Matrix Games Forums

Posted: **Fri Mar 27, 2009 1:38 pm**

In essence, this change from RAW may result in the same overall "average" but introduces MUCH higher variability in the process.
 
Under RAW, you may pull a number of "0" chits in a row, but eventually you WILL pull higher value chits, as the odds of pulling another "0" chit is reduced each time you pull one, and eventually reaches no chance of doing so. The same applies with any other particular value of chit. The RAW system leads to a covergence towards the mean over time, which narrows the range of the possible timing of US entry.
 
In the RAC system there is no such feedback loop. Each pull is independent of what has gone on before. This is a major change from RAW that will lead to a more variable US entry.  I'm not exactly sure how major the repercussions will be on the game. But I am concerned. Will this lead to a 1940 US entry into the war? Or 1944 comes around and the US is still sitting on the sidelines? Only time will tell.

Posted: **Fri Mar 27, 2009 2:17 pm**

ORIGINAL: KosMic

In essence, this change from RAW may result in the same overall "average" but introduces MUCH higher variability in the process.

Under RAW, you may pull a number of "0" chits in a row, but eventually you WILL pull higher value chits, as the odds of pulling another "0" chit is reduced each time you pull one, and eventually reaches no chance of doing so. The same applies with any other particular value of chit. The RAW system leads to a covergence towards the mean over time, which narrows the range of the possible timing of US entry.

In the RAC system there is no such feedback loop. Each pull is independent of what has gone on before. This is a major change from RAW that will lead to a more variable US entry. I'm not exactly sure how major the repercussions will be on the game. But I am concerned. Will this lead to a 1940 US entry into the war? Or 1944 comes around and the US is still sitting on the sidelines? Only time will tell.

Maybe in MWiF, the odds of picking a "0" chit can decrease each time a "0" chit is picked.
Likewise, maybe the odds of picking a "5" chit can decrease dramatically (going nearly to 0) each time a "5" chit is picked. Talking of the 1939 chits.

Posted: **Fri Mar 27, 2009 2:40 pm**

ORIGINAL: KosMic

In essence, this change from RAW may result in the same overall "average" but introduces MUCH higher variability in the process.

Under RAW, you may pull a number of "0" chits in a row, but eventually you WILL pull higher value chits, as the odds of pulling another "0" chit is reduced each time you pull one, and eventually reaches no chance of doing so. The same applies with any other particular value of chit. The RAW system leads to a covergence towards the mean over time, which narrows the range of the possible timing of US entry.

In the RAC system there is no such feedback loop. Each pull is independent of what has gone on before. This is a major change from RAW that will lead to a more variable US entry. I'm not exactly sure how major the repercussions will be on the game. But I am concerned. Will this lead to a 1940 US entry into the war? Or 1944 comes around and the US is still sitting on the sidelines? Only time will tell.

True, but I find the whole idea of a finite distribution distasteful.

For example, imagine that all the die rolls were done the same way. Having rolled a 1, you then know that you will not see another 1 until you have 'rolled' each of the other 9 numbers. The last number in each 10 wouldn't be random at all. If you find that to be an attractive idea, then we have very little in common concerning this topic.

In both cases the reduction in variation lends itself to 'gaming' the system.

WIF FE players are used to 'gaming' the US entry, and consider it an integral part of the game. Indeed, I have had people complain that disassociating the US Entry from the Nazi-Soviet neutrality pact chit draws is 'wrong', but MWIF does that too. I consider these both to be an artifact of WIF FE counter mix limitations (e.g., in WIF FE the Nazi-Soviet chits all have American flags on them).

That CWIF (an ADG game) was written to use an infinite distribution, instead of a finite one, is good evidence that Harry Rowland wanted WIF to use the infinite version.

Along the same lines, the change from a 6 sided die to a 10 sided die increased the variation of all random outcomes [I really liked the 6 sided dice simply for the aesthetics of how they bounced on a glass table top.] The change from the 1D10 CRT to the 2D10 CRT also increased the variation of the land CRT results. Any arguments advocating less variation in outcomes (in general) could be applied to both those changes in the WIF rules too.

