DCS: F-14 Development Update - AIM-54 Phoenix Improvements & Overhaul - Guided Discussion

[Harlikwin]

10 hours ago, Schmidtfire said:

After some more testing the main performance issues seems go be:

AWG-9 inability to hold tracks until missile goes active, combined with relatively low kinematics and easily defeated missile seeker.

Tested over and over in PvP multiplayer against other fighter sized aircraft. My verdict is that Phoenix is a near useless missile in that scenario. The pk under most conditions will be too low. From a realism standpoint it’s hard to reach a conclusion. If it suffers from DCS-isms, server lag, flawed missile code or if it’s actually right on the money in it’s representation. But unless running scripted PvE scenarios, it’s not something I would use if the target can A; get an RWR warning and B; pull >5g

I mean probably not really unrealistic. DCS PVP being what it is people have learned how to deal with the missile. If you take a hard look at the iran iraq war alot of those "fighter" kills were against non maneuvering targets that had no clue they were being shot at. So, a totally apples to oranges set of expectations. And yeah I would expect the AWG-9 to have issues holding tracks vs maneuvering fighters.

People may not like this since their expectations largely come from movies but it is what it is. The biggest realism flaw being that in DCS you know how to deal with it. IRL I think that level of understanding probably existed in the sense that there were likely recommended tactics (run!) if you got fired on by a tomcat. But we will never know what those were and how effective they were. The iraqis eventually did figure it out though as the losses to phoenixes in the late war were far lower. 

 

 

[captain_dalan]

11 hours ago, Tweety777 said:

The target was at 2k ft altitude, I suppose that qualifies as low flying?

Aye, i would consider everything under 5000ft to be very low. On co-altitude launches, i don't expect any hits at all from 50 miles away, not even on hot ones. Shooting from a look down may improve the chances, but then you are faced with the issue of radar dropping tracks due to less-then-perfect lookdown capability. Especially over land. But i hadn't tried this enough times to be able to make an educated guess. Anyways, against cold targets..... unless the bandit changes his mind and turns hot, 50 miles is a stretch, even at high altitudes. 

[captain_dalan]

1 hour ago, Harlikwin said:

The biggest realism flaw being that in DCS you know how to deal with it....

 

 

Realism may mean different things to different people. If we look at extremely basic, micro-tactical realism, then yeah. I would agree. But if we look at operational realism, then DCS open PvP servers, generally aren't a good example of it. I mean, air superiority is completely eschewed as a concept. So it's only naturally so that any associated metrics are rendered void as well. Spawn points tend to be just outside modern guided artillery range, so abundant and careless use of burner at low altitudes is non-issue. Had real life air operations been conducted like an average PvP brawl, there would hardly be any fighting at all. Reds would hide in the mountain, and blues would have uncontested control of the sky. At best some point defense actions might take place, but that's it. So the meta isn't exactly hospitable to some hardware. Nothing wrong with that. Fun is more important after all, and probably the main reason why the majority of the userbase is here.

[Tweety777]

5 hours ago, Harlikwin said:

I mean probably not really unrealistic. DCS PVP being what it is people have learned how to deal with the missile. If you take a hard look at the iran iraq war alot of those "fighter" kills were against non maneuvering targets that had no clue they were being shot at. So, a totally apples to oranges set of expectations. And yeah I would expect the AWG-9 to have issues holding tracks vs maneuvering fighters.

People may not like this since their expectations largely come from movies but it is what it is. The biggest realism flaw being that in DCS you know how to deal with it. IRL I think that level of understanding probably existed in the sense that there were likely recommended tactics (run!) if you got fired on by a tomcat. But we will never know what those were and how effective they were. The iraqis eventually did figure it out though as the losses to phoenixes in the late war were far lower. 

 

 

Could this not be somewhat neutralized by making the seeker a bit better then it really was? At the end of the day the results would then get closer to the real results described here. The improvement doesn't need to be huge, just a bit less vulnerable to notching. Just me thinking out loud.

3 hours ago, captain_dalan said:

Aye, i would consider everything under 5000ft to be very low. On co-altitude launches, i don't expect any hits at all from 50 miles away, not even on hot ones. Shooting from a look down may improve the chances, but then you are faced with the issue of radar dropping tracks due to less-then-perfect lookdown capability. Especially over land. But i hadn't tried this enough times to be able to make an educated guess. Anyways, against cold targets..... unless the bandit changes his mind and turns hot, 50 miles is a stretch, even at high altitudes. 

Thanks for all the information, I learned a lot here.

