Battlefront.com Posted January 20, 2011 Share Posted January 20, 2011 The testers and I are hard at work trying to track down some issues with long range gunnery that appear to be producing some unexpected results. Since the ranges are normally not seen in a game (or during the real war, in fact) it's not been a priority to track down until now. I set up a test which had 20x Tiger 1E vs. 20x Sherman M4A3 (75). The Tigers were neutered so they could't shoot back (we wanted a test of Sherman vs. Tiger only) and the test commenced. While checking out hits on the Tigers I noticed that none of the Tigers were even spotting the Shermans, which they should have been doing despite not being allowed to fire back. This wasn't expected! I unbuttoned the Tigers and bingo... they spotted the Shermans no problem. Hmmm... The results were passed onto Charles. He looked into it and found a very unexpected source of the error: The Tiger Tank Commander, when buttoned, was facing 90 degrees to his forward looking optics. In other words, his 3D polygonal ear was pressed up against the 3D polygonal optics' eye piece. That was the bug. Now, I can hear you all saying "WTF?!?" just like the testers said when I told them the news. Why should the graphical matter? The simple answer is it was the cleanest, easiest way to account for the peculiarities of each armored vehicle's unique strong/weak vision areas. Instead of hand coding special rules for each vehicle, it is handled "naturally" when Dan creates a model. Er, provided a crew member doesn't get all turned around For those of you who have followed the "design for effect" vs. "engineered" simulation philosophies, this would appear to show how "engineered" can trip over itself because in "designed" systems little issues like this wouldn't happen because they wouldn't even be considered. But this is not correct. Even the most simplistic simulations have mistakes in either execution (bugs) or design (logic flaws). While it is true that mistakes in execution are less likely with more simplistic systems, and those mistakes are usually easier to find, it is untrue that simplistic "designed" systems have fewer sources for flaws. In fact, our experience is that as "designed" systems for tactical battles at CM's scale produce far more flawed outcomes than engineered systems. We also found that increasing the complexity of the "designed" approach produced more sources of error (flaws) and increased execution error (bugs) to an unacceptable level. CMx2's engineered simulation approach is not perfect either in terms of execution or simulated outcomes. However, it's ability to simulate the subtle stuff that DOES MATTER, such as vehicle blind/weak spots, helps make CMx2 a much richer and realistic experience than any other tactical wargame out there. Steve Link to comment Share on other sites More sharing options...
Gpig Posted January 20, 2011 Share Posted January 20, 2011 Very Interesting. It really adds to the feel that each sub-element of a unit (and it's performance) matters. This can produce very believable results and outcomes. I assume a tank commander can still look behind, left and right . . . in opposition to whichever way the turret is pointed. Just like in CMSF? In WWII it may well be very difficult for a TC to see very much at all, eh? Cheers, Jon Link to comment Share on other sites More sharing options...
Wicky Posted January 20, 2011 Share Posted January 20, 2011 ...so a flanking attack of diversionary crack Canadian dancing girls (in the appropriate module) will distract Tiger crews leading to their utter devastation... *takes note* Link to comment Share on other sites More sharing options...
Battlefront.com Posted January 20, 2011 Author Share Posted January 20, 2011 Heh... and nose art on bombers increased crew morale Gpig, Yes, non-forward looking is also simulated abstractly based on what is supposed to be the optimal starting position. So if something comes up from behind there is a lower chance of Commander noticing (simulated turning his head around every so often) it than compared to front (which is primary focus) and sides (which are commonly checked on). When those side and rear checks are made they are made using whatever optics bonuses are available to the Commander. In WW2 there were very limited options, but modern vehicles often allow the Commander to view 360 degrees with all "sensors". Steve Link to comment Share on other sites More sharing options...
wolf66 Posted January 20, 2011 Share Posted January 20, 2011 Heh... and nose art on bombers increased crew morale Gpig, Yes, non-forward looking is also simulated abstractly based on what is supposed to be the optimal starting position. So if something comes up from behind there is a lower chance of Commander noticing (simulated turning his head around every so often) it than compared to front (which is primary focus) and sides (which are commonly checked on). When those side and rear checks are made they are made using whatever optics bonuses are available to the Commander. In WW2 there were very limited options, but modern vehicles often allow the Commander to view 360 degrees with all "sensors". Steve So when a grenade explodes outside the rear of the tank, the focus from the tank commander will immediately switch to the rear ? Link to comment Share on other sites More sharing options...
