THIS is META of Combat Width - HOI4 No Step Back Combat Width Guide

THIS is META of Combat Width - HOI4 No Step Back Combat Width Guide
Main Channel / davefeedbackgaming
Discord! / discord
Support me on Patreon! / feedbackgaming
Follow me on Twitter: / feedbackgaming
Follow me on Twitch: / feedbackgaming
-----------------------------------
FeedbackIRL is where feedback procrastinates with his hobbies and other interests. Warning REALLY BORING or REALLY INTERESTING

[@FeedbackIRL]

discord.com/channels/144006398743674881/685476336222994441/919286365403902012 Link to PDF for combat width meta. Think you got to me on my discord for that link to work. www.discord.gg/feedbackgaming

[@DSiren]

Why didn't this man model 60 widths???!!! One of my favorite templates was 15-10 inf-arty.

[@erwinvanleijenhorst7605]

It does feel like it confirms my first try in No Step Back as the Soviets. I had a lot of defensive 21 width divisions on the German front, not having enough equipment yet to start building 42 width divisions.. When they attacked I was absolutely thrashed by the attacking Teutonic hordes. No matter what I tried, I just kept seeing red dots as battle results. After I gave up on that game, I looked at these unstoppable German divisions and these were all 27 width (9 inf, 3 art).

[@aspielm759]

Another disadvantage of 10 width is Army size. As long as they’re not based off of manpower within them, it is going to be a change if the combat bonuses apply 24x 21 width or 24x 10 width

[@TheAndrew1987]

youre linking to the channel, not the file

[@mcmadow]

This guy really wrote a thesis over the combat with

[@blakedake19]

Shit got real when this dude decided to use Latex

[@killroy123]

Likely...instead of writing their actual Thesis. Lol

[@miguelrodriguezcimino1674]

Give the man his PhD

[@RipVanWhinkle]

@@killroy123 this was probably more interesting, plus it’s break finals we’re a week or two ago

[@warspite1995]

I for one am thankful!

[@ZinvictanGamer]

People using math to solve a math game? Preposterous!

[@dcred123]

Those bastards at paradox are secretly teachers! We're on to them!

[@wgoulding]

Dude, I wrote code to find the right proportion of factories to use for producing the equipment of my divisions.

[@dorusprime8037]

@@wgoulding where can i find this?

[@kurtshaw229]

In a way all games are math games

[@tiedeman39]

@@wgoulding Do you have a link to your code?

[@vincentschrama749]

Other disadvantage of 10width is you will need alot more generals and you will have more armies with low stat generals.

[@FeedbackIRL]

TRUE

[@Theo-vn9hm]

@@FeedbackIRL Also take into account the cost of support companies.

[@ElRealista.]

Exactly, I find more practical the 27 because of that and the suply out of Europe and Russia. Yes, you can build hubs but it takes too long to build them.

[@Rohit-tp7qx]

Best is 42 for attack 10 is only for costal defense or paratroopers

[@ElRealista.]

@@Rohit-tp7qx not really, there is a super complete thesis about that, and 10 is the most efficient... But I don't feel comfortable with it. Then 27 is the second best. 20 is one of the worst.

[@thomascrowther3203]

This man really made a dissertation over the hoi4 combat width meta... mad lad

[@infernosgaming8942]

His final project for his masters in Hoi4intology was analysing encirclements on the reddit

[@1Madlycat]

“For many small countries, however, it is very viable to build up a small military with many low width divisions.” I found this out the other day by accident when I played New Zealand. There’s a focus that lets you send volunteers with less divisions than normal, but you still have to spam out tiny divisions to get to the point where you have enough to send volunteers. I ended up shipping them to Africa, for whatever reason, and found that they were surprisingly effective.

[@stevepirie8130]

Bad supply regions perhaps suited small forces

[@DSiren]

@@stevepirie8130 dude all you need to make Ethiopia support 24 30 widths is upgrade 1 port and connect the railways to the supply zones. Ethiopia is STACKED

[@AHappyCub]

@@stevepirie8130 Central Africa is especially bad in terms of supplies, fought the Brits there as Fascist France and at one point no one can push, so i build multiple railroads and supply hubs to compensate that, hell i even manage to conquer British Raj from French Indochina by doing the same thing using only 24 Divs

[@ElRealista.]

@@AHappyCub and now we found 24 is not the best.

[@AHappyCub]

@@ElRealista. 24 as in the numbers not width

[@blakedake19]

10 width divisions is not really viable for multiple reasons. First, you "waste" a ton of support equipment and second, you do not have that many generals for all of your divisions. Remember that a good general, not an excellent one, just a good like a level 3 general with stats 3/3/3/3 gives a 15% bonus to many division's stats (without even considering general's traits). So, in fact, having a bigger division like 9 inf and 3/4 arty could be much much better than everything proposed here just because this guy did not take into account generals, traits and equipment cost.

[@dixonsider873]

you must not play MP cause 10 widths are usually banned because they're that good

[@blakedake19]

@@dixonsider873 I mean, the majority of MP games are modded. i was talking about vanilla here. Furthermore they are banned because vast amount of divisions can lag the game a lot

[@alexmannen1991]

10 width are probablu earlu pre 38 combat where u just have a support artu on it

[@zacharybray6200]

@@dixonsider873 they're banned because people think they're good, but they probably haven't even seen 15 or 27 width spam.

