I noticed you went by 2 for pure infantry. Could you possibly add 1 battalion of line artillery and run it again for the odd number widths? I know personally you said you'd done pretty well with 27 widths and with the combat width statistics chart it's been shown it's a peak in being more effective than basically anything else until roughly 41
Im at the point with 2000 hours in the game I just put field hospital and maintenance and logistics and just leave the computer I just don't have the patience to control the supply while the game runs at a slog
@@baileymash doing the same thing over and over expecting different results. (though in the case of this game its completely possible for different results to happen)
Makes the game more friendly for new players which is a good thing. For the longest time I was avoiding HoI4 because of the arcane knowledge it demands.
@@FairyRat Don't worry they added more shit to learn with supply lines. I like HOI but really to learn it I say play Germany or the US. One has all the advantages through the tree and the other has all the time in the world.
Makes you think maybe they wanted historical division to work better? I did a US playthrough recently, had no idea on the meta since the US fights world over and just went with the historical 1943 lay out. It did great against Japan and OK against Germany.
@@Xinamon98 the buff wasn't a direct buff to it specifically, but more that it can work well in most terrain types now. Before NSB 20 width was the obvious king
Also, Andrey Vlasov was leading kind of Russian anti-Stalin army in German Service. He was recently discussed in Czechia for his role in Prague uprising, where his army joined up resistance against Germans to be allowed to get away from Red army on time. Kinda controversial topic over here.
Many don't like him because he was a collaborator. but considering his other options namely life in a prison camp or returning to Stalin a failure it was really his only option. Plus, there is really no evidence that he was a national socialist himself he just worked with them.
@@nikoclesceri2267 I can't judge him, because I never was in the same situation, but many Russian/Soviet generals choosed death/prison over collaboration.
The amount of work put into this is commendable. I greatly enjoyed this and am always looking forward to the next video. Commenting for the algorithm because you deserve all the views and all the subs. Cheers friends.
Hi, Taureor. I’ve watched some Chinese Hoi4 content creators and their infantry templates for the new update. I feel like this is more or less related to this topic so I would bring it up here. As of 1.11.5 and for PVE (can work for PVP but need slight modification), the main strategy is as follow: The frontline consists of two types of divisions, purely defensive divisions called the ‘filler’ used to halt enemy advances and main offensive divisions called the ‘gundam’ which are literally Swiss Army knives which cut through the enemy lines like butter. The filler divisions can be any widths of pure infantry. No support companies are needed for them since their only job is to hold the enemy divisions and you don’t want your auxiliary divisions to use up your important equipments. If you run low on guns, you can even stop supplying guns to these filler divisions and let them hold the enemies with their flesh (more like a China strategy huh). An important note is that you would like to have some trained filler divisions ready to deploy. In case of enemy breaking through your frontline, you can immediately deploy these units to plug up the hole. Finally, the ideal filler width is 10/5 infs since low width fillers can not take much beating, and large filler units have less manuverability and tend to get focus fired. However, this is all up to the player and can vary based on the manpower situation. As for the more exciting gundam units, they are 42 width divisions comprised of: 4/6 artilleries, 1 anti-air medium tank and the rest are infantries. The support companies include engineering, supp art, supp flamethrower light tank and light tank recon. The final slot is free and can be anything of your choice. If you neglect building aircrafts in a game, you can switch out a couple of infantries for field anti-air guns for higher AA values. The differences between 4/6 arts is mainly soft attack value and organisation + hp. Having more arts will give higher soft attack but lowers org and hp, vice versa. The anti-air medium tank is nicknamed the ‘golden shield’ of the unit. You want to mount AA lvl 1 (1939) or AA lvl 2 (1940+) and stick on as much armour as you can into this thicc boi. Just make sure the speed is not lower than the infantries and reliability is not too shabby. On average for 1939 battles, this AA medium tank has ~100 armour value and can give your unit ~20 armour value, which is massive for infantry units. The sole purpose of flamethrower light tank is to give terrain modifiers, so just keep them as cheap as possible, and not slower than reg inf. Finally the light tank recons, these units will provide lots of breakthrough values, so design this tank with all the breakthrough modules. This is not necessary in most circumstances vs the AI, but is a must for PVP to penetrate enemy armour. Finally if you are playing on a poorer country, you can go for 21/22 width with half of the units listed above. They still have insane stats and can easily crush the AI. If you want to have armour divisions, still go for 42 width infantry division, train them to at least lvl 3 before switching to the armour template. As of now, you only lose division level if you need more manpower after switching the template. Since going from infantry to armour units you always use less manpower, there is no loss in division exp which save you the tanks to train them up. This exploit is especially broken if you have lvl 5 42 width infantry units... Finally, as for the strategy, you want to send volunteers to get as much army exp as possible in the early game. For example, as Germany, send volunteers to Ethiopia at the start of the game and slowly build them up to 42 widths. Since higher the widths the more exp they get from combat, it is absolutely worth it to upgrade these volunteer divisions to a higher widths. Then you should get tonnes of exp from Spain and Japan. When you go to war with other countries, put fillers on the frontline and gundams should focus enemy supply line. The gundams are so strong that one unit can easily punch through multiple enemy divisions. Well, I hope you enjoy this information and maybe this can help you with your future playthroughs.
The aggressiveness really just says how much Organization they will hold back for defense, it does not alter the AI's decision to take empty provinces at all. If you want to have the AI prefer land-grab, then the "Spearhead" battle plan is better then the Offensive line battle plan. (At least I think so, can't be 100% sure...)
This was an amazing video Taureror. I've been trying to play as China in NSB and just couldn't keep the Japanese out: they would roll through the North within a few months and I just could never push them out once they landed. After watching this I switched to ten widths with artillery support and everything just got easier, they reinforced faster, attacked faster, suffered less from supply issues and I finally defeated and conquered Japan! Can't wait to play as the Japanese now.
It would be interesting but even more timeconsumming to see it every time with same manpower in field (aka 10x more 2width divisions then 20 width to fulfill that)
Dear taureror! You mentioned wanting a "hyper agressive option, troops just pushing onwards without concerns for supply, taking every open province." That's what the spearhead order does. Just use that one for your troops. Works wonders in most games (especially with tanks). But in normal games make sure there is an army with normal orders nearby to reinforce the gaps the spearhead tends to leave behind so your troops don't get encircled.
This is not what spearhead does directly, spearhead just tells the divisions to find the closest tiles to the offensive line, and only go to them, not where the frontline is. It kinda has that effect, but it's very dangerous to use for large orders because they will just leave the sides wide open
I don't mind watching the testing content, it's pretty interesting. That said, I do also still really want to see a standard Stalin run and maybe even a modern version of the "Stalin: no army" run you did several patches ago, before espionage and the modern Soviet focus tree.
i enjoy this kind of content as much as i do the entertaining ones like savegame fixing or alt history mods. Its like the well made BBC Dokumentary about a topic you enjoy, keep up you kind of work mate.
So I've watched a few videos and done my own testing while playing around in single player. According to a very lengthy document from the community, which I saw in a Feedback video but read myself: 10, 15, 25 (Mountaineers) 27, 41, 42, and 44 are all the mathematically best widths Over all 15 and 30 along with 16/18 and 27, seem to be the best multiplayer and casual widths in different videos. And in my experience 15, 25 mountaineers, 26-8 and 30 are quite good. Then there's also the min-max'd ideal divisions based off of terrain widths.
Fbg released a new video a few days ago where he said after more testing 10-w and to an extent 15-w are not optimal because they take too many losses and shred ur equipment and manpower.
