Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

47 Excellent

About slug_yuugen

  • Rank

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. The spotting is amazing. Absolutely what I'd hoped for when I saw you were working on it. Big hugs to the dev team.
  2. The probability of that happening is 100% as it already did. You’re welcome. 😇
  3. I landed off field the other day with only half an elevator. It was super weird balancing throttle and using the bit of elevator left to slowly spiral down. I was super lucky to end up in that state relatively stable to start with. It’s a bit weird in hindsight that the elevator wasn’t lost as well given half the horz stab was gone. Probably to the contrary of it happening in real life it was pretty dull once I worked out what control I had left and meandered back down.
  4. Can you show your working? For me it sounds more plausible that if the second round is following nearly the same trajectory as the first it will be more likely to hit if the first round hit and more likely to miss if the first round missed. Yes that was the joke.
  5. Clearly you haven't had me wobbling erratically behind you. I'm thinking more from the point of view of modelling hit chances for bursts. Which presumably just adds more chunk to the results as for each burst since you wouldn't be rolling the worst case. The chunkiest being if the first round hits all subsequent rounds would hit as well. 😄 The hitbox fidelity is probably good enough that this isn't an issue though.
  6. You realise no where in that paragraph do you actually make a case that the assumption of independence is valid. There's a reason why the advice to pilots was to get as close as possible, be in the same plane of motion and shoot the bejeezus out of the pilot. The guns in FC are like laser beams as well. If you hop into multiplayer you'll meet some veritable snipers. Even I with my very mediocre performance can usually aim at and hit pretty well in and around the pilot and engine if not snap shooting.
  7. I think you grossly underestimate how good peoples gunnery is and typical engagement ranges if you think that's the case. O_o For fun because I'm at a loose end this afternoon I knocked up a monte carlo wing shooting simulator here is my patented "thicc boi" test wing going through its paces: This is if nothing else pretty fun. So that wing is 4.2515 m long, 1m across and 0.3m thick the front spar is through the full length and depth of the wing is and is 0.05m across the rear spar is the same but only 0.03m across. Over 100,000 shots it works out to roughly 30% chance of randomly hitting a spar from any angle although you're veeeery slightly less likely to hit the rear spar. Looking at shots that came within ten degrees of the vertical you get 7% for the front spar and 4% for a rear spar and for those within ten degrees of horizontal 53% chance of hitting the front spar and 46% chance of hitting the rear spar. This is without modelling penetration, essentially assuming the first spar contact stops the bullet. If anyone has any real dimensions to play with I can change them pretty easily and if I feel sporty over the next week I'll model both wings and struts/wires as well.
  8. Note that this is with a random distribution which is a fair enough assumption for the first shot of a burst but the following rounds will be decidedly not randomly distributed.
  9. 2D you're poking holes in the paper, 3D you're drawing lines on the paper except those lines also have depth and the wires are at different depths so there is more space to miss into. The 2D case is interesting though because it does show that if you could randomly fire into the wing area then there isn't an unreasonable chance of hitting something important. Also for example from the perspective of hitting spars the dead six is also not the best position a slightly higher or lower position gives more area to hit, for example in this exaggerated view.
  10. @unreasonable There is a mismatch between the contrived test and experience in game. Clearly there is a factor unaccounted for in the test. That might be unruly players, an error in the damage model not captured by the test or simply over-modelling also not captured. We don’t know and can’t really find out. What we do know is that people are having much less fun and as a result have abandoned the game in droves. I agree that a historically accurate recreation of WW1 combat is impossible in single or multiplayer right now. As such what outcome should we want from a WW1 combat game? I’d suggest that since history and a degree of accuracy is important to both the developers and players modelling towards known historical outcomes would give a better experience as the results of combat would match perceptions of what they should be. @DakkaDakkaDakka is right on the money. Failures can be fun but random failures with no or little feedback just feel like bullshit. Baer’s example of one stray long-range hit disabling his controls feels like bullshit. Having your controls disabled after a ferocious burst into the plane much less so. Losing a wing to damage that there is no indication of feels like bullshit (or the case directly above). Losing a wing to damage you’re clearly aware of much less so. Making a good game is very often about threading that failure, feedback and bullshit line. In comparison to a lot of people here I’m one of the hapless noobs since I only started playing combat flight sims again last October and I was pretty happy to have caught the rising wave of FC. It’s pretty sad to see what should be positive features that add to the game gut the community willing to participate. @HagarTheHorrible I believe the track recording just shows your view of the game so in multiplayer will capture the glorious inconsistencies latency creates.
  11. It could also be your CPU if low graphics settings don’t help.
  12. It is but the chances of flinging 26 rounds straight down the pipe are vanishingly rare and if you take depth into account for the same random distribution you need a lot more rounds. Let alone accounting for a distribution of fire that is less random (e.g. most people squeezing off a burst aren’t liberally spraying the entire wing section). The diameter question is also a good example in the 2D case diameters relationship to area is linear but in the 3D case it’s squared in relation to volume.
  13. Sorry I was talking about control cables that move the rudder, elevator and ailerons rather than the wires in the truss. I’d note though that assuming shots from the six (let alone other positions) have no depth and only considering area is going to give overly positive results. The important thing about volume is that there is another multiple of all the space you considered to miss into. The spatial organisation of bits is also why a probabilistic model needs more refinement than you might think. Essentially you want to know the probability for a bullet strike from the specific direction.
  14. This is the chance on a 1m cubed volume with a single piece of cable running through it just to illustrate the way volume makes these things end up probabilistically. Which is essentially the chance of a 14mm projectile hitting a 6mm cable in 1 sqm of air is pretty remote when both can be anywhere crossing the 1 sqm. I also accidentally used the diameter of the cables instead of the radius in my volume calculations so this is an over estimate. In essence it is the probability of a 1m run of 6mm crossing a 1m run of 14mm where they both lie on one of the principal axis. Obviously when you consider the wings and the fuselage the number of cables increases and the volume is different depending on the section a bullet hits. Then you have pulleys and other structural elements which if lost make the cable a floppy lost piece of string. A lot of these probably constitute more volume than the cable itself. Although we’re also not modelling secondary elements that might replace them albeit with increased friction and slop. Further the actual likelihood is enhanced by the normal attack positions because the cables are naturally laid out in certain directions. For example shooting from dead six up through the plane increases the chance of hitting elevator cables purely because their length is obviously going to be principally along it. As for redundancy it’s just the law of large numbers. Granting that mine are optimistic but if you are flying a bunch of planes getting shot at regularly where losing experienced crew and aircraft is important resilience is important. And further if the Dr1 and Bristol have such redundancy they should be far and away less susceptible to control loss than other aircraft.
  15. Any idea of how thick the control cables are? For example a 3mm thick cable passing straight through a 1m cube has a volume of 2.83x10^-5 m^3 volume so an over-modeled chance of control wire hit in that case would be approximately 0.00003% chance. It's over-modeled because the bullet itself will travel along a track through the volume so the actual chance of hitting depends on how the bullet enters the volume and must also be modeled as a cylinder. If for example we simplify and say any bullet hitting is going to go straight as well (we could pick the biggest dimension or calculate the actual straight path) then then volume of a 7mm track is 1.54x10^-4 or occupies 0.0002% of the volume. So the combined probability of them intersecting is 0.000000006‬% which is quite unlikely. NB: I'm doing this whilst drinking a beer and watching a Smurf's movie with my kids so things might be quite wrong and I think they probably are given we're looking for overlap and I'm just comparing volumes. Edit: Whoops made things twice as thick as they should be so this is still an overestimate for the example.
  • Create New...