[@deathdealer312]

@@zacharybray6200 15 width is probably the best. 9 10ws have a 2% stack penalty. Anything above that though, is not as efficient.

[@Iamaplatypus42]

Wait, he made a research paper with LaTeX, on hoi4's combat width. This is a whole new level of dedication!

[@TalonAshlar]

There are two major problems with the suggested meta: 1) A disproportionate number of battles will be fought on marshes, mountains and hills regardless of actual regional terrain. 2) A disproportionate number of battles will be fought on river lines which reduce combat width by 25% The result is that this is a single player or mobile warfare meta once a front has been broken.

[@t_y8274]

So the testing probably needs to be done on a usual eastern front frontline, dday line etc.

[@TalonAshlar]

@@t_y8274 Exactly its a praeto inequality where 80% of battles will be fought on 20 % of terrain. Should probably add France and El Alamain as well

[@tanthedreamer]

@@TalonAshlar and the more detail and the more accurate you want it to be, the closer you are to the old conclusion: “there’re no universal best width, it depend on your nations and your front”

[@TalonAshlar]

@@tanthedreamer I disagree defensive lines are the same in virtually every historical game the only times that need specialised divisions are for taking islands, breaking el Alamein and possibly Norway and Greece

[@0rdnajela664]

Why would more battles be fought on mountains, marshes and rivers?

[@hedgeearthridge6807]

There's few things that excite me and impress me more, than math nerds taking little and obscure problems, calculating the best ways of solving these problems, writing it up in LaTeX, and publishing an entire thesis/lab report on it. Thank you guys for doing this, keep going!

[@orsobruno1980]

Paradox has the merit to overcomplicate things to the point that the META becomes an overly-simplified use of it, like sub spam, or corvette spam in Stellaris. Now let's welcome the 10 cw divisions spam.

[@artruisjoew5473]

Corvette spam has been dead for like a year now

[@orsobruno1980]

@@artruisjoew5473 Of course, the METAs change over time, this is also what patches are for, but this is through nerfing and buffing other things rather than changing and slightly simplifying the system. There was also an artillery spam META in Victoria 2 and they had to nerf the artillery. My point is that it would be better to keep things a little simpler, rather than making them more and more complex in order to make the game more "realistic" and as a result having the community playing unrealistic METAs, such as corvette, sub, or artillery spams.

[@GlobstersMessenger]

@@artruisjoew5473 and largely replaced with battleship spamming and adjusting your proportion of artillery ships (mega-cannon, neutron launchers + 1 kinetic artillery) to your number of carrier battleships (though still usually with an X-slot because they're so potent) to have insurance against corvette fleets. Of course the actual meta is to out-tech and out-produce enough that it becomes largely about efficiency late game, where the game is slow as shit unless you're on a tiny map. Paradox games never change I suppose.

[@cyrusol]

@@artruisjoew5473 For like... a couple years. Instead we have unadulterated arty battleship spam in Stellaris!

[@512TheWolf512]

HOI4 really needs something like special ranger divisions for fighting in forests and jungles

[@grandmagertrude6358]

There is in R56

[@apexinstinct]

There's a mod called special forces which adds desert infantry, jaeger forest infantry, shock infantry, jungle infantry and flamethrowers

[@noanyabizniz4333]

There were no rangers in WW2. They were created in Vietnam.

[@512TheWolf512]

@@noanyabizniz4333 that's just a relevant name that came to mind. I am sure there were specialized troops or specialized training for forest fighting during WW2, no matter what it was called

[@noanyabizniz4333]

@@512TheWolf512 No.

[@digitaljungle7154]

To add something, 18 widths have a slightly smaller penalty than the 27 width. This could mean that the default 18 width divisions are another alternative but still weaker than the 27 because of the stat difference

[@ufuker5754]

Paratroopers Man smaller less supply

[@ESG1]

In Grisha's example he also stacked infantry and special forces advisors. Thanks for also pointing out that the tanks do better on the attack.

[@FeedbackIRL]

TRUEEE

[@dixonsider873]

groundbreaking thought that tanks have better stats than infantry!!! I think the whole point of his video was that 10 widths don't perform that much worse plus they will allow you to dominate the air war

[@ESG1]

@@dixonsider873 I agree, I just don't want people to think that you have to go 10 width in every case.

[@tobiaswagner5496]

A few mistakes here . . . The first mistake is assuming every division should have the same unit width. Only a small number of provinces will make 40w to 45w not viable, that's 11% mountain and 2% marshland (discussed in 2nd paragraph). It's a bad strategy to adjust your entire army to account for 13% of combat situations. The second mistake, setting the 1,2,1 weights is not only subjective, but it calculates a single specific case and it actually converts all the values into one attack case where 2 zones attack 1 zone. Example, 1,2,1 weights applied to Hills with 80w base and 40w extra attacking from 1, 2, or 3 zones gives an expected value of 120w, which is the 2 zone to 1 zone case; its the same for all the others. 1,2,1 should have been variables to examine different attack patterns. Also, I rarely ever attack from only 1 province, 25% is way too high. And I always try to attack from multiple provinces so 50% for 2 provinces is way too low. 3rd (potential) mistake, adding a square on the penalty formula and the reason given was that the penalty affects attack and breakthrough. You can't just arbitrarily assign a square because it seems right. You would have to investigate battle mechanics to see if that's appropriate to do so. This might be ok, but might not be ok. And a final note, this calculation does not calculate optimal unit width. It's an attempt to minimize the size penalty. Minimizing the size penalty doesn't mean the unit does optimal damage to the enemy. Mountains were 11% of the provinces. ONLY EVER fight on mountains with specialized mountaineer units. This should not change how your standard armies look and only affect your mountain divisions. A quick look shows between 31w to 33w are good choices for mountain units. This should be calculated separately. Marshland accounts for 2% of all of the provinces. 2%. You don't alter every single army to account for 2% of possibilities. You handle marshland by encircling it. Don't include marshland in the calculation because it's insignificant in large scale. Now that mountains and marshes have been taken care of, run analysis on the rest. But again, minimizing the penalty doesn't optimize damage, so battle mechanics must be included before an optimal unit width can be proven. Knowing that larger units do damage more optimally than smaller units, 42w to 43w appear to be the best choices for your main armies. I'll take a look at the battle calculations and see how damage is calculated. If the optimal widths are different from about 32w for mountaineer and about 43w for everything else, I'll come back and edit this comment to give you a concrete answer. I'll also look into divisions with different sizes, such as 50w + 50w + 20w for a 120w battle. Intuitively it appears having them all the same size is optimal, but I'll double check and make sure.