27w is bad in actual side by side testing with 7/2s, which have the best performance again ironically enough. I might make a video on it and why, but the bottom line is as soon as the line fills up they have *crippling* overwidth problems in plains and desert that were not in any way accounted for in the post that claimed they are "mathematically the best." They end up performing slightly worse than a 7/2 division that's ~25% more expensive. The rest of those widths aside from 10w have similar issues, but 27w is probably the worst offender.
I have been constantly trying different army loadouts similar to this but you went farther than I ever could, so this is extremely helpful. Especially the conclusion/thoughts at the end. Thank you.
i watched the whole video, it was really informative and interesting! I'd love to see the 10 width strategy in an actual full game maybe with a smaller country and no allies because, yea, being in a war with many allies and enemies definitely distorts the results a lot
i always use 10 width as minor however even though it did not show here you tend to loose units in 10 width which as minor is really bad cause it makes a hole in the small frontline you use
Small units with support companies are proportionately worse. Support companies have very low HP and Org, so smaller units with support companies can get weighed down by these penalties. Some support companies, like support artillery, add to pure combat stats so having more units with them would be better, but others provide different bonuses. Like signal companies; the bonuses to initiative and coordination aren't going to be additive if you have more units with the, you'll just require more support equipment
I think the reason why some smaller divisions would be better than their larger counterpart (in some cases) is because of the support company. 2 artillery support company for 2 10-width division would have more impact than 1 artillery support company from 1 20-width division. Although that would cost more total manpower and total production cost so you might want to consider that. With that in mind, choose an arbitrary total production cost or total manpower, maybe 120 Division for 45-width with full support companies (as it is really doable for major countries to get that number and to remove the unfairness of not having enough divisions to cover the frontline as in most cases, you should be able to). Calculate their total production cost or manpower (depending on which you choose), and put it all in an excel to easily calculate the number of divisions per combat width. Also, you might want to check the paper some dude released on reddit who did the math and take maybe the top 5 most effective combat width and top 5 least effective combat width to test if that is really the case.
Maybe try it with +1 Artillery, so 2, 4, 5, 6, 7... and so on. The biggest question is probably the 15 to 21 for those who play smaller manpower countries. Also, a good mountain and jungle test is Peru and form Grand Columbia... Probably some of the worst terrain in the game.
I had once a WC attempt as Germany. I was invading the entirety of South America through the amazon in the 50s. Tried naval invasions but the AI spammed like 100 divisions to guard its ports. It was a lot of pain pushing through both mountains and the amazon while soldiers were eating tree bark and grass. Some of the worst pain in my life
It is the worst terrain, I'm currently slogging it out with fascist Peru rn in the mountains, I have 40 combat with and neealy 40 divisions in two different armies. Imma switch them to 18 or 20 to see how it goes. Also Peru has CAS so that's an issue.
For some reason to me 18w has been new 20w and had best success with it. Also 15w with special forces have been really solid for me too. Anything above 21 has been just sad, though in all fairness I mostly play weak nations so that way I have been able to get higher unit counts which seems to be more important now.
I ran some numbers myself and found A few width options depending on the part of the world you're playing in. 42 width and 28 width are good if there aren't mountains or marshes. 15 width is really good if there aren't forests or jungles. 21 and 12 width are the best all-terrain options. If you want units specifically made for fighting in mountains, 37 width, 25 width and 18 width are all great options.
Yeah, I totally hate it when the speeches start playing, too. Even though I disable the whole radio pack. I've disabled the DLC from the launcher, it was the only way to stay sane.
I would say try it again to get a more accurate results, but I do feel ether 18 or mid to high 20’s are the most effective templates. It would all come down to Industry and Manpower that would be the best option for templates.
@@lancesavage5539 Yea ''the thunder roars alone'' is much better than ''Seek alignment with Germany" because of "Dievturība", but I found a trick how can you do "The Thunders roars alone" while being in axis. Basically, you have to decline the anti comintern pact and you shouln't trade with germans. Then do "Seek alignment with Germany" and like 95% of the times they decline (because you declined the anti comitern pact and they have a bad opinion about you) And then you have a button/event to automatically do "The thunder roars alone". But you shouldn't press it yet (you have 13 days until it automatically presses itself) and just improve relations with Germany. And like within 6 days of improving relations with Germany they will accept the invite to the axis. Then when you are in the axis press the button and you'll be in the axis while having the thunder roars alone focus. But for me, this works only when you have done "The old ways" and "Latvian military complex".
@@Latziks thats actually Interesting, see when Ive played thunder cross Latvia, I managed to unite the baltics and create an army of nearly 600k, (underequipped), the Germans wont accept you in their faction, nor will accept the guarentee from one the axis investment focuses… youre on your own against the soviets and although I havent tried falling back on Riga, you just get crushed fast… air and logistics is a nightmare and the soviet ai will constantly navally invade you too… So Taureor, Play Latvia, fully upgrade Dievturība, and create an empire if you dare
To make the test more reliable, I advice the follows : - testing it with infinite number of guns and manpower (using cheats) - no supplies issues (however I don't really know ho to implement this (maybe infinite transport planes ?), so let's say as few supplies issues as possible) - keeping roughly the same number of men wich implies starting at 1 and deleting division as the width increase - deleting the airforce of both nations - and no allies obviously
I think the intention of this patch was to introduce variance -- in MP games for example i have been v successful with 45 width AT divs for holding the belgian border and haveing a few armies of 21 or 27 widths in reserve
40 width infantry with aa, engineers, field hospitals (grand battleplan infiltration branch) are doing absolutely amazing with air support for me. i love this build cause 40width pure infantry have so much hp that with filed hospitals after year of figthing you will have veteran 40widths which is aboslutely devastating
Excellent idea for something that I would really love to see more of - people actually testing the stuff in action! Maths are cool tho. I would personally suggest, in the future, perform tests in multiple phases - you've spent the same amount of time analysing combat widths that showed promise as those that flopped quickly or had no other supporting data. If you were to do this again, I'd suggest doing a quick version of the test for a large amount of widths (like you did with 2 to 50), and then pick only, say, top 5 performers, and maybe a couple extra widths that, perhaps, didn't do well but some people did the math blah blah blah - interesting points, basically - and play around more with those. Run several tests on them each or maybe change the terrain they're fighting in, that kinda stuff. Also a very interesting point about support companies, especially with superior firepower!
Thank you for the video! I've seen the combat width Stat charts at this point but it is nice to see it actually play out. If you do release a more controlled version, I'd certainly enjoy it. Not sure if the sentiment is shared, though.
In my experience so far, a 15W infantry (6inf 1 art) with basic support (art, engi, recon) works well. Especially in areas with less than optimal supply. 15 works well with a lot of the width limits, and the artillery allows for additional soft attack, to safe time, and lives. “Our manpower is precious, bullets are cheap.” One thing this video did not really talk about that I have found frustrating, is the effect of supply on units. It makes certain stead’s nearly impossible to attack, and the supply hub is very expensive. So not always an option for smaller nations. At least in my head, 25W should be optimal, because the supply limits are now 75-96. It follows the principal of the 20/40 meta which is concentrating force in your army. You know that on any given tile you can always put 3 (and 4 if the limit is higher-will take small minus ) units and +1 for each adjacent attacking tile. I’m not sure the best comp for such a 25W perhaps 3 arty? Idk open for conversation.
