Large scale IJN airgroup resizing question

Large scale IJN airgroup resizing question (Full Version)

All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition >> The War Room

Message

Psaeko -> Large scale IJN airgroup resizing question (12/6/2020 3:54:29 PM)
I searched the forum for a question I had on the subject of resizing IJN airgroups, but while I did find a discussion on whether it is "gamey" or not, I still had some questions left. The backstory: I'm currently playing DBB-C scenario 30 against an opponent from the forum here. I have only played against some friends in the past so if not that familiar with several of the more arcane strategies people use. After some disastrous results sweeping with Oscars I talked with my PBEM opponent about how I was more conservative with the IJN pilots than IJA ones since the IJA has a large amount of restricted Nates available for combat training, while the IJN has a much more limited training program. He then suggested taking several floatplane groups off the IJN ships that have 2 of them, and resizing them using the CS to increase training capacity which is a good idea. However, he also suggested resizing some of the fighter units to 81 planes for large training/sweeping units. Regardless of whether that is "gamey" or not, we didn't set up any HR about it and he's obviously fine with me doing it, so that left me with the following question. For those of you who do resize to those very large size units, how many of those do you resize and what impact does that have on the economy? While I am playing scenario 30 which is ironman Japan in terms of economy, going to 81 planes is nearly doubling unit size and would require a large increase in navy plane production (and pilot training). The effects on AF support and supply usage when the planes fly are much more direct, but I'm not very sure what effects the large amounts of extra production will have down the line. I will probably convert two, maybe three groups for that purpose, which already requires over a hundred planes to fill out. Any more seems dangerous to me, but maybe some of the people more experience with that can weigh in on this.

Nomad -> RE: Large scale IJN airgroup resizing question (12/6/2020 4:31:15 PM)
One thing to keep in mind is that he can resize his USMC air groups to size 90.

Evoken -> RE: Large scale IJN airgroup resizing question (12/6/2020 4:39:24 PM)
I mean setting up production to be higher shouldnt be too hard in scenario 2 , you got more of everything. I got 3 of the supersized squadrons in my game and didnt notice any negatives about it

mind_messing -> RE: Large scale IJN airgroup resizing question (12/6/2020 5:39:25 PM)
Not sure if I can claim to be the creator of this approach, but fairly sure I was the first to AAR it used on a large scale. As you point out, it's a double edged sword, but one that I fairly consistently use as I think it adds a lot of flexibility for Japan (when used right). Some reflections from me on this would be: - The combat impact of having 81 sized squadrons is not to be underestimated. I've had great success with size 81 Jill/Judy squadrons bashing through Allied CAP. - On balance, I think the optimal way to leverage resizing is to turbo-charge the IJN training programme in preparation for the late war. With careful manipulation of frontline IJN strength and resizing of suitable squadrons, you can really focus on quality pilots for the IJN in the same way as you can for the IJA. - Of notable use is the resizing of torpedo bomber squadrons to provide pilots for the IJN 2E squadrons - the core skills can be trained in resized Kate/Jill groups. - Production strategy needs to mesh with this strategy of resizing for full effect in combat, but you should be doing this anyway in anticipation for the late-war Gotterdammerung of dive and torpedo bombers. In practice, that means lots of Jill/Judy/Grace for the bombers and Zero/IJA land-based fighter of choice. - Size 8/9 airbases become much more important for enabling massed attacks, both for co-ordination reasons and for the double aviation support bonus. Plan accordingly. - The collective supply costs of all these actions are not trivial, so be sure to factor that into consideration. In my view, they're worthwhile - the impact of having a large strategic reserve of well-trained IJN pilots is worth a great deal of supply. In short, going all in with resizing is an effective way to boost your IJN training regime early, crank out decent pilots en-masse, and then transition into massed attacks on Allied shipping for the late game. It works, but it has costs, so think it out fully.

GetAssista -> RE: Large scale IJN airgroup resizing question (12/6/2020 6:33:53 PM)
The only downsize of resizing is that you increase your consumption of supplies noticeably. Increased production is negligible cause you need to fill the air groups once to make them useful for eternity as e.g. training units. Benefits from resizing to training are enormous in case of Japan, especially for the IJNAF which has a consistent lack of training space for pilots especially fighter ones. And float planes air groups' resizing is a goldmine of training space, covering all your needs in bombers, ASW and xp training with no actual strike groups diverted from the front.

rustysi -> RE: Large scale IJN airgroup resizing question (12/7/2020 2:20:39 AM)
quote: After some disastrous results sweeping with Oscars Not really a great sweeper plane. Firepower is too low and air-frame durability is crap. quote: going to 81 planes is nearly doubling unit size and would require a large increase in navy plane production You could use your excess Claude's if the unit is only for training and in a rear area.