[@RubberFacee]

Should try out Mass Assault doctrine too. You can stack more into the template hence better stats than other doctrines, in theory of course.

[@ruukinen]

But that works equally for all widths. In essence it's like adding a support company to the template. Higher cost, more stats, same width. For this discussion however it doesn't matter since it works out equally over all the combat widths.

[@CM-db5cg]

Yes but it's only infantry you can put in there. So it's only for the defensive, and only when manpower isn't a concern. So basically only when you're playing Russia or China (Warlords probably want superior firepower for max soft attack for border wars)

[@WaluigiIV]

So wait, you are staying that the historical soviet doctrine promotes the very historical soviet organizational structure which favors a large amount of small units rather than a small amount of large units? I like it.

[@ruukinen]

@@WaluigiIV I mean depends entirely on what you mean with 'units'. Lowering the width of infantry brigades means you can stuff more of them into a division. Thus you have fewer but larger divisions. On the other hand you can then make your divisions smaller with the same amount of firepower/manpower thus you can have more 'smaller' divisions. In either case the manpower and equipment cost is the same. The combat width is the only thing that changes.

[@TheMelnTeam]

@@ruukinen The one wrench mass assault does throw in is that you can set a preferred tactic that increases the width once you have it, and it's not hard to trigger that.

[@daedalusmedia]

If you bring divisions that are roughly historically sized you should end up around 26-27 width. My 27 width divisions have been working out well for me.

[@artruisjoew5473]

But why 27? I don’t see it working well for plain and forests. 27 seems like it is appeasing mountain and marsh tiles which isn’t very common for Europe.

[@daedalusmedia]

@@artruisjoew5473 Theory and practice. The 27 width works out well. They hit hard and balance the reinforce rate tends to effect most fights more than penalties for being over width.

[@haukionkannel]

The best size for plains are 30, so 27 is really close, so it work almost as well in plains. 27 also fits perfectly to the forest (3*27=81 vs 84) so there is a point in there!

[@ojdavies7699]

What do you use? 9,3 or something else?

[@papastalin531]

@@ojdavies7699 I personally use 9/3, it's a great division in my opinion, well rounded for everything, and that's what it's meant to be

[@Mar_Marine]

Glad to see my prediction a handful of videos about thinking 15 Width was a good starting spot! It's cheap enough that most any country can field it, while being strong enough to warrant training over the 10 width. I give myself a pat on the back for seeing this ahead of time!

[@pasanaator9874]

I found out that if making a tank and its reliability is greater than 100% then you can use the extra ammunition storage because it takes away reliability but doesnt add production cost.

[@FeedbackIRL]

Who's ready to smash the Soviets with 44 width infantry??

[@aussieosborne116]

Other way around ;))))

[@martinaustin6230]

What I been doing is using 21 and 41 width (I'll replace 21 with 27), and then make specialized units if I can't push, usually mountains. That has been working well for me.

[@JoeBurner1720]

9mot 8mot art is insane with mobile warfare, i’ve used it since the update and it’s cheaper than tanks, Also i guess i got lucky and it turned out to be 42 width

[@a2e5]

it's the Answer to the Ultimate Question of Life, the Universe, and Everything after all!

[@guzza5375]

Yes I agree. I used a 5/4 Motorised, Motorised Artillery template as Romania, and I was able to cut through the Russians like nothing.

[@Anton43218]

@@guzza5375 i am a romania exclusive player i have played 1k hours of romania,so how do you make mot divisions as romania?

[@thi.]

idk what you do but they cant even defeat normal 20 witdh inf division without being killed instantly

[@hedgeearthridge6807]

That's what iSorrowProductions has been doing, is motor/mech with Motorized Rocket Arty. He says the tanks are just too expensive and Katyushi are cheaper and do the job just fine. (If my Russian is correct, the plural of Katyusha is Katyushi. Be sure to bring that up at the dinner table)

[@anastasiolabropoulos]

I actually do enjoy looking at this and finding the small 18w peak before the 27w, since that is the standard division size of many majors. It could potentially be a small way of improving a new player experiences.

[@yulu803]

Now we need more analysis on how attacking from multiple tiles affect the average combat width/reinforce width and how that affect the weighting.

