Late game modelling. Might it be improved?

[stockwellpete]

I wasn't quite sure what to call this post, but in my last 2 games against the AI (at Veteran level easy, and Veteran level default), where I have played as Central Powers, I have managed to get Stalemates both times. In this latest game, I think I would have lost had I continued. A-H was down to about 15% National Morale and Germany was at about 38%, while Turkey was knocked out; Bulgaria was at about 95%. The UK was at 70%, France at 45% and Russia was reinvigorated at about 45% too after defeating the Ottomans. The USA had yet to join the war. I was going to knock Italy out and I would have got a NM boost from that.

But when the game ended and all the units were revealed, there was (again) a solid mass of 6 rows of British and French troops waiting to attack on the Western Front. In the game I had modified the Artillery to represent just the heavier guns. So the number of Artillery units was halved, the number of shells they had as a maximum was halved, and the range was increased to 3 hexes to enable counter-battery fire. This seemed to work very well and the Germans were able to hold their lines, despite very heavy losses to themselves and the French (this is why both German and French morale was around 40%).

These 6 rows of British and French units waiting to rotate into the front line seem extraordinary to me. And they make US entry into the war rather pointless. In Stevenson's book on WW1 he makes the point that all the combatant nations (except the USA) were facing war weariness to some extent by 1917 and they were struggling to maintain recruitment levels. There were serious mutinies and the "Spanish flu" also started to make an impact in 1918.

I need to get some hard figures from Stevenson to make my argument stronger, but I am just wondering about the late war modelling here. For example, when a unit is "destroyed" it can be "re-purchased" at about 60% cost so it can return to frontline duties relatively quickly. What if this couldn't happen from 1917 onwards to represent war weariness? If you wanted to replace a particular unit then you would have to pay full price for it. Maybe the USA would be exempt from this rule as they joined the war much later? There may be other things, such as making the impact of the mutinies more severe, but I would need to research that issue properly.

[The Land]

Don't forget on high difficulty levels the AI gets not only a big industry bonus, but stacks of free units. The 'simulation' quality of the game really breaks down to increase the challenge level for the player.

[stockwellpete]

ORIGINAL: The Land

Don't forget on high difficulty levels the AI gets not only a big industry bonus, but stacks of free units. The 'simulation' quality of the game really breaks down to increase the challenge level for the player.

OK, does anyone know a bit more detail about how the game allocates resources to the AI as you progress up through the levels? I started playing at Intermediate and so far have usually achieved Stalemate as Central Powers and Victory as Entente. I lost a couple of Central Power games.

Notwithstanding the fact that the AI has to have certain bonuses as the Difficulty level is increased, are there any more historical factors that can be modelled into the game to represent war weariness? For example, the maximum number of units a particular nation can have might start to taper down after 1916.

The other thing is, at the end of my game against the AI, the Entente Powers had huge amounts of MPP's unused, so they could have built far more units if they had not reached the maximums allowed already.

What happens in MP? Do players generally use all their MPP's each turn, maybe holding back a small kitty for each nation to deal with emergencies?

[The Land]

ORIGINAL: stockwellpete

The other thing is, at the end of my game against the AI, the Entente Powers had huge amounts of MPP's unused, so they could have built far more units if they had not reached the maximums allowed already.

So in the difficulty settings the AI gets +10% of +20% MPP. I suspect that actually has a much larger effect than advertised, so to speak. Why? Because each country already labours under a MPP penalty that is as high at 70% for the Russians. So a 20% bonus to the Russians increases the effective MPP rate from 30% to 50%. The effect is less pronounced for Britian and France who have a lower initial penalty but will still be pretty big.

For the "AI Helper" units you'd need to look in the Units event file, but from memory there are a good few of them.

[ThisEndUp]

Not for me. Unless you're referring to a smaller power like Serbia, Italy, or the Ottomans, there really isn't a need for it, since the per turn income is more than enough for most expenses. Maybe in anticipation of large expenditures, such as for infantry weapon upgrades, it might make some sense. But otherwise no.

[stockwellpete]

I am just starting to look at this in a bit more detail using David Stevenson's excellent "!914-18". In Chapter 8 called "Manpower and Morale" he gives the following figures . . .