RangerJoe -> RE: Large scale IJN airgroup resizing question (12/7/2020 11:57:16 PM)
If you also "save" you 81+ experienced pilots in TRACOM and supersize your air units, then you can also save on HI from the sped up pilot training along with pulling pilots out of the training schools early. Those experienced pilots are needed later on in night fighters and when the enemy has better aircraft.

Psaeko -> RE: Large scale IJN airgroup resizing question (12/11/2020 10:19:49 AM)
Alright, thank you all for the replies, I've decided to resize a lot of the shipborne floatplanes, and increase the size of a few rear area naval airgroups and set them up for training. I'll also increase once dedicated sweeper unit to see how they fare in air-to-air compared to the 45 units. quote: One thing to keep in mind is that he can resize his USMC air groups to size 90. I didn't even think about that, that could be rough indeed, but at least in the early war airframes shouldn't be plentiful enough to use those in combat duties, for training purposes might be another thing. quote: The only downsize of resizing is that you increase your consumption of supplies noticeably. Increased production is negligible cause you need to fill the air groups once to make them useful for eternity as e.g. training units. quote: You could use your excess Claude's if the unit is only for training and in a rear area. Good point, if used for training the squadrons only have to be filled out once and can be done with outdated frames, and I can always assess the economic situation later when deciding to use them as combat units or not. quote: - The combat impact of having 81 sized squadrons is not to be underestimated. I've had great success with size 81 Jill/Judy squadrons bashing through Allied CAP. - On balance, I think the optimal way to leverage resizing is to turbo-charge the IJN training programme in preparation for the late war. With careful manipulation of frontline IJN strength and resizing of suitable squadrons, you can really focus on quality pilots for the IJN in the same way as you can for the IJA. - Of notable use is the resizing of torpedo bomber squadrons to provide pilots for the IJN 2E squadrons - the core skills can be trained in resized Kate/Jill groups. - Production strategy needs to mesh with this strategy of resizing for full effect in combat, but you should be doing this anyway in anticipation for the late-war Gotterdammerung of dive and torpedo bombers. In practice, that means lots of Jill/Judy/Grace for the bombers and Zero/IJA land-based fighter of choice. - Size 8/9 airbases become much more important for enabling massed attacks, both for co-ordination reasons and for the double aviation support bonus. Plan accordingly. - The collective supply costs of all these actions are not trivial, so be sure to factor that into consideration. In my view, they're worthwhile - the impact of having a large strategic reserve of well-trained IJN pilots is worth a great deal of supply. In short, going all in with resizing is an effective way to boost your IJN training regime early, crank out decent pilots en-masse, and then transition into massed attacks on Allied shipping for the late game. Yeah, I'll go with this approach, increase training by a lot and don't really use them in combat until lategame, ensuring a large supply of decent skill pilots.

inqistor -> RE: Large scale IJN airgroup resizing question (1/8/2021 9:01:10 PM)
You don't need planes to train. Full complement increases speed of training by IIRC 1/3rd, but your pilots will train even in units with no planes at all. Any skill.

GetAssista -> RE: Large scale IJN airgroup resizing question (1/8/2021 10:26:23 PM)
quote: ORIGINAL: inqistor Full complement increases speed of training by IIRC 1/3rd No, planes are much more important. This thread has some statistics https://www.matrixgames.com/forums/tm.asp?m=4215656

mind_messing -> RE: Large scale IJN airgroup resizing question (1/9/2021 1:08:54 AM)
quote: ORIGINAL: GetAssista quote: ORIGINAL: inqistor Full complement increases speed of training by IIRC 1/3rd No, planes are much more important. This thread has some statistics https://www.matrixgames.com/forums/tm.asp?m=4215656 There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject.