[@aspielm759]

I’m glad that I wasn’t completely off when choosing my new combat width. (I used 21, sometimes 27 and tanks were 42, I think) What I had realized is that my Italy playing friend accidentally therefor discovered the best combat width

[@lochnessmonster5149]

In all my NSB games, I've found that the German and Soviet AI exclusively use 27 width infantry divisions and 18 width armor divisions. The US and Uk will use 20 width infantry and 20 width armor. Everyone uses 8 width garrison divisions. Both the Germans and Soviets will retire all other templates.

[@tbeller80]

So if I'm reading the spreadsheet that goes with this report: 10w: Good for defense if you can afford that many of them. 27w: Good all around for everything but hills. 20 or 25w: hills and mountains I'm not seeing where the high marks of the 40s came from. It seems from the numbers that you'd always need an attack from more than one side in order to avoid exceeding combat width.

[@hotwaff]

My man did more research on HOI4 combat-width than anyone on Twitter who says they "do [their] own research" about literally anything.

[@VarenvelDarakus]

underlying subject is considering XP is now used for doctrines AND other things you preferably want to spend less xp then more , getting just 1 template of 27 with support companies is like ugh...80-100 xp who is equivalent of 1 doctrine , so if you want to have good inf , tanks and maybe 1 extra , thats equivalent of being 300xp behind of somone just using starting template with few touches , this is even more if you switch to medium tanks from light or heavy tanks , your again looking at ether getting spirit who makes them "free" for 50xp or spending a lot , there is so many other things to consider then "penalties" , not to say editing tanks also...eats your xp

[@quedtion_marks_kirby_modding]

I hope he does another test with 3d graphs (2 independent variables), my strategy for now is mixing 40s and 25s. 2 40s for hills, 3 25s for mountains, mixing them for deserts and plains, etc.

[@marcus7564]

I'm interested in the viability of the smaller divisions to reduce supply demands, especially for special forces. I also hope small special armored units can work in sp for role play. I want the small AT or heavy tank regiments.

[@ethancox4572]

Personally had done some similar math though mine was based off of efficiency of filling combat width (ie how close a width comes to full combat width without exceeding) as this would also mostly account for exceeding combat widths and got a similar conclusion to this person. I got 10w being best followed by 12 and 15 width divisions tied for second place. Given both what I did and this person it seems to depend entirely on resources to spend. If you can afford more equipment and generals for an equivalent amount of regiments, the smaller widths will be better. If you are unable to afford the extra cost for whatever reason (perhaps you only have one tank general) then the larger widths would be far more efficient.

[@Crusader-eh2cv]

When you write a paper in Latex on game stats and meta.....next level shizzle

[@ryankolbe365]

The crazy part is that 8-12 widths with a bunch of support companies is basically the modern brigade combat team. So wow maybe they could plug this change into a modern day hoi 5

[@jip5529]

One thing that should be taken into account when looking at tank divisions is that in a meta situation they are not used in marches, mountains, forest and urban for pushes. This means that when designing a tank meta there should only be looked at the better terrain types. This is a bit of personal preference but as we are speaking about the meta I think this should be taken into account.

[@Luredreier]

5:16 I assume that the assumtion made in this paper is that *all* units are of equal size? I however favour having two different sizes of the divisions. One that's 30 and one that's 24 *not* 25. I'd love to see the numbers for *mixing* different combat widths. That said, yeah, I think I'll probably be spamming a lot of 15 width too on some fronts.

[@monkeydog8681]

I've been rocking 15 width lately 6 infantry + 1 artillery (SPG preferred). I felt it was performing good lately nice to have a confirmation.

[@FeedbackIRL]

You do you. This video is about min-maxing.

[@nellowyt7417]

I tagged the German AI just to see what they were making because their units (infantry) always seem quite strong and they were using 27w so it is possible that 27 works well

[@mattevans5236]

Great paper, great video. The current paper approach appears to be to minimize the average over-stacking penalty for a region. The methodology appears to be to count the territories of a given type (hills, marsh, etc) for a given region of the world. The more frequently a territory type appears (e.g. there are more mountains than marshes), the higher its "weight" becomes in the analysis. E.g. if you have a region of the world with lots of mountains, the ideal combat widths for that region will strongly skew towards those that minimize the stacking penalty for mountain territories. Three additional levels of detail that might improve the analysis come to mind, and all have to do with how the weighting function works. My understanding is that in the current paper, the weight is simply a count of territory type ["territory type frequency"] This weighting function does not necessarily appropriately weight the actual battles, however. Firstly, not every territory has an identical number of bordering territories. The actual effective combat width, and therefore, the over-stacking penalty, is determined by the relationship between a territory and the specific territories that surround it. Suppose we have a territory of width 20, and which borders exactly one other territory. The combat width there will be 20 (i think), and to minimize stacking penalties you would probably want 20 width divisions, and the more you deviated from 20w, the worse life would be. But suppose you have a different 20w territory, but this one is bordered by 8 other territories, all of which you can attack from. The CW is now much larger. The math for stacking penalty and indeed the overall impact of stacking penalty I believe are quite different than the 20 vs 20 case. How many of the territories in a region have 1 neighbor for 8 (or any other number?) The weighting function could potentially be improved by counting the specific total combat widths that exist for each specific territory vs. its specific neighbors. Secondly, the above argument applies, but goes one step further, and considers the specific territories and alignments upon which battles are likely to actually happen. Since we have some control over where battles will take place, we should weight the territory CW analysis to the battles that we expect to actually have. The frequency of a territory by type and by count of neighbors in a generic sense is not as good of an estimate as a system of weights based on the specific battles that are expected to take place. E.g. suppose that a cw 20 territory is likely to be attacked, and has 8 neighbors, but in the real world, only 5 of them would ever be held by an attacker. This produces a new simulation CW and changes the weights of what the divisions will be expected to deal with. Thirdly, we move from the realm of static analysis to dynamic simulation. The first two options are simply analysis of map tiles. The third option I believe requires simulation. Consider a specific campaign: the german/ussr eastern front, from the POV of germany on the offense. The map tiles are what they are, and, the starting conditions are what _they_ are. If we consider the perspective of germany wanting to advance east ward, what _specific_ attacking vs defending CW scenarios will germany encounter? The answer very much differs from playthrough to playthrough, and depends on the specific crenellations/pockets of the front line that emerge as a given territory is gained. If one were to run this simulated east ward march thousands of times, what would the frequency of specific combat CWs be, from Germany's POV? Using _this_ as a weighting function would indicate which unit CWs would be minimize average CW penalty if one were building a german army to move east across Russia. Interestingly, one simulation run could be compared to another simulation run. Simulation might reveal that there are specific battle paths [sequence of territories attacked] that are especially advantageous, in terms of optimizing every battle for Germany's CW, whereas there might be other battle paths that are especially disadvantageous. The difference between best, worst, and average case battle CW frequency for a given sequence of territory attacks would shed some light on the maximum advantage that a player could gain by clever microing, and conversely, would also quantify how much difference CW might make (e.g. sensitivity analysis) in the overall scheme of the eastern front.

