I agree, Rescuer is a very good fighter order for dogfights to counter Last Stand. Now, in a general fleet action, if this makes the fighters target cruisers instead of fighters, that’s cool with me. Enemy fighters are a distraction from the real goal of killing big ships, which is what ultimately determines victory. If the enemy has to make his cruisers fighter-proof, that’s less ordnance he can toss at my big ships, so all to the good.
Well, the thing is in a LEASHED fight (escort vs escort)
slow heavy armed fighter -> fast fighter -> laser fighters -> slow heavy armed fighters.
So yes, if I arm my serenity with 2 lasers, it will slaughter your fighter in this “slow vs slow” fight where I have a huge DPS advantage. But it doesn’t make it good in other dogfights because it can’t hit anything past 3 speed.
If they are unleashed, then rocket will always win against laser as long as the rocket guy set engagement range to 1000. Rocket fighters are both faster and has longer range, so the rocket will always run away and the laser chase them, never get to fire a shot while getting hit.
There are many many triangular relationship in this game and to find “the best setup” is often misleading.
Although yes, Tribe is the best race with by far the most OP bonus, which makes them the best in pretty much everything. Their HP bonus is so huge that you can out survive other ship with just 1 reflect shield and no armor. This means you are also getting a cost reduction, speed bonus, power bonus, and crew bonus on top of your extra survivability for not having to buy extra shield/armor that other race needs. The only thing they are missing are armored tanks.
And that, I think sums GSB up best. Rock, paper, scissors. For everything there is an antidote, even though it doesn’t appear so on the surface. So I’ll quit trying to figure out what’s “best”.
Still, there’s room for experimentation. Some things are probably better in a wider range of circumstances than others.
Well certainly, especially once you start designing fleet for posting rather than just beating something, then you gotta play an handicap match where you have to account for everything while trying to hide the weakness you very well know is there.
WOW, Really impressive work man, loved reading it, GJ!
Oh man, i really want to see that, but i think that Uni-T wins
OK, to continue…
My next question was, which fighter laser is the best dogfighting weapon? The Fighter Laser Cannon (FLC), the Fighter Pulse Laser (FPL), or the Nomad Dogfight Laser (DFL)? To determine the answer, I created mobs of the fastest FLC, FPL, and DFL fighters I could, then fought those of the same race against each other. Thus, same hull, same orders, same numbers. They only differed in weapons, plus the associated speed and hitpoint differences due to the different weapon weights and the different engine and power combinations needed to make them as fast as possible.
The question thus became, for any given race, which laser weapon made the best dogfighter? Note that I did NOT address how the laser fighters of different races compared, nor how lasers compared to rockets. This was just to see whether a given race’s best FPL could beat that same race’s best FLC or DFL.
Some may find the results surprising. The FLC is hands-down the best laser dogfight weapon. The FPL is decidedly inferior to the FLC, but vastly superior to the DFL. This is exactly the opposite order of their respective tracking speeds. There, FLC is the worst with 2.8, then FPL with 2.9, and DFL with 3.1. So, what’s up with that?
Well, there are 2 factors involved: the ability to get a shot off, and then the chance to hit with that shot. But I’ll address the chance to hit first.
Per Cliffski, if the target’s speed is greater than the weapon’s tracking speed, then there is only a 2% chance to hit. But if tracking speed is greater than target speed, the chance to hit is calculated by the following formula:
% To Hit = (1 - (S/T)) * (0.5 + (L/256) * 0.5)), where
S = target speed, T = weapon tracking speed, and L = target length
Now, all laser-armed fighters are slower than 2.8 (the FLC’s tracking speed) because they all need a power generator. Thus, the formula applies and all to-hit chances are > 2%. Also, the FPL is both heavier than the FLC and uses more power, so usually needs a bigger generator as well, especially to get it as fast as possible. The net result of the extra weight is that any given race’s fastest FPL is significantly slower than its fastest FLC. This speed difference is so great that it overwhlems the FPL’s slight advantage in tracking rate, so the FLC always has a significantly better chance to hit. Given that the FLC and FPL have the same rate of fire, the advantage goes to the FLC.
For example, the fastest Federation FLC has a speed of 2.48, their fastest FPL has a speed of 2.15, and both are 12m long using the Leopard hull. The FLC’s tracking rate is 2.8, the FPL’s is 2.9. Plugging these values into the formula, we see that the FLC has a 12.2% chance of hitting the FPL, while the FPL has only a 7.6% chance to hit the FLC. So all other things being equal, the FLC should win most of the time. And that’s what happened. I ran 10 tests. The FPL barely beat the FLC the 1st time (20-10) but lost badly in all other fights (average 21-43). So the FLC definitely superior at dogfighting to the FPL.
But then I tested the DFL against the best Nomad FPL. Both were 11m long. The DFL had a speed of 2.09 tracking 3.1 and the FPL a speed of 2.11 tracking 2.9. This gave the DFL a 16.7% chance to hit compared to the FPL’s 14.6%. uIOW, the DFL’s higher weight and slower speed, which increased the FPL’s chance to hit, was exceeded by the DFL’s better tracking speed[/u]. So you’d think that the DFL would win in a close fight. This, however, was not the case. The FPL stomped the DFL every time at an average of 37-75. WTF?
