Jump to content

WW2 BAD-T IV BDAc AI Dogfight Tournament


 Share

Recommended Posts

11 hours ago, SuicidalInsanity said:

The final match of the third heat, between @dundun92's Tallyhawk Mk.XIVb and @aleksey444's AF-516 Benign Lurker:

 

Very good battle.  I did better than I expected.  Tallyhawks accelerate faster than me (because two engines) and they turn (pitch) quite a bit faster, too.  Their guns are not very powerful, and they're not very accurate (and I'm not either), but speed and turning are already enough to make them deadly.

I ran some test fights to see how this fight would turn out.  1v1 because my laptop isn't fast enough for 2v2.  Usually, Tallyhawk would out-turn me and land a few shots.  Once did not do any damage, because these are only 15mm guns. But 4-5 times would kill my plane.

I guess my plane's propensity to collide worked in my favor. :D

BTW, this is the first tie of the tournament. Although I suppose technically @dundun92 won 2/3 fights.

Link to comment
Share on other sites

On 12/11/2018 at 11:08 AM, aleksey444 said:

Very good battle.  I did better than I expected.  Tallyhawks accelerate faster than me (because two engines) and they turn (pitch) quite a bit faster, too.  Their guns are not very powerful, and they're not very accurate (and I'm not either), but speed and turning are already enough to make them deadly.

I ran some test fights to see how this fight would turn out.  1v1 because my laptop isn't fast enough for 2v2.  Usually, Tallyhawk would out-turn me and land a few shots.  Once did not do any damage, because these are only 15mm guns. But 4-5 times would kill my plane.

I guess my plane's propensity to collide worked in my favor. :D

BTW, this is the first tie of the tournament. Although I suppose technically @dundun92 won 2/3 fights.

In hindsight, 2 30mms or 3 23mms probably would have been a better armament. Great battle anyway! 

Link to comment
Share on other sites

3 hours ago, dundun92 said:

In hindsight, 2 30mms or 3 23mms probably would have been a better armament. Great battle anyway! 

30mm is deadly, but it has low muzzle velocity and somewhat low rate of fire.  It's a close range weapon.

23mm looks good on paper, but I don't think it works out in practice.  It seems to be less accurate than other guns (as in, bullets like to fly in different directions).  It seems to overheat quickly.  And, while you'd think it would do a lot of damage, it's rather average.

I think for Tallyhawk, three 20mm's (Hispano) would be the best option.  Hispanos are accurate, have better rate of fire and muzzle velocity than most guns, and do very decent damage.  Instead of taking 5-6 passes to kill my Lurker with the 15mm's, you would do it in 2-3 passes with the 20mm's.

Link to comment
Share on other sites

The YVa-23 has one of the highest accuracy and muzzle velocity values of AviatorArsenal weapons, but AI aiming quirks (accurate stream of bullets all 20m to the left), lower RpM (650), and perhaps most of all, odd point cost do make it a less attractive option. Though I may have a solution for that...

Anyway, aesthetics question. So far, all battles have been filmed using stock vessel UI markers for Red vs Blue Magenta vs Cyan team colors, which work, but the Cyan does get lost against the sky at times. However, in my latest bout of BDA hacking, I've added in Team Icons:
qS0mO3d.png
So, for the final 6 matches, would you prefer to stick with the current method, or try out the new ones?

Link to comment
Share on other sites

30 minutes ago, SuicidalInsanity said:

The YVa-23 has one of the highest accuracy and muzzle velocity values of AviatorArsenal weapons, but AI aiming quirks (accurate stream of bullets all 20m to the left), lower RpM (650), and perhaps most of all, odd point cost do make it a less attractive option. Though I may have a solution for that...

Could you expand on that a little?  The part about it having the highest accuracy, and the part about shooting 20m to the left.

I ask because from my observations, 23mm is not accurate.  20mm shoots in a line. 23mm sprays in different directions.

Link to comment
Share on other sites

IMO the 30mm are probably the best gun considering AI aiming characteristics and that parts usually die from being hit with small numbers of rounds, increasing the amount of actually-useful damage tremendously. 40mm and 75mm also kill parts and usually get 1-shot-kills, but their rate of fire is absolutely abysmal. A part that takes 2-3 20mm rounds is still gonna be functional until its HP is zero, which means so is the plane it's attached to. Any damage that doesn't actually destroy parts is cosmetic damage. Not actually meaningful unless you hit the same part again.

Doing 150 damage to 20 random areas on a plane is not likely effective. Not when it takes 4 hits to kill weak parts, 7 to kill average, 10 to kill strong, and probably 17 to kill anything deliberately reinforced. Even if the guns are tightly packed in salvo mode each trio of bullets most- likely cannot quite destroy a weak part.

Doing 500 damage to 6 random areas is not only effective, it's a likely kill against anything that isn't overwhelmingly made of 1500+ hp parts.

Then again maybe I'm biased since all my planes so far either collided with other aircraft or got hit with 30mm. It is possible that 20 mm can laser through just one part of a plane, or that the plane can have a weakspot, such as the Shadowbird's tail.

Edited by Pds314
Link to comment
Share on other sites

21 hours ago, aleksey444 said:

Could you expand on that a little?  The part about it having the highest accuracy, and the part about shooting 20m to the left.

I ask because from my observations, 23mm is not accurate.  20mm shoots in a line. 23mm sprays in different directions.