[@AppleSauceGamingChannel]

Much of that variance would come down to good strategy. Minimizing the encounters in unfavourable terrain for specific forces and maximizing the opposite. Essentially, a sound military strategy will come out on top if all else is equal.

[@mudadamuda294]

a downside of low width divisions is that they take up a lot more generals for example 24 20 width divions take 1 general those divisions have the same value as 48 10 width generals which take up 2 generals so if you have too many 10 widths you can potentially run out of good generals also 10 widths are way more micro intesive and cost you and arm and a leg worth of support equipment

[@randomperson8337]

The advantage 10 width infantry has is support artillery, for example two 10 widths have double the soft attack bonus from support artillery than one 20 width with support artillery, this way if you research the artillery tree and superior firepower you can get about 100 soft attack on ONE 10 width infantry

[@FeedbackIRL]

Good luck attacking

[@bodives]

high combat widths have an issue with supply, though, since you just cannot stack many divisions into a same tile or you run out of supplies

[@TricaGamer]

If a 40 width attacks 4 10 width now, the damage is divided by all four of them in no step back

[@nebojsag.5871]

Honestly, I don't bother much with exact widths, I have a priority list: -Good all-round combat stats. -Make them as small as possible to have more org overall -Make them still big enough to make support companies viable, and to avoid stacking penalty, which can go up to 70% if you use single battalions. Yes I tested that.

[@AG3624C]

I think you also need to bear in mind in 40 vs 4x10, that 40 is going to dump a full damage into one... which breaks then it's 40 vs 3x10, and so on, and so on. A 40 can typically hold the fight longer than smaller due to continuous depreciating combat power due to smaller units being broken more easily. On paper yes it works on smaller width, perhaps though in field 20s and 40s still have a place for the sheer tank ability.

[@artruisjoew5473]

27 seems really odd given the width, seems to get screwed with because of mountain and marshes. Would be more interesting to see the optimum by theater - what’s best for Europe, North Africa etc.

[@huntersorensen5000]

"eeeeh I don't know forty width is still good, like 14-4s man" -TommyKay Probably

[@aletoledo1]

I tried to upload a spreadsheet link, I think it got caught in youtubes spam folder. It shows that 27 and 28 are tied. 28 does better in plains and forests, while 27 does better in urban and hills. The problem with 41-45w numbers is that they are averaged. If you look at the performance of a 41w for example, it's perfect in forest, but horrible on plains. Vic versa for 45, it's perfect on plains but garbage in forests. The 27/28 is more of a middle ground.

[@Normalguy1690]

Basically 10 width inf stack to defend and hold the line with your tanks or motorised\mech divisions on the attack. So same meta basically before the update.

[@DTDEVO12]

When he said submit to reddit to get peer-reviewed I laughed pretty hard

[@dsjaks6983]

2 width meta confirmed??

[@spamfilter32]

In Chemistry they would call these the Islands of Stability. I like it. 10, 15, 27, 41-42 and 44

[@haukionkannel]

Well I have been using now 21, 42 for normal divisions and 35 for mountain divisions. It has bee working quite nicely. But 21 vs 25 is not big difference so maybe I put mountain units at 21 too... All in all when size does matter also, I think that 10 wide is too small even it is the most efficient size. Effiency is not everything, you have to think also stats. So we come back to 21 and 42 wide. But what is good in here is that they did manage to suffle combat width META really well! It is still impossible to tell what is the best choise! And that is good. Do you take the most effieciet or maby something that is not quite as efficient, but offer better stats! We need some one to run many simulations using different size units at different terrains. See the real result and compare those to the costs of producing those units!

[@CarlosGomez-tt4jh]

Lol and here I was using 10 width units just because they were cheaper than larger Units and took less supply (because I suck at the new logistics system) and in no way was trying to min-max anything xD

[@DarkAngel-vz3zx]

Mega helpful!