I think 2 variables you did not try with this is 1. Baltics capitulating to USSR messing with frontlines 2. Airpower - USSR AI could have used different planes for each run, which could have had adverse affects on each different run. CAS, tac bombers, etc.
The idea to re-do some templates (the ones that performed the best) while keeping the same number of soldiers in the field sounds like a pretty interesting prospect, would give a clearer picture to each templates strength. You could even try them out on smaller countries, thus being able to cover the whole front without the need to rely on allies to fill the gaps..
I would recommend redoing the test as Japan vs. China. Compared to Germany moving through the European Plain into Soviets. There is such a variety of terrain and difficult terrain to push through in China. I think it better test the army combat widths and penalties.
Cos 10 widths are trash; they lose more equipment, are easily destroyed, and need a lot more equipment. They are legit only best in this exact scenario against an ai soviet union simply because 10 width fits into the most widths because it is so tiny.
Taureor I would love to see more testing vids, I really enjoy people different takes and strategies. Besides in the no step back dlc, I’m not really sure if what I’m doing is ever right.
I really appreciate the effort and content! I don't know if there is a 'creative mode' but that might prove better testing results. You could leave 2 full army groups to be able to fill gaps and you could manually add equipment to always be supplied. For example, when you get up to 48 width but add in a bunch of guns, it would allow you to compare against 12 or 24 width with full supply and the same number of units.
Changing the number of divisions in the field (units), will have a very big impact, if you only want to test combat width it will have a very big influence on it and how the line progresses. It's ok I suppose for an anecdotal test, but there are so many things that have to be controlled for
I think that 1) you should have more manpower in the field, like 2-2.5M or something. 2) test with the mod that gives you 300 division per general, if it's better to have more generals or just if the more units, the better. 3) cheat to get what you need right, like infinite guns 3 or whatever you choose. Put a set up that can make a lot of conquest but not annihilate the soviet union. A instawin army strength will not tell us anything, but also a stallmate army strength will have very little variance between good and bad units. 4) Maybe a reverse test could be interesting, to see the difference between offense and defence: tag into the soviet Union at the right, cheat to get the army you need, (maybe use all "green" division if have them trained is too hard), and see how well you can hold with different divisions.
fun "fact" (i guess) the term Theory is to describe something that has been tested so many times that it cannot be disproven, like gravity is a theory, relativity is a theory etc, somewhere along the lines of history people started to use theory in place of hypothesis which is what people mean when they are saying "well in theory xyz". just thought you might find that interesting :p
German civil war might be an easier spot to test this on. You have a pretty contained 1v1 with no chance of outside interference. You could just test mean time until Capitulation. You could probably use console commands to spawn and destroy your units to keep your army at the same size.
The biggest advantage of small units 6-10w is the stacking of support arty. Since support arty is a massive soft attack buff for only 12 arty. Especially when using superior firepower right right.
I think testing would have been cleaner if you were doing this in multiplayer with someone working with you to get consistent results, though I'd pity anyone that had to sit through to hours of content you trimmed out of this. Thanks for going through all this effort.
Great video, thanks Taureor! I wonder if you got a chance to take a look at the casualty numbers on these saved games? I vaguely got the impression (judging from the manpower reserves, admittedly not the best of indicators) that small divisions were taking greater numerical losses. In particular the 10-widths seem to be taking about 400k more casualties than the 16- and 18-widths, if my reverse engineered napkin-math is correct. If this is true, that difference is remarkably large despite the one-sample size of the test.
Kinda at the same point with Allied speeches pack...It was a neat little gimmick DLC the first couple times they played but if I have to have my eardrums burst by Stalin one more time I'm gonna go crazy
Taureor I don’t understand. How is canceling Mefo bills a good thing the construction bonuses are crazy, you can just destroy your enemy with shear economy alone.
As thorough as this test might be within its parameters, I feel as though the premise of it is flawed. Plain infantry isn't normally used this way. When people talk about what combat width is best they're usually thinking about it in two distinctly different ways: what is best for defensive infantry templates (like the ones that you're testing, but you're not using them in a primarily-defensive capacity) and what is best for offensive templates (huge infantry/arty templates, armor templates, etc). I don't know a single person who does pure infantry slogs like this, because it's a good way to waste a lot of manpower and equipment. I don't think that the conclusions you draw at the end are valid for the broader game because you're ignoring the broader game entirely. It might be true that 10w pure infantry is best *in comparison to other widths of infantry with exactly the same amount of overall manpower and purely on the offensive*, which is the only case you've tested here, but they're also going to be less efficient in terms of manpower/equipment losses, and also not as effective as properly-designed offensive templates used properly. This would be much better tested in a controlled multiplayer game. At least then you could completely eliminate all of the random AI factors.
The reason for the "randomness" is probably that better widths manage to break through to Moscow better than the others, which requires more troops to focus on that part of the frontline. That's why Leningrad's not falling in these approaches. Lower Widths don't manage to capture Moscow, so they have enough troops to allocate to capture Leningrad
42:50 actually you can ban specific songs or speeches when you go to the radio. There is like a small red X when you hover over the left? side of the song so you can block it if you dont like it
he showed at 19:32 that he has blocked it, but either it's bugged or hardcoded to play that speech at some point, no matter if you have it disabled or not
You went into organization vs. HP choice. Organization is averaged, and HPs are summed. That mean that each of your divisions have 63 organization points, but HP are 25 x number of infantry battalions, or 12.5 x combat width. Once, many moons before, I checked, and I got 2.4 ish org point loss for each HP loss. So I assume that 70% of hits goes to organization, and 30 % to HP pool. It is random of course. Division will stop fighting when it runs out of either one or another. So for 63 organization points, you will lose 27 HP points, and I am going to round down to 25. So every time your division loses, or losses all organization points, it will have one battalion of casualties as well. So for 10 width, 5 battalion, it will be 20% casualties, and for 20 width, 10 battalions, it will be 10% casualties. So two 10 width division will have twice as many organization points, and they will fight twice long, but they will lose 2 times more people. Maximum combat penalties is 33% and it apply to all modified stats (so it applies to both base stats and to bonuses). For each percentage of overstocking, 1.5 % of penalty applies. 26 width is last width in which you can squzee 4 division on 90 width, so I guess that was difference maker. I generally like 22 width all infantry divisions, but I use them mostly for defense.
When using decent infantry (with like 2 extra support companies), how much would a level 2 field hospital help you death attacking through the Soviets?
I'd say for the interest of time, have 25 different people running this experiment with one specific division template. That way each person can try any sort of deviations to account for other factors such as the actual number of men in the field
While the test provided a lot of interesting data, one thing you completely overlooked, I think, is that during most but not all of your runs, Italy makes an invasion in the black sea at some point. Naval invasions are quite difficult for the ai, just because they tend to scramble their frontlines quickly redeploying troops and such, so it makes me wonder how much that made your attack more successful.
I feel like you should start with no generals, all in one army, and after the stockpile changes (due to template switch), delete or create units to get the same stockpile (and manpower i guess), and then stop producing equipment (apart from trucks and trains for supply). Also NO ALLIES.
You're free to do that but it's one of the absolute worst combat widths you could use for stats... 20 and 40w were hit hard, probably because they were previous meta.
@@pagatryx5451 20w is still one of the best in the testing I've done. It underperformed here because pure infantry is a bad use of it, but the old 7/2 is the most cost-effective offensive infantry. 10w pure inf is better at raw DPS as shown in this video, but two or three times more expensive in manpower and IC losses.
