Discussion on CEAW-GS v2.10

[Cybvep]

Testing first, changes later. The new naval rules will probably change balance in some way.

[pk867]

Maybe, but it will be more fun like BattleShip

A lot more strategies and tactics concerning naval combat. Weather is taken into account along with tech levels of the various combatants.

[Peter Stauffenberg]

That will force the UK navy to run for port and stay there until they have enough DD's and ASW tech to deal with the subs. That's not historical at all.

I have the opposite feeling. Wasn't it that Allies didn't have enough vessels or inferior ASW techs to escort their ships properly till 1943?

Right now Germany can build 5 SUBS by 1940 and if UK isn't overwhelmed on ground by 1941 you can have 10+ DDs so Germans don't have any chances to get any superiority, even with sub evasion on place.

In 1941 Germany can have 7 subs without build penalty and can easily get 10 subs with some slightly build penalty. The PP build penalty is not high if you get slightly above the limit. I usually get slightly above the limit when I'm building subs.

I rarely see the UK having 10 DD's in 1941. They need to build labs, air units for Egypt, maybe some CV's and land units as well. They can have 6-8 DD's with normal play. If you decide to just go for DD's then you can do it, but you have to give up something else. No air units for Egypt means Libya will hold longer. Italy won't fall as early etc.

[Peter Stauffenberg]

Testing first, changes later. The new naval rules will probably change balance in some way.

No. I must be the other way around. You first make the changes so you have something to test on. When the changes are done you release the updates to a selected few people for testing. Then you tweak the values until you're happy.

The changes will certainly change the game balance in some way. Particularly if you play the way you used to. That means you need to find new ways to play to get the upper hand again. What we need to find by testing is that both sides benefit from the changes so they almost cancel each other out. That's why you have sub evasion and sweep as a pair.

The air and naval retreat rules won't affect normal game play much. They will more likely reduce the success of exploits we see.

[Cybvep]

No. I must be the other way around. You first make the changes so you have something to test on.

I meant sth like this: we have to do some testing, because some changes have already been made. There is no reason to make more changes if even the recent ones are untested.

I rarely see the UK having 10 DD's in 1941

I think that it's normal for the UK to have 8-10 DDs in 1941.

They need to build labs, air units for Egypt, maybe some CV's and land units as well.

Just move some air units to NA, if necessary. They are not really needed in GB in 1941 if the enemy is not doing Sea Lion. I won't even mention the fact that most Axis players ignore NA altogether by either moving units back to Italy or adopting a defensive posture with the Italians and maybe sending 1-2 German unit(s). You have plenty of time to defeat these forces.

Note that unless you want to be super-aggressive with the UK, you don't have that much to do in 1941.

[Peter Stauffenberg]

I meant sth like this: we have to do some testing, because some changes have already been made. There is no reason to make more changes if even the recent ones are untested.

We're not going into a long beta phase with changes and fixes coming all the time. People don't like having to install updates all the time. So instead we want to check what needs to be fixed in GS v2.1 and make sure we get them all fixed before we release any files for people to test. That means updates will only be bugfixes. Most of the changes can be controlled in general.txt so it's easy to tweak values by just changing in this file. That means we don't need to send a lot of class files etc. except the first time.

Many of the updates we make are not intended to alter the general play. They're intended to fix a particular exploit. Like Free French units spawning as US and being impossible to move if they spawn prior to US activation. Such a change won't affect game balance, except plugging a hole where the Axis snuck up on these units and DoW'ed USA the turn before they would otherwise join.

[Peter Stauffenberg]

We're also discussing a minor change regarding convoy sizes. The current rules is that convoys can
have a size from 12 to the max size (northern = 90, central = 110, southern = 90).

Max size is modified with the average war effort between USA and UK. That means smaller convoys in the beginning and larger after both countries get above 100 war effort.

With such a rule you can get huge convoys up to about 150 and very small convoys like 15-20, even in 1945. That doesn't feel right. The spread is linear and not bell shaped or whatever. Convoys with strength less than 12 won't spawn at all.

Instead we think about altering the rule so you get a minimum convoy size dependent upon game year like this:
1939: 10
1940: 10 + 5 = 15
1941: 10 + 10 = 20
...
1945: 10 + 30 = 40

The convoy size is currently = random (max size) * average war effort (UK and USA)

A suggested change could be:
convoy size = minimum size + random (max size - minimum size)* average war effort (UK and USA)

The end result would be fewer huge convoys and fewer very small ones. That means all convoys will be worth protecting. That means more hunting opportunities for the Axis subs.

I need help with someone with number crunching so we ensure the average convoy output with the old and new rule stay the same. I have a feeling the minimum convoy size adds more to the average size than reducing the max size with the minimum size when finding the random number.

So please make comments to adjustsments to the calculation to ensure we get the same average convoy size.

[Peter Stauffenberg]

I've done some math here and it's hard to find a single formula that works that will give the same results with different war efforts.

Another and maybe simpler solution could be to calculate the convoy size 3 times and pick the average of the 3 results. That would probably give the same result, i. e. fewer very small and very big convoys.

[Cybvep]

Instead we think about altering the rule so you get a minimum convoy size dependent upon game year like this:
1939: 10
1940: 10 + 5 = 15
1941: 10 + 10 = 20
...
1945: 10 + 30 = 40

Wouldn't sth like this work? Sounds simple enough.

