F-5E AI FM Flies Like A UFO

[LowRider88]

Anyone can set this up and try.

DCS Version = 2.7.9.18080

Comparison between the AI for the F/A-18C and the F-5E, both set to Ace level, and shows the inaccuracy of the F-5E AI flight model

The F/A-18C is like a injured pidgeon cowering from the F-5E that is flying overhead like a hawk.

It is unrealistic that the Northrop predecessor can outfly the successor.

The F/A-18C has the advantages of thrust/weight ratio, flaperons, and fly-by-wire/relaxed-static-stability.

And yet, the F-5E AI is able to climb higher and get on the tail of the F/A-18C several times.

This affects the use of AI as opponents and wingmen in single player and campaigns.

 

[razo+r]

Well, what do you want to hear? 

They are developing a new FM for the AI which hopefully is going to fix all of this. But until then, the AI is going to be borked as they always have been. 

Maybe they can tune it a bit until that happens, but I doubt they'll do that though. 

Also, in the video, the F-18 is crawling around on the floor at high AoA never maintaining high speed. And once he gained speed, he immediately depleted it again to outturn the F-5, which at this point can only go up again to preserve energy. So how exactly is this UFO?

Edited January 16, 2022 by razo+r

[Zeagle]

7 hours ago, LowRider88 said:

Anyone can set this up and try.

 

The F/A-18C is like a injured pidgeon cowering from the F-5E that is flying overhead like a hawk.

It is unrealistic that the Northrop predecessor can outfly the successor.

 

 

you crack me up...seriously? lol 

 

[LowRider88]

5 hours ago, Zeagle said:

you crack me up...seriously? lol 

 

Seriously what?  Don't understand what you are implying.  That was a useless comment.

[LowRider88]

8 hours ago, razo+r said:

Well, what do you want to hear? 

They are developing a new FM for the AI which hopefully is going to fix all of this. But until then, the AI is going to be borked as they always have been. 

Maybe they can tune it a bit until that happens, but I doubt they'll do that though. 

Also, in the video, the F-18 is crawling around on the floor at high AoA never maintaining high speed. And once he gained speed, he immediately depleted it again to outturn the F-5, which at this point can only go up again to preserve energy. So how exactly is this UFO?

 

You are asking me what I want to hear, and then try to just brush off what my concerns.  If they are finding something, great.  They fix a lot of things and somethings items are not included because they are not aware of it.  They can fix the FM engine, but what about the differences between the planes.  In each plane lua file there are some FM details.  If they fix the engine, they should look at these as well, and how they related across planes.  You doubting whether they do this is of no use to this request and adds no value.

Yes,  the F-18 is reacting much like a coward in this video, and should also be looked at.  So there's nothing wrong with the F-5E flight model?  How do you explain it getting behind the hornet twice?  I posted 2 prior videos.  Compare which of these whips past in a head on pass in those videos. I have tested all the planes.  The F-18 flys past at a rate in line with all other planes.  Only the F-5, AV-8, and VIggen AI fly past like a UFO.  Like I said anyone can try this for them selves.

Edited January 16, 2022 by LowRider88

[razo+r]

I have a great idea.

Let two fight again but this time, you try keep up with the AI F-5 in your F-5. Then you share the track and your findings.

That should give a much better understanding at how much the AI F-5 is a UFO than saying the AI F-5 is a UFO because it beats a newer plane.

 

[LowRider88]

On 1/11/2022 at 10:24 PM, LowRider88 said:

Below are 2 videos I posted showing a comparison of the Ace AI FMs of the F-5E and the F/A-18C, against a player flown MiG-19P.

In both cases, afterburner was not used, and at most Max dry thrust was used.

The F-5E video shows the AI is able to meet every 1 or 2 circle fight with a head on pass, while the second video shows it is actually easier to get on the tail of the Hornet.

This should not be the case, as the MiG-19P has both better wingloading, and thrust/weight ratio, over the F-5E.

The F-5E is retaining too much speed in turns, recovering from dives better, and climbing better than both the F/A-18C and the MiG-19P.