This brings me to the other factor in success: getting an opportunity to fire. The chance to shoot is a function of 3 things: 1) the weapon’s range, 2) the relative speeds, with the faster fighter able to get in range more often, and 3) the rate of fire, so that when you get in position, your gun isn’t recharging. The FPL is superior to the DFL in all respects here: longer range, slightly faster fighter speed, and 2.27x the rate of fire. These advantages proved to be WAY more significant than the slight difference in to-hit chances. While the FPL won’t hit as often per shot, it gets the chance the fire many more shots and a fair number of these will be when the DFL is unable to reply due to range or reloading.
So, the bottom line is that if you want to use laser-armed fighters to dogfight, the FLC gives you the best chance. It’s got the longest range, it’s the fastest (thus hardest to hit) laser fighter you can make, it’s tied for 1st in rate of fire, and it’s also the cheapest. The FPL is decidedly worse in range, fighter speed, and cost. The DFL is total crap. It’s the most expensive by far, yet makes the slowest, most vulnerable fighters with the shortest range and a MUCH lower rate of fire.
I can thus see no reason at all to own an FPL. Anything it can do, the FLC can do better. There’s even less reason to use the DFL.
EDIT: fixed an incomplete sentence.
This bit is because the FPL fires more than twice as quickly as the DFL (every 15 times the DFL fires, the FPL fires 34 times, and with the hit chances as they are, the FPL makes about 4.96 hits per 34 shots while the DFL makes 2.51 hits per 15 shots). On top of this, the FPL has a higher theoretical DPS (20) than the DFL does (11.76). The combined effect of these two factors is that the FPL deals between 14.89 and 29.78 damage in the time that the DFL deals 10.02 to 20.04 damage. This means that, assuming fighters have an average health of about 20, the FPL will kill 0.75 to 1.5 fighters in the amount of time it takes a DFL to kill 0.5 to 1 fighter.
The DFL also has a lower theoretical DPS than does the FCL, which does 13.33 theoretical DPS. This combined with the above paragraph indicates to me that there no reason to use the DFL, since when firing on unarmored targets that the FCL, FPL, and DFL have similar chances to hit, the higher rate of fire of the preceding two weapons will produce a greater number of hits for the FCL and FPL as compared to the DFL, and both the FCL and the FPL have a higher theoretical DPS than the DFL, so the damage differential will probably be magnified. If the DFL had significantly higher chances of hitting than either the FPL or the FCL, then there would probably be situations where it was better. Right now, if any such situations exist they are probably quite rare. Only advantage the DFL might have is that it is more likely to disable a fighter system with each successful hit than the FCL and FPL are, but hitting half as frequently is a very strong argument against trying to make use of that.
I would say that having a mix of FPL and FLC fighters for strafing armored cruisers is probably the best use of the FPL. Why? When the average armor rating is greater than 8, both the FPL and the FCL have the same chances of making critical hits, and the FPL will erode the armor more rapidly since the FLC and the FPL have the same rate of fire but the FPL has a higher damage per shot. Once the average armor is between 3.8 and 8.0, the FCL becomes temporarily superior to the FPL and erodes the armor rapidly due to being able to penetrate the armor while the FPL cannot. Once the average armor falls below 3.8, the FPL enters into its own and deals damage at a 50% greater rate than the FLC, which results in the cruiser dying rapidly. If you’re facing unarmored cruisers for whatever reason, then the FPL is strictly superior to the FCL as a strafing weapon.
Where does this leave the Swarm’s Fighter Fusion Gun (FFG)? I would say that the FFG is unarguably the worst laser-type weapon in the game for a dogfighter, with a very low tracking speed of 2.0, and half the rate of fire of either the FLC or the FPL. Thus, against almost any fighter (except for torpedo fighters), the FFG will have the lowest hit chance of all possible dogfighter weapons, and doesn’t make up for it with an unusually high DPS (14.17 theoretical maximum, only slightly better than the FLC, and much worse than the FPL. The DPS is much higher for the FFG than for the DFL, but this is probably the one case in the game where the DFL’s high tracking speed makes up for its low rate of fire and low DPS). Where the FFG ‘shines’ is in strafing cruisers - it’s sort of a compromise between the FLC and the FPL, with slightly more DPS than the FLC and slightly more armor penetration than the FPL. The FFG is also more expensive than either of the two weapons that it is a compromise between. Thus, it is second-best to FLC/FPL combinations at stripping armor from highly armored cruisers, it is second-best to the FPL at killing things that have lost their armor or never had armor (though it is superior to the FLC here, just not by much), it won’t help an FPL strafing squadron much since it will not start stripping armor rapidly as soon as the FLC would, and will not improve the DPS of an FLC strafing squadron as much as supporting FPL fighters would, and still won’t save any money (in fact, it will cost more while doing less). Thus, it’s essentially a second-best strafing option, slightly outperforming FLCs once armor is broken and slightly outperforming FPLs when the average armor is between 3.8 and 5.0 (not a very big window, is it?), and the worst dogfighter option.
I’d say that the ordering of strafing groups is roughly: FPL/FLC mixes > pure FPL >= pure FLC ~= pure FFG >> DFL (a ~= b means a is approximately equal to b, a > b means a is better than b, a >= b means a is mostly better than b, a >> b means a is much better than b). Naturally, there will be differences in the hierarchy of effectiveness depending on the scale of the armor and the presence and effectiveness of strafing prevention measures (cruiser antifighter weapons, escorting fighter groups, escorting antifighter frigates), with pure FLC strafing groups becoming better as the effectiveness of the antifighter defenses increases.