[DD_Fenrir]

1 minute ago, Tweety777 said:

Could this not be somewhat neutralized by making the seeker a bit better then it really was? At the end of the day the results would then get closer to the real results described here. The improvement doesn't need to be huge, just a bit less vulnerable to notching. Just me thinking out loud.

The notch susceptibility is not the problem per se; it's the automatic pitch up that the missile does on losing track of a target in the notch. This pulls the missile seeker field of view away from the target and negates any chance to reacquire the target if it pops out of the notch again.

Let's be clear, the DCS AIM-54 will reacquire if given the opportunity. I had a rare case last night where a 40nm shot against a MiG-25 was notched when the MiG cranked to defend but as the missile commence it pull-out, the MiG popped out of the notch just within the seeker field of view and the Phoenix reacquired. Only just got the MiG as the -25 then defended with a descending slice-back and I - on the edge of my seat - was willing it all the way in. It seemed an eternity till it finally connected with only about +25 knot overtaking speed!

The trouble is finding any data to corroborate whether any of these behaviours are right or wrong.

Would the real AIM-54 pull up as soon as it is notched? Seems a bit unlikely that the guidance engineers would miss the opportunity to allow the missile a chance to reacquire a target that might only notch momentarily, but this then begs the question:

Would the real AIM-54 seeker and guidance system be able reacquire a target emerging from a notch? Or is the technology of the Phoenix's guidance tracking system of just too early a generation to have that abilty?

 

[draconus]

1 hour ago, Tweety777 said:

Could this not be somewhat neutralized by making the seeker a bit better then it really was?

This is not the way. Fix any bugs, make the missile specs and behavior as realistic as possible according to any known data - that's how it's done.

[Dannyvandelft]

Flew with a 4-ship of Tomcats last night, loaded 2-2-2 across the board, AIM-54C mk47 motors, all 60nm+ shots against targets ranging from MiG-21 to MiG-31. Out of 8 missiles 6 hit their targets, 1 was defeated, and 1 was an extra shot on a target that already had an incoming missile that was thought defeated but did indeed hit. So of 7 single missiles shot at targets, 60nm+ distance, 6 hit. Mix of TWS and STT.

The missile works good, BUT it's a lot more hands on than an AMRAAM. You have to know your launch windows. Speed, altitude, of both you and your target, to make those long distance kills.
The AI cheats. At 10nm they detect the missile and go defensive, even if the missile isn't Pitbull yet. That is what needs fixing. A lot of misses we see in tacview afterwards are because the AI just decides to go defensive when there's no possibility of it knowing it was shot at.

Sent from my SM-G960U using Tapatalk

[draconus]

20 minutes ago, Dannyvandelft said:

...the AI just decides to go defensive when there's no possibility of it knowing it was shot at.

Them flying stupidly straight at us wasn't good and realistic either.

[Dannyvandelft]

Them flying stupidly straight at us wasn't good and realistic either.

Sure, but it should be possible to meet somewhere in the middle and let the AI detect the missile when it goes pitbull I would think.

Sent from my SM-G960U using Tapatalk

[Schmidtfire]

@Dannyvandelft I aggree. Against AI the Phoenix is good enough. For some reason it does not work well against other players. If you try the exect same shots against a human player, they will miss (unless that player forgot to turn on the RWR)

3 hours ago, draconus said:

This is not the way. Fix any bugs, make the missile specs and behavior as realistic as possible according to any known data - that's how it's done.

As realistic as possible, yes. But known data is just part of all data needed. Gaps need to be filled by the devs or the overall accuracy will suffer. Going only by known data will result in an inaccurate implementation. If devs need to add 20% qualified guesswork to get a more realistic outcome, I think it’s better than to leave things out.

 

 

Edited May 10, 2023 by Schmidtfire

[Dannyvandelft]

@Schmidtfire I think that has more to do with players being more educated. There's so many articles, posts, youtube videos about how to defeat each missile. They know the limits of the simulation and use that to their advantage to defeat the threat.

Sent from my SM-G960U using Tapatalk

[draconus]

49 minutes ago, Dannyvandelft said:

Sure, but it should be possible to meet somewhere in the middle and let the AI detect the missile when it goes pitbull I would think.

Either react at the moment of RWR warning or evade some time before that, like the human would do, expecting that Tomcat might have taken a long shot already. That should work not only for taken shots obviously and should be based on AI difficulty.

