SE V: Unique Abilities And Ideas

[President_Elect_Shang]

@SJ: Sorry but I didn’t get that at all? Was that meant for me?

@SpectarofDeath: Man I have got to come up with something shorter for you. Typing out your name is like typing mine out! Nutz! Do you have a preference? Anyway I don’t think SJ was saying making the mod was a dumb or pointless idea. He was just arguing that optimal fighter designs in each category will surface. While I don’t agree and feel a middle of the road design will come to dominate, everything is still good. I respect his opinion and feel enriched for having read it. I’m sure he will agree that he too took something positive out of this.

Now as for the mod Atrocities and I have spoken a little by PM and he has expressed an interest but lacks the time to dedicate right now. Did I get that right Atrocities? Anyway I am interested in this and have even gone so far as jotting down a few short notes focused on ship classes and what role they may best serve. Kana has posted this excellent graph which would create a wide range of flexibility, maybe even outside the scope of a mod like this. I would envision his graph best applied to one or two limited hull sizes and those few would be specialty classes. If another person is willing to head up this mod; with Atrocities blessings of course, I would certainly be willing to work on it. Currently I have another project underway and fear getting into this as a solo project.

[Suicide Junkie]

The empirical evidence in CBmod does show that you *can* make the specialized designs come to dominate... but its not nessesarily easy.

The important point is that you have to set up the underlying physics carefully; all the roles have to be of roughly equal importance. (Focus on balancing point defense, leaky PD too, in order to keep missiles fighters and ships all useful throughout the game.)
The weapons that are good for one should be mostly incompatible with use in the other roles. Not only the weapons, but engines, and defenses too.
It helps to allow things to take up huge amounts of space. A good anti-ship weapon could take up 80% of a fighter, naturally leaving you with a slow, lightly armored bomber design.
The best anti-missile weapons could be short ranged, thus requiring powerful engines to intercept successfully. Engines taking up half the space on your fighter, plus the guns, leaves little armor.
And if you want a fighter that can stand up to return fire, you'll need moderate guns and a moderate engine, giving you a dogfighter.

In SE5, you have the additional dimension of shot speed to play with, but you lose the weapon-stacking mechanism.


Its not easy, but IMO, it is worth it.

PS:
Would you like to play a game of CBmod to see the theory in action?

[President_Elect_Shang]

SJ I just realized we are talking apples and oranges here. Your observations are based on MP and mine on Solo. From each of our respective POV’s we are correct. Thanks for the offer but I don’t play online games, never have, given it some thought but haven’t been persuaded yet.

[Phoenix-D]

President_Elect_Shang said:
SJ I just realized we are talking apples and oranges here. Your observations are based on MP and mine on Solo. From each of our respective POV’s we are correct. Thanks for the offer but I don’t play online games, never have, given it some thought but haven’t been persuaded yet.

The same thing applies in solo if the AI is smart enough to recognize it.

[Spectarofdeath]

Spec or Spectar is fine, it's what other ppl use.

[Suicide Junkie]

Phoenix-D said:

The same thing applies in solo if the AI is smart enough to recognize it.

The AIs that Rollo made for CBv1.5 should fit that bill. As AIs, they are pretty limited in flexibility, but they should be able to show you how it works.
(Use no bonus setting. Low bonus = steamrolling)

Humans are the ultimate opponents, though some more than others. Myself, I'm a good ally, but a weak opponent

[Suicide Junkie]

How about a simple idea:
Leaky point defense.

Since SE5 dosen't have any concept of multiplex tracking or autofire, we need a new way to make PD leaky.

1) Make each PD shot capable of taking out a missile in one hit, maybe 2 for a heavy missile.
2) Make the PD refire rate 50ms.
3) Give the PD -999% chance to hit.
4) PD shots should be very fast. Beams are good too. There should only be one or two shots from each gun on screen at any one time.
5) For varing PD resistance on your missiles, use *speed* rather than hitpoints or ECM.
6) For varying PD techs, reduce size and cost, increase range.

This gives you a poisson distribution of PD hits...
One missile can get lucky and penetrate 5 PD guns, or one PD gun could get lucky and shoot down 3 missiles at once.

