CWiF Sequence of Play

[Greyshaft]

This is just an idea I posted in a previous thread many moons ago .
It's not endorsed by Matrix, Harry, Robert Crandall or even me when I'm sober.
It's just a starting point for what might be possible to streamline the cardboard WiF sequence.



**************
One of the biggest challenges for the CWiF design team will be the PBEM sequence of play. If you go to www.a-d-g.com and download the WiF:FE rule set then you will see that the standard sequence of play allows over 100 successive interactions between the Axis and Allies players within a single impulse. Multiply this by (say) half a dozen impulses for each of the 36 turns within a campaign game and you have a couple of thousand emails passing between the Axis and Allied teams in order to finish a game. Of course, if you are playing a team game then it will be necessary for the teams to communicate amongst themselves before they send their move to the other side. However even if the teams could commit to always agreeing on their combined response and returning their move to the opposing side within 24 hours of receiving an update, it still gives the distinct possibility of CwiF being the first real-time strategy game i.e. it takes as long to play as the original war took to fight… six years, give or take a week or so. Clearly there will need to be a severe rationalisation within the CwiF PBEM turn sequence and the following proposal is meant as a starting point for discussion of that topic.

In order to prepare a Draft Sequence of Play it is necessary to assume certain points. That is not to say that those points should be implemented without further discussion (Heaven Forbid!), but rather that in order to address the topic in manageable chunks it is preferable that they should be the subject of their own Thread. (I am certain that my suggestion re: auto Naval intercept and Combat will generate many kilowatt hours of electronic discussion.)

* The CW and the Ge players will send out the combined move for their own team. All other players will send their own move to the CW and Ge players for consolidation and calculation.
* Naval interception and combat will be handled by the CPU although players will retain the choice of whether to intercept for each sea area. This specification can be changed at any time. (eg Until further notice the CW will intercept any and all Axis units moving through the North Sea)
* Air Combat (includes decisions re: Clearing Through and choice of Aborts and Kills) will be handled by the CPU.
* Combat results are not visible to the player who created the combat in order to avoid the temptation to reload the file and redo the moves. In a similar way, a player never knows if their Impulse triggered an End-of-Turn event.

The Game starts…
SETUP
Allied players determine their default Naval intercepts for each sea area
Allied Players email their setups to CW player. CW consolidates files. <… alternatively email the game file back and forth between players to do a standard Setup sequence>
Cw mails Game file to all Axis players
Axis do Setup

First Impulse starts…

ATTACKER (Axis) AIR/NAVAL PHASE
Axis declare any DOW
Axis choose Actions (Air, Naval, Ground, Combined)
Axis fly all Port Attack, Naval Air, Strategic Bombardment, Carpet Bombing and Ground Strike missions. < …this includes determining Order of planes for Air-to-Air combat.>
Axis sail all Naval missions. < …each mission will be preprogrammed with destination, path, which sea box to move into in case of combat, choice of combat type if intercepted etc>
Axis provisionally allocate air units to support any Naval Combats <…these units will only be committed if a combat occurs. >
Axis Players email their moves to Ge player. Ge consolidates files.
<CPU calculates Naval combats but results are not visible to Axis players>
Ge mails Game file to all Allied players


DEFENDER (Allied) AIR/NAVAL PHASE
<Allies see Axis Naval moves>
Allies fly all Port Attack, Naval Air, Strategic Bombardment, Carpet Bombing and Ground Strike defensive intercepts. <…this includes determining Order of planes for Air-to-Air combat.>
Allies provisionally allocate air units to support any Naval Combats <… these units will only be committed if a successful search occurs.>
Allies can nominate discretionary Naval combats.
Allies place CAP against anticipated Axis Air Transport and Axis Ground Support and nominate other Air units available for Defensive intercept
Allies nominate Air units to provide defensive Ground support …Air units are placed on standby but actual hexes are not nominated.
<CPU calculates results of Axis Port Attack, Naval Air, Strategic Bombardment, Carpet Bombing and Ground Strike missions but results are not visible to Allied players>
<CPU calculates results of Naval combats but results are not visible to Allied players>
Allied Players email their moves to CW player. CW consolidates files.
Cw mails Game file to all Axis players