[@electricVGC]

10 defensive 42 offensive looking like the rule of thumb for single player - but with smaller fronts in SP you can really micro your width

[@ruukinen]

If you are germany your defensive units should be bigger than 10 widths since you have a limited amount of decent generals. Also the more armies you put on a line the bigger the chance that the frontline system fucks up somehow.

[@111tubefan]

Tank are best at fighting in lowlands, so you should adapt their combat width to plains's combat width. This paper seems as a great aproximation of the meta but I suspect that the meta will be more complicated than having just one combat width in your army per average terrain type of an area.

[@thevoxdeus]

It really depends how much you want to micro-manage your armor, or how much you already do so, in order to keep it out of common but unfavorable terrain like forests and hills.

[@CM-db5cg]

I'd disagree. Tanks are almost always going to kick ass on plains anyway, because all being fair a well made tank division will push a defensive division. You need to optimise for the situations where your tanks don't push well, the situations where because of the enemies terrain advantage you need to cram every brigade into the fight you can just to break through. So 21 width is the way to go, because the make or break moment for tanks is always going to be in forests. Plains you roll anyway, cities you can stack more in or go around, forests are where you NEED optimum width. Of course the sane applies to hills, but outside of Asia I find those to be just less common and generally near mountains, which, just don't, if you're trying to push mountains with tanks you're doing it wrong.

[@hunterwotruba774]

The meta depends on the nation you play and which theatre you play in. I did a Finland game the other day, and since you only fight in forest, 21 width had 0 combat penalty, except in Leningrad. I suggest holding ports with 32 wide, 30 wide tank, 21 or 30 infantry, 25 mountaineers. Tends to work quite well. The meta is going static on your combat width will punish you.

[@Vani_Vozrob]

What tank width did you use for Finland ?

[@TheWeedIsland]

why would you use mountaineers for port garrisons, esp when you're going to have little of them as Finland?

[@hunterwotruba774]

Not all of this is about Finland lmao

[@hunterwotruba774]

@@TheWeedIsland as finns, your ports are forest, just use the 21 width. I'm not using mountaineers as garrison, you misunderstood

[@odycz]

Support companies can be beneficial on 10w like support arty and racked arty. When you go superior firepower.Because they do not have combat width so you can have more stats in max combat width. i tested it last night as USSR against max out Germany now elite difficulty and I have 5to1 K/D ratio. But 42w are best If you want to go space marines whit one HTD

[@crowbarviking3890]

Was coordination and engagement width included in the calculations? Because smaller divisions are better to focus their firepower.

[@isaacrobinson9911]

Mayhaps they could make maps of effectiveness for select divisions? I don't even know how one would go about constructing such a graphic, but it could be super cool

[@TeleologicalConsistency]

Lower combat width is more effective but downside is more generals needed and more support equipment of all types needed.

[@spiderhouse9809]

Btw 27 from what i tested world conquest multiple infact it is Way better than any other in af as me eu na sa And for tanks gone for 43 41 and33 all did the best for me

[@emperorjosh559]

I've literally never played this game, nor have I any idea wtf comabt width is... but I love numbers and this logic is solid.

[@at0mic282]

10 widths use the equip for support companies way less effective... bigger template for flatlands rule supreme and assigning different templates to different regions will need micro management but would be optimal... what paradox wanted

[@robertalaverdov8147]

After a lot of games I've come to the conclusion that your combat width depends on the terrain you're fighting inn + your industrial capacity to create/support said divisions. I know, shocking! As the Soviets I've been using 45 width and it's been working ridiculously well on the river line, which is mostly plains. I've also tried 30, 27, 21, 15 and 10 and they're ok for defense but not as great on the offense. 10 and 15 are bad with low air support since you have less HP and CAS bombing will hurt more, thus more equipment lost. On top of needing way more support equipment. 42 is decent as well, maybe the best all arounder for Germany, especially when you're fighting in the forests of northern Russia. And as the Soviets it might be sensible to switch to 42 as your reach the forests on the Polish German border. But even then combat width penalty for forests vs plains is minimal for the extra width vs being dominant on plains. Mountains and swamps are definitely a challenge. Although why the hell are you fighting in that terrain, go around! Final thing with the Soviets is that they have access to masterful Blitz which gives +10% combat width when using tanks or generals with +1 skill overmatch. So plains become 99 width, forests 92, hills 88. Allowing you to use a 45 width with minimal issues. With all that being said if you're playing Japan the best width might be 27 as it fits into mountain 75, hills 80 and jungle 84 with the least amount of leftover space and combat width penalty. The most common terrains in Asia, plus the supply in China/Indo is hard to come by and bigger divisions just can't be supplied. I've also done well with 30 width as the UK or Italy in North Africa as it's mostly desert. Last thing; the 10 width meta, aside it's weakness to CAS and the quadruple cost for support companies. You'll never have enough trained generals to make the most use of it as a major. A level 6 general with panzer/infantry expert gives a division 40% better stats than a level 3 general. Though it might have some use for Minors with low manpower and decent factories who can collaborate with Majors on certain fronts. Aka Hungary, Romania, Canada and South Africa to name a few. That's just my two cents, with what I've seen work in MP and SP.

[@unbindingfloyd]

I've been using 21 as my all around since launch but ima give 27 a try.

[@MaFo82]

Unless I'm mistaken the historical german infantry division is 27-wide if created in the HoI 4 divsion designer. It could be coincidence but I don't think it is.