@@joshmakescontent It's not though. Multiple statistics have been shared by user testing within the community and all I've seen have 20w as an OK division. It's OK and not nerfed into the ground but isn't meta... 18w and 42w both perform better on average and should always be used over a 20w. The ONLY time 20w outperforms such divisions is on a handful of terrain types with only 1 attacking tile (less overall width). 20w especially suffers at extended combat widths. i.e. performs slightly worse on ~80w battles but ~120 and 160 it really starts to suffer. Meaning they don't work very well for inf that often rely on massing their assault. 20w also loses out to 42 width which has improved combat effectiveness and even less loss which is probably irrelevant for inf but definitely preferred for tanks and mech. 20w isn't horrendous. It's one of the better widths but it's still not optimal or as go too as it was before. I would pretty much always go for 18w (6/2) unless I'm somewhere like asia where I can adjust my expected combat width requirements.
@@pagatryx5451 I'm aware of the huge amount of hypothetical posts about ideal width and the like, I used them to inform my testing. I'm just sharing what I've seen play out with testing conducted very similarly to this video, but focused on division composition rather than width alone. If the results of the best width don't match the results of best full comp, the logical conclusion is that width is not the be-all end-all of division design anymore, and composition and role are just as important. 9/1 is the new king of infantry IMO, it has the punching power to de-org attackers rather than simply hold, which prevents the enemy from cycle attacking your org down. They can do good work attacking once you've denied enemy supply with your tanks, which is more effective than trying to go for big pockets with the supply changes. 7/2 won the trials I've done by a relatively small margin on 14/4 replacer divisions like 9/2, 6/3, 6/1 etc. at battle planning, and a significant margin over 9/3 and the 15/4. I prioritized losses with that because I'm of the mind that you should only run offensive infantry if you can't afford tanks. If you play like Taureor though you should do 10w with supports on SF R+R, those have the best punch by a large margin but burn IC and manpower 2 or 3 times as fast as 7/2. 42w runs into problems with punching power because of support stacking and is significantly more likely to get screwed by width in a specific tile you really need to win on, even if it works well on average. Big divisions tend to get demolished by smaller divisions in actual combat testing as well, again because support stacking. The most important factor IMO is that they get punished heavily by supply once you're a few tiles into enemy territory. They're too supply hungry to take advantage of local sources and too few in number to spread consumption out across multiple tiles. This is especially important with tanks, you can check out Feedback's USA tank video to see what I'm talking about. 21w tanks are way better at dealing with the stop-and-go and lose fewer tanks because they take attrition less often.
My strat: A bunch of 10 width infantry to defeat Poland, Denmark, Norway, France, UK, etc. 12 width infantry for Barbarossa and to defeat usa. Leave 1 50 width infantry at the capital. Those will be for in case things go VERY bad
Speaking from a realism point of view. 18-24 width would probably be most appropriate. Most WW2 divisions (or the average infantry division, at least) would consist of three combat regiments of three battalions each.
Great video thanks. Quick question out of interest that wasn't mentioned in the video. Did you remove air force (or atleast CAS) from both yourself and AI before war declaration? I just feel that could possibly awnser as to why some of those runs were so random (Soviet ai concentrating their cas in a particular area on different runs and saves) if you didn't tag switch to delete them
Since this has been plaguing the community ever since no step back came out, I would like to see more strenuous testing. Use this as a preliminary to see what unit widths may be good, than start with the same number of troops every time. Not the same number of divisions mind you, but the same number of actual men. You will need to math a bit. Than run each one of those multiple times, lets say 3 times each. Leave out any allies, increases variables. 12/16/18/22/24/26/28/30 and that should be enough. Record stuff so you can make sure you do the same thing every time. Than record deaths, equipment stockpiles and enemy casualties. Since you're using the same amount of men every time, the equipment and casualties will matter, and the supply requirement should stay the same. Edit-wrote this about halfway through the video. Thank you for immediately taking my advice, lol.
Imo 10 widths are better if u can support it as if u have support arti/support equipment you can have more of it(example, u can have 2 support arti with 2 10 widths instead of 1 support arti with 1 20 width).
I think if you don't want to min-max to 100% just don't build smaller than 10 or larger than 44, and anything between 22 and 39. So basicly 10-21 and 40-44
great video, thank you for your service. If you were to repeat the experiment, i think starting with a larger normalized force of say 20w, and sticking to the same manpower would better. keeps the front line filled on larger widths and we get to see if having 4army groups of 10w is better than 2army groups of 20 etc
TIMESTAMPS FOR TEMPLATES AFTER 6 MONTHS (original testing method):
16:20 - 2 width
17:09 - 4 width
17:45 - 6 width
18:29 - 8 width
19:10 - 10 width
19:59 - 12 width
21:04 - 14 width
21:48 - 16 width
15:48 - 18 width
22:55 - 20 width
25:17 - 22 width
26:24 / 27:13 (only one army group instead of two) - 24 width
28:03 - 26 width
29:01 - 28 width
29:41 - 30 width
31:57 - 32 width
32:29 - 34 width
33:11 - 36 width
none - 38 width
none - 40 width
none - 42 width
none - 44 width
none - 46 width
none - 48 width
none - 50 width
TIMESTAMPS FOR NEW TESTING METHOD AFTER 6 MONTHS (no allied nations, same amount of troops in the field for every width: 1.2 million):
34:37 - 10 width
36:26 - 12 width
37:41 - 14 width
39:11 - 16 width
40:38 - 18 width
41:40 - 20 width
43:48 - 22 width
44:20 - 24 width
45:37 - 26 width
46:03 - 28 width
47:46 - 30 width
49:06 (one month longer) - 32 width
49:49 - 34 width
50:21 - 36 width
50:49 - 38 width
51:17 - 40 width
52:46 - 42 width
53:24 - 44 width
54:30 - 46 width
54:55 - 48 width
56:06 - 50 width
CONCLUSION:
56:25
Wow thanks!
Thank you good sir
Could you add a conclusion timestamp?
Pinned
I noticed you went by 2 for pure infantry. Could you possibly add 1 battalion of line artillery and run it again for the odd number widths?
I know personally you said you'd done pretty well with 27 widths and with the combat width statistics chart it's been shown it's a peak in being more effective than basically anything else until roughly 41
Constantly invading the USSR over and over as Germany in HOI4 is the definition of insanity. Thank you for your service.
Dealing with those supply lines consume my soul each time
Im at the point with 2000 hours in the game I just put field hospital and maintenance and logistics and just leave the computer I just don't have the patience to control the supply while the game runs at a slog
Have you ever heard... the definition... of insanity?
@@baileymash doing the same thing over and over expecting different results. (though in the case of this game its completely possible for different results to happen)
Only if you continue to get the same result haha
I'm convinced Paradox was sad that people were constantly changing from the standard 18 width units before NSB so they gave 18 width a massive buff
Makes the game more friendly for new players which is a good thing. For the longest time I was avoiding HoI4 because of the arcane knowledge it demands.
@@FairyRat Don't worry they added more shit to learn with supply lines. I like HOI but really to learn it I say play Germany or the US. One has all the advantages through the tree and the other has all the time in the world.
Makes you think maybe they wanted historical division to work better? I did a US playthrough recently, had no idea on the meta since the US fights world over and just went with the historical 1943 lay out. It did great against Japan and OK against Germany.
I haven't watched the video yet and I don't know if he will discuss it later but what buff did they receive exactly?
@@Xinamon98 the buff wasn't a direct buff to it specifically, but more that it can work well in most terrain types now.