44 minutes ago, Schmidtfire said:

As realistic as possible, yes. But known data is just part of all data needed. Gaps need to be filled by the devs or the overall accuracy will suffer. Going only by known data will result in an inaccurate implementation. If devs need to add 20% qualified guesswork to get a more realistic outcome, I think it’s better than to leave things out.

It's probably what they do already, if needed, but educated guess is not the same as "let's make it a little better in the name of the gameplay".

[Harlikwin]

22 hours ago, captain_dalan said:

Realism may mean different things to different people. If we look at extremely basic, micro-tactical realism, then yeah. I would agree. But if we look at operational realism, then DCS open PvP servers, generally aren't a good example of it. I mean, air superiority is completely eschewed as a concept. So it's only naturally so that any associated metrics are rendered void as well. Spawn points tend to be just outside modern guided artillery range, so abundant and careless use of burner at low altitudes is non-issue. Had real life air operations been conducted like an average PvP brawl, there would hardly be any fighting at all. Reds would hide in the mountain, and blues would have uncontested control of the sky. At best some point defense actions might take place, but that's it. So the meta isn't exactly hospitable to some hardware. Nothing wrong with that. Fun is more important after all, and probably the main reason why the majority of the userbase is here.

I mean I agree. All I'm saying is that compared to real life where you don't have unlimited lives to figure out how to "game" the missile defense. And you know there is an actual risk of dying. You are gonna fly very differently than what DCS looks like in PVP. 

[Billy]

Compared to IRL, people in DCS are too good to exploit the game's flaws, but too bad at avoiding mistakes. 

[Kageseigi]

1 hour ago, Billy said:

Compared to IRL, people in DCS are too good to exploit the game's flaws

I think this is probably the most accurate statement.

There are three ways to fly in DCS:

1. By the book. The way you are "supposed" to fly.
2. Pushing the limits. "Cheating" by doing things that you aren't "supposed" to do, but could be done (in reality).
3. Gaming the system. "Cheating" by doing things that should not be possible (exploiting flaws in the simulation).

At least against modern fighters, it seems that the F-14/AIM-54 combination relies heavily on #2, but gets hard countered by #3.

[eatthis]

2 hours ago, Kageseigi said:

I think this is probably the most accurate statement.

There are three ways to fly in DCS:

1. By the book. The way you are "supposed" to fly.
2. Pushing the limits. "Cheating" by doing things that you aren't "supposed" to do, but could be done (in reality).
3. Gaming the system. "Cheating" by doing things that should not be possible (exploiting flaws in the simulation).

At least against modern fighters, it seems that the F-14/AIM-54 combination relies heavily on #2, but gets hard countered by #3.

if half the stries from flight crew are half true and plenty of them actually did number 2 to some extent irl, me personally im british so im a sneaky bugger looking for ways to make things work but i dont flat out cheat. example being i had to find a strike group without knowing where they were coming from.
i figured theyd be on a boat so cycled through the standard tacan freqs, boom there they are (where theyre coming from any  way) 

 

i know if you fire the c in stt its fox 1 all the way and will go duymb if you snip it. is the same true for the c or will it wake up if the awg9 loses the track?

3 hours ago, Billy said:

Compared to IRL, people in DCS are too good to exploit the game's flaws, but too bad at avoiding mistakes. 

agreed, i try to milsim (or mostly) and every single flight has errors in it! 

Edited May 11, 2023 by eatthis

[draconus]

12 minutes ago, eatthis said:

is the same true for the c or will it wake up if the awg9 loses the track?

The C fired in PD-STT should go active only when losing track but in DCS it will always go active.

[Temetre]

Am 11.5.2023 um 09:16 schrieb eatthis:

if half the stries from flight crew are half true and plenty of them actually did number 2 to some extent irl, me personally im british so im a sneaky bugger looking for ways to make things work but i dont flat out cheat. example being i had to find a strike group without knowing where they were coming from.
i figured theyd be on a boat so cycled through the standard tacan freqs, boom there they are (where theyre coming from any  way) 

Real war is all about improvisation, reacting to the enemy strategy, and changing your own approach faster and more effectively than the opponent. "Cheating" is a game term, a concept of fairness that doesnt exist in war. The only time you ever do things "by the book" is if theres no alternative, or if you are in such a superior position, that doing things by the book is guaranteed to be succesfull anyway. The more serious a fight is, the more you get willing to throw over previous rules.

Thats why eg Desert Storm saw less improvisation, compared to Vietnam, which basically just rewrote "the book". Arguably the USAF even took some wrong lessions in Vietnam, because while their upgrades to F-4Es & co were meaningful, they were not nearly as effective as the Navys focus on "just" better training.