Battlefront.com Posted January 20, 2011 Author Share Posted January 20, 2011 It should get his attention, yes, but it is neither guaranteed nor necessarily immediate. Tanks are noisy internally, especially when moving, so coupled with the natural sound barrier of steel it could very well be that nobody inside is even aware of the grenade. Vision from within is also not very good even with good optics. Steve Link to comment Share on other sites More sharing options...
Rocky Balboa Posted January 20, 2011 Share Posted January 20, 2011 WTF!! indeed .... This puts a whole new importance on making sure that I issue that end of move facing command to my dismounts as well as vehicles. Link to comment Share on other sites More sharing options...
Erwin Posted January 20, 2011 Share Posted January 20, 2011 Re facing command and covered arcs, would either/both of those have affected the Tiger CO spotting ability in this situation? In the CMSF I thought I read someplace that covered arcs were not as necessary (as in CM1) most of the time, as units did a better job of spotting than in CM1(?) - I interpreted that as implying that arcs in CM2 were more for preventing a unit from firing outside of that arc, and/or to have one's best armor facing in the desired direction. Is that accurate, or do arcs improve spotting within that arc? If an arc increases spotting ability within the arc, presumably, spotting doesn't end at the limits of the arc as fire would, but extends to the limits of visibilty(?) Link to comment Share on other sites More sharing options...
Battlefront.com Posted January 21, 2011 Author Share Posted January 21, 2011 Even in CMx1 there was superior spotting to the front, medium to sides, and poor to the rear. But it was abstract and not really customized for a particular vehicle's unique quirks. With CMx2 this is now simulated specifically and therefore the results are more "natural". Still not 100% simulated (we don't have the Commander's head turning around using complex Human like behavior modeling), but more historically nuanced. Arcs do focus the unit's attention more in one area than another. However, the primary purpose is to decrease the changes of targeting outside of the arc. Orientation is always optimized for the arc, which means it does have an impact on Facing. Spotting is a separate function. Units with a forward facing arc can still, theoretically, spot things behind them. They are just less likely to because their attention is mostly forward most of the time. Anything that is spotted is registered by the unit and then, usually, ignored unless it drifts into the arc. Not guaranteed as poorer units might ignore the arc, or a big threat comes up, but the general concept is to not fire at things outside of the arc. Steve Link to comment Share on other sites More sharing options...
Canada Guy Posted January 21, 2011 Share Posted January 21, 2011 I noticed in some of the DAR pics that the green unit icon under the Shermans had 1-2 little arrows pointing in different directions. Is one for AFV front and the other for turret front? Also, can we see any pics of the covered arcs or is that something that either has not been used or in the NDA? Link to comment Share on other sites More sharing options...
panzermartin Posted January 21, 2011 Share Posted January 21, 2011 Its exciting to know how "alive" the code acually is! But does this mean you have to check again all tanks/vehicles in the game to make sure commanders dont stare at nude playboy posters on the side of their turret interiors? Link to comment Share on other sites More sharing options...
Battlefront.com Posted January 21, 2011 Author Share Posted January 21, 2011 The "base" under the units has three little indicators pointing in the direction the unit is facing. There are three because some vehicles can sit right on top of the middle one and obscure it from the player's view. We have already initiated a "Hey, which way are you facing?" campaign to realign disorientated vehicle commanders Steve Link to comment Share on other sites More sharing options...
dan/california Posted January 21, 2011 Share Posted January 21, 2011 I have read that the Russians were known for conducting their realignment campaigns with Makorov's of various calibers. I hope you feel bad for the widows and orphans. Link to comment Share on other sites More sharing options...
Lanzfeld Posted January 21, 2011 Share Posted January 21, 2011 I hope this doesnt throw off all your original spotting values if you tweaked them to "feel right"! Link to comment Share on other sites More sharing options...
Sivodsi Posted January 21, 2011 Share Posted January 21, 2011 Interesting! But wouldn't the driver have a good chance of spotting the shermans? He doesn't have the optics, but I would expect him to notice the flash and then bang of a 75, even if it is a couple of km away. Unless of course the Tigers were hull down. Link to comment Share on other sites More sharing options...