Posted: **Fri Mar 27, 2009 3:43 pm**

The finite chits in the paper game have the following effects:

1. If you draw bad chits, those are gone and cannot be drawn again (assuming no recycling).
2. If you draw good chits, those are gone ...
3. If the USSR has really good chits (to prevent GER from breaking the Pact), then the US must have worse chits.
4. If the Pact is held until 1942, as deliberate USSR or GER strategy, the US will draw much better chits since the earlier pools will be exhausted.

3 and 4 above are abolished by having, effectively, no interaction between the pools. This is fine and perhaps how HR would have designed it given no constraints. However, there will be no acceleration of chit additions, so US entry will, on average, slow.

I am concerned that the effect of past draws has no impact at all on future draws.

I had thought this was designed so that you get infinite chits in 1939 using a 1939 distribution. Then 1940 using a 1940 distribution. This is not the case. Instead it is 1940 using a mixed 1939 and 1940 distribution.

I know in the past Patrice has run Monte Carlo simulations to show average US entry and gear up and the ability of GER to break the Pact. These have also included a standard deviation.

Would it be possible to run the same sort of simulation with the infinite chits?

My concern is that if a "5" is drawn, there is still the same chance of drawing a "5" next. Same for a "0".

I think this part of the program should be reconsidered. I favor the seperate chits pools, but a decreasing chance of drawing a chit if you have previously drawn a chit. For that matter, if there is no need to ever add the next year's pool, why not just use the original pools?

Posted: **Fri Mar 27, 2009 4:15 pm**

ORIGINAL: sajbalk

The finite chits in the paper game have the following effects:

1. If you draw bad chits, those are gone and cannot be drawn again (assuming no recycling).
2. If you draw good chits, those are gone ...
3. If the USSR has really good chits (to prevent GER from breaking the Pact), then the US must have worse chits.
4. If the Pact is held until 1942, as deliberate USSR or GER strategy, the US will draw much better chits since the earlier pools will be exhausted.

3 and 4 above are abolished by having, effectively, no interaction between the pools. This is fine and perhaps how HR would have designed it given no constraints. However, there will be no acceleration of chit additions, so US entry will, on average, slow.

I am concerned that the effect of past draws has no impact at all on future draws.

I had thought this was designed so that you get infinite chits in 1939 using a 1939 distribution. Then 1940 using a 1940 distribution. This is not the case. Instead it is 1940 using a mixed 1939 and 1940 distribution.

I know in the past Patrice has run Monte Carlo simulations to show average US entry and gear up and the ability of GER to break the Pact. These have also included a standard deviation.

Would it be possible to run the same sort of simulation with the infinite chits?

My concern is that if a "5" is drawn, there is still the same chance of drawing a "5" next. Same for a "0".

I think this part of the program should be reconsidered. I favor the seperate chits pools, but a decreasing chance of drawing a chit if you have previously drawn a chit. For that matter, if there is no need to ever add the next year's pool, why not just use the original pools?

If you are interested in this there is no need to run a simulation.

The code excerpt I gave shows the means and calculating the standard deviation from the given distributions can be done by anyone who took an introductory statistics course (and passed[;)]). [I used to teach statistics.]

Posted: **Fri Mar 27, 2009 6:54 pm**

ORIGINAL: Shannon V. OKeets

I really liked the 6 sided dice simply for the aesthetics of how they bounced on a glass table top

If you roll your 10-sided dice out of a cup into the game boxtop, they'll sometimes spin like a top for 30 seconds before coming to rest.

Now that's exciting !!

Posted: **Fri Mar 27, 2009 7:05 pm**

ORIGINAL: paulderynck

ORIGINAL: Shannon V. OKeets

I really liked the 6 sided dice simply for the aesthetics of how they bounced on a glass table top

If you roll your 10-sided dice out of a cup into the game boxtop, they'll sometimes spin like a top for 30 seconds before coming to rest.

Now that's exciting !!