This is a perfect example of the Fighter Paradox, which states that fighter effectiveness is inversely proportional to fighter cost. With fighters, speed is the dominating factor, and you only get a slow fighter by spending money to add stuff to it. Fighter modules are structured so that the more they cost, the heavier they are, and so the slower your fighter is. And the extra money and the loss of speed come nowhere near justifying themselves in terms of being more likely to hit the enemy.
I’m sure Cliffski did not intend for things to be this way. I mean, pulse lasers on frigates and cruisers are anti-fighter weapons, so you’d think the FPL was supposed to be an anti-fighter weapon, too. But because the FPL weighs more and uses more power than the FLC, any fighter carrying the FPL will cost more but will be inferior to the same hull with an FLC. It’s even worse for the DFL, whose name implies that it was supposed to be THE BEST weapon laser for dogfighting. The way it is now thus cannot be what was intended.
To fix this, I’d switch the weight and power stats of the FLC and FPL. If that didn’t work, next I"d up the FPL’s rate of fire and/or tracking speed, or perhaps lower its energy needs so FPL fighters didn’t need a power generator but could make due with just an engine. THAT would really make the FPL a killer in dogfights. Then, after I’d made the FPL markedly superior to the FLC in dogfights, I’d give the DFL the same stats as the new FPL but with a higher tracking speed and/or rate of fire and/or lower weight, so that it was markedly superior to the new FPL.
Strafing effectiveness is something I intend to test in the near future. While as usual you have some nice on-paper numbers to back you up, it’s been my experience so far that an FLC will eventually gnaw through a non-tank’s armor but an FPL never will. I have no idea why this is, but it’s something I’ve observed in numerous battles. I intend to find out the answer.
I can say, however, that most FLC fighters have speeds from about 2.45-2.58, which compare favorably with the Cruiser Pulse Laser’s tracking speed of 2.60. Even at a relatively slow FLC speed of 2.45, the CPL only has about a 3% chance to hit, barely above the minimum. With faster FLCs, it’s the minimum 2%. Meanwhile, the FPLs mostly go 2.20-2.00. At an average FPL speed of about 2.10, the CPL has about a 10% chance to hit. So, the longer it takes to “lucky hit” the armor into oblivion, which is at the whim of the Dark Gods, the less likely there’ll be any FPLs left to do it. However, given that each FLC shot has about the same odds of causing a “lucky hit” as the CPL has of hitting an FLC, and that the FLCs of the whole squadron shoot way more often than the handful of CPLs available to oppose them, the odds favor the FLCs devouring the cruiser before the CPLs kill very many FLCs.
We shall soon see. I have already determined that the FFG utterly sucks in dogfights but remain unconvinced that it has any virtue as a strafing weapon. My opinon right now, supported only by unscientific observations, is that the FPL and FFG are both significantly inferior to the FLC in strafing. If that’s true, it doesn’t matter whether the FFG is better or worse at strafing than the FPL, because neither is as good as the FLC. Because the Swarm can use the FLC, I predict that they will have no reason to use the FFG. But this remains to be determined.
For the nonce, I am testing the effectiveness of the Micro Target Booster in Swarm vs. Swarm fights. I shall post the results shortly but can say already that things aren’t looking good for the MTB .
The FPL’s description indicates that it’s more of a very close in weapon than the FLC, and specifically mentions strafing runs. I’d say it was meant to be about on par with the FLC as a dogfighting weapon and superior as a strafing weapon, but that’s just an impression. One challenge might be to try to make a Fighter Pulse Laser that results in a roughly 50-50 win/loss ratio against same-race FLC fighters (or maybe a 40-60 split, in the FLC’s favor) while maintaining the partial superiority that I think the FPL has in strafing. I’d sooner leave the FLC as-is and reduce the weight of the FPL than see the FLC become significantly worse as a dogfighter. Something as simple as giving the FPL the same power consumption as the FLC would help here, since this would give the FPL a speed boost through being able to use the same power generator as the FLC. It would probably still need to have some weight reduction in addition to the power consumption reduction, though.
The DFL, from its name, really should do much better in dogfights than it does, but as you’ve pointed out it’s terrible at it, and with its lower rate of fire and poor theoretical DPS it isn’t worth having while strafing because of the pitiful armor penetration.
Having just tested the relative strafing effectiveness of Imperial Phalanx Fighters equipped with Pulse Lasers (speed 1.98) against Imperial Phalanx Fighters equipped with Fighter Laser Cannons (speed 2.39), with both fighter forces having 100% priority 100 range Attack Cruiser orders and no orders to attack fighters, and with both fighter groups arrayed in 7X7 squadrons in opposite corners of the map, with 3X3 Imperial Praetorian Tank Cruisers (~73 average armor, engine 1, minimal crew, 1 MWM launcher (to avoid hitting fighters), and 1 Nanobot Repair System) in the other two corners, spaced widely enough to avoid damage from shockwaves, I can say that the Pulse Laser strafing group consistently defeated all nine tank cruisers in the time it took the Laser Cannon fighters to defeat six tank cruisers, despite being able to afford only 47 full squadrons and ~5 extra fighters armed with the FPL, as compared to 49 full squadrons of FLC fighters, and despite the FLC fighters engaging their targeted Tank Cruisers before the FPL fighters were able to reach their targeted Tank Cruisers. Pulse Lasers win this contest handily.