Lanzfeld Posted January 21, 2011 Share Posted January 21, 2011 I THINK there is only one spotting point per vehicle. The height of that point can be a one of 4 or 5 levels to give a slight variation due to vehicle construction. This is how it is in CMSF anyway....I think. If BFC changed this for CM:BN (I hope!!!!) than maybe the drivers were hull down. Link to comment Share on other sites More sharing options...
Battlefront.com Posted January 21, 2011 Author Share Posted January 21, 2011 At the range I was testing at the best chance of spotting was by the Commander. The other guys would have eventually figured it out, but without optics (gunner's sight is only good if you know what to look for) it should be much slower. When the Tigers were moved closer to the Shermans the spotting problem went away because the other crew DID see the Shermans. Which is why we've only seen this problem reported intermittently because at normal combat ranges the Tigers did spot things at a reasonable speed. It was only when the range was extended to the point of the Commander's optics being very important that we ran into problems. For those interested in how 3D polygons interact with the code, there are hot spots in the models which tell the system where special things of interest are. Hatch locations, firing ports, entrance/exit points, seats for particular passengers, extra equipment stowage on the outside, etc. In this case there is one hotspot that says "here are the optics" and another that says "here are the eyes". The code understands what these hotspots are and their 3D spacial relationship to each other. This allows the system to make calculations based on the information in the 3D model. They don't have to be exactly lined up, but 90 degree error was way too much. Steve Link to comment Share on other sites More sharing options...
Rocky Balboa Posted January 21, 2011 Share Posted January 21, 2011 For those interested in how 3D polygons interact with the code, there are hot spots in the models which tell the system where special things of interest are. Hatch locations, firing ports, entrance/exit points, seats for particular passengers, extra equipment stowage on the outside, etc. In this case there is one hotspot that says "here are the optics" and another that says "here are the eyes". The code understands what these hotspots are and their 3D spacial relationship to each other. This allows the system to make calculations based on the information in the 3D model. They don't have to be exactly lined up, but 90 degree error was way too much. Steve I had no idea that your models were this detailed. Does it work this way for CMSF as well or have you fine tuned it for CM:BN ?... Link to comment Share on other sites More sharing options...
MikeyD Posted January 21, 2011 Share Posted January 21, 2011 I've never seen Steve so proud of a game bug before! This was an excellent opportunity, I admit, to give everyone a show of what's meant when they use the phrase "under the hood". It means more than any of us imagined. ...Unless of course the Tigers were hull down. Yup, a perfect description of the test map that discovered the problem. A shooting gallery with hull-down Tigers at one end and Shermans at another. Link to comment Share on other sites More sharing options...
dan/california Posted January 21, 2011 Share Posted January 21, 2011 The other thing it really points out is the importance of understanding why something isn't working. MANY game companies would have ham handed in some kind of modifier and then been bitten on release when many thousands of "playtesters", I mean customers, found out where that broke something else. Link to comment Share on other sites More sharing options...
scottie Posted January 21, 2011 Share Posted January 21, 2011 In this case there is one hotspot that says "here are the optics" and another that says "here are the eyes". The code understands what these hotspots are and their 3D spacial relationship to each other. This allows the system to make calculations based on the information in the 3D model. They don't have to be exactly lined up, but 90 degree error was way too much. Steve Seriously impressive level of detail , thanks for explaining. Link to comment Share on other sites More sharing options...
MeatEtr Posted January 23, 2011 Share Posted January 23, 2011 Interesting read, thanks for sharing! Now that everybody enjoyed Jons/Elvis excellent AAR/DAR, how about a video now? Pretty please with gunpowder on top! Oh and any timeframe on release date? Link to comment Share on other sites More sharing options...
BFCElvis Posted January 23, 2011 Share Posted January 23, 2011 Interesting read, thanks for sharing! Now that everybody enjoyed Jons/Elvis excellent AAR/DAR, how about a video now? Pretty please with gunpowder on top! Enjoyed? It ain't over yet! Link to comment Share on other sites More sharing options...
dan/california Posted January 23, 2011 Share Posted January 23, 2011 Your down to the odds and sods, he has two corporals,and no one has any morale or command to speak of. Its looking kind of over, to me at least. Link to comment Share on other sites More sharing options...
MeatEtr Posted January 23, 2011 Share Posted January 23, 2011 Enjoyed? It ain't over yet! We're gonna need some video for proof! Link to comment Share on other sites More sharing options...
Recommended Posts