These comparisons are even more extreme when the AIs are set to the lower Trained level, as the F-5E seems to have less turning ability and decides to turn less, opting for even greater climbs, while still retaining too much speed, with such tiny engines.

 

 

 

Watch how fast each one flys past in the head on passes.  Notice the difference in rates.

Please provide a better reason why the older plane beats the newer plane.

[razo+r]

According to your video, the F-18 after the pass has it's nose on the F-5 before the F-5 has it on the F-18. And after that, the F-5 keeps turning while the F-18 goes almost straight towards the ground with M1.0 , recovers and only then starts turning. And in the third incidence, the F-18 almost points its nose on the F-5 but at the cost of bleeding all it's energy while the F-5 stays higher and faster.

And the fact that everything is from the perspective of the F-18 doesn't help a lot.

Maybe you should provide some reasons why an older aircraft flown better is not supposed to beat a newer plane. Or maybe even how comparing it to a different aircraft should explain inaccuracies.

[LowRider88]

35 minutes ago, razo+r said:

I have a great idea.

Let two fight again but this time, you try keep up with the AI F-5 in your F-5. Then you share the track and your findings.

That should give a much better understanding at how much the AI F-5 is a UFO than saying the AI F-5 is a UFO because it beats a newer plane.

 

If you want to explore this, be my guest.  Maybe that will show the player FM also has a problem.  I will do this when I have a free cycle.  But this "great idea" means nothing to what I posted about, and is off topic.  My concern is the difference between planes.

[LowRider88]

24 minutes ago, razo+r said:

According to your video, the F-18 after the pass has it's nose on the F-5 before the F-5 has it on the F-18. And after that, the F-5 keeps turning while the F-18 goes almost straight towards the ground with M1.0 , recovers and only then starts turning. And in the third incidence, the F-18 almost points its nose on the F-5 but at the cost of bleeding all it's energy while the F-5 stays higher and faster.

And the fact that everything is from the perspective of the F-18 doesn't help a lot.

Maybe you should provide some reasons why an older aircraft flown better is not supposed to beat a newer plane. Or maybe even how comparing it to a different aircraft should explain inaccuracies.

How does seeing it from the F-5's perspective change the end result?  Like I said, anyone can set this up.  Try it yourself.

Yeah, we already went over that the F-18 FM is flying like a pansy.  That doesn't confirm the F-5E AI FM is not inaccurate.  What logic is that?  Look at the flybys I mentioned.

To quote the rather lame comment above, you crack me up, seriously? Lol.

How is the F-5 flown by a better pilot?  They are both set to Ace.  

I already posted why the F-18 should beat the F-5 in my original post.  Why make a hornet if the F-5 is better?  That's the most ridiculous thing I have heard.

Edited January 16, 2022 by LowRider88

[razo+r]

28 minutes ago, LowRider88 said:

How does seeing it from the F-5's perspective change the end result?  Like I said, anyone can set this up.  Try it yourself.

Yeah, we already went over that the F-18 FM is flying like a pansy.  That doesn't confirm the F-5E AI FM is not inaccurate.  What logic is that?  Look at the flybys I mentioned.

To quote the rather lame comment above, you crack me up, seriously? Lol.

How is the F-5 flown by a better pilot?  They are both set to Ace.  

I already posted why the F-18 should beat the F-5 in my original post.  Why make a hornet if the F-5 is better?  That's the most ridiculous thing I have heard.

 

Your logic does not make sense.

Let me demonstrate:

Ace I-16 vs Ace F-18. The I-16 shot down the F-18. Does that mean the I-16 FM is inaccurate too? Because according to you, since both pilots have equal skill, the F-18 should win as it is superior in almost every aspect (and I'd agree with that).

So why make jets when the I-16 is clearly superior in a 1v1?

Oh right. Because you can clearly see that the F-18 "Ace" pilot has no idea how to handle the airplane, in both your and my case.

So the only true issue you discovered here is how shitty the Ace F-18 pilot is.

example.trk

Edited January 16, 2022 by razo+r

[LowRider88]

8 minutes ago, razo+r said:

Your logic does not make sense.

