|
Post by hoffmads on Jun 25, 2019 17:31:29 GMT -6
Shoho went down after being hit by 13 bombs and 7 torpedoes. She was a 12000 ton CVL. And the only reason she didn't get hit by more was because there simply weren't more planes around to drop stuff on her. And Shoho wasn't travelling alone either. Granted, I do think the AI is FAR overzealous in kicking the dead horse (particularily so with gunfire), and the AI engagement logic may be also flawed in that sense. But while agreeing that a look is necessary in that department, I don't necessarily thing it's as outrageous as you seem to think . But you also have to remember that the Shoho was the object of the strike, not some derelict the strike found burning along the way to the target location. Overkilling the target, especially if it's a carrier or capital ship seems acceptable to me. Some early attacks seem reasonable but it seems to me that there should be around an 80% chance of a strike flying to the target location before executing a strike even if there are available targets along the way. This was sort of my thinking too. Shoho was a fully functional carrier when the strike arrived, and also my pilots were more experienced than Lexington's and Yorktown's were at that point in the war. Another thing which stuck me as odd was that it was an invasion battle and my transports were mostly destroyed while I was focused on the enemy's main force, leading to me not achieving the objective, but the invasion was deemed successful after the battle anyway (I wiped out the Soviet fleet after my transports were destroyed, perhaps this is why). I thought about posting this on the bug thread, but I'm not sure whether or not it's a bug.
|
|
|
Post by dizzy on Jun 25, 2019 17:36:43 GMT -6
We will be looking at adjusting the AI &/or employing some other method(s) to alleviate this issue - of course keep in mind that a small degree of 'dumping stuff on the first viable target' is not unrealistic based on history, but I do agree it should not be to such a degree as that which the game currently does. Thanks! I'll disagree with that, williammiller. Let's take the 1st, most important naval battle involving carriers, The Battle of Midway. The Japanese carrier Akagi looked like she would have made it away without getting bombed, but at the last moment, three dive bombers of the squadron pummeling Kaga broke off, commanded by **** Best, and they made a run on Akagi. An interview later revealed he realized the squadron's mistake of going after only one target, which would seem to back up what your AI does, but this Lieutenant knew better. His two wingmen missed Akagi, having one jam Akagi's rudder with a stern water hit, but ****'s bomb penetrated deep into the upper deck and exploded in the hangar where Nagumo's carrier strike were waiting to spot and set off a chain reaction of fuel, bombs and torpedoes that doomed the ship. I made THIS POST about this issue a while back and I'm glad you finally decided to take a look at this problem. I can't tell you how frustrating it is to not be able to actually have a true carrier battle in this game because bombers will not fly to their destination without unloading everything on a KE in between them and their enemy carrier target. Don't forget to look at THIS POST regarding similar ship behavior. Be nice to get them both fixed.
|
|
|
Post by dizzy on Jun 25, 2019 17:43:18 GMT -6
Shoho went down after being hit by 13 bombs and 7 torpedoes. She was a 12000 ton CVL. And the only reason she didn't get hit by more was because there simply weren't more planes around to drop stuff on her. And Shoho wasn't travelling alone either. Granted, I do think the AI is FAR overzealous in kicking the dead horse (particularily so with gunfire), and the AI engagement logic may be also flawed in that sense. But while agreeing that a look is necessary in that department, I don't necessarily thing it's as outrageous as you seem to think . But you also have to remember that the Shoho was the object of the strike, not some derelict the strike found burning along the way to the target location. Overkilling the target, especially if it's a carrier or capital ship seems acceptable to me. Some early attacks seem reasonable but it seems to me that there should be around an 80% chance of a strike flying to the target location before executing a strike even if there are available targets along the way. I had a merchant ship get hit with 20 aerial torps. So what's the cut off for the AI? I dont think it has one. What is the historical cut off for a pilot to NOT torpedo a sinking ship and move on to another one? Seriously? Was this something in a TB or DB manual? Is it the Titanic with her stern sticking straight into the air like in the James Cameron flick? Would they pass on that? Or could it be something more reasonable like ship isn't moving, is listing in the water to a degree making any turret fire impossible? I mean, I know it's not easy in realism, you can't see torpedo damage from the air, thru the flak, and dodging AA, and often times ships burn and still fire at you, but what's the cut off? Seriously, is there one, do pilots know IRL back in WW2 when to go for another target?
|
|
|
Post by ramjb on Jun 25, 2019 18:18:19 GMT -6
But you also have to remember that the Shoho was the object of the strike, not some derelict the strike found burning along the way to the target location. Overkilling the target, especially if it's a carrier or capital ship seems acceptable to me.