[Peter Stauffenberg]

Using the minimum convoy size means that the Allies will on average get slightly bigger convoys. We don't want to boost the Allied economy. Averaging the convoy strength rolls means that the size over time won't increase.

[Cybvep]

You can always increase the minimum size and decrease the maximum size.

[GogTheMild]

Personally I feel that averaging 3 results would give fairly standard sized convoys. That is they would bunch close to the mean, or the curve would be fairly tall and narrow. How about averaging 2 results? That should give a sigma which makes convoys near the extremes very much the exception whilst giving a reasonable spread around the mean for the majority.

[pk867]

I was unaware that the southern Lend lease came from Britain's on board income. It should represent off-board income from the Pacific side that is off map.

[Cybvep]

I was unaware that the southern Lend lease came from Britain's on board income. It should represent off-board income from the Pacific side that is off map.

I think that it represents the income coming from India, South Africa, Australia etc. Most likely, it also covers standard trading with South American nations.

[GogTheMild]

I was unaware that the southern Lend lease came from Britain's on board income. It should represent off-board income from the Pacific side that is off map.

If you mean the southern lend lease to Russia, it represents the aid shipped through Iran. (See especially the 'statistics' section.) There was some discussion of it here. Approximately the same amount of lend lease came through Iran (the Persian Corridor) as through Vladivostok (the Pacific Route). Edit: another source states that half as much 'lend lease' went via Iran as Persia, with the balance being made up of non-lend lease supplies. (I am not sure what, if any, difference that makes.)

[GogTheMild]

Back to Stauffenberg's original request for a critical review of 2.1: if Archangel is lost to the Axis (or possibly even if it is cut off, or if its port is bombed to zero) it would make sense for the northern lend lease convoy to revert to the UK. After all, if it is going to Russia, just where is it going to?

[_Augustus_]

Greetings,

Just a quick comment from shadows.

1. Retreats for air units being attacked by ground units.

Retreat is now set to 100 and will happen even if the air unit occupies a city or fortress. So you can't anymore hold any defense lines with air units unless the defense line is double so retreat is not possible.

This has one interesting side effect. I assume paratroopers are classed as a ground units regards doing a ground attack. Often one uses an air unit located in a high production (German) city to limit the strategic bombing attacks to one per turn, right. When the enemy is knocking at the gates and paratroopers are in jump range you're basically forced to choose between accepting two strategic bombing attacks per turn or allowing paratroopers land next to the city and automatically take over your high production city. Or worse yet your capital if you're not careful.

Cheers,

Augustus //Who hasn't played a game for well over the year yet CEaW has a strange appeal and I still keep up with the forum to see what's going on with the game. Big thanks for all taking the time to produce the marvelous AARs, BTW!

[ianc]

We moved Glasgow and this is how it looks now. By doing so we had to move the 1 production from Edingburgh to Aberdeen to avoid text overlap between Glasgow and Edinburgh at zoom level 2. This makes sense because Edinburgh is a financial city with income from trading etc. while Aberdeen is a city with several mills, quarries and after the war oil production. Production in GS is factories and other manufacturing facilities and not shops, banks etc.

Fixing the terrain and moving Glasgow was nice. You should really also move Rosyth as well. Its NW of Edinburgh over the opposite side of the R. Forth. I would also add Loch Ewe as it was a large natural harbour used for setting up convoys (to Russia and aso to USA/Canada?) and a base for warships on the west coast.

Ian from Dundee (just below Aberdeen on the map)

[Peter Stauffenberg]

The main problem with moving Rosyth is that Rosyth on the map is not a land hex, but a sea hex. It's a sea port. So if we move Rosyth to 1x NW of Edinburgh it means we get a sea port there and the current Rosyth hex will be open sea. That will make Scotland land just 1 hex wide (Glasgow). That doesn't feel right.

What we see here is the limitation of the game where sea ports can't be in the same hex as a city. This is because of the 1 unit per hex limitation. So naval units have to be in separate hexes from land units.

What we need is to have Rosyth adjacent to Edinburg. So the current hex is the least inaccurate. It indicates Rosyth is north of Edinburgh and that there is a fjord north of Edinburgh where the naval base is.

Many hexes on the map is partially land and partially sea. The entire hex is either land or sea based on the majority shown in the hex. There are a few exceptions regarding seaports where they hex indicates land, but where we have a seaport because the city had a seaport and very narrow access to the port. Rosyth is such an example.

If he map resolution had been higher then placing Rosyth NW of Edinburgh is right.

We have to look at Loch Eve. A sea port needs to be attached to a resource like a city, rail depot, mine or whatever. Was it possible to rail units to Loch Eve? If not then I think the use of the loch eve will be limited. There are quite a few ports that are not on the map because we have other ports nearby. In fact almost all coastal cities had a port. In GS we need to ask ourselves if the port had any significance regarding military troop transport, naval repair etc. If the port was mainly for trading purposes then it's not necessarily on the map. Norwich and Bilbao are such examples.

I also see you named the Glasgow port Clydebank. Should we add that on the map?

[GogTheMild]

Loch Ewe is 40km from a miserably poor single track railway. To which it is connected by a hair raisingly bad single track road. 70 years ago it must have been much worse. Loch Ewe was served entirely by sea and was a convoy collection point. Nothing came in by land. Not much came in by sea as it had almost no port facilities. The main settlement, Poolewe, has (today) less than 200 inhabitants.