Let me demonstrate:

Ace I-16 vs Ace F-18. The I-16 shot down the F-18. Does that mean the I-16 FM is inaccurate too? Because according to you, since both pilots have equal skill, the F-18 should win as it is superior in almost every aspect.

So why make jets when the I-16 is clearly superior in a 1v1?

Oh right. Because you can clearly see that the F-18 "Ace" pilot has no idea how to handle the airplane, in both your and my case.

So the only true issue you discovered here is how shitty the Ace F-18 pilot is.

example.trk 46.62 kB · 0 downloads

You are the one with the faulty logic.  You take it way off topic.

So you proved the F-18 AI is making bad choices, and so flys low and slow all the time.  In this case, of course it will lose to a prop plane.  Why are you comparing between eras?  That is the ridiculous thing here.

My comparison is more meaningful, as they are the same era, and from the same manufacturer.

You still haven't explained why the F-5 fly bys are mor extreme than all other planes.

Test all the jets and see the difference.

You still haven't explained why the F-5 wins against the F-18 in this scenario.

This is why DCS really needs a dedicated bug reporting system, rather than using a forum.  Answers should come from flight manuals and not from biased multiplayer teams, trying to protect certain advantages, whether accurate or not.

[razo+r]

8 minutes ago, LowRider88 said:

You are the one with the faulty logic.  You take it way off topic.

I'm argumenting by your logic that you came up with. Good to see you ackowledging it's faulty.

8 minutes ago, LowRider88 said:

So you proved the F-18 AI is making bad choices, and so flys low and slow all the time.  In this case, of course it will lose to a prop plane.  Why are you comparing between eras?  That is the ridiculous thing here.

Just doing what you are doing. You compare the F-5 to the F-18. I just took it to the extrem for a showcase.

But yes, this is a very important point. I proved the F-18 AI is bad. I took your example and went further with it. You even see that the F-18 is doing wrong things against the I-16, why can you not see it against the F-5? The F-18 is literally doing the same thing, fly low and slow. You even recorded the F-18 stalling a few feets above the ground while looking at the F-5 flying well above and not stalling. 

You even recorded the F-18 outturning the F-5 once the F-18 was not stalling.

You recorded evidence that the AI cannot handle a F-18, not that the F-5 is UFO-like. If that would have been the case, that the F-5 is UFO-like, so would the I-16 because it's older and managed to beat the F-18 (by actually shooting it down and not by getting him to eject).

If you want to prove that the F-5 AI is like a UFO, do what I told you to do. If you can do what the AI can do, it means it's performance is close to real. If it deviates, you then know the inaccuracy. But if you just compare AI vs AI at the same skill level, the only thing you'll show is how the AI handles these planes differently.

 

But anyway, here's another example that will hopefully finally convince you, as I made it with your conditions.

F-14A on ace vs F-18C on ace. F-14 wins.

They are from the same era and manufacturer. The F-14 is the predecessor of the F-18. So by your logic, it should not be able to beat the F-18, but in DCS does. So what does this mean then? Is the F-14 also a UFO? Or is it again because the AI doesn't know how to fight in the F-18? Or perhaps the F-18 is having an inaccurate FM that needs to be "buffed"?

Almost the same thing happened. F-18 got low and slow while the F-14 was higher than the F-18 and faster.

8 minutes ago, LowRider88 said:

My comparison is more meaningful, as they are the same era, and from the same manufacturer.

Nope. It's just less extreme. 

But anyway, my second experiment above should be as meaningfull if not more meaningfull than yours.

8 minutes ago, LowRider88 said:

You still haven't explained why the F-5 fly bys are mor extreme than all other planes.

What? What "more extreme"? And why should I explain it?

8 minutes ago, LowRider88 said:

Test all the jets and see the difference.

Why don't you test all of them?

8 minutes ago, LowRider88 said:

You still haven't explained why the F-5 wins against the F-18 in this scenario.

I did actually, even before this post.