Before NSB 20 width was the obvious king
Maybe go with 3M troops in each run since that's closer to historical, but also, it will reduce gaps we see with the bigger units.
This^
Well, historically the Germans weren't attacking in 1939 either.
Also, Andrey Vlasov was leading kind of Russian anti-Stalin army in German Service. He was recently discussed in Czechia for his role in Prague uprising, where his army joined up resistance against Germans to be allowed to get away from Red army on time. Kinda controversial topic over here.
Many don't like him because he was a collaborator. but considering his other options namely life in a prison camp or returning to Stalin a failure it was really his only option. Plus, there is really no evidence that he was a national socialist himself he just worked with them.
Vlasov based
@@nikoclesceri2267 maybe he shouldn't go to the Nazi side, than it wouldn't be needed for him to flee the red army after
@@nikoclesceri2267 I can't judge him, because I never was in the same situation, but many Russian/Soviet generals choosed death/prison over collaboration.
@@JIopemUncym well he also wasn't very fond of Stalin as he blamed and his purges for him being captured in the first place
The amount of work put into this is commendable. I greatly enjoyed this and am always looking forward to the next video. Commenting for the algorithm because you deserve all the views and all the subs. Cheers friends.
The way you said, with such incredulous inflection, that Radom was the capital of Poland had me howling. Radom ain't that bad! Better than Płock lmao
That was hilarious. I don't know anything about Radom, but I love that he kept that in the video. Beef with Radom is tasty ;-)
Meh, rest of Poland here. Płock is (again) "meh" nvm etc. Radom is a meme
Hi, Taureor. I’ve watched some Chinese Hoi4 content creators and their infantry templates for the new update. I feel like this is more or less related to this topic so I would bring it up here. As of 1.11.5 and for PVE (can work for PVP but need slight modification), the main strategy is as follow:
The frontline consists of two types of divisions, purely defensive divisions called the ‘filler’ used to halt enemy advances and main offensive divisions called the ‘gundam’ which are literally Swiss Army knives which cut through the enemy lines like butter.
The filler divisions can be any widths of pure infantry. No support companies are needed for them since their only job is to hold the enemy divisions and you don’t want your auxiliary divisions to use up your important equipments. If you run low on guns, you can even stop supplying guns to these filler divisions and let them hold the enemies with their flesh (more like a China strategy huh). An important note is that you would like to have some trained filler divisions ready to deploy. In case of enemy breaking through your frontline, you can immediately deploy these units to plug up the hole. Finally, the ideal filler width is 10/5 infs since low width fillers can not take much beating, and large filler units have less manuverability and tend to get focus fired. However, this is all up to the player and can vary based on the manpower situation.
As for the more exciting gundam units, they are 42 width divisions comprised of: 4/6 artilleries, 1 anti-air medium tank and the rest are infantries. The support companies include engineering, supp art, supp flamethrower light tank and light tank recon. The final slot is free and can be anything of your choice. If you neglect building aircrafts in a game, you can switch out a couple of infantries for field anti-air guns for higher AA values. The differences between 4/6 arts is mainly soft attack value and organisation + hp. Having more arts will give higher soft attack but lowers org and hp, vice versa. The anti-air medium tank is nicknamed the ‘golden shield’ of the unit. You want to mount AA lvl 1 (1939) or AA lvl 2 (1940+) and stick on as much armour as you can into this thicc boi. Just make sure the speed is not lower than the infantries and reliability is not too shabby. On average for 1939 battles, this AA medium tank has ~100 armour value and can give your unit ~20 armour value, which is massive for infantry units. The sole purpose of flamethrower light tank is to give terrain modifiers, so just keep them as cheap as possible, and not slower than reg inf. Finally the light tank recons, these units will provide lots of breakthrough values, so design this tank with all the breakthrough modules. This is not necessary in most circumstances vs the AI, but is a must for PVP to penetrate enemy armour. Finally if you are playing on a poorer country, you can go for 21/22 width with half of the units listed above. They still have insane stats and can easily crush the AI.
If you want to have armour divisions, still go for 42 width infantry division, train them to at least lvl 3 before switching to the armour template. As of now, you only lose division level if you need more manpower after switching the template. Since going from infantry to armour units you always use less manpower, there is no loss in division exp which save you the tanks to train them up. This exploit is especially broken if you have lvl 5 42 width infantry units...
Finally, as for the strategy, you want to send volunteers to get as much army exp as possible in the early game. For example, as Germany, send volunteers to Ethiopia at the start of the game and slowly build them up to 42 widths. Since higher the widths the more exp they get from combat, it is absolutely worth it to upgrade these volunteer divisions to a higher widths. Then you should get tonnes of exp from Spain and Japan. When you go to war with other countries, put fillers on the frontline and gundams should focus enemy supply line. The gundams are so strong that one unit can easily punch through multiple enemy divisions.
Well, I hope you enjoy this information and maybe this can help you with your future playthroughs.
Great write up
The aggressiveness really just says how much Organization they will hold back for defense, it does not alter the AI's decision to take empty provinces at all. If you want to have the AI prefer land-grab, then the "Spearhead" battle plan is better then the Offensive line battle plan. (At least I think so, can't be 100% sure...)
Yes so technically they take land faster because they do not have to wait as long to hit the org threshold
This was an amazing video Taureror. I've been trying to play as China in NSB and just couldn't keep the Japanese out: they would roll through the North within a few months and I just could never push them out once they landed. After watching this I switched to ten widths with artillery support and everything just got easier, they reinforced faster, attacked faster, suffered less from supply issues and I finally defeated and conquered Japan! Can't wait to play as the Japanese now.
Thank you so much for the tests, Taureor!
"It costs stuff to produce stuff"
-Taureor, 2022
It would be interesting but even more timeconsumming to see it every time with same manpower in field (aka 10x more 2width divisions then 20 width to fulfill that)
I do that too later in the video
obv thanks for hard work tho
@@Taureor ah sorry, commentedd before seeing the finish :D
Dear taureror! You mentioned wanting a "hyper agressive option, troops just pushing onwards without concerns for supply, taking every open province." That's what the spearhead order does. Just use that one for your troops. Works wonders in most games (especially with tanks). But in normal games make sure there is an army with normal orders nearby to reinforce the gaps the spearhead tends to leave behind so your troops don't get encircled.
This is not what spearhead does directly, spearhead just tells the divisions to find the closest tiles to the offensive line, and only go to them, not where the frontline is. It kinda has that effect, but it's very dangerous to use for large orders because they will just leave the sides wide open
It’s always a good day when the polish man uploads
I don't mind watching the testing content, it's pretty interesting. That said, I do also still really want to see a standard Stalin run and maybe even a modern version of the "Stalin: no army" run you did several patches ago, before espionage and the modern Soviet focus tree.
i enjoy this kind of content as much as i do the entertaining ones like savegame fixing or alt history mods. Its like the well made BBC Dokumentary about a topic you enjoy, keep up you kind of work mate.
So I've watched a few videos and done my own testing while playing around in single player.
According to a very lengthy document from the community, which I saw in a Feedback video but read myself: 10, 15, 25 (Mountaineers) 27, 41, 42, and 44 are all the mathematically best widths
Over all 15 and 30 along with 16/18 and 27, seem to be the best multiplayer and casual widths in different videos. And in my experience 15, 25 mountaineers, 26-8 and 30 are quite good.
Then there's also the min-max'd ideal divisions based off of terrain widths.
Fbg released a new video a few days ago where he said after more testing 10-w and to an extent 15-w are not optimal because they take too many losses and shred ur equipment and manpower.