[@Xinamon98]

If you wanted a historical template it would be: 9/3/1 at with engineers, recon, hospital and signal. If you wanted one with correct equipment it would be: 11/1/2 at with engineers, recon, hospital, signal and support art.

[@juvandy]

Thanks for this, very interesting. Unrelated question: in the tank designer do you now have to research AT guns to upgrade the high velocity guns in tanks? If so, is there any reason now to focus production on AT guns at all or does the AI still just not really build many tank divisions? I'm mainly thinking from a perspective of playing as a minor on the defensive against a bigger power. The meta of encircling enemy armor with your own armor optimised for soft attack makes sense... if you can afford tanks yourself and can build up offensive capability.

[@FeedbackIRL]

Yes

[@stevepirie8130]

Most single player play throughs I see the AI doesn’t replace armoured divisions. Perhaps earlier templates would benefit more from piercing and hard attack but mid to late game you could go more into soft attack to melt AI infantry.

[@juvandy]

@@stevepirie8130 early game AA also has enough piercing to defend against tanks too, at least before NSB. It makes sense historically too given early AT guns were basically rifles up to 20 mm.

[@artruisjoew5473]

@@stevepirie8130 no no. As I learnt the hard way, AI build tanks now. First game after the DLC my dumbass thought not much changed. Was a nasty surprise when the Italians showed up with tanks in North Africa.

[@joaovilaca1436]

I build a infantry division for each type of terrain in fighting in. Of course, SP.

[@robpeacock6287]

While the content of this is very good & insightful #THANKYOU FEEDBACK IRL i would love to see a video on what divisions the AI spams for each country as the war makes progress. For example i have noticed Japan spams more light tanks divisions/brigades aganist China.

[@154Kilroy]

I had been using 24 width mountain divisions, and 30 width infantry. The mountain troops were surprisingly ineffective. And that seems to line up with the data. 30 width tanks and inf seemed to be OK. Apparently 27 is the way to go. Edit: 25 width mountaineers aren't possible without adding artillery or something like that, which screws up their terrain stats defeating the purpose of mountaineers. I suspect the 15 width might be good for them.

[@pocketgroyper9301]

I think the minor loss in terrain bonus is worth the massive increase in soft attack a couple of artillery give you.

[@buysideprussian]

They even used a LaTeX font for the analysis. Impressive

[@taro9289]

Correct me If I’m wrong, the new meta is small country’s= 10 and 15 with(27 if affordable) and for Mp= 40 and 45 with right?

[@Strait_Raider]

I'm going to maintain that 45-width is the superior general offensive division. In general you will try to focus your breakthroughs on areas with the least attack penalties (plains), so the ability to fit perfectly in will maximize your power right where you want it most. On average it loses out very slightly to 41, 42, and 44 (1-2% only) but in exchange it will also be marginally more efficient in raw combat strength and support resources. As a sidenote, does having huge numbers of 10-width divisions cause noticeable lag?

[@fanis1414]

Isn't that kind of backwards thinking though? As you said plains have the least attack penalties so wouldn't it make sense to NOT specialize your divisions for them since you'll probably be able to push them anyways and try instead to make your tanks fit a bit better in forests/marshes/hills where you will be having trouble pushing?

[@Strait_Raider]

@@fanis1414 When you are on the offensive, you get to choose where the fight happens. You choose to break through where the line is most vulnerable (typically plains provinces), while the defenders holding better terrain are forced to fall back or risk being encircled. If the enemy is weak enough that you can push a wide-front offensive across, and/or you are not going to micro individual attacks, then 41/42/44 may make more sense than 45-width but again, only by 1-2% on average, with some fraction of a percent gained back in raw efficiency. Put simply, your offensive divisions should avoid rough terrain, so why optimize them around it? As the article mentions, the presented numbers are most relevant to defensive divisions, who have less initiative to choose their terrain. Although... in the defensive posture your logic might make more sense. If the enemy is most likely to attack into your plains provinces, a division best suited to plains defence would help cover your weakest areas. The 27-width is very handy here, dividing perfectly into 135 (a plains province attacked from two sides), and likely is one of the reasons it has such an overall efficiency spike.

[@jimmysundberg2376]

It is call reinforced meme as the hole stack being push out

[@dakelpoandgames2593]

I have been using 10 and 27 widths since the update and have not lost a single game on hard yet. Day 1 me and my friend were messing around and he said lol what if i use 9/3s (27 widths). I said ill use 10 bc im a loser. We dominated an ai game on hard that we normally lose. So yes, i agree with this 100%.

[@silverhost9782]

Found that 25 width infantry with a heavy tank battalion will completely break the AI in my last run, you don't really need to worry about combat width that much tbh

[@FeedbackIRL]

True. Min-maxing could've made your attacks 30% more effective

[@silverhost9782]

@@FeedbackIRL I'm thinking about doing it again on veteran difficulty so i'll switch to 27 widths for that. Honestly seems like CAS fixes any inefficiencies anyway though🤣

[@ParanoidMaster]

@@silverhost9782 well you can smash the braindead AI with literally any width division. Thats not the piont. The discussinon is more about whats better when you fight someone who actually knows what he is doing ;)

[@Aiphiae]

@@ParanoidMaster Braindead AI always smashes me :(

[@CesarAugustoRocabadoVillegas]

The 10W would be effective if you go superior firepower and have a lot of the best support companies. It would need a good economy tho.