8 minutes ago, LowRider88 said:

This is why DCS really needs a dedicated bug reporting system, rather than using a forum.  Answers should come from flight manuals and not from biased multiplayer teams, trying to protect certain advantages, whether accurate or not.

Actually, this is the nice thing about a forum. This way people can discuss everything. If someone is coming to a wrong conclusion, a forum can point it out.

 

[LowRider88]

43 minutes ago, razo+r said:

I'm argumenting by your logic that you came up with. Good to see you ackowledging it's faulty.

Just doing what you are doing. You compare the F-5 to the F-18. I just took it to the extrem for a showcase.

But yes, this is a very important point. I proved the F-18 AI is bad. I took your example and went further with it. You even see that the F-18 is doing wrong things against the I-16, why can you not see it against the F-5? The F-18 is literally doing the same thing, fly low and slow. You even recorded the F-18 stalling a few feets above the ground while looking at the F-5 flying well above and not stalling. 

You even recorded the F-18 outturning the F-5 once the F-18 was not stalling.

You recorded evidence that the AI cannot handle a F-18, not that the F-5 is UFO-like. If that would have been the case, that the F-5 is UFO-like, so would the I-16 because it's older and managed to beat the F-18 (by actually shooting it down and not by getting him to eject).

If you want to prove that the F-5 AI is like a UFO, do what I told you to do. If you can do what the AI can do, it means it's performance is close to real. If it deviates, you then know the inaccuracy. But if you just compare AI vs AI at the same skill level, the only thing you'll show is how the AI handles these planes differently.

 

But anyway, here's another example that will hopefully finally convince you, as I made it with your conditions.

F-14A on ace vs F-18C on ace. F-14 wins.

They are from the same era and manufacturer. The F-14 is the predecessor of the F-18. So by your logic, it should not be able to beat the F-18, but in DCS does. So what does this mean then? Is the F-14 also a UFO? Or is it again because the AI doesn't know how to fight in the F-18? Or perhaps the F-18 is having an inaccurate FM that needs to be "buffed"?

Almost the same thing happened. F-18 got low and slow while the F-14 was higher than the F-18 and faster.

Nope. It's just less extreme. 

But anyway, my second experiment above should be as meaningfull if not more meaningfull than yours.

What? What "more extreme"? And why should I explain it?

Why don't you test all of them?

I did actually, even before this post.

Actually, this is the nice thing about a forum. This way people can discuss everything. If someone is coming to a wrong conclusion, a forum can point it out.

 

Again with the faulty logic.  I said your logic is faulty, not mine.  Way to reach for a conclusion.

There is no need to take it to the extreme in your off topic case.  Like I said they are not the same era, and are not comparable.  That is poor scope on your part, and rather ridiculous.  I am not even going to address anything about interwar planes vs 4th gen jets.

You say I don't see the F-18 doing wrong things.  Read my earlier posts.  It is making bad decisions, already said that.  Don't rephrase me.

The F-5 is not only retaining too much speed, but can pull more Gs in the turn than the F-18.  That is why it gets behind the F-18 even when both are flying low and slow.  Watch the video again.

And no, you still did not explain why the F-5E reasonably wins here.  

I find it ridiculous.  The video clearly shows a discrepancy and you try to so hard to make sense of it.

Like I said, your "great idea" does not prove anything.  It may show the player F-5 FM may also be flawed.  And like I said, I will do this when I have time, even though this is irrelevant to the comparisons between planes AI FM.

Yet another off topic example.  You have a problem with scope.

Wrong, the F-14 and F-18 are from different manufacturers.

And why shouldn't the F-14 win?  The F-14 has much better wing loading.  It has the extra wing area between nacelles.

You just focus on the word predecessor, without understanding the peculiarities of era and within design house evolution.  I-16 is a different era, F-14 is a different design house.  The F-18 is an evolution of the F-5.  Yet another waste of an off topic comparison.

No your experiments are not useful and are off topic.

Why should you explain it?  Because you are challenging with bad data.

Why don't I test it?  I did.  All of them.  Why don't you since you still have not given any worthwhile data.