In Tommykay great game 15 width with anti air was meta
Also flamethrower support companies are meta af
@@MoonIronTR I think Paradox JUST nerfed flame units in the latest patch.
27w is bad in actual side by side testing with 7/2s, which have the best performance again ironically enough. I might make a video on it and why, but the bottom line is as soon as the line fills up they have *crippling* overwidth problems in plains and desert that were not in any way accounted for in the post that claimed they are "mathematically the best." They end up performing slightly worse than a 7/2 division that's ~25% more expensive. The rest of those widths aside from 10w have similar issues, but 27w is probably the worst offender.
Please try to remember to look at the casusties because I’ve noticed that higher width gives less casualties. Great Vid though
10w inf for battleplan.
42w pzr/art for micro.
I have been constantly trying different army loadouts similar to this but you went farther than I ever could, so this is extremely helpful. Especially the conclusion/thoughts at the end. Thank you.
i watched the whole video, it was really informative and interesting!
I'd love to see the 10 width strategy in an actual full game maybe with a smaller country and no allies because, yea, being in a war with many allies and enemies definitely distorts the results a lot
i always use 10 width as minor however even though it did not show here you tend to loose units in 10 width which as minor is really bad cause it makes a hole in the small frontline you use
Small units with support companies are proportionately worse. Support companies have very low HP and Org, so smaller units with support companies can get weighed down by these penalties. Some support companies, like support artillery, add to pure combat stats so having more units with them would be better, but others provide different bonuses. Like signal companies; the bonuses to initiative and coordination aren't going to be additive if you have more units with the, you'll just require more support equipment
I think the reason why some smaller divisions would be better than their larger counterpart (in some cases) is because of the support company. 2 artillery support company for 2 10-width division would have more impact than 1 artillery support company from 1 20-width division. Although that would cost more total manpower and total production cost so you might want to consider that.
With that in mind, choose an arbitrary total production cost or total manpower, maybe 120 Division for 45-width with full support companies (as it is really doable for major countries to get that number and to remove the unfairness of not having enough divisions to cover the frontline as in most cases, you should be able to). Calculate their total production cost or manpower (depending on which you choose), and put it all in an excel to easily calculate the number of divisions per combat width.
Also, you might want to check the paper some dude released on reddit who did the math and take maybe the top 5 most effective combat width and top 5 least effective combat width to test if that is really the case.
Maybe try it with +1 Artillery, so 2, 4, 5, 6, 7... and so on. The biggest question is probably the 15 to 21 for those who play smaller manpower countries.
Also, a good mountain and jungle test is Peru and form Grand Columbia... Probably some of the worst terrain in the game.
I had once a WC attempt as Germany. I was invading the entirety of South America through the amazon in the 50s. Tried naval invasions but the AI spammed like 100 divisions to guard its ports. It was a lot of pain pushing through both mountains and the amazon while soldiers were eating tree bark and grass. Some of the worst pain in my life
It is the worst terrain, I'm currently slogging it out with fascist Peru rn in the mountains, I have 40 combat with and neealy 40 divisions in two different armies. Imma switch them to 18 or 20 to see how it goes. Also Peru has CAS so that's an issue.
For some reason to me 18w has been new 20w and had best success with it. Also 15w with special forces have been really solid for me too. Anything above 21 has been just sad, though in all fairness I mostly play weak nations so that way I have been able to get higher unit counts which seems to be more important now.
i use 21w for infantry (1 w is aa or at usually) and 42 for tanks and heavy infantry including special forces
I use 27w, 3 artillery in them. They work pretty well considering I can get the soft attack up
I ran some numbers myself and found A few width options depending on the part of the world you're playing in. 42 width and 28 width are good if there aren't mountains or marshes. 15 width is really good if there aren't forests or jungles. 21 and 12 width are the best all-terrain options. If you want units specifically made for fighting in mountains, 37 width, 25 width and 18 width are all great options.
Why those numbers?
Very interesting, thanks for all the hours you put into it!
My evening just got even better, thanks Taureor!
Yeah, I totally hate it when the speeches start playing, too. Even though I disable the whole radio pack.
I've disabled the DLC from the launcher, it was the only way to stay sane.
Definitely do other maybe on the V.S map where it's just Germany and USSR so that randomness is lowered .
I would say try it again to get a more accurate results, but I do feel ether 18 or mid to high 20’s are the most effective templates. It would all come down to Industry and Manpower that would be the best option for templates.
4:40 Radom is the capital :D As a Pole I know exactly why it is funny and absurd :D
Could you explain:)
Will you make a video about Latvian new focus tree someday?
sure!
@@Taureor thank you. And thank you for the quality content.
@@Taureor Play Latvia, go down the “thunder roars alone” path… and just create an empire
Or perhaps just survive
@@lancesavage5539 Yea ''the thunder roars alone'' is much better than ''Seek alignment with Germany" because of "Dievturība", but I found a trick how can you do "The Thunders roars alone" while being in axis. Basically, you have to decline the anti comintern pact and you shouln't trade with germans. Then do "Seek alignment with Germany" and like 95% of the times they decline (because you declined the anti comitern pact and they have a bad opinion about you) And then you have a button/event to automatically do "The thunder roars alone". But you shouldn't press it yet (you have 13 days until it automatically presses itself) and just improve relations with Germany. And like within 6 days of improving relations with Germany they will accept the invite to the axis. Then when you are in the axis press the button and you'll be in the axis while having the thunder roars alone focus. But for me, this works only when you have done "The old ways" and "Latvian military complex".
@@Latziks thats actually Interesting, see when Ive played thunder cross Latvia, I managed to unite the baltics and create an army of nearly 600k, (underequipped), the Germans wont accept you in their faction, nor will accept the guarentee from one the axis investment focuses… youre on your own against the soviets and although I havent tried falling back on Riga, you just get crushed fast… air and logistics is a nightmare and the soviet ai will constantly navally invade you too…
So Taureor, Play Latvia, fully upgrade Dievturība, and create an empire if you dare
This was really interesting. I'd like to see more videos like that, maybe with artillery and support companies, normalising production cost
Well someone had to do it and as such I thank you for your service you madlad
To make the test more reliable, I advice the follows :
- testing it with infinite number of guns and manpower (using cheats)
- no supplies issues (however I don't really know ho to implement this (maybe infinite transport planes ?), so let's say as few supplies issues as possible)
- keeping roughly the same number of men wich implies starting at 1 and deleting division as the width increase
- deleting the airforce of both nations
- and no allies obviously
Italy’s naval invasion changed it every time
I think the intention of this patch was to introduce variance -- in MP games for example i have been v successful with 45 width AT divs for holding the belgian border and haveing a few armies of 21 or 27 widths in reserve
Thank you for video! That's was informative would like to see more of that tupe of contet
40 width infantry with aa, engineers, field hospitals (grand battleplan infiltration branch) are doing absolutely amazing with air support for me. i love this build cause 40width pure infantry have so much hp that with filed hospitals after year of figthing you will have veteran 40widths which is aboslutely devastating
Excellent idea for something that I would really love to see more of - people actually testing the stuff in action! Maths are cool tho.
I would personally suggest, in the future, perform tests in multiple phases - you've spent the same amount of time analysing combat widths that showed promise as those that flopped quickly or had no other supporting data. If you were to do this again, I'd suggest doing a quick version of the test for a large amount of widths (like you did with 2 to 50), and then pick only, say, top 5 performers, and maybe a couple extra widths that, perhaps, didn't do well but some people did the math blah blah blah - interesting points, basically - and play around more with those. Run several tests on them each or maybe change the terrain they're fighting in, that kinda stuff. Also a very interesting point about support companies, especially with superior firepower!
