge line like I ordered him to). However, the problems created by this are far less likely to be fatal to the units in question. The current TAC AI often errs by leaving slow traversers extremely vulnerable to attacks from the very direction we perceive the greatest threat, which is presumably the direction we ordered them to face. At least with the slow travering units, it should be set to err in the other direction - ignoring low priority targets which, if acquired would render the slow traversers unable to effectively engage targets appearing in front of them during the execution phase. If this tweak of the AI is too extreme a coding task (Charles must love when we use the word tweak to describe something which probably takes him days to code), perhaps what could be done it is to expand the trigger radius of the ambush markers laid down by AFV and guns. I assume that there is now an imaginary circle around these ambush marker which, when entered, triggers target acquisition. If this circle were instead an ellipse, with the foci placed along a line passing through the unit and the ambush marker (which, if I recall my high school geometry, will create an ellipse stretched out along this axis), the ambushing unit would be less likely to fire at targets which require a major weapon traverse. If the distance between the foci of the ellipse was based upon the speed of the ambusher's traverse (slower traverse = greater distance between foci), this would result in narrower ambush zones created by the slowest traversing weapons - which overall would be a good thing. Obviously not a perfect solution, but at least it will allow us to create a narrower corridor in which slow traversing guns and AFVs will tend to limit themselves to when acquiring targets, making it less likely that they will traverse 45 degrees to fire at half track crews only to be blasted by the tank that we knew was about to crest the hill right in front of them. Even if low threat targets enter the ambush zone thus triggering an unwanted target acquisition, at least the units main guns will still be pointing in more or less the direction we want. Of course, they will also be less likley to target real threats appearing outside the ellipitical ambush zone, but the current ambush markers used by AVF's create this same danger. Looking forward to hearing feedback on these ideas and finding out if any of them are desirable or feasible.

Has anyone in NY gotten their pre-orders? I read a couple of years ago that the postal service in Westchester County (where I live) has just about the lowest on time delivery rate in the country. Guess they haven't corrected the problem yet.

Ordered in January. Don't remember whether I shipped to my office in NYC or to my home in the burbs. Either way, no CM yet. However, yesterday I did receive a USPS Priority Mail box. Upon inspection (i.e. tearing it to shreds) I discovered that it contained a Little Mermaid stencil that my wife had ordered from the Disney Store. I'm still trying to recover.

Before running out and buying a new mouse, try disabling the Intellimouse mouse scroll features by turning off the mouse app which is running in the sys tray. This worked for me when I had problems running the demo on a system using a Logitech wheel mouse. Good luck. [This message has been edited by bazooka10165 (edited 06-18-2000).]