GetAssista -> RE: Large scale IJN airgroup resizing question (1/9/2021 2:57:58 AM)
quote: ORIGINAL: mind_messing There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject. Sure there is statistics. post #20 is exactly that, there are also others, in more aggregated form. Alfred was wrong on several occasions over the years wrt how things end up in game. In that particular thread post #25 shows that he does not understand the particular experiment design and how to refute it. Namely he presents "training happens with no planes" to refute "training is slower with no planes" statement - which is obviously logically wrong. My own experiments also coincide with InfiniteMonkey results - approximately linear relation between the speed of training and the complement of airframes. Speed also depends on the level of skill, rookies will train faster. Sure you can pick whatever belief you like in the end, it's not my intention to lecture anyone on scientific method and data analysis

RangerJoe -> RE: Large scale IJN airgroup resizing question (1/9/2021 3:04:16 AM)
quote: ORIGINAL: GetAssista quote: ORIGINAL: mind_messing There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject. . . . Namely he presents "training happens with no planes" to refute "training is slower with no planes" statement - which is obviously logically wrong. . . . Sorry, but you are incorrect. The "training happens with no planes" and the "training is slower with no planes" statements are not mutually exclusive.

GetAssista -> RE: Large scale IJN airgroup resizing question (1/9/2021 3:09:56 AM)
quote: ORIGINAL: RangerJoe quote: ORIGINAL: GetAssista quote: ORIGINAL: mind_messing There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject. . . . Namely he presents "training happens with no planes" to refute "training is slower with no planes" statement - which is obviously logically wrong. . . . Sorry, but you are incorrect. The "training happens with no planes" and the "training is slower with no planes" statements are not mutually exclusive. Did I say they were? I spoke about appropriate refutation of the statement. Formal logic opposite for "training is slower with no planes" is "training is same or faster with no planes as compared with some/full planes". Alfred did not include the "as compared with.." part

mind_messing -> RE: Large scale IJN airgroup resizing question (1/9/2021 1:10:14 PM)
quote: ORIGINAL: GetAssista quote: ORIGINAL: mind_messing There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject. Sure there is statistics. post #20 is exactly that, there are also others, in more aggregated form. Post 20# is not a statistical test as not all variables are controlled for enable the relationship between the predictor variable to be changed to explore its relation with the outcome variable. quote: Alfred was wrong on several occasions over the years wrt how things end up in game. Care to list those occasions? quote: In that particular thread post #25 shows that he does not understand the particular experiment design and how to refute it. Namely he presents "training happens with no planes" to refute "training is slower with no planes" statement - which is obviously logically wrong. That's completely false. Post #30 (particularly the first point Alfred makes) highlights that he understands just fine. quote: My own experiments also coincide with InfiniteMonkey results - approximately linear relation between the speed of training and the complement of airframes. Speed also depends on the level of skill, rookies will train faster. Sure you can pick whatever belief you like in the end, it's not my intention to lecture anyone on scientific method and data analysis No, I would like to hear more about your use of "scientific method and data analysis" so I can have a chuckle. Questions to assess your methodology: 1. Please list all variables that the experiment controlled for. 2. What assurances do you have that the list of variables that were controlled for the experiment was complete? 3. Please provide a full methodology for your "tests"? All these so called "tests" fall foul of the first two points outlined in Alfred's post #30. Does anyone actually know how training is coded? Posts elsewhere indicate it is not as simple as some would like to think - https://www.matrixgames.com/forums/tm.asp?m=4539162&mpage=1&key=pilot&#4539968