I also tested a modded Pulse Laser whose power requirements were the same as the Fighter Laser Cannon’s, using Imperial Phalanx Fighters. Simply reducing the Pulse Laser’s power consumption to that of the Fighter Laser Cannon allowed me to boost the fighter’s speed from 1.98 to 2.15, and resulted in the Pulse Laser-armed fighters consistently beating the Laser Cannon-armed fighters (with victory being at about 20% for FPL, 10% FLC, using the above mentioned test scenario). However, these were Imperial Phalanx Fighters, and I did not test other types of Imperial fighters, nor did I test fighters of other races, so results may vary depending on the quality of the fighter used. This indicates that, if there is a fighter in the unmodded game where you can switch out a FLC for a FPL without having to upgrade the power generator, that fighter would be better off using an FPL than an FLC in laser fighter versus laser fighter dogfights.
If you want to play with the scenario, I can post it as a challenge. Just be warned that the defending fleet uses 794 pilots, of which only ten are for cruisers (I put a cruiser behind each fighter group to encourage the block of nine tank cruisers to stay at their starting map edge, not that it mattered since they’re so slow that they weren’t going anywhere, anyways, and the fighters never chose to run across the center of the map to target the far-corner cruiser, opting instead to go for the near-corner group of nine cruisers). Gratuitous Space Battles, at least on my computer, suffers from slight graphical glitches when approximately 1500 fighters are present on-screen, and replaces the engine trails with white boxes while slowing down the game slightly. The fighters in the defending fleet are Imperial Phalanx fighters armed with Fighter Laser Cannons, so the Pulse Laser-armed fighters are something you’d need to provide.
Next up, the skinny on the Swarm’s Micro Target Booster (MTB). This gizmo, for a small bit of weight and power consumption, adds 0.10 to the tracking speed of the fighter’s weapon. Is this worth the loss of speed compared to fighters without the MTB? The answer is no, although it does make things more interesting.
The speed loss imposed by the MTB is greater than its relatively low weight and power consumption would lead you to believe at first. This is because the Swarm is one of the races whose fastest laser fighters use 4-slot hulls with 2 engines. So, not only do you have the weight of the MTB itself, but you lose an engine’s worth of thrust.
So, after much tinkering with all combinations of hulls, engines, and power generators, I arrived at the following designs. All use the Seth hull, selected because, with the same modules, it gives exactly the same speeds and HP as the Amun hull but costs less, and is even the same length (12m) so provides no defensive size bonus. So what use is the Amun? Anyway…
Modules: FLC, PG2, E2, E1
Speed: 2.54, Cost: $80.60, HP: 21.85
Tracking speed 2.8
Modules: FLC, MTB, PG3, E2
Speed: 2.05, Cost: $81.40, HP: 20.90
Modules: FPL, PG2, 2x E1
Speed: 2.21, Cost: $81.40, HP: 20.90
Modules: FPL, MTB, PG3, E2
Speed: 1.87, Cost: $86.20, HP: 20.90
NOTE: I did not try the MTB with rockets because the FRL’s tracking speed of 2.0 is too low for the MTB to help it hit an FLC. FRLs can only hit painted fighters, and then only when the rockets don’t have to turn while en route. The same can be said of not testing the FFG, with or without the MTB (which I’ve already done anyway); it’s a non-starter as a dogfighter.
Anyway, I set the presumptive overall winner, the standard FLC, as the challenger and attacked it with the other 3 variants. As usual, same numbers, same orders (other than engagement ranges for the various weapons), same deployment, same hull. Here are the results:
NOTE: I am unsure how target boosters affect tracking speed and thus can’t figure out the to-hit chances of the MTB fighters. Do you add the MTB’s 0.10 to the weapon’s base, do you add 10% of the base tracking speed, or something else? I await word from the experts. But in the meantime, I will show the to-hit chances of the MTB fighters using both methods, simple addition first, and remember, both could be wrong ;). And anyway, neither method results in a higher to-hit chance than the FLC has standing alone.
FLCMTB (6.5% or 9.2%) vs. FLC (14% to hit)
The FLCMTB lost 5 straight matches with an average score of 26-52. The 0.49 loss of speed overwhelms the better tracking rate, no matter how you figure it.
FPL (6.5% to hit) vs. FLC (11% to hit)
The FPL lost 5 straight with an average score of 25-50, which is essentially the same as the FLCMTB. The conclusion, therefore, is that if you want your FLC fighters to be as bad as FPL fighters, put an MTB on them.
FPLMTB (8.0% or 10,7%) vs. FLC (17.4% to hit)
I did 10 matches with this pairing because the lead always switched back and forth rapidly for the 1st 40% or so losses to each side, and the FPLMTB actually won the 2nd match. However, the FPLMTB lost 8 out of 10 times with an average score of 24-47. It won the other 2 matches on average 28-10. The overall scores for all 10 matches averaged 24-40 against the FPLMTB. But more about this match below…
The MTB is useless. However you figure its boost to tracking speed, it doesn’t make up for the speed penalty of using it. The Swarm would be better served by having their own versions of the standard FLC and FPL that incorporate the tracking speed increase without needing to carry the MTB. Now, if the Swarm was one of the races that doesn’t benefit from twin-engine laser fighters, this might be a different story.