ATTACKER (Axis) GROUND PHASE
<Axis see results of previous phase Air Combats>
<Axis see results of previous phase Naval Combats>
Axis Air RTB for missions from Attacker Air/Naval phase
Axis Rail move
Axis Land move
Axis Air Transport <Air Combats may occur but results will not be visible to Axis player>
Axis Ground support
Allied Ground support allocated by CPU <…air Combats may occur but results will not be visible to Axis players>
Axis Land Combat <Results will not be visible to Axis players>
Axis Players email their moves to Ge player. Ge consolidates files.
Ge mails Game file to all Allied players


DEFENDER (Allied) GROUND PHASE
<Allies see results of Air/Naval/Ground combats from previous phases>
Allies retreat/remove units as required by Combat results
Allies Air RTB from all previous phases.
Allies place CAP against anticipated Axis Air Supply and nominate other Air units available for Defensive intercept
Allies place Provisional CAP for next turn <…this will be implemented only if Turn Ends and other side gains initiative for next turn>
Allies commit for rerolling if Turn Ends and other side gains initiative for next turn.
Allied Players email their moves to CW player. CW consolidates files.
Cw mails Game file to all Axis players


ATTACKER (Axis) REORG PHASE
<Axis see results of previous phase Ground Combats>
Axis advance after combat
Axis Air Rebase
Axis re-org
Axis Air supply <Air Combats may occur but results will not be visible to Axis players>
Axis place CAP against next impulse (or next Turn) Port Attack, Naval Air, Strategic Bombardment, Carpet Bombing and Ground Strike missions.
Axis commit for rerolling if Turn Ends and other side gains initiative for next turn.
<CPU determines if Turn ends but does not advise Axis.>
Axis Players email their moves to Ge player. Ge consolidates files.
Ge mails Game file to all Allied players


<We now go onto the second impulse with the Allies as the Attacker and the Axis as the Defender>

ATTACKER (Allied) AIR/NAVAL PHASE
DEFENDER (Axis) AIR/NAVAL PHASE
ATTACKER (Allied) GROUND PHASE
DEFENDER (Axis) GROUND PHASE
ATTACKER (Allied) REORG PHASE

<…and then onto the third impulse with the Axis as the Attacker and the Allies as the Defender etc. etc.>


At some point the Turn ends and we go into a whole bunch of other Production and End-of-Turn stuff, but the important thing is that we have shrunk the 100 or so interactions down to five. I still have a whole bunch of unresolved problems with this idea. Here are some of them…
* How to spend Surprise points in Naval combat
* How to choose the destination for Naval Aborts
* How does the Defender get to choose whether to use a notional defender and/or defensive shore bombardment against an enemy amphibious invasion
* How to deal with the immense frustration of having no tactical control over my air and naval combats.

But hopefully I can now finish a PBEM game within a single lifetime.

[macgregor]

But hopefully I can now finish a PBEM game within a single lifetime.

I see no need to argue with you. What you have will work for what you want -provided you inject alot more AI to handle your unresolved issues(which I'm a little surprised you still have-with all that work). You want asynchronous play. That's fine. I see no reason why they can't make a game like that. I just want my WiF-which should be alot less work.

I'm assuming that you're not suggesting a TCP/IP would take a lifetime -but a pure pbem w/ non-phasing decisions-which I agree could take a long time to finish. What you have is leaving me wanting more control over my units however.

[coregames]

ORIGINAL: Greyshaft

I still have a whole bunch of unresolved problems with this idea. Here are some of them…
* How to spend Surprise points in Naval combat
* How to choose the destination for Naval Aborts
* How does the Defender get to choose whether to use a notional defender and/or defensive shore bombardment against an enemy amphibious invasion
* How to deal with the immense frustration of having no tactical control over my air and naval combats.