Taureor, great vid I would love to see more testing
"Radom is the capitol, OK xD" ~Taureor 2022
Thanks for all the hard work! 😎
Thanks for the video. It was interesting. I would be interested in seeing more.
Thank you for the video! I've seen the combat width Stat charts at this point but it is nice to see it actually play out. If you do release a more controlled version, I'd certainly enjoy it. Not sure if the sentiment is shared, though.
you need a physical clipboard with a pre war testing CHECKLIST of did id do all the things before hand other wise love this kinda stuff
Taureor's best division: 10 width pure infantry + Support companies (Art + Recon + Eng + Anti-air + Signal)
In my experience so far, a 15W infantry
(6inf 1 art) with basic support (art, engi, recon) works well. Especially in areas with less than optimal supply. 15 works well with a lot of the width limits, and the artillery allows for additional soft attack, to safe time, and lives. “Our manpower is precious, bullets are cheap.”
One thing this video did not really talk about that I have found frustrating, is the effect of supply on units. It makes certain stead’s nearly impossible to attack, and the supply hub is very expensive. So not always an option for smaller nations.
At least in my head, 25W should be optimal, because the supply limits are now 75-96. It follows the principal of the 20/40 meta which is concentrating force in your army. You know that on any given tile you can always put 3 (and 4 if the limit is higher-will take small minus ) units and +1 for each adjacent attacking tile. I’m not sure the best comp for such a 25W perhaps 3 arty? Idk open for conversation.
I think 2 variables you did not try with this is
1. Baltics capitulating to USSR messing with frontlines
2. Airpower - USSR AI could have used different planes for each run, which could have had adverse affects on each different run. CAS, tac bombers, etc.
Thank you Taureror this is so helpful
Taureor setting the new meta
The idea to re-do some templates (the ones that performed the best) while keeping the same number of soldiers in the field sounds like a pretty interesting prospect, would give a clearer picture to each templates strength. You could even try them out on smaller countries, thus being able to cover the whole front without the need to rely on allies to fill the gaps..
I would recommend redoing the test as Japan vs. China. Compared to Germany moving through the European Plain into Soviets. There is such a variety of terrain and difficult terrain to push through in China. I think it better test the army combat widths and penalties.
I saw a Video where they said 27 is best so you need artillery
Cos 10 widths are trash; they lose more equipment, are easily destroyed, and need a lot more equipment. They are legit only best in this exact scenario against an ai soviet union simply because 10 width fits into the most widths because it is so tiny.
Taureor I would love to see more testing vids, I really enjoy people different takes and strategies. Besides in the no step back dlc, I’m not really sure if what I’m doing is ever right.
I really appreciate the effort and content! I don't know if there is a 'creative mode' but that might prove better testing results. You could leave 2 full army groups to be able to fill gaps and you could manually add equipment to always be supplied. For example, when you get up to 48 width but add in a bunch of guns, it would allow you to compare against 12 or 24 width with full supply and the same number of units.
Changing the number of divisions in the field (units), will have a very big impact, if you only want to test combat width it will have a very big influence on it and how the line progresses.
It's ok I suppose for an anecdotal test, but there are so many things that have to be controlled for
I think that
1) you should have more manpower in the field, like 2-2.5M or something.
2) test with the mod that gives you 300 division per general, if it's better to have more generals or just if the more units, the better.
3) cheat to get what you need right, like infinite guns 3 or whatever you choose. Put a set up that can make a lot of conquest but not annihilate the soviet union. A instawin army strength will not tell us anything, but also a stallmate army strength will have very little variance between good and bad units.
4) Maybe a reverse test could be interesting, to see the difference between offense and defence: tag into the soviet Union at the right, cheat to get the army you need, (maybe use all "green" division if have them trained is too hard), and see how well you can hold with different divisions.
fun "fact" (i guess) the term Theory is to describe something that has been tested so many times that it cannot be disproven, like gravity is a theory, relativity is a theory etc, somewhere along the lines of history people started to use theory in place of hypothesis which is what people mean when they are saying "well in theory xyz". just thought you might find that interesting :p
German civil war might be an easier spot to test this on. You have a pretty contained 1v1 with no chance of outside interference. You could just test mean time until Capitulation. You could probably use console commands to spawn and destroy your units to keep your army at the same size.
10 width divisions need a lot of support equipment though.
I am really in love with 15w
I heard that 27 width is like a good average like middle ground for infantry.
The biggest advantage of small units 6-10w is the stacking of support arty. Since support arty is a massive soft attack buff for only 12 arty. Especially when using superior firepower right right.
I think testing would have been cleaner if you were doing this in multiplayer with someone working with you to get consistent results, though I'd pity anyone that had to sit through to hours of content you trimmed out of this. Thanks for going through all this effort.
another epic vid keep up the good work
Great video, thanks Taureor! I wonder if you got a chance to take a look at the casualty numbers on these saved games? I vaguely got the impression (judging from the manpower reserves, admittedly not the best of indicators) that small divisions were taking greater numerical losses. In particular the 10-widths seem to be taking about 400k more casualties than the 16- and 18-widths, if my reverse engineered napkin-math is correct. If this is true, that difference is remarkably large despite the one-sample size of the test.
Kinda at the same point with Allied speeches pack...It was a neat little gimmick DLC the first couple times they played but if I have to have my eardrums burst by Stalin one more time I'm gonna go crazy
Another test. How will work defense with different combat widths. That is very important if you play as small nation who need defend a lot of time.
Taureor I don’t understand. How is canceling Mefo bills a good thing the construction bonuses are crazy, you can just destroy your enemy with shear economy alone.
Thank you! This is so nice
As thorough as this test might be within its parameters, I feel as though the premise of it is flawed. Plain infantry isn't normally used this way. When people talk about what combat width is best they're usually thinking about it in two distinctly different ways: what is best for defensive infantry templates (like the ones that you're testing, but you're not using them in a primarily-defensive capacity) and what is best for offensive templates (huge infantry/arty templates, armor templates, etc). I don't know a single person who does pure infantry slogs like this, because it's a good way to waste a lot of manpower and equipment.
I don't think that the conclusions you draw at the end are valid for the broader game because you're ignoring the broader game entirely. It might be true that 10w pure infantry is best *in comparison to other widths of infantry with exactly the same amount of overall manpower and purely on the offensive*, which is the only case you've tested here, but they're also going to be less efficient in terms of manpower/equipment losses, and also not as effective as properly-designed offensive templates used properly.
This would be much better tested in a controlled multiplayer game. At least then you could completely eliminate all of the random AI factors.
Thank you, great information
The reason for the "randomness" is probably that better widths manage to break through to Moscow better than the others, which requires more troops to focus on that part of the frontline. That's why Leningrad's not falling in these approaches. Lower Widths don't manage to capture Moscow, so they have enough troops to allocate to capture Leningrad
42:50 actually you can ban specific songs or speeches when you go to the radio. There is like a small red X when you hover over the left? side of the song so you can block it if you dont like it
he showed at 19:32 that he has blocked it, but either it's bugged or hardcoded to play that speech at some point, no matter if you have it disabled or not
@@Firesword05 ok
You went into organization vs. HP choice. Organization is averaged, and HPs are summed. That mean that each of your divisions have 63 organization points, but HP are 25 x number of infantry battalions, or 12.5 x combat width.