HOWEVER, I noticed something very interesting in the FPLMTB vs. FLC matches. Note above that based on the official to-hit formula, the FPLMTB has only an 8.0% (or 10.7%, or something else) chance to hit the FLC. This is better than any FPL without the MTB but still well below the FLC’s 17.4% chance to hit it back, which is markedly the highest so far in any pairing because the speed loss in the FPLMTB is the most extreme. But the actual after-battle statistics told a very different story. The stats showed that the FPLMTB always hit more often than the FLC, even when losing. The actual hit rates were usually about 22% for the FPLMTB vs. about 10% for the FLC.
Given that both weapons have the same rate of fire, and that the FPL does 50% more damage than the FLC, this means that the FPLMTB must have fired SIGNIFICANTLY fewer times than the FLC, to out-hit it by that much and still lose most of the time. And I don’t see how the FLC’s tiny advantage in HP could make such a difference. So, I’m left to conclude that the opportunity to fire a shot is more important than the chances of hitting with it. In this case, the FPLMTB is way slower than the FLC and has a slightly shorter range, although rate of fire is the same. This is the only explaination I can come up with.
But I’m still left with a question. Where did the > 20% hit rate of the FPLMTB come from? It implies that the MTB does considerably better than even adding 10% to the base tracking speed.
This also begs the question of how the FPL managed to fire significantly fewer times than the FLC, if that is the reason for the disparity in performance. The difference in range isn’t significant enough to give large differences in firing opportunities using standard fighter orders, as frequently the fighters will make head-to-head passes, fly roughly parallel to one another, or be in one giant clump of fighters for significant portions of a fighter battle. Assuming you were using the standard fighter engagement orders of ‘engage fighters at 200 range’, the majority of the fighters should have spent most of their time in firing range of whatever weapons their enemy was carrying. Even if you used an increased engagement range, you couldn’t have increased it much or your laser fighters would spend most of their time completely out of range of their targets, which isn’t at all useful.
I also wonder about FPL versus FPLMTB engagements, as it would seem from the FPLMTB versus FLC engagement that the FPLMTB might actually beat the FPL fighter group (not that that’s a real advantage, at present, but still).
Another thing: did the FLCMTB fighters show a similar unexpectedly high hit rate, or was it only the FPLMTB fighters that demonstrated a higher-than-expected hit rate?
Brilliant! Thanks for taking the time.
OK, so you’ve answered 2 key questions: 1) FPLs can actually devour cruisers, and 2) can do so better than FLCs, at least in the absence of opposition. I owe you a beer :).
Now, we need to try them against something with anti-fighter weapons and, very importantly, shields. The shield is an important variable because it will stop all fighter shots fired from outside it. I was planning on making siccing 1 squadron of fighters against 1 normal, fully equipped cruiser complete with 1 CPL, then 2. And also CDLs in their turn.
Note: I only tested each of the following once.
I tested 2 pulse laser + 1 supercharged tractor beam Imperial Legion Cruisers, with Nanobot Repair and Fast Recharge Shields, Engine 1, and minimum power and crew necessary to run the cruiser, then switched out both pulse lasers for cruiser defense lasers while keeping everything else the same, and then replaced 1 CDL with an armor plate, moving the remaining CDL to the just forward of center hard point of the Legion cruiser. The Supercharged tractor beam occupied the centermost hardpoint on all cruisers, and when two weapons were carried, I used the leftmost and rightmost hardpoints of the legion cruiser, to give roughly equal coverage all around the cruiser.
Fighters tested were Phalanx hulled laser and pulse laser fighters, described previously.
Fleet set-up is the same as described previously, with the exception that I dropped the tenth cruiser which was sitting behind each fighter group. Each fleet contained 49 full fighter squadrons this time around; eliminating the tenth cruiser freed up enough money for the last few FPL fighters, I guess (the target cruisers might have been cheaper this time around, too).
Pulse Laser fighters performed consistently better, killing all nine target cruisers before the Laser Cannon fighters killed more than 5 cruisers.
Results at the time the last cruiser on one side died:
…1x CDL, STB…2x CDL, STB …2x Pulse, STB
FPL-targeted Cruisers Killed:…9 … 9…9
FLC-targeted Cruisers Killed:…3… 4 … 5
Results at time of Victory:
FPL-side Cruisers Surviving:…3…5…3
Score (FPL:FLC):… 63:32…51:26…53:26
I decided not to test other combinations (such as no tractor beams, normal tractor beams, or higher densities of CDLs/Pulse Lasers) since, aside from higher densities of Pulse Lasers, I don’t believe that any of those combinations are very likely (not that I think that a fleet consisting of 9 cruisers armed exclusively with a pair of CDLs or Pulse Lasers and a Supercharged Tractor Beam is very likely). I also decided not to repeat the tests done above, as the minimum number of surviving cruisers on the side of the FPL fighters at the time the last of the cruisers on the side of the FLC fighters died was 4, none of which had yet taken hull damage, and cruisers take a very long time to kill when they have a reasonable amount of armor remaining. Even just one of the cruisers described above is incredibly survivable to fighter swarms, though, and that in all cases at least three cruisers on the side of the FPL fighters survived without significant hull damage (even though the repair systems generally ran out by the end of the fight even for the last surviving cruisers) despite being targeted by roughly 700 laser fighters (I believe that the CDLs and Cruiser Pulse Lasers killed fewer than 100 fighters in all cases, although it is very difficult to assess fighter losses in non-campaign settings) is quite impressive to me. Granted, all of these cruiser started the battle with average armor ratings 56.82 (2 CDL/CPL, 1 tractor) or 61.x (1 CDL, 1 tractor), but I still would have expected to see more damage from having 700 or more fighters constantly strafing the cruisers.