France - its army on the Western Front peaked at 2.234 million men in July 1916. By October 1917 it had dropped to 1.888 million men.

Britain - the BEF peaked at 1.801 million men in October 1917 and then began to fall after that.

Russia - the figures are less clear but Russia mobilised 4.5 million men in 1914, suffered great losses in 1915 and by late 1916 its losses were 5.5 million.

Italy - no numbers given beyond the assertion that Italy reached its peak at about the same time as Britain.


Germany - its army peaked at 5.8 million men in mid 1917 and then declined to 4.9 million men in mid 1918.

Austria-Hungary - no figures given beyond the assertion that it reached its peak size in 1915.

Ottoman Turkey - its army peaked at 0.8 million men in early 1916 and by March 1918 had collapsed to just 200,000 men.


Stevenson says that by Spring 1917 most of the combatant nations had reached their maximum efforts on recruitment and after this needed to substitute firepower for men, by which he means that technological advances in weaponry were needed to offset the decline in numbers and quality of new recruits. He writes,

"Neither side could sustain the peak of mobilisation attained in 1916 . . . they needed to pause . . . in 1917 all three main armies on the western front began to shrink . . . following the Germans' lead, first the French and then the British reduced the number of batallions in each division. Although trying to compensate through increased firepower, one army after another converted from an offensive to a defensive posture in response to changed strategic priorities, shortage of men, and ebbing morale."

David Stevenson, Chapter 12 - "The Third Phase Spring 1917 - Autumn 1918" (pp297-8).

[Dazo]

@ stockwellpete:

One thing you can try is change the XP level of AI units while staying at intermediate difficulty level.
That way you won't have rows of units but AI ones will be harder to kill, max xp (+2) is a bit too much especially if you play Entente as Germans already have more XP at start but +1 makes it quite good in my opinion as you'll really have to focus your attacks while AI offensive are clearly more threatening. Trenches are an absolute must when giving AI more XP as usual solo player strategies arn't working or at least not as well as before .

[Platoonist]

ORIGINAL: stockwellpete



But when the game ended and all the units were revealed, there was (again) a solid mass of 6 rows of British and French troops waiting to attack on the Western Front. In the game I had modified the Artillery to represent just the heavier guns. So the number of Artillery units was halved, the number of shells they had as a maximum was halved, and the range was increased to 3 hexes to enable counter-battery fire. This seemed to work very well and the Germans were able to hold their lines, despite very heavy losses to themselves and the French (this is why both German and French morale was around 40%).

These 6 rows of British and French units waiting to rotate into the front line seem extraordinary to me. And they make US entry into the war rather pointless. In Stevenson's book on WW1 he makes the point that all the combatant nations (except the USA) were facing war weariness to some extent by 1917 and they were struggling to maintain recruitment levels. There were serious mutinies and the "Spanish flu" also started to make an impact in 1918.

I has the same thoughts in my game when as the Central Powers I achieved victory in May 1918. I turned off the fog of war and found myself staring at what looked like the hosts of Mordor aligned against me on the Western Front and wondered how I was winning this thing. This was a game on the normal setting with no advantage for the AI.

I have a feeling the one unit stacking limit may be to blame here too. There's too many units and not enough front.

[stockwellpete]

ORIGINAL: Platoonist

I have a feeling the one unit stacking limit may be to blame here too. There's too many units and not enough front.

Yes, that is the issue I guess. I am not sure how many soldiers the Infantry/Cavalry Corps and Detachments are actually meant to represent (roughly). Stevenson gives the following numbers where an army Corps usually consisted of 2 Divisions . . .

Germany - a division was 17,500 soldiers, 72 field artillery and 24 machine guns, so a German Corps was around 35,000 men

France - a division was 15,000 soldiers, 36 field artillery and 24 machine guns, so a French Corps was around 30,000 men

Britain - a division was 18,073 soldiers, 76 field artillery and 24 machine guns, so a British Corps was around 36,000 men

Presumably Detachments are roughly the same size as Brigades, are they? 4-500 soldiers? Or are they Regiments at 2-3,000 men?

Note: These numbers declined as the war progressed into 1917/18. USA divisions were 28,000 men strong.

Are these numbers that are being used in the game?