How stable is the firing craft? The 23mm has a dispersion of 0.16 degrees (everything else is on average around 0.25 ), so fired rounds are not going to go very far off-bore; the videos with the DUMMY Strikes Back are a good example of this (armament: 3x Vya-23); when it fires, it's a line of bullets. The 20m to the left is hyperbole, but something commonly seen in the battles - one plane will fire an accurate stream of gunfire, all of which misses the target by the same amount, as if the firing plane was actually targeting a point a little ways off from the target plane, rather than the plane itself; this is due to the BDAc 1.2.2.2 AI aiming code, summarized, being:
{maxAutoFireCosAngle = 0.9993908; if TargetCosAngle >= maxAutoFireCosAngle, fire}.
In layman's terms, the AI will fire when the target is within 2 degrees off-bore from the weapon.

For some examples of both, and since it's a convenient segue, here's Match 19 between @Kunza's K-1 and @Pds314's Shadowbird Mk IV. Since there seemed to be positive response to the new icons, here's those in action as well:

I may need to play with the color of the B team Icons a bit.

Edited by SuicidalInsanity
Link to comment
Share on other sites

Spoiler

That wasn't really the result I was expecting. But I'm not sure what I was expecting to begin with.

The new icons look good. My only (minor) issue is that they don't display the distance like the old one does.

Link to comment
Share on other sites

4 hours ago, SuicidalInsanity said:

here's Match 19 between @Kunza's K-1 and @Pds314's Shadowbird Mk IV. Since there seemed to be positive response to the new icons, here's those in action as well:

 

Yay. Finally a match that doesn't result in my plane getting lazor'd by 30mm.

On an aside, anyone else notice how at one point my plane practically stalls even though it's in level flight not pulling any aggressive maneuvers? I thought I eliminated this behavior in testing but.. nope. It gets stuck in high-yaw positions like that and reduces speed considerably.

 

Also noticed that it still can't aim well most of the time. =P

Edited by Pds314
Link to comment
Share on other sites

Currently trying to optimize a craft for durability in the 1.5.1 meta.

I will say this: most stock parts have consistent, reasonable-ish numbers for hitpoints.

MOST.

Not some though. There is at least one aerodynamic part with >30000 HP.

...

I'm on my phone so I don't have a screenshot, but suffice to say the craft is nuts. It can do sustained turns that will not only knock Val unconscious but keep her unconscious until she crashes. Sustained 11 G turns with immediate G-force being around 18 while it has indestructo-surfaces in some vital areas. It's also just a nice-looking dual engine plane.

Edited by Pds314
Link to comment
Share on other sites

Match 21, between @Van Disaster's D.A.C. Deimos Mk X and @53miner53's Delta:

Trying a different UI layout here. The team icon nametags are a togglable feature, so if the VS GUI window is preferred, I can go back to that.
...
Three matches left.
With the tourney's end in sight, each craft has gone up against three different opponents, allowing a better appraisal of strengths and weaknesses of each design than in a single-elimination ladder; but these match-ups were as determined by the Swiss rules rankings. As in BAD-T III, I'm game for some post-tournament exhibition matches/bonus vids, so are there any custom match-ups that you'd like to see?

Link to comment
Share on other sites

Technical question: why are BAD-T props currently operating on constant ISP and variable fuel burn with speed? It seems to me like the propellant burn ought to be constant and the ISP should vary with speed.

I could understand if this were not technically possible, or required in-depth code mod to do, but this does not seem to be the case. I just did a little experiment with the Widshed engine's .cfg file. All I added was this:

useVelCurveIsp = True
velCurveIsp
{
	key = 0 1.5
	key = 0.08 1 0.6 -1.9
	key = 0.35 0.6
	key = 0.78 0.2
	key = 1.15 0.1 0 0
}


and then changed the velCurve to this:

velCurve
{
	key = 0 1.0
}


The results:
ML37Nef.png
p7aEgzk.png

Exact same performance except the fuel burn does not vary with speed, as is true for real piston engines.
This is much preferable and much more realistic.
We can of course mess with the actual ISP if we want a specific fuel burn rate, rather than the burn rate being exactly what it was at Mach 0.08 before.  (2.04 kg/minute for the Widshed, 11.88 kg/minute for the Falcon).

Edited by Pds314
Link to comment
Share on other sites

The reason the props' fuel use is the way it is is due to the props being ported from, what, KSP 1.0? From BAD-T 1 or 2, and since I couldn't recall an occasion when fuel quantity was ever an issue (save for BAD-T III Round 2), modifying fuel burn as part of the engine tweaks for IV just never came up.

Link to comment
Share on other sites

5 minutes ago, SuicidalInsanity said:

The reason the props' fuel use is the way it is is due to the props being ported from, what, KSP 1.0? From BAD-T 1 or 2, and since I couldn't recall an occasion when fuel quantity was ever an issue (save for BAD-T III Round 2), modifying fuel burn as part of the engine tweaks for IV just never came up.

Yeah I suppose the most anyone needs is still less than a Mk0 tank.

Link to comment
Share on other sites

I would LOOOOOOOOOOOOOOOOOOOOOOOOOOOOOVE to do this, but one problem - I'm still stuck on 1.2.2 

 

I believe the props will work though.  I'm going to put one on my Kustard and see how it works.  aaaand I may have to edit the craft file to get it to load in the new one.

Link to comment
Share on other sites

This thread is quite old. Please consider starting a new thread rather than reviving this one.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...