Attack fighter orders were deleted from the fighters, attack cruiser orders were deleted from the cruisers. Fighters still went after other fighters (selected fighters as driver targets) once all potential cruiser targets were cleared from the battlefield. FLC fighters continued their track record of reaching the target first and killing it more slowly, and shield-equipped cruisers live longer against strafing runs than do pure armor tanks (largely due to all the laser fire bouncing off of the shields). All fighters were ordered to engage cruisers at 100 range (minimum possible engagement range), but this still resulted in the fighters spending about half their time outside the shield bubble.
Actually, I put the FPLs on a range of 193 (because it wouldn’t let me do 190). I was trying to make it close to within the same fraction of its maximum range as the FLC’s default of 200, but in hindsight I see that should have been 180. Too much beer I guess. Anyway, the reason for setting the FPLs to close in more was to compensate for their shorter range and slower speed. I didn’t want them to turn away from their faster targets too soon to get a shot off. Both sides had Stick Together, Last Stand, and Rescuer.
Anyway, now that I’ve realized these things, I’ve refought the battle numerous times with the FPLMTBs set for different ranges (180, 200, and 250) and with and without rescuer. This made no discernable difference in the results and the FPLMTBs never won again. So orders don’t seem to make any real difference, and the FPLMTBs always out-hit the FLCs but still lose.
Then, I refought the FLCMTB vs. FLC a few times just to look at the stats, which I hadn’t done before. As you suspected, the FLCMTB hit more often than the FLC but the difference was not as pronounced as in the other fight. Here, the FCLMTB hit about 20% of the time and the FLC about 15%. So again, the FLCMTB fired significantly less times than the FLC, despite both sides having the exact same weapon and the exact same orders (200 range, Last Stand, Stick Together, and Rescuer). So, the only differences were speed on 1 side and MTB on the other.
Wow. So then I looked at the stats from a rematch of non-MTB FPL vs. FLC. Again, the losing FPL had a significantly higher hit percentage while still losing. Can’t blame this on the MTB because it wasn’t there, so it must have to do with the speed difference (as the above rules out range being a factor).
I suspect that it all comes down to the AI, or rather the AIs because there are definitely 2 different AIs at work here: the movement AI and the shooting AI. Unless put on a leash with Escort or whatever, the movement AI is a runaway horse. It selects a target to move towards and will keep moving toward it until some trigger event occurs (the target dies, a ship reaches its specified range and doesn’t have Keep Moving, etc.). It seems that the movement AI hardly ever changes targets on its own beyond these triggering events. It does not care at all that it can’t catch its chosen target, it will futilely chase it anyway, always trying to reach the specified range to that target. In a situation where the only available targets are all faster, such as here, this is happening all the time.
The shooting AI is more controllable with orders (Cooperate, Retaliate, Rescuer, etc.) and also frequently changes targets on its own. For these reasons, it will often shoot at different targets than the movement AI is heading toward. However, I get the impression that the movement AI’s target holds a special place in the shooting AI’s heart. After all, the movement AI usually picks it to start with because it’s the closest example of the highest priority target type, which is exactly what the shooting AI is most interested in by default. I therefore think it possible that the when the shooting AI needs to change targets, the 1st thing it looks at is the movement AI’s current target, on the assumption that this is still the closest example of the highest priority target type. And I figure the shooting AI must keep its eyes on this target for a measurable amount of real time before deciding it can’t hit it and looking for a different target on its own.
This all seems a plausible way to do the AI for big ships that don’t move very fast or at all, and so don’t change their relative positions very often. But in a dogfight, it would cause problems. The movement AI of the slower fighter latches onto a target it can’t catch and charges futilely after it, ignoring the fact that it’s passing within spitting distance of dozens of equal-priority targets, and the shooting AI is dragged along with it. The shooting AI does take potshots at passing targets of opportunity but because none of them stay in range very long, it’s always having to change targets. If each time it changes targets it looks to the movement AI’s target for a while, then for that time period it won’t shoot at any of the other targets that are in range right then. And because this is the slow fighter, it probably won’t be in range of the movement AI’s target, so it won’t shoot at it, either. IOW, it doesn’t shoot at all for a while until it decides to look for its own target. But in the swirl of a dogfight, that new target quickly goes out of range and so it wastes more time looking at the movement AI’s target.
Applying these rules to the faster fighter avoids these problems. Because this is the faster fighter, it WILL come within range of the movement AI’s target. Once it’s dead, the movement AI will pick another target which, in the middle of a dogfight, is probably already in range and the shooting AI will shoot at it immediately. Thus, no wasted shot opportunities.
Anyway, this is all speculation but it does match observations and explain why the losing fighters fired so much less despite having an equal rate of fire.
In all 3 scenarios, the losing fighters (FPL, FPLMTB, and FLCMTB) ALL had higher hit percentages than the non-MTB FLC. The difference was greater with the FPLs, whether equipped with MTB or not, than with the FLCMTB.
There’s still the question of the higher-than-expected hit rates. I can only guess. But I have noticed in many cases with big ships that their 1st shot almost always hits, even though the next 4 or 5 might miss. Perhaps there’s something (the RNG or some factor we don’t know about?) that gives a bonus to the 1st shot. In a dogfight, there are a lot of 1st shots because the shooting AI is constantly changing targets. So if there’s some 1st shot bonus, this would be the place you’d see it most.