Some early attacks seem reasonable but it seems to me that there should be around an 80% chance of a strike flying to the target location before executing a strike even if there are available targets along the way.Oh, don't take me wrong, as I said I do think the AI has a very very nasty tendence to kick the dead horse until is not only dead, but has been erased out of any possible afterlife and then some. And aircraft attacks share this tendence. As for attacking a derelict, I'd be OK with it...to a point. The same I'd be OK to a point with my DDs launching a couple torp spreads to make sure that the ship that seems to be doomed is INDEED doomed...but when it's two full destroyer divisions dropping so many torps from so many directions that they end up hitting each other (yep, have had it happening myself), yeah, well...not so much. So I know where you come from. All I'm trying to point out is that a certain degree of overkill in the game, and a certain degree of wrong target selection and a certain degree of things that tended to happen aswell in real life is perfectly fine. My only gripe with things as they currently stand is that the AI tends to overdo it up to ridiculous levels. So, you see, I'm agreeing with you, I'm just commenting that while things should be adressed, the pendulum should not be swung in the totally direct opposite direction, which would be as unrealistic (if not more) than what the AI tends to do right now .
|
|
|
Post by ramjb on Jun 25, 2019 18:29:29 GMT -6
I had a merchant ship get hit with 20 aerial torps. So what's the cut off for the AI? I dont think it has one. What is the historical cut off for a pilot to NOT torpedo a sinking ship and move on to another one? Seriously? Was this something in a TB or DB manual? IN general pilots of the day would rather drop on a present target than holding on for one that he didn't even know he'd get to see to begin with. Naval strikes of the time weren't flown by GPS to contacts pinpointed by radar or some other means of advanced location - they usually were sent from ships in movement to intercept enemies in movement, based on information (maybe accurate, maybe not, more times not) collected well before they took off from their carriers, and usually on contact reports of wrongly IDd vessels to begin with. If during the time that took to assemble a strike force and for that strike force to fly to the target, said target changed his course, speed, or either in any drastic way, there was a huge chance those planes would end up having nothing to attack. And -THAT- if the initial position&heading&speed contact report was accurate (which again, more times than not, wasn't too much), and that the own navigation of the strike group was flawless (Which again, wasn't always the case). It wasn't unheard of for airstrikes to completely miss their mark. So, in general, it was always better to attack a clear and present target that's in front of you than wait for one that might never appear. Then comes a whole different chapter about wrong ship ID, and pilots who thought they were looking at much bigger ships than what they actually had in front of them. You also have to keep in mind, coming back to the CV and landing couldn't be safely accomplished with a bomb or a torpedo slung under your fuselage (lest you have an accident and end up with an exploding warhead on the deck of the ship you're landing at). If you had to come back to your carrier having attacked nothing, your ordinance would be dropped, and some of that ordinance (particularily so, torpedoes) was VERY expensive. So before running the risk of returning to base without having dropped ordinance, planes would drop it on anything at hand. How does that english saying go?...."A bird in the hand is worth two in the bush"?. Again I'm not stating that the AI doesn't overdo it, because it does - but the idea of an "AI cutoff" or "AI holding weapons just in case they happen to find something bigger or better" is even more unrealistic than the "look guys, a KE, let's give it some torpedo love" that happens now and then right now. IF you're unsure about your strike being able to find the correct target, just load it with bombs instead, so you don't waste torpedoes on wrong targets.