But hopefully I can now finish a PBEM game within a single lifetime.

By consolidating the activities like this, can't scripting continencies in advance make up the difference? For instance, a task force of boats could have a standard pop-up menu that asks how the player wants to spend surprise points (if any) and if they want to stay in any combat for mulitple rounds, or abort at the earliest opportunity. This could include abort destination preferences, as well as conditions for deciding when to use notional units and defensive shore bombardment, and which combat chart to use when defender has choice.

This is essentially giving your units orders in advance for how to react during the enemy's turn. Granted, it doesn't allow as much fine control tactically as would being available to make those decisions when it's not your turn. What I like is that your proposed sequence change does not substantially change the overall flow of WiF. That means a TCP/IP mode could still be available that gives players that extra control and interactivity.

In a different thread, I suggested that, if such a dual-mode approach was used, perhaps the AI could be developed to only operate in the streamlined mode. Wouldn't the reduced number of interactions make an AI easier to program?

I'ts encouraging to see you are thinking so productively about these issues Greyshaft.

[macgregor]

Yes I suppose you can script your way through the non-phasing decisions. Is it just me (and Erwin Rommel) who feel that flexibility is too important to be reduced to 'a script'.

[SamuraiProgrmmr]

Nice Job Greyshaft!

It will take some time to digest.

[coregames]

ORIGINAL: macgregor

Yes I suppose you can script your way through the non-phasing decisions. Is it just me (and Erwin Rommel) who feel that flexibility is too important to be reduced to 'a script'.

No, not just you (and the Desert Fox)... I am only trying to help the developers brainstorm for possible ways to make this game PBEM-suitable without stripping it of the interactivity I enjoy. I am interested in hearing other ideas about how to accomplish this. I would probably not use the PBEM mode nearly as much as the TCP/IP mode, assuming both are available.

[macgregor]

I have no problem cracking open a bottle of cognac, Putting on a good movie, (I'd like to see this movie 'Downfall') maybe a jar of nuts, and at the pre-arranged time of 2130 cranking up the game. And with the magic of the internet connect two people from Philadelphia, one from Florida,and one from Hawaii .(Where it's 1630? You can eat dinner and play this game at the same time.) If I could do this....

[Greyshaft]

ORIGINAL: coregames
By consolidating the activities like this, can't scripting continencies in advance make up the difference? ...This is essentially giving your units orders in advance for how to react during the enemy's turn.

Actually I think scripting is more realistic than the cardboard WiF. I'm just rereading Churchill's History of WWII vol.II and the issues faced in co-ordinating the first allied advance against the Italians in 1940. In WWII there was no way you could run a battle in Africa from a war room in London yet that's what happens every day in WiF.

It's far more realistic to provide scripts for the troops and ships... eg if they're bigger than you then run like crazy!!!

I'm just concerned how big the decision tree needs to be to cover a Naval units move

[coregames]

ORIGINAL: Greyshaft
Actually I think scripting is more realistic than the cardboard WiF.

This is especially true for the Soviet player, since their command structure was (and still is) based on a battle drill; field commanders for the Western Allies were given much more freedom in pursuing their general objectives.

[macgregor]

Ehh...I don't think so. We're talking about scripting the entire navy and air force pretty much .(unless you're fortunate enough to use them all in your impulses) And not too much on the land . And how many times would a player be able to change this script before he realizes it's actually faster to make this decision on the fly? Or you wont be able to change the script and focus will ultimately change to 'who knows how to identify and manipulate his opponent's AI' the fastest. Also if you do this, like I said before, there's simply not enough detail to the land combat (which now becomes 80% of your game). What you want is TOAW with production and US entry and strategic warfare.

[Greyshaft]

ORIGINAL: macgregor
We're talking about scripting the entire navy and air force pretty much.