This theory is plausible to me, although it would be a royal pain to prove using only end-of-battle fighter duel data. It would also be the first target on the list, and then the fighter would have to spend some time re-evaluating local targets to choose who to fire on when it realizes that no, it can’t shoot the driving target.
Also, the testing of varied ranges is interesting, though I suspect that varying the engagement range on the faster fighters is more likely to affect the outcome, since the faster fighters are more able to control the engagement range and ‘choose’ whether to spend any time at all inside the range of the opposing force’s guns (more or less, anyways - maneuvering takes a little time to respond, and a fighter’s turning radius and response time is large enough that I would expect that the slower fighters would still manage to get off some shots).
Which is why I don’t dignify it with the name “theory”. Theories make predictions that can be proved or disproved by experiment and observation. Because this idea can’t be, it’s an hypothesis :). I call it:
The Slow Fighter Hypothesis: In the constant swirl of a dogfight, the AI of the slower fighter will waste many firing opportunities at available targets due to focusing its attention on the uncatchable target it’s currently chasing.
When I was doing rocket-painters earlier, I noticed that when everything else was the same (fighter design, numbers, deployment, and all other orders), having different ranges made a significant difference. In such fights (using rockets), IIRC a range of 400 will beat all shorter ranges. However, ranges above about 450 will always lose to shorter ranges. In between 400-450, it’s a toss-up. I therefore think it possible that there is a similar optimum range setting for dogfighting with the other weapons. In fact, the next thing on my list is testing this for the FLC and FPL. Then the question will be whether ranges that work best in same vs. same fights also work best when facing fighters with different weapons, but that’s something far down the list.
So I’ve been playing around with the targeting calculations a bit, and the only way that I have come up with that an MTB (which ‘adds’ 0.1 to tracking … somehow) would produce a >20% hit rate for the fighter speed and size that you’ve listed is if the tracking speed gets modified by 0.1 (by both multiplying the tracking speed by 1.1 and by adding 0.1 to the tracking speed, to get a roughly 12% chance for an FPLMTB Seth to hit an FLC Seth; the order in which the multiplication and the addition is performed does not make a noticeable difference in the calculations), and then the MTB adds 0.1 to the decimal (10 to the percentage) final hit chance. This would seem to be a ridiculously complicated way in which to calculate the bonus from a target booster, and would also make me wonder why the cruiser target boosters never seemed worthwhile if this really is the way the calculation is done. If the MTB only added 0.1 to the final hit chance, then I would expect to see around 16.5% of the shots fired by the FPLMTB Seths hit the FLC Seths. Given, though, that the actual hit rate of the FLC Seths was about 57.5% of the expected hit rate, having a 16.5% expected hit rate jump to 22% would be somewhat surprising.
Possibly the FLC Seths were more negatively affected by the hits they sustained, as they have two engines, a power generator, and a weapon, which gives a 50% chance for their engines to be damaged by hits sustained, which would lead to the fighters being slowed more rapidly on average than the MTB fighters, which have only one engine, a power generator, a target booster, and a weapon, giving a 25% chance for the engine to be affected by the fighter sustaining hits. It’s well-known that damage to a ship’s engines will cause its maximum speed to be reduced, and the FLC and FPL Seths are much more likely to sustain engine damage with each hit than either of the MTB fighters are, which would likely result in the FLC and FPL Seths losing more speed per hit than the MTB fighters would.
This, however, doesn’t explain why the FLC Seths saw a reduction in accuracy from that which was expected, since they should still benefit from the slowing effect of damage on their opposition, even if not by as much as their opponents gained from damage to the FLC Seths.
Is there an optimum dogfight range for fighter laser weapons, and if so, what is it? The answers are “yes” and 100. HOWEVER, the actual rule is that having a shorter range is advantageous, and the greater the difference, the better it is. So, if both sides use the same range, it’s a push.
My baseline enemy was the Swarm Seth FLC with the default range of 200. I attacked this with an exactly identical force except using different ranges. Here are the results:
FLC 250 vs FLC 200: FLC 250 lost every time, average score 17-34.
FLC 150 vs FLC 200: FLC 150 lost 3 out of 5 games but the overall average score of the 5 games was 25.2-24.6 in favor of range 150. So in general, I think the shorter range has the overall edge and it was just luck of the draw that it lost the majority of games. Had I played more games, I think the long-term trend would have been in favor of 150 range. Still, a range difference of 50 isn’t enough of an advantage to really be significant.
FLC 100 vs FLC 200: The FLC 100 won every time at an average of 41-10. So, having a range 100 shorter than the enemy is a decisive advantag.
Note that the effect of a 50-unit range different are not consistent. At the longer ranges of 250 vs 200, having the shorter range is decisively good. At the shorter ranges of 150 vs 200, having the shorter range arguably makes no difference, although it might be slightly advantageous.
I got the same results when I tried FPL vs FPL. Once again, a range of 100 decisively beats a range of 200, a range of 150 isn’t much different from 200, and anything over 200 loses to 200.
Having now set up such challenges, I decided to go ahead and test whether using a range of 100 made any difference in fights between different types of lasers. The results are as follows:
FPL 200 vs. FLC 100
Test to see if th FLC would lose its previous advantage by spending more time inside the FPL’s shorter range. The FPL managed to win 1 game 28-10 but lost the other 4 by an average of 24-48. So… Setting a range of 100 for the FLC does make it more vulnerable to the FPL but not a lot.