[stockwellpete]

ORIGINAL: Dazo

@ stockwellpete:

One thing you can try is change the XP level of AI units while staying at intermediate difficulty level.
That way you won't have rows of units but AI ones will be harder to kill, max xp (+2) is a bit too much especially if you play Entente as Germans already have more XP at start but +1 makes it quite good in my opinion as you'll really have to focus your attacks while AI offensive are clearly more threatening. Trenches are an absolute must when giving AI more XP as usual solo player strategies arn't working or at least not as well as before .

How would I do that please? I want to play as Central Powers versus Entente.

[BillRunacre]

ORIGINAL: stockwellpete
How would I do that please? I want to play as Central Powers versus Entente.

You can amend the settings in the Options screen when playing the AI, as the MPP, experience and spotting bonuses the AI might receive can be raised or lowered independently of setting the difficulty level. This is really there to allow more fine tuning.

[BillRunacre]

ORIGINAL: stockwellpete

ORIGINAL: Platoonist

I have a feeling the one unit stacking limit may be to blame here too. There's too many units and not enough front.

Yes, that is the issue I guess. I am not sure how many soldiers the Infantry/Cavalry Corps and Detachments are actually meant to represent (roughly). Stevenson gives the following numbers where an army Corps usually consisted of 2 Divisions . . .

Germany - a division was 17,500 soldiers, 72 field artillery and 24 machine guns, so a German Corps was around 35,000 men

France - a division was 15,000 soldiers, 36 field artillery and 24 machine guns, so a French Corps was around 30,000 men

Britain - a division was 18,073 soldiers, 76 field artillery and 24 machine guns, so a British Corps was around 36,000 men

Presumably Detachments are roughly the same size as Brigades, are they? 4-500 soldiers? Or are they Regiments at 2-3,000 men?

Note: These numbers declined as the war progressed into 1917/18. USA divisions were 28,000 men strong.

Are these numbers that are being used in the game?

Essentially yes, but in a more abstracted way. Detachments appeared in the Orders of Battle right from the start of the war for a number of countries, and appear to have generally been larger than Brigades. But unlike Corps I'm not aware of an official structure/OOB for them, i.e. troops might be grouped together to form a Detachment to carry out a task such as guarding a location.

[stockwellpete]

I mentioned the "Apocalypse" documentary series in another thread. It talks about the "Spanish flu" epidemic and its impact in 1918. It gives the following figures . . .

UK - 200,000 affected
France - 400,000
Germany - 500,000

It does not give numbers for other countries but the USA must have been hit very hard too. Presumably, I will be able to find numbers for other combatants.

[stockwellpete]

Short excerpt on the impact of Spanish flu on German and US armies in 1918 . . .

https://www.bing.com/videos/search?q=sp ... &FORM=VIRE

[Chernobyl]

That's a good point. Is there no Spanish Flu event? There are typhus epidemic events so maybe there should be a Spanish Flu event in late 1918 seeing as it's a major historical event of the period.

[stockwellpete]

ORIGINAL: Chernobyl

That's a good point. Is there no Spanish Flu event? There are typhus epidemic events so maybe there should be a Spanish Flu event in late 1918 seeing as it's a major historical event of the period.

It was mentioned in another thread recently but I cannot find which one at the moment. The mechanism for having a Spanish flu event (two actually, in Spring and then the Autumn of 1918) is already in the game (e.g. typhus, mutinies) so it would be straightforward to introduce. The more that I read about the later stages of the war the more the various countries resemble "punch-drunk" fighters who are desperately hoping for a knock-out so the fighting can stop. According to Part 4 of that "Apocalypse" documentary I put up yesterday the French were incapable of mounting a major offensive for much of 1917 even though the first American soldiers had arrived in France, their commander Pershing didn't want to put his soldiers under Allied command or get them involved in trench warfare. Instead he was preparing them for a first offensive in 1919 based on a "war of movement".

I do think the general readiness value of units in all the war weary nations needs to be looked at as the war progresses. Their recovery should perhaps take a bit longer. Many countries faced severe manpower shortages so perhaps their ability to repurchase "destroyed" units for 135 MPP instead of the 225 MPP (and longer time frame to recruit them) should be curtailed. Instead of having a campaign map full of units at 10 strength points (and some with elite reinforcements) it should be more usual to see units with 7 or 8 strength points and lower readiness values just seeking to hold their positions than attack.