While it is fun to point out you wrong ideas, it is a waste of time.  Something more efficient is to have the bug reporting system based on flight manuals, rather than your missing data and misinterpretations.

According to you, the ridiculous end result in the video of my OP would just illogically explained away and nothing would get addressed.

Edited January 16, 2022 by LowRider88

[=475FG= Dawger]

There is no problem with AI FM for the F-5 that doesn't exist in every other module. The entire premise of this thread is faulty.

[LowRider88]

Just now, =475FG= Dawger said:

There is no problem with AI FM for the F-5 that doesn't exist in every other module. The entire premise of this thread is faulty.

Coming from one who knee jerks and jumps into experiments with all the wrong premise to begin with.

What expert are you to state that, with zero evidence?  Still see the wrong result in the OP video.  

Other AI are more realistic to play against, so your first sentence is useless.

[razo+r]

So if you tested it with all jets, please provide your data. Against which jets does the ace F-5 win? And same for the F-18.

And saying the F-5 overperforms, where do you present your data? Where are the charts? You have yet to present concrete data in this topic.

They cannot adjust things if you don't provide data. A video from the PoV of the F-18 while it's about the F-5 is not enough.

[LowRider88]

4 minutes ago, razo+r said:

So if you tested it with all jets, please provide your data. Against which jets does the ace F-5 win? And same for the F-18.

And saying the F-5 overperforms, where do you present your data? Where are the charts? You have yet to present concrete data in this topic.

They cannot adjust things if you don't provide data. A video from the PoV of the F-18 while it's about the F-5 is not enough.

You are questioning me, so you provide the data.  I will provide more findings when I have time, even thought the OP video speaks for itself, and anyone can reproduce this.

Are you a devleper?  If not, how do you know it is not enough?  There is a freaking video, which anyone can reproduce.  Ridiculous.

I prefer to provide those details to actual people who can fix this and would be objective about it.

[=475FG= Dawger]

This thread and the OP's insistence on problems with the F-5 AI FM remind me of this very funny scene.

 

[razo+r]

19 minutes ago, LowRider88 said:

You are questioning me, so you provide the data.  I will provide more findings when I have time, even thought the OP video speaks for itself, and anyone can reproduce this.

Are you a devleper?  If not, how do you know it is not enough?  There is a freaking video, which anyone can reproduce.  Ridiculous.

I prefer to provide those details to actual people who can fix this and would be objective about it.

No, you are reporting a bug so you have to provide the data.

And how do I know it's not enough? Read the "How to report a bug" post by 9L.

4 minutes ago, =475FG= Dawger said:

This thread and the OP's insistence on problems with the F-5 AI FM remind me of this very funny scene.

 

Haha, I feel like this right now

[LowRider88]

5 minutes ago, razo+r said:

No, you are reporting a bug so you have to provide the data.

And how do I know it's not enough? Read the "How to report a bug" post by 9L.

Haha, I feel like this right now

No, you are challenging it, and requesting it despite being far removed from being a developer or expert.  Case in point, you think Northrop Grumman, means F-14 and F-18 are designed by the same company.  So you provide what rebuttal you want.  How many times do I have to tell you, I will provide more details, when time permits?  Anxious much?

Don't care whether you think it is enough or not.  You are not the authority.  Will wait for an objective developer response.

Didn't bother watching your off topic linked video.

OP video end result speaks for itself.  Horse is beaten dead already.

[razo+r]

I'm not the authority, you're right. But don't cry when the moderators tell you the same thing I did. And by the way, you only said you will post your findings about my suggested experiment when time permits but okay. We shall wait and see.

OP Video does speak for itself, but not in the way you think it does.

[LowRider88]

2 minutes ago, razo+r said:

I'm not the authority, you're right. But don't cry when the moderators tell you the same thing I did. And by the way, you only said you will post your findings about my suggested experiment when time permits but okay. We shall wait and see.

OP Video does speak for itself, but not in the way you think it does.

I don't cry when interacting with objective developers.

No I disagree, the video end result is an end result, and not what you try to explain it away to be.