|
|
|
Post by seawolf on Jun 25, 2019 18:39:31 GMT -6
I had a merchant ship get hit with 20 aerial torps. So what's the cut off for the AI? I dont think it has one. What is the historical cut off for a pilot to NOT torpedo a sinking ship and move on to another one? Seriously? Was this something in a TB or DB manual? IN general pilots of the day would rather drop on a present target than holding on for one that he didn't even know he'd get to see to begin with. Naval strikes of the time weren't flown by GPS to contacts pinpointed by radar or some other means of advanced location - they usually were sent from ships in movement to intercept enemies in movement, based on information (maybe accurate, maybe not, more times not) collected well before they took off from their carriers, and usually on contact reports of wrongly IDd vessels to begin with. If during the time that took to assemble a strike force and for that strike force to fly to the target, said target changed his course, speed, or either in any drastic way, there was a huge chance those planes would end up having nothing to attack. And -THAT- if the initial position&heading&speed contact report was accurate (which again, more times than not, wasn't too much), and that the own navigation of the strike group was flawless (Which again, wasn't always the case). It wasn't unheard of for airstrikes to completely miss their mark. So, in general, it was always better to attack a clear and present target that's in front of you than wait for one that might never appear. Then comes a whole different chapter about wrong ship ID, and pilots who thought they were looking at much bigger ships than what they actually had in front of them. You also have to keep in mind, coming back to the CV and landing couldn't be safely accomplished with a bomb or a torpedo slung under your fuselage (lest you have an accident and end up with an exploding warhead on the deck of the ship you're landing at). If you had to come back to your carrier having attacked nothing, your ordinance would be dropped, and some of that ordinance (particularily so, torpedoes) was VERY expensive. So before running the risk of returning to base without having dropped ordinance, planes would drop it on anything at hand. How does that english saying go?...."A bird in the hand is worth two in the bush"?. Again I'm not stating that the AI doesn't overdo it, because it does - but the idea of an "AI cutoff" or "AI holding weapons just in case they happen to find something bigger or better" is even more unrealistic than the "look guys, a KE, let's give it some torpedo love" that happens now and then right now. The issue is more of them hitting ships that are halfway between the target and the carrier. They hadn't even reached their destination, had plenty of fuel, and still waste all of their ammunition on tiny little sinking ships. In situations like the sinking of Shoho or Musashi or Yamato they were at least the intended target when they got obliterated. And in the sinking of Neosho and Sims they were the intended targets they were just misidentified. I don't know any historical situations of aircraft ignoring a target to sink a ship they met along the way, especially one surrounded by their own ships
|
|
|
Post by jwsmith26 on Jun 25, 2019 19:00:08 GMT -6
I think you have to put the expected target into the equation. If a strike is going after a reported carrier force I find it highly unlikely that the squadron leader would settle for attacking a corvette halfway to the target location. Even more valuable targets might be rejected. At Santa Cruz one American strike group spotted the Tone and rejected it as a target because they were searching for the enemy carriers. When they reached the end of assigned range without finding the carriers they then turned back and attacked the Tone. Carriers were special targets and in most cases pilots did not give up the chance to attack one until they had no other choice.
|
|
|
Post by axe99 on Jun 25, 2019 19:07:02 GMT -6
Just to make things clear. The limit is 2 torpedoes per torpedo capable plane (TB and in some cases MB). This is pretty close to historical limits, in fact it might be generous for land bases. I agree it would be desirable to have a limit on heavy AP bombs as well, but torpedoes were a much easier case to handle. We will see about bombs later. One cannot simulate everything in a game. A limit on heavy AP bombs might be a good idea if feasible, but one idea (although I suspect not feasible) could be 'max munitions on a carrier' - so instead of a 'torp limit' or a 'heavy AP bomb' limit, there's a tab relating to preferred carrier loadout (as there is for AP/HE/SAP shells), with options to preference torpedoes, heavy bombs, lighter bombs, etc., and then in the design stage for a carrier, it makes it clear how much space there is for munitions for aircraft (and perhaps even fuel). This might be getting a bit over the top in terms of detail, though, and would probably need some new in-battle UI elements so players knew how close to exhaustion they were pushing their carriers as offensive weapons (but it would bring carriers to the same level as gun-based ships, which do have ammo and fuel limits on their effectiveness as offensive units). None of the above suggestion should be seen in any way, shape or form as a criticism - RtW2 is sensational, thank you very much for all off your most excellent work .
|
|
|
Post by dizzy on Jun 25, 2019 19:22:13 GMT -6
Just to make things clear. The limit is 2 torpedoes per torpedo capable plane (TB and in some cases MB). This is pretty close to historical limits, in fact it might be generous for land bases. I agree it would be desirable to have a limit on heavy AP bombs as well, but torpedoes were a much easier case to handle. We will see about bombs later. One cannot simulate everything in a game. A limit on heavy AP bombs might be a good idea if feasible, but one idea (although I suspect not feasible) could be 'max munitions on a carrier' - so instead of a 'torp limit' or a 'heavy AP bomb' limit, there's a tab relating to preferred carrier loadout (as there is for AP/HE/SAP shells), with options to preference torpedoes, heavy bombs, lighter bombs, etc., and then in the design stage for a carrier, it makes it clear how much space there is for munitions for aircraft (and perhaps even fuel). This might be getting a bit over the top in terms of detail, though, and would probably need some new in-battle UI elements so players knew how close to exhaustion they were pushing their carriers as offensive weapons (but it would bring carriers to the same level as gun-based ships, which do have ammo and fuel limits on their effectiveness as offensive units). None of the above suggestion should be seen in any way, shape or form as a criticism - RtW2 is sensational, thank you very much for all off your most excellent work . My apologies if someone asked already, but the 2x Torps per plane... what if your plane can carry two torps? Is it out of ammo after one sortie?
|
|
|
Post by jwsmith26 on Jun 25, 2019 19:43:31 GMT -6
This issue of torpedo limitations seems intimately tied to the issue of target selection. With this limitation it become even more important to get that part of the equation right. I have a few comments.