To some degree that even happens in DCS PvP: Phoenix come out, they are super powerful, till people start notching and learn how to defend when F-14s are around. In return, the F-14 pilots develope strategies how to use Phoenix' against opponents with those strategies. Or they just abuse the fact that they can force opponents to defend. Even if the Phoenix is easily defeated, its an opportunity cost to do so.

As to how you wanna fly the plane, follow specific eras or situations rules... thats then more of a RP-thing, where you and others put limitations on themselves to create a certain kind of scenario. And ofc the AI has the issue that it doesnt know how to adapt to strategy.

 

Thats why Im almost skeptical by principle when people talk like combat in DCS is only realistic if youre doing 100% by the book. Its more like, the issue is that we only got the book for how to use planes. And we dont even have the full book, because the miltary keeps everything beyond the basics secret. So its very difficult to actually create "realistic" scenarios. Thats not a dealbreaker, but I wish a lot of people were more conscious about that.

Edited May 12, 2023 by Temetre

[CarbonFox]

I'll try to upload more examples but this is something I setup quickly last night. Tried the AIM-54As in STT this time against a Tu-95 and got the same results with the missiles flying a straight trajectory.

No Loft.trk

[draconus]

15 hours ago, CarbonFox said:

I'll try to upload more examples but this is something I setup quickly last night. Tried the AIM-54As in STT this time against a Tu-95 and got the same results with the missiles flying a straight trajectory.

Indeed PD-STT shot fly straight while TWS-A lofts, but that is because Tu-95 starts jamming immediately after locking. If you set him to no reaction to threats, the missile lofts in PD-STT too.

Edited May 13, 2023 by draconus

[eatthis]

19 hours ago, Temetre said:

Real war is all about improvisation, reacting to the enemy strategy, and changing your own approach faster and more effectively than the opponent. "Cheating" is a game term, a concept of fairness that doesnt exist in war. The only time you ever do things "by the book" is if theres no alternative, or if you are in such a superior position, that doing things by the book is guaranteed to be succesfull anyway. The more serious a fight is, the more you get willing to throw over previous rules.

Thats why eg Desert Storm saw less improvisation, compared to Vietnam, which basically just rewrote "the book". Arguably the USAF even took some wrong lessions in Vietnam, because while their upgrades to F-4Es & co were meaningful, they were not nearly as effective as the Navys focus on "just" better training.

To some degree that even happens in DCS PvP: Phoenix come out, they are super powerful, till people start notching and learn how to defend when F-14s are around. In return, the F-14 pilots develope strategies how to use Phoenix' against opponents with those strategies. Or they just abuse the fact that they can force opponents to defend. Even if the Phoenix is easily defeated, its an opportunity cost to do so.

As to how you wanna fly the plane, follow specific eras or situations rules... thats then more of a RP-thing, where you and others put limitations on themselves to create a certain kind of scenario. And ofc the AI has the issue that it doesnt know how to adapt to strategy.

 

Thats why Im almost skeptical by principle when people talk like combat in DCS is only realistic if youre doing 100% by the book. Its more like, the issue is that we only got the book for how to use planes. And we dont even have the full book, because the miltary keeps everything beyond the basics secret. So its very difficult to actually create "realistic" scenarios. Thats not a dealbreaker, but I wish a lot of people were more conscious about that.

 

thats pretty much how i look at it

[CarbonFox]

21 hours ago, draconus said:

Indeed PD-STT shot fly straight while TWS-A lofts, but that is because Tu-95 starts jamming immediately after locking. If you set him to no reaction to threats, the missile lofts in PD-STT too.

 

Thanks. Did another test against an H-6 and set it so it wouldn't employ ECM. The AIM-54C lofted as it should. When locking in TWS, how close do I need to get to burn through any jamming the target aircraft may be using?

[eatthis]

1 hour ago, CarbonFox said:

Thanks. Did another test against an H-6 and set it so it wouldn't employ ECM. The AIM-54C lofted as it should. When locking in TWS, how close do I need to get to burn through any jamming the target aircraft may be using?

iv seen 30 miles mentioned but take that with a big pinch of salt as im not 100%

[Callsign JoNay]

1 hour ago, eatthis said:

iv seen 30 miles mentioned but take that with a big pinch of salt as im not 100%

Should be 29 nm for most fighters.

23 nm for some blessed fighters, and most of the attackers.

12 nm for strategic bombers.

[Karon]

More info: https://flyandwire.com/2022/12/07/electronic-countermeasures-missiles-employment-part-iii/