[MiG21bisFishbedL]

8 hours ago, razo+r said:

So if you tested it with all jets, please provide your data. Against which jets does the ace F-5 win? And same for the F-18.

And saying the F-5 overperforms, where do you present your data? Where are the charts? You have yet to present concrete data in this topic.

They cannot adjust things if you don't provide data. A video from the PoV of the F-18 while it's about the F-5 is not enough.

They wouldn't anyways. Especially now that the GFM is in the pipe.

[LowRider88]

For Developers:

 

DCS Version = 2.7.9.18080

 

This video shows a comparison of the Ace AI FMs for both the DCS F-5E and the F-14A.

 

The F-14 has:

Gross Weight = 27,669 kg

Engine Thrust with Afterburner = 125 kN

Thrust/Weight = (125 * 2 * 1000) / (27,669 * 9.807) = 0.921

So, it has a better T/W ratio over the F-5's ratio of 0.598

Also, when taking into account the wing area between its nacelles, the F-14 has an effective wing loading of about 230 kg/m^2, while the F-5 is worse at 407.919 kg/m^2.

 

Yet, in this video, the F-5E is out-climbing the F-14 and turning tighter without losing speed. Here are some examples:

 

*Time: 2:30 - 3:15 2 circle fight, with 2 head on passes, in both, the F-14 starts with better angle off nose (e.g time 2:45), but at time of head on pass closure, F-5 swings unrealistically and gets a better angle off nose over the F-14 (times 2:54, and 3:15 ). This should not happen when the F-14 has the double advantages of the stats described above.

This happens again at timeframe: 8:54 - 9:14.

 

*At 3:46, F-5E pops brakes to prevent itself from overshooting the F-14 in a climb. Again should not be possible by a plane with weaker engines and smaller wings.

 

*At 4:57, the F-5 pops breaks again, then is immediately about to climb straight up to catch up to F-14 at the latter's climb peak and stall. The F-5E should stall well before the F-14, especially when it starts a climb with brakes, is climbing straight up, and only turns on afterburner halfway through the climb.

 

* At times 6:44, 7:31, 8:15 F-5E once again wins 2 circle fight head on pass with better angle off nose.

 

* At time 11:10 - 11:12, the F-5 pulls 7.1 Gs at 13739 ft with only a speed lose from 398kn to 382kn, 16 kn.

 

* At time 15:40 the F-5 is climbing with brakes again, at 15:56, the F-14 stalls and F-5 still able to have nose authority to shoot at it.

 

* Fight ends with F-14 bailing out due to lack of fuel. This also should not happen. The F-14 has turbofan engines while the F-5 has turbojets. Turbofans are more efficient at sub and transonic speeds like these during a dogfight. The F-14 should also have way more fuel capacity. This occurred in the video because the F-5E is retaining too much speed and can climb straight up with brakes on and still catch the F-14 stalling.

 

Given the F-5E's small wings and weak engines, the only way it should be able to shoot at the F-14 is if it can force the F-14 to overshoot through a scissors fight. Instead, the F-5E is able to win a 2 circle fight despite weaker engines and smaller wings.

 

 

 

For Developers:

 

DCS Version = 2.7.9.18080

 

This video shows a comparison of the Ace AI FMs for both the DCS F-5E and the F-14A.

Specifically, it is meant to provide a comparison point from the DCS F-5E AI FM to the F-5E Turn Performance Chart FA8-13 (Sheet 2) on page A8-41 (or page 466) of the F-5E NATOPS Flight Manual.

 

In this test, the F-5 and the F-14 are bot set to 50% fuel, so that the F-5 can best match the chart. The F-5 is also loaded with 2 missiles as indicated in the chart.

 

As can be seen in the video, the F-5 is pulling a sustained turn at time 7:19.

At this time, the F-5 is showing the following stats:

IAS = 352 kn

ALT = 5328 ft

G = 6.6

 

Looking that the manual's chart, this should indicate a point in the graph to the right of the 330 KIAS vertical line, and between the 6 G and 7.33 G curved downward sloping lines. This point is well above the Sustained curve.

So the DCS F-5E AI FM is able to pull way too much G.