[@Scy]

I just have to say infantry division of 3 brigades of 3 infantry+1 artillery... plus a support arty and engineer... its plain OP. 27 width for those wondering.

[@jeremyundsoo1872]

rly interesting i rly often think about the combination of combw div it would be so nice if sombody calculat it i am to lazy :D, probtly one is 15 and the other 27 or somelike this becaus you need a straight and a non_straight nummber also how it look if you atk from 2 sides. but yey he get me there too :D best wishes from germany

[@dakelpoandgames2593]

9 infantry and 3 art. I have been using it since day 1 of the update. That's 27 width and it's amazing. The hoi4 subreddit shit on me for it when i tried recommending and i saw Tommyk use them too. So now i think people will listen and use them.

[@kevindoyle1884]

My man has made a bigger pdf than the project for my engineering class😅😅

[@magni5648]

SO what would you say would be good division widths for Japan, considering the amount of fighting in hills, jungles and naval landings they tend to get into?

[@micheal6898]

30 and 25 widths are still good because they fit the combat widths better on mountains 3 25 withs will beat a 2 44 widths because only one of them can get into battle . so basically there is no meta to combat withs anymore it just depends , there are some that are better than others but it all depends on terrain and nation size

[@Bal1]

8:30 what about the 18 widths, that looks good, actually better than 27

[@adamhorridge1100]

if you make a division that is all infantry at 41 width verses mix of infantry and arty at 41?

[@notabyzantophile]

The way that attack and defense retroactively influence each other makes larger divisions better anyway because a single large attack vs many small defenses that counter the attack less because the attack is larger make the larger attack better, this is also why coordination is good

[@0witw047]

With the way targeting works now, large divisions don’t really make “a single large attack”

[@sudo4598]

except that the one single attack is spread over the many defenses so, no it isn't. Coordination is low, and while it definitely does factor into the equation, this doesn't mean your 1400 SA division is ever going to do 1400 SA to just one division. It's never going to be 100%.

[@notabyzantophile]

@@sudo4598 yeah I forgot they changed the game mechanics in this latest patch, but beefy 40 width divisions could just stomp all over 20 width ones in the previous versions of the game, but the whole defense thing still counts, you would take less casualties I guess.

[@FishyNipples]

Imagine if this guy put this same effort into something that could help him in life lmao

[@HeyHeyHayden]

Given that 27 width seems to be a good all-rounder, what do you guys recommend I put in the infantry division? I've tried 9 inf 3 arty but ive heard people have good stats with AA in their divisions.

[@goldgreatdane1]

thats kind of what im thinking. I dont usually build SPAA or towed AA so no idea there. I dont play a lot of multiplayer though so no strong opinions on it

[@HeyHeyHayden]

@@goldgreatdane1 I primarily play singleplayer, so i'm more focused on a division to beat A.I.

[@bennettflynn5278]

I still think combat width, at least in single player, is much better managed by theater specific templates with a greater emphasis on stats rather than nailing the PERFECT combat width

[@bodhisfattva7462]

i thought damage was now divided between all divisions in the fight now so 4x10w would outperform a 40w if the same support companies were used...that is assuming there is no issue with reinforcement.

[@12gark]

This guy did an amazing work. The only issue I have with it, is that he's averaging terrain types with zero human intervention: for example, a decent player would micro some bad terrain areas (avoiding mountains or marshes for example), but at least for battle planning, this document seems very, very, very well made. 10 widths imho have another big issue (on top of losses): you need a large number of generals and field Marshalls to keep everything under control. But for low supply areas they are great. So maybe a good mix of 10, 27 and 42W may be the way to go. I'm happy because I got the 27 and the 42. I'm unhappy because I thought that the 21W was the great and it's not, but I thought about the mixed width, and the 21W was good especially in that scenario, together with 18W, so maybe I can still be happy 😂

[@dakelpoandgames2593]

My friend guess 9/3's on day one and we started using them. 27 width and 10 width is just op right now.

[@12gark]

@@dakelpoandgames2593 and if you make the classic 40W 13/7 tank division, just swap 2 tanks for TD, and you have the same monster division you were used to.

[@fishpig4391]

I'm new to the game and have a question: why is it that 42 width is very effective but 21 width not so much? I thought that since it was a multiple would, therefore "fit in" to the more effective combat width.

[@thatonekid6576]

I love how feedback actually reads comments lmfao

[@themarcy4517]

Guy just wrote latex thesis in hoi4...love it

[@Luredreier]

10:09 Noticed the part about the simplification at the end about him not considering combinations of different combat widths? Yeah, that's where I think this approach is flawed.

[@stephenandrzejewski6765]

Where did you publish this white paper? Was this part of your thesis? Also, seems to ignore Engagement Width.

[@serg5136]

Unless I’m forgetting something, don’t big divisions split up their damage over several small divisions now? I haven’t heard that mentioned (unless I’m deaf). Isn’t this important now? Please correct me if I’m wrong.

[@wp12mv]

45 width together with 27 width combines really well into the different terrain!!!!!

[@Andromeda_Ada]

What you didn't mention is that 18w is more effective than 27w. What this means is that 18w is viable to hold lines. So, as a Major, what i would do is to use 18w to hold the lines (they are cheaper for support Companies too), 44w tanks to make breakthroughs and 27w Motorised to exploit those holes and to create a massive frontline, wich should be hard to mann by the enemy. This would also make it easy to get encirclements (especially for lone divisions trying to hold tge Frontline)