I think that in most cases naval pilots greatly overestimated the amount of damage they had inflicted. They often claimed hits that did not occur, they thought bombs exploding in water were torpedo hits, they assumed heavy smoke meant the ship was doomed, etc. It is much more likely that pilots would assume a damaged ship was not a worthy target because - surely it must be sinking if it looks like that. Pilots got one shot and they were risking their lives to get that shot, so they wanted to make sure they were hitting something important. A sinking ship was no longer important. In that respect I think that RTW2 is currently pretty far off the mark. Historically, sinking ships dead in the water were not primary targets for carrier pilots.
However, there were exceptions.
Carriers were one. If there was just one carrier visible it would be pummeled relentlessly while still afloat. If a carrier was obviously foundering the squadron leader would sometimes assign planes to attack other valuable targets and some section leaders would make their own choice as to which ship to attack. From the outside it could look like a fair amount of apparent randomness in the choice of target, but the great majority of a squadron would almost always strike a carrier if one was visible, regardless of its apparent state.
If there was more than one carrier at the target location, priorities changed. The primary job was first to mission kill every visible carrier, preferably by sinking them, but every carrier needed to be at least damaged to prevent it from operating its air group. That was standard policy in every navy from the 1920s on. Even a battleship should not distract a strike targeted on a carrier sighting report. Look at The battle of Santa Cruz. The South Dakota was right there waiting to be hit. She was hit just once. Unfortunately, I have seen so few actual aerial attacks on carriers in RTW2 that I can't say for sure how the game handles this type of situation.
Battleships were another exception. Battleships were always tough targets and usually required the kind of swarming attacks we regularly see being rendered against any damaged or sinking ship in RTW2. Historically these types of wave after wave attacks were required to sink battleships. Unlike carriers, which generally succumbed in a very visible fiery death, it was extremely difficult for a pilot to tell the state of damage of a battleship. To attacking pilots, a battleship could look completely demolished topside and still be perfectly capable of steaming home. So swarming attacks against battleships are reasonable in RTW2. However in RTW2, even for battleships I think there should be a limit for gameplay purposes if only to reduce the extreme frustration these attacks cause to the player. I think a reasonable cut off point would be when the ship goes dead in the water. At that point the number of attacks should drop off substantially. Again, I suggest this for gameplay purposes. Historically, attacks would certainly continue until the ship plunged.
|
|
|
Post by dizzy on Jun 25, 2019 19:56:32 GMT -6
Just to make things clear. The limit is 2 torpedoes per torpedo capable plane (TB and in some cases MB). This is pretty close to historical limits, in fact it might be generous for land bases. I agree it would be desirable to have a limit on heavy AP bombs as well, but torpedoes were a much easier case to handle. We will see about bombs later. One cannot simulate everything in a game. Just to be clear, the 2x torps per capable plane is generated at the start of the mission, does it include ALL planes or exclude those damaged, destroyed, what happens when a plane can carry two torps, is it good for just one sortie, are the torps tracked in case we launch with a medium load or one torp in the case of planes capable of carrying two torps and will the plane be able to launch four sorties, and finally, do partial load-outs work when having fewer than the number of torpedoes necessary for a squadron of fewer than full strength planes? Edit: If torpedo stores are going to be a thing, I'd like an up to date ammo counter that works unlike THIS POST. Knowing I have only so many, I'd probably opt to launch some strikes with AP bombs instead, given the circumstances. Edit2: Dive bombers seem to be king in late game, they get 2-3x more hits than torp planes, so knowing torps are limited, I might swap a squadron or two over to DB's.
|
|
|
Post by oldpop2000 on Jun 25, 2019 20:16:21 GMT -6
Historically, there were 156 MK 13 torpedoes built for the four pre-war carriers of the Yorktown class which means each carrier carried 39 torpedoes, two for each torpedo bomber plus 12 spares. That's 12 spares total, so each carrier got 3 spare torpedoes. The Essex were upgraded Yorktown class carrier so they probably carrier the same amount.
