Discussion thread - BD Armory AI tournaments

[EvenFlow]

The laser you want to add is on a turret, right? That's the problem. We don't want turreted weapons in the challenge.

It indeed is, and, ironically i was the one to suggest it's banning, but the one I'd be using would have all kinds of turret movement locked. The only problem i see with it is lack of projectile speed which can make it way more accurate than any other weapon. It would be compensated by it's great power usage and nerfed damage (after last BD Armory patch). I guess i will just do some testing and report on the results sometime soon.

Thanks for considering the idea! c:

[RocketTurtle]

It indeed is, and, ironically i was the one to suggest it's banning, but the one I'd be using would have all kinds of turret movement locked. The only problem i see with it is lack of projectile speed which can make it way more accurate than any other weapon. It would be compensated by it's great power usage and nerfed damage (after last BD Armory patch). I guess i will just do some testing and report on the results sometime soon.

Thanks for considering the idea! c:

Yeah tell me what's it's like. I'm considering it. It would look cool, but you know, I don't want to ruin the dogfighting aspect.

[colmo]

I'm trying to think of an alternative to the very close starting positions we've been using currently, which favour the plane to get in the air and fire first.

How about two distant bases, out of range of each other, but in range of a 'lure', a simply probe armed with a weapon manager (or two, one for each team!) which gets both pairs of planes to a common place. It can be self-destructed or given a minimal battery so it runs out of power (and becomes dead to guard mode) before getting shot at. The result - two sets of planes already flying at their preferred altitude get to dogfight.

[Slam_Jones]

^^ I usually set mine up as a 1v1 (can be adapted to 2v2), facing away from each other on either side of the runway (due north and due south), with 500+meters between them. This usually allows the jets to get about 3.5km between them before missiles are launched.

Another one of mine is a variation of the one in use now, but instead of taking off in the same direction, they take off in opposite directions (team A heads west, team B head east) so that, again, there's at least some (usually about 2)km (rather than <1km) between them when the battle "begins".

From Top Gun AI Tourney Thread

I don't like the no rcs rule. What is the point? It isn't op in atmosphere, and to stop wing spam you could also say no more than 2 true wings( largest surface area vs control surfaces, which have moving parts. If the wing can move individually it must be locked).

So would canards, for example, have to be locked? I use them frequently as tail fins and rear wings... is this bad form or "cheaty" or something?

Edited August 4, 2015 by Slam_Jones

[Alphasus]

No, the canards are control surfaces. I mean main wings, like if you use the 455 tail plane as a wing it would need to be locked.

[Slam_Jones]

No, the canards are control surfaces. I mean main wings, like if you use the 455 tail plane as a wing it would need to be locked.

oh, okay I get what you're saying now. My most recent designs (which are proving to be a worthy foe to the XMF ) have included a "body wing" that kinda wraps around the fuselage of the jet, to provide a little extra lift all around, and that definitely uses more than 2 wings. That's why I think something like wing area vs. mass would make a little more sense. But, of course, I could be completely wrong.

Edited August 4, 2015 by Slam_Jones

[Alphasus]

Well, those aren't main wings, again. My designs are trending that way too, and we already have planes that do that. The surface area is tiny in comparison to the main wing, maybe half of the surface area. So, that wouldn't be disqualified because it generates little lift in comparison to the main wing. If you covered the plane in that, though, that would be wing spam.

[SuicidalInsanity]

RE: wing area vs mass, if I recall correctly the FAR editor analysis window has a readout for wing loading and wing area that could be used; admittedly, this would require switching to FAR, which would be a boon and a bane. On one hand, stock aero is remarkably lenient and tolerant of, shall we say, unorthodox airframe designs, but also open to abuse of wing spam and so forth. FAR, on the other hand isn't, but requires more skill in building planes that fly well.

[Alphasus]

No, far is something I do not want to use. The more mods, the worse I see it. Especially if I don't have the option for the mod to be disabled or not used when I don't want it without an uninstall.

[Slam_Jones]

Yeah I'm undecided about FAR. On one hand, it's not as dramatic a difference between stock and FAR (post-1.0 that is) as it used to be, but it's still a lot more challenging to build with.

I wouldn't quit the competition if the rule were introduced, but I wouldn't be excited either, exactly.

Well, those aren't main wings, again. My designs are trending that way too, and we already have planes that do that. The surface area is tiny in comparison to the main wing, maybe half of the surface area. So, that wouldn't be disqualified because it generates little lift in comparison to the main wing. If you covered the plane in that, though, that would be wing spam.

Oh, hm. I'm not 100% sure what you mean then... sorry.

Edited August 4, 2015 by Slam_Jones

[Alphasus]

Far interferes with everything else that I build not related to this.

[Three1415]

RE: wing area vs mass, if I recall correctly the FAR editor analysis window has a readout for wing loading and wing area that could be used; admittedly, this would require switching to FAR, which would be a boon and a bane. On one hand, stock aero is remarkably lenient and tolerant of, shall we say, unorthodox airframe designs, but also open to abuse of wing spam and so forth. FAR, on the other hand isn't, but requires more skill in building planes that fly well.

I admit that my design relies mostly on brute-forcing the matter of maneuverability with control surfaces; personally, I would be open to matches with FAR to make such exploitation of in-game aerodynamics more difficult. However, we must realize that, as this is a game, not real life, the optimal designs may not be the ones we would expect from personal experience or knowledge of actually extant aircraft; although we can attempt to alleviate this issue, at some point we will simply have to deal with it.