My comments are based on what I am seeing in games against the AI at Veteran level rather than what might be happening in MP.

[mdsmall]

The Strategic Command game system takes into account many factors, but the devs have chosen (up till now) not to introduce further limits on strategic resources, such as manpower or oil. So the only strategic constraints to production in the game are the MPP limits and the build limits per unit type for each Major.

For all the reasons cited above by stockwellpete, I think it would make a big difference to the WW1 game, especially in 1917-18, to start to introduce manpower constraints on new or rebuilt units. It would make the game not only more realistic but also more interesting to play. It would certainly make the arrival of fresh U.S. troops that much more significant a factor on the Western Front.

In theory, there are lots of different ways that manpower constraints could be introduced into the game once a manpower threshold had been reached. This could include:
- Increasing the MPP cost of reinforcements for existing units
- Reducing the maximum amount which units can be reinforced to levels below 10, even if fully supplied
- Increasing impact on unit morale of strength point losses and reducing the morale benefit of new replacement points added
- Increasing the MPP cost of rebuilding destroyed units
- Raising the supply threshold below which destroyed units can only be rebuilt at full cost (i.e. above 5 supply)
- Starting rebuilt units at a lower initial morale level
- Lengthening the production time for rebuilt or new units

An interesting way of triggering some of these measures would be to introduce for each Major a "whole campaign" build limit for infantry and cavalry units. This limit would track the total number of units built or rebuilt over the duration of the game, and not just the number on the board or in production at any given time. Once a Major had reached its campaign build limit, new conscription measures would have to introduced via a DE to increase that build limit, which would come with significant loss of National Morale, increased risk of demonstrations and strikes, along with several of the above constraints on building or rebuilding units. If the increased build limit was reached, there could be further rounds of conscription proposed in subsequent Decision Events (similar to the DE for expending MPPs to raise National Morale), triggering even collateral bigger effects.

I'd be interested to hear other players views on these ideas - especially people who have designed or played mods in any of the Strategic Command games where manpower constraints were a factor.

[Chernobyl]

In my games against the computer I've never seen a front build up like that. I am too busy smashing units with artillery and advancing. This is with +20% MPP for the computer, but not +20% experience, so maybe if you're playing with extra AI unit exp then they might hold out longer?

But the main question I have for you is: how many corps did you destroy by what date? I have a save game against the Entente AI where it's early May 1916 and I've destroyed 145 regular corps (not counting cavalry, hq, artillery, detachments, mountain, ships, etc). Their front is completely busted in Italy France and Russia. They don't have enough units to form a coherent line anywhere except the Caucasus (the ai tends to send too many units to the caucasus for some reason)

I do think there is a bit of "MPP bloat" and an issue where particularly Russia can elect to sit back and tech up with their 1600 max tech pool which bizarrely is the same as England and Germany. And their Industry tech which gives a massive +125 MPP per turn (compare to the second best in the game Germany which gets roughly +65 per turn)...

In addition to changes to Russia's technological and economic potential (reduce the max Russia can spend on tech to something closer to Austria)... well I won't go into more detail than that cause it's kind of outside the scope of this thread.

But one change I'm implementing in my personal scenario/mod that I think applies to the issues raised in this thread: increasing the global buyback time for Corps/Detachments/etc.

Not sure about increasing buyback MPP cost but the time is ridiculously short. You get these corps back way too quickly (it's practically instant) and there's very little early pressure for certain nations to buy early to get "ahead of the losses curve" if you will.

Problem #1: The quick buyback time punishes attackers who destroy units and take losses to do so. Yes you get a NM swing for killing a unit, but this is only a long term advantage. In the short term the benefits of taking say 8 total damage to 3 corps for killing one enemy corps is only 47MPP (135 for their buyback - 88 for your repairs). For certain nations with weak economies and low NM pools (Ottomans, Italy) such a tradeoff isn't even necessarily advantageous at all. These nations desperately need their MPP as early as possible for technology and overpowered weapons (artillery) and it's often disadvantageous for them to trade even at positive ratio. Moreover, spending full movement points on 3 corps to repair is arguably more costly in momentum than is losing a corps for 2 turns awaiting buyback.