Note: A total of 17,000 MK13's were built through the entire war, but they were used by other aircraft.
|
|
|
Post by zabieru on Jun 25, 2019 23:33:52 GMT -6
To be clear, what I'm suggesting is not "TB spots KE, looks around for bigger target, doesn't see it, casts Lvl 3 Remote Rearm and switches to bombs" but simply "TBs carrying torpedoes will not attack targets they identify as smaller than XX."
And yes, the AI does seem to try and target bigger ships (it's definitely not perfect, but this is fine: if I send a sortie against a group of 3 DD, 2 CA, and 1 CV, I do expect some planes to make runs at each of those ships.) The game already appears to handle misidentifications, so if I set the torp limit to BC/BB/CV and my squadron spots a CA they think looks awfully battle-ish, they should try and sink her.
The scenario I have an issue with is when your sortie runs into something tiny on its way to the target and decides to blow it out of the water. THAT seems like one where your pilots ought to know better. (And, for that matter, if they get to the target and there's nothing there, or only a destroyer, same deal.)
Really it's less "I don't want to see anything smaller than XX with an aircraft torpedo hole in it" so much as "I don't want to see torpedo bombers making an attack run unless there's something bigger than XX in the area." As you say, fog of war and all. I have no issue with any of that. But very few pilots came back to the ol' Kido Butai bragging about how they were on their way to the navy base but they saw this police boat so they blew it up instead, and I can't imagine they'd have had long careers if they had. There's a difference between hitting the wrong battleship-shaped object in the middle of a major naval installation, and turning back early because there was this coast guard cutter and it was just SO JUICY, sir, SO JUICY AND IT NEEDED MY TORP LOVE.
Woul you like to send your torpedo bombers with torpedoes back to land on carrier? :-))
You really like your torpedoes more than your carrier. :-) Landing is always dangerous and anything can happen. It is enough that plane itself has flammables.
Until the landing-accident model incorporates the risk of aircraft munitions going off in the crash... Yes I would. Even then, I can certainly see situations where I'd gamble on the fairly small possibility of one torpedo exploding on the flight deck of my carrier (landing accidents aren't THAT common and only a minority of those would detonate the ordnance) against the damage ten TBs could do if they did manage to find the enemy battle/carrier division on the next sortie.
If nothing else, dumping ten torps against a KE/TR before you even reach the target location is an issue. You're right: there's an argument to be had that dumping torps on whatever you find at the target even if it's small potatoes or dumping them on any target at all on the way back is correct behavior just so you don't have to risk landing with them... but none of that applies to launching early against something worthless.
|
|
|
Post by dorn on Jun 26, 2019 0:25:31 GMT -6
We will be looking at adjusting the AI &/or employing some other method(s) to alleviate this issue - of course keep in mind that a small degree of 'dumping stuff on the first viable target' is not unrealistic based on history, but I do agree it should not be to such a degree as that which the game currently does. Thanks! I'll disagree with that, williammiller. Let's take the 1st, most important naval battle involving carriers, The Battle of Midway. The Japanese carrier Akagi looked like she would have made it away without getting bombed, but at the last moment, three dive bombers of the squadron pummeling Kaga broke off, commanded by **** Best, and they made a run on Akagi. An interview later revealed he realized the squadron's mistake of going after only one target, which would seem to back up what your AI does, but this Lieutenant knew better. His two wingmen missed Akagi, having one jam Akagi's rudder with a stern water hit, but ****'s bomb penetrated deep into the upper deck and exploded in the hangar where Nagumo's carrier strike were waiting to spot and set off a chain reaction of fuel, bombs and torpedoes that doomed the ship. I made THIS POST about this issue a while back and I'm glad you finally decided to take a look at this problem. I can't tell you how frustrating it is to not be able to actually have a true carrier battle in this game because bombers will not fly to their destination without unloading everything on a KE in between them and their enemy carrier target. Don't forget to look at THIS POST regarding similar ship behavior. Be nice to get them both fixed. On the other hand some planes just attacked escorts and did not even tried to attack carriers.
|
|
|
Post by dorn on Jun 26, 2019 0:27:32 GMT -6
I think you have to put the expected target into the equation. If a strike is going after a reported carrier force I find it highly unlikely that the squadron leader would settle for attacking a corvette halfway to the target location. Even more valuable targets might be rejected. At Santa Cruz one American strike group spotted the Tone and rejected it as a target because they were searching for the enemy carriers. When they reached the end of assigned range without finding the carriers they then turned back and attacked the Tone. Carriers were special targets and in most cases pilots did not give up the chance to attack one until they had no other choice. Completely agree, you can even look at the Mediterranean and Axis attacks to British forces. If there were carrier, carrier was target objective number 1.
|
|