[Alphasus]

Why not just hardcap the number of wings allowed on planes?

[RocketTurtle]

Why not just hardcap the number of wings allowed on planes?

I don't want to do that, because I'm fine with a lot of wings as long as it's done in a creative way. I'm cool with biplanes, intricate wing designs, just not spamming the wings in a non creative and boring manner.

[Slam_Jones]

My new jet is turning out a lot better than the little Falcon 1-A. Certainly a lot more competitive. I should have it ready for the next tourny

Download link if anyone wants to test it. But keep in mind, it's still a work in process.

Edited August 5, 2015 by Slam_Jones

[t3hJimmer]

I admit that my design relies mostly on brute-forcing the matter of maneuverability with control surfaces; personally, I would be open to matches with FAR to make such exploitation of in-game aerodynamics more difficult. However, we must realize that, as this is a game, not real life, the optimal designs may not be the ones we would expect from personal experience or knowledge of actually extant aircraft; although we can attempt to alleviate this issue, at some point we will simply have to deal with it.

I totally agree. We shouldnt try and force planes to look like real fighter planes, but we do need rules to keep things competitive. You're very good at finding holes in the rules that allow unconventional designs to dominate, but this spirit of this challenge is elegant design not exploiting glitches in KSP's physics engine. So far we've found movable turrets, SAS, and excessive clipping to be overpowered and have banned them.

I vote for banning rcs (this is a fighter plane challenge) and setting a maximum number active control surfaces per plane. I think that will prevent control surface spamming, while leaving room for creative designs. I think 8 control surfaces is reasonable.

This would allow:

2 rudders

2 elevators

2 canards

2 ailerons

or

3 rudders

4 elevons

1 elevator

[RocketTurtle]

I totally agree. We shouldnt try and force planes to look like real fighter planes, but we do need rules to keep things competitive. You're very good at finding holes in the rules that allow unconventional designs to dominate, but this spirit of this challenge is elegant design not exploiting glitches in KSP's physics engine. So far we've found movable turrets, SAS, and excessive clipping to be overpowered and have banned them.

I vote for banning rcs (this is a fighter plane challenge) and setting a maximum number active control surfaces per plane. I think that will prevent control surface spamming, while leaving room for creative designs. I think 8 control surfaces is reasonable.

This would allow:

2 rudders

2 elevators

2 canards

2 ailerons

or

3 rudders

4 elevons

1 elevator

Okay. I'm thinking a general no wing spamming rule, and then the 8 control surfaces rule.

No RCS I can agree with.

[Slam_Jones]

Any chance you can up it to 10? 8 just doesn't seem like enough imo.

[RocketTurtle]

Any chance you can up it to 10? 8 just doesn't seem like enough imo.

Yeah you're right. 10 seems more reasonable

[Slam_Jones]

Yeah you're right. 10 seems more reasonable

Much appreciated

[tetryds]

Guys, do you see what is happening here?

By limiting the ammount of wings you go directly against the purpose of this challenge concept, you end up forcing a specific design to be made under mechanics that don't privilege such design choice.

The best performing stock jets will not look like real life best performing jets because they are situated on a different environment with different rules.

It's like trying to limit the ammount of fuel on a challenge to see who has the most deltav with the least weight.

There are two things you can do about it:

1- move to FAR, then you will be having designs that look like real life ones.

2- get over it and try to make something that works on KSP without any relation to what would happen in real life.

Since FAR is quite complicated, even more when going so fast, I think it's time to rethink all stock design logic and work with what you have.

In resume, clipping is bad, many wing parts can't be bad.

Edited August 5, 2015 by tetryds

[colmo]

I took on the Double Down with my best jet and the major clincher was that big GAU 8 gun.

My plane had almost the same turn rate, was much, much faster and I think, with comparable armaments, could take it. It introduced an interesting stick or twist scenario - if I set min altitude to low (150m), I was vulnerable to missiles, if set high (580m), the gun got me.

It was annoying but I always felt I was close to cracking it. I just think the GAU and 6 missiles is a bit much. It feels comparable in damage to twin Vulcans, at 5 points less cost.

The Double Down needs wing spam because it's so heavy. I'm for letting wingspan go for now, it feels like a compromise for slow, heavy planes.

- - - Updated - - -

Actually

2*(11+4) = 30

13 + 4 = 17

So 1 GAU is vastly cheaper than twin hidden Vulcans.

[Three1415]

I took on the Double Down with my best jet and the major clincher was that big GAU 8 gun.

My plane had almost the same turn rate, was much, much faster and I think, with comparable armaments, could take it. It introduced an interesting stick or twist scenario - if I set min altitude

The increased speed of your airframe is what allows the Double Down to win. Due to the latter's biplane configuration, it has high drag but massive lift, and it uses its low stall speed but high turn rate to get inside of other craft (as its turn radius is and can remain small); faster planes "waste time" decelerating to optimal turning speeds, whereas the Double Down does not.

It was annoying but I always felt I was close to cracking it. I just think the GAU and 6 missiles is a bit much. It feels comparable in damage to twin Vulcans, at 5 points less cost.

The GAU-8 is probably too effective for its point value; because it uses 30mm rounds, even a single hit will destroy something, and given its high fire rate, anything caught in its weapon arc will be subjected to immediate vaporization, which is why the Double Down can end battles so quickly.

[Alphasus]

We can let wingspan go for now then, but it shouldn't get too crazy.

[Alphasus]

We should decrease the points needed for the hidden vulcan, as it is too many. It should be 8, same as the turret.