Lets define this into two classes of script.
Player Script is the decision tree set up by the player which says things like:
* if the CW player moves a non-CV fleet of less than 24 attack factors into the north sea then the Bismark and Tirpitz attempt to intercept.
* if I get 4 or more surprise points then I allocate them like this...
* if I get 3 or less surprise points then I allocate them like this...
AI Script handles combat interactions. Which aircraft is aborted/ killed etc

And how many times would a player be able to change this script before he realizes it's actually faster to make this decision on the fly?

Au contraire mon ami... if you are playing PBEM then it is ALWAYS faster for the AI to make the decision. The alternative is for your opponent to mail you the file and let you make the decision then you mail it back to him. Easy if you're both on line, but a serious drag for asynchronous play.

Or you wont be able to change the script and focus will ultimately change to 'who knows how to manipulate his opponent's AI' the best.

When you say 'opponent's AI' I presume you mean outguessing the opponents script... but isn't that what wargaming is all about? The only difference I see here is that you want the experience to be realtime and I'm suggesting that you write a script for the combat. Either way the winner is the one who outguessed his opponent.

Also if you do this, like I said before, there's simply not enough detail to the land combat.

Land combat is certainly easier to script. Priority of unit losses and whether or not to advance after combat

[macgregor]

The alternative is for your opponent to mail you the file and let you make the decision then you mail it back to him.

You mean automatically sent TCP/IP message that announces it's arrival and portrays the decision as a number of options. Email is just for the big stuff -like updating the map at the end of the entire impulse. Of course with TCP/IP I don't believe you would even need email for that.

* if I get 4 or more surprise points then I allocate them like this...
* if I get 3 or less surprise points then I allocate them like this...

This is getting to be alot of script -though it could work. I'm assuming you wont be able to change it.

The only difference I see here is that you want the experience to be realtime and I'm suggesting that you write a script for the combat.

Yes.

Either way the winner is the one who outguessed his opponent.

'If you've beaten Rommel's plan, you've beaten Rommel'. I'm not defending my version of WiF.(though at one point I was) I just want to see the best asynch/pbem global WW2 game now. For this I'd like to see something more like a global WiTP. (from what I've seen of it)

[macgregor]

Eventually what they could do -is add 'Leaders in Flames' to the game. Each leader would have his own scripted AI (plus their own modifiers). HQs could also affect AI. I think they'd have to add more possible leaders(at least air force and naval) to the pool. If this is the case, 'fog of war' would have to substantially conceal their placement -though not completely. Don't get me wrong. I still want to play the way I've been 'whining' about. Just a thought. This way, the protocol for acquiring and changing leaders could be preserved.

[Greyshaft]

ORIGINAL: macgregor

The alternative is for your opponent to mail you the file and let you make the decision then you mail it back to him.

You mean automatically sent TCP/IP message that announces it's arrival and portrays the decision as a number of options.

Therin lies the big difference. You see default interaction in terms of synchronous TCP like Instant Messanger/Sametime while I see it in terms of email. There is room for both, but there is a class of player who can't devote the regular scheduled time to sitting down and playing every Wednesday night. Do you expect them to accept that their only play for that email is a single Kill/Abort decision and now they have to wait till their opponent reads their reply and follows up with the next combat... boring!!!

[coregames]

ORIGINAL: Greyshaft

Therin lies the big difference. You see default interaction in terms of synchronous TCP like Instant Messanger/Sametime while I see it in terms of email. There is room for both, but there is a class of player who can't devote the regular scheduled time to sitting down and playing every Wednesday night...

What I envision is a game that keeps enough information in its "buffer" to simulate the flow of interactivity, even when the interactivity is toggled to the off position.