Once, many moons before, I checked, and I got 2.4 ish org point loss for each HP loss. So I assume that 70% of hits goes to organization, and 30 % to HP pool. It is random of course. Division will stop fighting when it runs out of either one or another.
So for 63 organization points, you will lose 27 HP points, and I am going to round down to 25. So every time your division loses, or losses all organization points, it will have one battalion of casualties as well. So for 10 width, 5 battalion, it will be 20% casualties, and for 20 width, 10 battalions, it will be 10% casualties. So two 10 width division will have twice as many organization points, and they will fight twice long, but they will lose 2 times more people.
Maximum combat penalties is 33% and it apply to all modified stats (so it applies to both base stats and to bonuses). For each percentage of overstocking, 1.5 % of penalty applies. 26 width is last width in which you can squzee 4 division on 90 width, so I guess that was difference maker.
I generally like 22 width all infantry divisions, but I use them mostly for defense.
When using decent infantry (with like 2 extra support companies), how much would a level 2 field hospital help you death attacking through the Soviets?
I'd say for the interest of time, have 25 different people running this experiment with one specific division template. That way each person can try any sort of deviations to account for other factors such as the actual number of men in the field
While the test provided a lot of interesting data, one thing you completely overlooked, I think, is that during most but not all of your runs, Italy makes an invasion in the black sea at some point. Naval invasions are quite difficult for the ai, just because they tend to scramble their frontlines quickly redeploying troops and such, so it makes me wonder how much that made your attack more successful.
I feel like you should start with no generals, all in one army, and after the stockpile changes (due to template switch), delete or create units to get the same stockpile (and manpower i guess), and then stop producing equipment (apart from trucks and trains for supply). Also NO ALLIES.
I am still sticking to 20 with.
purely for the balance of stats and suply.
You're free to do that but it's one of the absolute worst combat widths you could use for stats... 20 and 40w were hit hard, probably because they were previous meta.
@@pagatryx5451 20w is still one of the best in the testing I've done. It underperformed here because pure infantry is a bad use of it, but the old 7/2 is the most cost-effective offensive infantry. 10w pure inf is better at raw DPS as shown in this video, but two or three times more expensive in manpower and IC losses.
@@joshmakescontent It's not though. Multiple statistics have been shared by user testing within the community and all I've seen have 20w as an OK division. It's OK and not nerfed into the ground but isn't meta...
18w and 42w both perform better on average and should always be used over a 20w. The ONLY time 20w outperforms such divisions is on a handful of terrain types with only 1 attacking tile (less overall width). 20w especially suffers at extended combat widths. i.e. performs slightly worse on ~80w battles but ~120 and 160 it really starts to suffer. Meaning they don't work very well for inf that often rely on massing their assault. 20w also loses out to 42 width which has improved combat effectiveness and even less loss which is probably irrelevant for inf but definitely preferred for tanks and mech.
20w isn't horrendous. It's one of the better widths but it's still not optimal or as go too as it was before. I would pretty much always go for 18w (6/2) unless I'm somewhere like asia where I can adjust my expected combat width requirements.
@@pagatryx5451 I'm aware of the huge amount of hypothetical posts about ideal width and the like, I used them to inform my testing. I'm just sharing what I've seen play out with testing conducted very similarly to this video, but focused on division composition rather than width alone. If the results of the best width don't match the results of best full comp, the logical conclusion is that width is not the be-all end-all of division design anymore, and composition and role are just as important.
9/1 is the new king of infantry IMO, it has the punching power to de-org attackers rather than simply hold, which prevents the enemy from cycle attacking your org down. They can do good work attacking once you've denied enemy supply with your tanks, which is more effective than trying to go for big pockets with the supply changes. 7/2 won the trials I've done by a relatively small margin on 14/4 replacer divisions like 9/2, 6/3, 6/1 etc. at battle planning, and a significant margin over 9/3 and the 15/4. I prioritized losses with that because I'm of the mind that you should only run offensive infantry if you can't afford tanks. If you play like Taureor though you should do 10w with supports on SF R+R, those have the best punch by a large margin but burn IC and manpower 2 or 3 times as fast as 7/2.
42w runs into problems with punching power because of support stacking and is significantly more likely to get screwed by width in a specific tile you really need to win on, even if it works well on average. Big divisions tend to get demolished by smaller divisions in actual combat testing as well, again because support stacking. The most important factor IMO is that they get punished heavily by supply once you're a few tiles into enemy territory. They're too supply hungry to take advantage of local sources and too few in number to spread consumption out across multiple tiles. This is especially important with tanks, you can check out Feedback's USA tank video to see what I'm talking about. 21w tanks are way better at dealing with the stop-and-go and lose fewer tanks because they take attrition less often.
@@joshmakescontent Source? Dude just trust me. And ignore other sources.
The Stalin Speach keeps playing because you have Weighted Shuffle on which makes music play when fitting.
My strat:
A bunch of 10 width infantry to defeat Poland, Denmark, Norway, France, UK, etc.
12 width infantry for Barbarossa and to defeat usa.
Leave 1 50 width infantry at the capital. Those will be for in case things go VERY bad
Speaking from a realism point of view. 18-24 width would probably be most appropriate. Most WW2 divisions (or the average infantry division, at least) would consist of three combat regiments of three battalions each.
Thank you Taureor
ı use 9/3 27 wıdth ım pretty ok wıth that. I would recommend that. for tanks ı use 6 med 5 mot/mec 1 aa. for late game 42 widths are best rn i guess.
The Soviet Union is way too weak in this game. You can just click a button and capture Moscow.
I really appreciate the effort but I don't think this test shows how to design divisions or which width is best. Respectfully.
Great video thanks. Quick question out of interest that wasn't mentioned in the video.
Did you remove air force (or atleast CAS) from both yourself and AI before war declaration? I just feel that could possibly awnser as to why some of those runs were so random (Soviet ai concentrating their cas in a particular area on different runs and saves) if you didn't tag switch to delete them
I did not use air myself, and I did not delete anything from USSR. Maybe I should have
that's something , to break the 20/40 wicth meta...
Since this has been plaguing the community ever since no step back came out, I would like to see more strenuous testing. Use this as a preliminary to see what unit widths may be good, than start with the same number of troops every time. Not the same number of divisions mind you, but the same number of actual men. You will need to math a bit. Than run each one of those multiple times, lets say 3 times each. Leave out any allies, increases variables. 12/16/18/22/24/26/28/30 and that should be enough. Record stuff so you can make sure you do the same thing every time. Than record deaths, equipment stockpiles and enemy casualties. Since you're using the same amount of men every time, the equipment and casualties will matter, and the supply requirement should stay the same.
Edit-wrote this about halfway through the video. Thank you for immediately taking my advice, lol.
Almost a month passed and finaly video came.
Another great video very helpfull too
Imo 10 widths are better if u can support it as if u have support arti/support equipment you can have more of it(example, u can have 2 support arti with 2 10 widths instead of 1 support arti with 1 20 width).
when taureor doesnt give hungary transilvania: happy romania noises
I think if you don't want to min-max to 100% just don't build smaller than 10 or larger than 44, and anything between 22 and 39. So basicly 10-21 and 40-44
great video, thank you for your service. If you were to repeat the experiment, i think starting with a larger normalized force of say 20w, and sticking to the same manpower would better. keeps the front line filled on larger widths and we get to see if having 4army groups of 10w is better than 2army groups of 20 etc
Thank you for your hoi4 service