GetAssista -> RE: Large scale IJN airgroup resizing question (1/9/2021 8:37:40 PM)
quote: ORIGINAL: mind_messing quote: ORIGINAL: GetAssista quote: ORIGINAL: mind_messing There are many things in that thread, "statistics" is not one of them. Post #30 from Alfred should be seen as the definitive word on this subject. Sure there is statistics. post #20 is exactly that, there are also others, in more aggregated form. Post 20# is not a statistical test as not all variables are controlled for enable the relationship between the predictor variable to be changed to explore its relation with the outcome variable. What particular variables that can affect the outcome are you referring to? List them please and explain why they should behave this way. Also please include any data supporting those claims ) See, two can play this game. quote: Alfred was wrong on several occasions over the years wrt how things end up in game. Care to list those occasions? kamikaze bomb load effect on damage, bonus from HQc/y on adjusted AV, movement of AI TFs with zero fuel. Search those yourself please. quote: In that particular thread post #25 shows that he does not understand the particular experiment design and how to refute it. Namely he presents "training happens with no planes" to refute "training is slower with no planes" statement - which is obviously logically wrong. That's completely false. Post #30 (particularly the first point Alfred makes) highlights that he understands just fine. post #30 does not include anything of substance except "you did not control for everything" and usual Alfred's rants towards everybody else not understanding anything. Science does not work like that, general appeal to unspecified controls is not sufficient to refute the hypothesis with this much visible data effect. Present your refuting data (which Alfred failed to do there) quote: My own experiments also coincide with InfiniteMonkey results - approximately linear relation between the speed of training and the complement of airframes. Speed also depends on the level of skill, rookies will train faster. Sure you can pick whatever belief you like in the end, it's not my intention to lecture anyone on scientific method and data analysis No, I would like to hear more about your use of "scientific method and data analysis" so I can have a chuckle. Questions to assess your methodology: 1. Please list all variables that the experiment controlled for. 2. What assurances do you have that the list of variables that were controlled for the experiment was complete? 3. Please provide a full methodology for your "tests"? Now, sometimes I do enjoy proving people wrong on the Internet, but that depends on my mood and opportunity costs, right. Do you really think I'll rush and do all that additional work for that precious chuckle of yours? Like, seriously? I don't see the desire to learn here for my work to be justifyied. [8D] I have my results, I'm fine with them, they coincide with what other people obtained from actual tests instead of "you don't understand anything" intellectual laziness. You can run your own tests, they are pretty obvious, just labor intensive. If you don't understand how the tests should be done (did I mention they are strikingly simple in setup?), then probably you should not pick off the statistics other people are doing. This all does not mean that I do not care about what Alfred says about the game. He's a walking encyclopedia all right and very rarely wrong. His rants are endearing too. Edit: ah, an important addendum. JFBs that are using resizing of float plane squadrons on a large scale and are attentive enough, can see the effect in the regular games. Many of those FP squadrons are doing the same training on fresh rookies for several months, like ASW or NavB, and some of those squadrons remain unfilled for quite some time. There is a noticeable difference in the accumulated training in the end between full and unfilled squadrons, especially on the higher level of skills. Fresh low-skill rookies are popping rather eagerly in both types.

mind_messing -> RE: Large scale IJN airgroup resizing question (1/10/2021 12:01:05 AM)
quote: kamikaze bomb load effect on damage, bonus from HQc/y on adjusted AV, movement of AI TFs with zero fuel. Search those yourself please. You're off to a great start considering that Alfred was quite right regarding kamikaze bomb load not being a factor worth considering. http://www.matrixgames.com/forums/fb.asp?m=4591155 You might want to have a deeper think of the implications of #5 before spouting "aLfReD is wRoNg!" quote: post #30 does not include anything of substance except "you did not control for everything" and usual Alfred's rants towards everybody else not understanding anything. Science does not work like that, general appeal to unspecified controls is not sufficient to refute the hypothesis with this much visible data effect. Present your refuting data (which Alfred failed to do there) Almost as if not controlling for something might...just might...be important. But nah, this is science! Who needs proper controls! [8\|] quote: Now, sometimes I do enjoy proving people wrong on the Internet, but that depends on my mood and opportunity costs, right. Do you really think I'll rush and do all that additional work for that precious chuckle of yours? Like, seriously? I don't see the desire to learn here for my work to be justifyied. I have my results, I'm fine with them, they coincide with what other people obtained from actual tests instead of "you don't understand anything" intellectual laziness. You can run your own tests, they are pretty obvious, just labor intensive. If you don't understand how the tests should be done (did I mention they are strikingly simple in setup?), then probably you should not pick off the statistics other people are doing. Why am I not surprised that you'd be reluctant to provide the requested information? Is that the norm in science; to publish the results but no methodology and add a note saying "I am quite satisfied with the results, so no!"? I, for one, have no doubt that you've managed to effectively de-code how the entire pilot training codebase works thanks to your "results". You probably know how it works far better than the developers did! How amazing! quote: This all does not mean that I do not care about what Alfred says about the game. He's a walking encyclopedia all right and very rarely wrong. His rants are endearing too. It's almost....almost...as if he might have a deeper understand of things than you? No, that can't be possible. Your "results" said he's wrong. The results can't be wrong, so Alfred must be right about everything else and wrong about this... [8\|]

Page: [1]

Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI

0.859375