For example, if PD guns have a range, fire rate and incoming missile speed such that they can each fire 50 times, then:
- One PD gun has a 40% chance of shooting down at least one missile. A 9% chance of shooting down two or more missiles, and 1.5% chance of shooting down three missiles.
- Two PD guns have a 63% chance of shooting down at least one missile. 26% chance of shooting down two or more, 8% chance of shooting down three or more, and 2% chance of shooting down 4+
- Four PD guns have a 87% chance of shooting down at least one missile. 59% chance of shooting down at least 2, 32% of shooting down 3 or more, 14% for 4 or more, 5% of 5 or more, etc.

Longer range gives more shots before the missile hits, until the point where the second wave starts coming into PD range before the first wave of missiles has hit.
Smaller PD guns, lets you fit more guns to increase your fire rate for more shots total again.

Key features of this system:
- One missile CAN sneak in occasionally. (13% chance vs 200 enemy PD shots)
- Graceful degradation of PD as the number of missiles increases (unlike stock where the first X missiles are shot down, and the rest all hit)

Downsides:
- Lots of shots make the CPU work a little harder
- ECM and combat sensors do not affect PD. (But there are other ways to do it; speed, range, etc)

[Kana]

I love it...

Suicide Junkie said:
How about a simple idea:
Leaky point defense.

Since SE5 dosen't have any concept of multiplex tracking or autofire, we need a new way to make PD leaky.

1) Make each PD shot capable of taking out a missile in one hit, maybe 2 for a heavy missile.
2) Make the PD refire rate 50ms.
3) Give the PD -999% chance to hit.
4) PD shots should be very fast. Beams are good too. There should only be one or two shots from each gun on screen at any one time.
5) For varing PD resistance on your missiles, use *speed* rather than hitpoints or ECM.
6) For varying PD techs, reduce size and cost, increase range.

This gives you a poisson distribution of PD hits...
One missile can get lucky and penetrate 5 PD guns, or one PD gun could get lucky and shoot down 3 missiles at once.

For example, if PD guns have a range, fire rate and incoming missile speed such that they can each fire 50 times, then:
- One PD gun has a 40% chance of shooting down at least one missile. A 9% chance of shooting down two or more missiles, and 1.5% chance of shooting down three missiles.
- Two PD guns have a 63% chance of shooting down at least one missile. 26% chance of shooting down two or more, 8% chance of shooting down three or more, and 2% chance of shooting down 4+
- Four PD guns have a 87% chance of shooting down at least one missile. 59% chance of shooting down at least 2, 32% of shooting down 3 or more, 14% for 4 or more, 5% of 5 or more, etc.

Longer range gives more shots before the missile hits, until the point where the second wave starts coming into PD range before the first wave of missiles has hit.
Smaller PD guns, lets you fit more guns to increase your fire rate for more shots total again.

Key features of this system:
- One missile CAN sneak in occasionally. (13% chance vs 200 enemy PD shots)
- Graceful degradation of PD as the number of missiles increases (unlike stock where the first X missiles are shot down, and the rest all hit)

Downsides:
- Lots of shots make the CPU work a little harder
- ECM and combat sensors do not affect PD. (But there are other ways to do it; speed, range, etc)

[President_Elect_Shang]

I haven’t seen it yet Phonix-D. Then again as SJ pointed out (man am I glad I read all the new posts first) the AI is only as smart as the modder who tells it how to build those designs. For me I just find the hole in the design and build ships to deal with it. Then you are right back to an optimal middle of the road design. Sorry Phoenix-D but I’m not buying it till I see it. “It” being an AI that can recognize the flaw in my middle of the road design and adjust its designs to counter and exploit that flaw.

SJ at a -999% chance to hit I’m not seeing how you are calculating such good odds for the P-D to hit. Can you drop your formula here? I might try the CBMod again when it comes out for SE5.

Maybe I should dust off SE4 and try a MP game after all… MP experience allowed Kana to point out a flaw in one of my component ideas and I haven’t even finished the TC yet! Or maybe when I get far enough I can do a MP to track errors in the TC!

[Phoenix-D]

The odds are good because the fire a lot. At 50 fire rate you get 20 shots a second. That means even with a -999 hit chance you still get an average of one hit every 2 seconds.

I see a few problems, though:

-In Starfury, very fast projectiles had a tendency to randomly explode before hitting. This might have carried over to SE5.
-SE5 might not calculate very low hit possibilities correctly