Problem #2: I think an even bigger problem is that it (the low cost in time and possibly also MPP to buyback a unit) contributes to Russia (and sometimes France England or others) to simply turtle and tech up. To illustrate I will elaborate on Russia: It's unclear to me what the advantage is for Russia to press hard in 1914/15 with their units on any front except the Austria/Lemberg area and the Caucasus if there's an opening. They have the greatest advantage in terms of quantity of units anywhere on the map, the potential to throw their forces in bulk to stretch the capability of Austrian/German defenses and force German corps to early rail transport to the East, and yet the current optimal strategy for Russia may very well be for Russia to turtle and tech, using your numerous corps as meat shield timewasters not to be bought back until their research is maxed out at 1600/1600 and their artillery is purchased. Yes, Russia who has the greatest numbers advantage also has the greatest potential for economic growth and their tech growth potential is quite high too. Due mostly to their advantage in numbers, but also partially to the fact that buybacks are cheap and quick, there's little urgency to purchase back lost Russian corps until their numbers approach double digits. Combine this with a passive/turtle strat where Russia is not attacking into Germany at all, and they simply aren't losing many corps. Germany doesn't have enough spare forces to push towards Warsaw until early 1915 and against a shrewd opponent who retreats only the most exposed Russian corps every turn, Germany won't do much damage until mid 1915. By which point Russia has made significant tech/economic progress. Russia's MPP per turn actually surpasses Germany's after four Industry tech levels (I am aware there is a tech speed penalty for Russia/Austria etc and some cost increases for Russia but the point isn't that Russia has slightly worse tech research than Germany, it's that Russia's tech ability is almost as good as Germany's while her economic growth far outclasses everyone else in the game (+125 per turn compare to say Austria who get about +40-45 MPP per Industry Tech advancement)).

Worst of all, in the case of Russia, there is NO COUNTER to this strategy. The Central Powers might breathe easier due to fewer Russian advances in the early game, but they also have no way of stopping the proportionally stronger Russian tech/economic growth from such a turtle strategy. The Entente simply has a unit advantage that they can cash in to buy time and there's no way for Germany to throw an early money wrench into this strategy. Germany can defeat Serbia and have more units in France, but Russia's tech progress won't be slowed by a single MPP until you kill say 8 or more Russian corps (which is far too late to put a dent in their research, not to mention the costs and risks involved in maximizing early Russian losses).

This problem applies most strongly to Russia because of their superior numbers, but also applies to the other powers to a lesser extent. As Germany or Austria I often judge (correctly) that I can wait an extra turn or two to click buyback on lost corps because they arrive so damn quickly.

Okay so as I said that all probably deserves its own thread, and multiple changes as I have alluded to, but the point relevant to this thread is that it relates to our discussion about problems with late game unit/MPP bloat and buyback costs and time:

For starters, I think one easy change that would improve the game is increase the buyback time for Detachments/Corps/etc* from 60% of 3 turns to something like 100% of 5 turns. It would force tougher and more interesting economic dilemmas (try planning ahead what your front will look like 5 turns from now instead of 2) and partially weaken the ahistorical turtle/techup strategy.

*Also probably increase the buyback time of ANZAC corps by quite a bit because they come from the other side of the world!

[USGrant1962]

When I looked at Platoonist's screenshot, the BEF seemed way to big to me. I looked in the editor and both France and UK have a corps build limit of 24 in this scenario. Looking at the 2017 scenario setup there are only 10 UK corps. Perhaps adjusting the UK build limit down to say 16 will keep that AI from overbuilding (and probably better represent the manpower available).

That doesn't solve the apparent surplus of MPPs late in this scenario, they will flow to other builds or research, but perhaps Spanish Flu event(s) can absorb some of that by forcing spending on reinforcements. A bump in Reformation cost and Reformation delay also seems to make sense.

[Chernobyl]

ORIGINAL: USGrant1962
That doesn't solve the apparent surplus of MPPs late in this scenario

Change Industry Tech from +15% MPP to +10% MPP and limit investment to 2 chits instead of 3 would change the late game economy quite a bit