FPL 100 vs. FLC 200
Would the range difference make the FPL competitive with the FLC? Damn if it didn’t. The FPL won 7 out of 10 games with an overall average score of 31-18.
FPL 100 vs FLC 100
So now the real test. With both types at their best ranges, who would win? Answer: FLC 100, 4 games out of 5, total average score 26-42. IOW, the FLC doesn’t do quite as well against the FPL as when both are at 200, but still has a decided advantage.
FPLMTB 100 vs FLC 100
The FPLMTB lost all 3 games by an average of 21-43, a slight improvement over when both are at 200 but not enough to matter.
So, bottom line:
- In dogfights, a range of 100 optimizes the performance of all laser fighters. A range of 100 provides enough of an advantage over the default of 200 that non-FLC laser fighters will usually beat FLCs this way.
- The FLC is still the best laser dogfighter, provided it uses a range of 100. If it keeps the default of 200 and other types use 100, the other types will usually win. However, if the FLC has range 100, it will almost always beat the other types, although will lose occasionally, and will take more losses than if both are at 200.
While I was doing this, I also made some observations concerning the Slow Fighter Hypothesis. What I did was click on both fast and slow fighters so their modules appeared in the lower right, then observed what the weapon modules were doing. As I understand things, if there’s a white dot in the upper left corner of the module, it’s tracking a target. Then there’s a white bar across the bottom. When this is full, the weapon is ready to fire. When it’s black or partially white, the weapon is recharging. Anyway, here’s what I saw happening.
Most of the time, the modules had a white dot. While they had a white dot, the white bar was almost always in constant motion with the weapon firing and recharging apparently at its maximum rate of fire. Only very, very rarely did I see a white dot and a full white bar remain in place for any length of time, which I assume means that the tracked target was out of range and the shooting AI was waiting to see if it would come into range. As I framed the Slow Fighter Hypothesis above, this is a condition I was expecting to see quite frequently and last a long time. But it hardly ever happened at all, and only lasted a split second when it did happen.
HOWEVER, I also saw times where there was no white dot and the bar remained totally black. I take this to mean that the weapon had no target, and thus saw no reason to recharge after its last shot. While this happened to both fast and slow fighters, it appeared to happen slightly more often and last slightly longer for the slow fighters.
So, I’m forced to come up with a new basis for the Slow Fighter Hypothesis. I still believe that the slower fighters take less shots and that the root cause is the AI, but instead of being fixated for a while on an unhittable target, it appears that instead the shooting AI can’t decide what to shoot at and wastes time floundering in indecision.
Because the FPL will achieve about 20% against the FLC even without the MTB involvd, I don’t think you need to concoct weird ways for the MTB to work. I think you explained the situation with the loss of 1 engine, something that hadn’t occurred to me. When you lose and engine, I assume the thrust goes away but the you keep the weight. So, I opened the Seth FLC up in th Ship Designer and swapped out each engine with some other module that weighs the same. Here are the results, without worrying about the MTB:
Loss of the Engine 2 (weight 2.5, 17 thrust): Speed drops from 2.54 to 1.59, giving FPL a 23.6% chance
Loss of the Engine 1 (weight 2.0, 14 thrust): Speed drops from 2.54 to 1.94, giving the FPL a 17% chance.
But the FLP has 2x Engine 1s and the loss of 1 drops the speed from 2.21 to 1.53, giving the FLC a 23.7% chance. So, the same thing applies, although the increase in hit chance isn’t as dramatic because the FLC has a pretty good chance to start with.
The Engine 1 has 5 HP and the Engine 2 has 6 HP. Thus, the FPL can destroy either with 1 hit, and each hit on a cherry FLC has a 50% chance of doing 1 or the other. Hits on engines don’t decrease the FLC’s outgoing fire so the orders I’ve been using will keep the FPLs shooting at it, and now it’s much easier to hit. It will take 3 more hits to finish it off, all registered at the higher to-hit chance. Thus, the final hit rate stats will show a hit rate greater than expected.
The other 50% of the time, however, the FPL’s 1st hit will the weapon or power plant. It can also destroy these with 1 hit. These hits will stop or greatly reduce the FLC’s outgoing fire so suddenly this targets assumes low priority under the orders I’m using. It might therefore even survive the battle. So the only hit registered on it will be at the expected lower rate.
But OTOH, the FLC can’t destroy an Engine 1 with 1 hit and, as I understand things, damaged parts keep working at full efficiency until destroyed. So, the FLC requires at 2 hits on the same engine before its hit chance improves, and this will only happen 1/16 of the time for the 1st 2 shots. Thus, the majority of hits scored will be at the original, lower rate, keeping the post-battle stats low.
NOTE: The above assumes all hits happen at optimum range. As I understand things, laser damage varies from 50% at minimum range to 100% at optimum range back to 50% at max range. Thus:
FPL FLC Range Damage Range Damage 15 3 20 2 80 4 110 3 145 5 240 4 210 6 270 3 230 5 300 2 250 4 270 3
So, in the general course of events, it will usually take the FPL has a better chance of taking out the FLC’s Engine 1 with 1 shot than it does the Engine 2, but the FLC will usually always just need 2 hits to destroy an engine.
Thus, it looks like the Slow Fighter Hypothesis was a theory after all, and has now been decisively disproven. I owe the AI an apology.