When using the PBEM asynchronous mode (also useful for hotseat), each player would leave orders for their forces on how to handle themselves during the other side's turn. Often these settings could probably remain constant for most units from one impulse to the next. During each impulse, the settings of the opponent's forces can be reflected by custom animations on the map, as the flow of the impulse progresses. Obviously, there would not be as much tactical freedom as would be provided by concurrent play, but the disadvantage would be to both sides equally. This mode would obviously be easier to program an AI for. The downside is that each side's impulse might take a bit longer at the end, when it came time to set or change orders for the other side's impulse.

If the interactivity was toggled to the on position for synchronous play, the game could closely resemble WiFFE RAW 7, except that there would be cool animations and sounds (if they are turned on) to reflect each player's decisions and the outcomes of engagements. If the ability to set orders for units is optional on a unit-by-unit basis, players could have plans for their forces in advance even in a TCP/IP game, probably speeding play of the game beyond what is possible in table-top play without taking away any of the tactical fine-control. Those who like to play with a time control would especially appreciate this. If a unit has orders during TCP/IP play, the player could still have the option to step in, making the decision themselves; this would remove any of the tactical disadvantage.

Chris Marinacci's beta had a rudimentary function along these lines -- units could be placed on "sentry" mode, so that when changing between selectable units quickly, they were skipped until the sentry status was turned off.

[coregames]

ORIGINAL: Greyshaft

* if the CW player moves a non-CV fleet of less than 24 attack factors into the north sea then the Bismark and Tirpitz attempt to intercept.
* if I get 4 or more surprise points then I allocate them like this...
* if I get 3 or less surprise points then I allocate them like this...

IMHO, Matrix should not implement a "Player Script" feature that tries too hard to reproduce the fully interactive experience of synchronous play. I have two reasons for saying this:

1. The resulting scripts are just too complex and time-consuming (as you alluded to in your post Greyshaft). It's just too much to expect to have contingencies upon contingencies, and to expect moreover that the computer will be able to recognise all of them.

2. If the number of options is reduced substantially, an AI that is able to play well will be much easier to accomplish. AI as it applies to games still uses a tree of possibilities, and the number of branches and sub-branches on such a tree will correlate to how far the AI can see ahead in the game.

I am certain that TCP/IP play would be my preferred method if it is available for MWiF; however, if it's going to have PBEM suitability without losing the option for synchronous play, I think players choosing asynchronous play will want fairly simple options that don't take more than 5 or 10 minutes at the end of their impulse to decide upon.

[coregames]

ORIGINAL: Greyshaft
Land combat is certainly easier to script. Priority of unit losses and whether or not to advance after combat

... and combat table choice when applicable. Actually, since advance-after-combat is something that happens only for the phasing side's forces, scripting it would not be required for PBEM.

[Greyshaft]

ORIGINAL: coregames
Actually, since advance-after-combat is something that happens only for the phasing side's forces, scripting it would not be required for PBEM.

In the model I proposed the phasing player doesn't see the results of his combats for that phase:
Axis Land Combat <Results will not be visible to Axis players>
This prevents people reloading the game and trying for a better result. Therefore Advance after Combat is also delayed unless a player script is added.

[coregames]

ORIGINAL: Greyshaft

In the model I proposed the phasing player doesn't see the results of his combats for that phase:
Axis Land Combat <Results will not be visible to Axis players>
This prevents people reloading the game and trying for a better result. Therefore Advance after Combat is also delayed unless a player script is added.

True, if that model is implemented then each impulse would be broken up into multiple "chunks". I can see how consolidating activities does solve a lot of the PBEM issues, but there are so many different ways they could do this. I personally still hope for a game that closely resembles WiFFE when played synchronously, and if they take that course then it might influence how they group the activities into chunks for asynchronous play.

[Greyshaft]

ORIGINAL: coregames
I personally still hope for a game that closely resembles WiFFE when played synchronously...

I can't see why TCP or hotseat games can't have the full sequence of play.
There could also be two PBEM modes... one for the full Sequence of Play and one consolidated.
Getting our own preference doesn't mean the other guy is automatically deprived of his own choice.