Ah, but I own over 100 of the small 6 sided dice and they bounce like Mexican jumping beans on a 1/8 inch glass tabletop.[:D]

Posted: **Fri Mar 27, 2009 7:14 pm**

ORIGINAL: Shannon V. OKeets

ORIGINAL: sajbalk

The finite chits in the paper game have the following effects:

1. If you draw bad chits, those are gone and cannot be drawn again (assuming no recycling).
2. If you draw good chits, those are gone ...
3. If the USSR has really good chits (to prevent GER from breaking the Pact), then the US must have worse chits.
4. If the Pact is held until 1942, as deliberate USSR or GER strategy, the US will draw much better chits since the earlier pools will be exhausted.

3 and 4 above are abolished by having, effectively, no interaction between the pools. This is fine and perhaps how HR would have designed it given no constraints. However, there will be no acceleration of chit additions, so US entry will, on average, slow.

I am concerned that the effect of past draws has no impact at all on future draws.

I had thought this was designed so that you get infinite chits in 1939 using a 1939 distribution. Then 1940 using a 1940 distribution. This is not the case. Instead it is 1940 using a mixed 1939 and 1940 distribution.

I know in the past Patrice has run Monte Carlo simulations to show average US entry and gear up and the ability of GER to break the Pact. These have also included a standard deviation.

Would it be possible to run the same sort of simulation with the infinite chits?

My concern is that if a "5" is drawn, there is still the same chance of drawing a "5" next. Same for a "0".

I think this part of the program should be reconsidered. I favor the seperate chits pools, but a decreasing chance of drawing a chit if you have previously drawn a chit. For that matter, if there is no need to ever add the next year's pool, why not just use the original pools?

If you are interested in this there is no need to run a simulation.

The code excerpt I gave shows the means and calculating the standard deviation from the given distributions can be done by anyone who took an introductory statistics course (and passed[;)]). [I used to teach statistics.]

No need for a statisitics course to know there is a great deal of difference between the two methodologies. Although I think I'd be more comfortable with doing it the way the board game does it, your point about the design of CWiF is interesting. I guess I wonder just how much input Harry had into this part of it.

There have been very detailed studies of the U.S. entry chit distributions and their effect on when the U.S. can come in - done by people on the Yahoo list, but they did not address Steve's Points 3 & 4. (BB Iowa Steve I mean.)

My personal experience is that the distributions in all the pools stay close to the means, but where real good or bad luck can occur is when the U.S. loses a chit for an Allied action or when it moves a chit from the Entry Pool to the Tension Pool. Just last night I had one zero in the Ge/It pool with six other chits, any one of which would, when moved to Tension, put me at the level needed for War Appropriations next turn. You can guess what happened. (Really the most annoying thing about luck like that is you can't bitch and whine about it without giving away too much info to the Axis.)

Posted: **Fri Mar 27, 2009 7:17 pm**

ORIGINAL: Shannon V. OKeets

ORIGINAL: paulderynck

ORIGINAL: Shannon V. OKeets

I really liked the 6 sided dice simply for the aesthetics of how they bounced on a glass table top

If you roll your 10-sided dice out of a cup into the game boxtop, they'll sometimes spin like a top for 30 seconds before coming to rest.

Now that's exciting !!
Ah, but I own over 100 of the small 6 sided dice and they bounce like Mexican jumping beans on a 1/8 inch glass tabletop.[:D]

You've been playing too many of those Columbia games from Canada. "O.K. Boys, this next sand pail full of dice will be for my attack just south of Leningrad..."

Posted: **Sun Mar 29, 2009 6:49 am**

ORIGINAL: paulderynck

ORIGINAL: Shannon V. OKeets

ORIGINAL: paulderynck

If you roll your 10-sided dice out of a cup into the game boxtop, they'll sometimes spin like a top for 30 seconds before coming to rest.

Now that's exciting !!
Ah, but I own over 100 of the small 6 sided dice and they bounce like Mexican jumping beans on a 1/8 inch glass tabletop.[:D]
You've been playing too many of those Columbia games from Canada. "O.K. Boys, this next sand pail full of dice will be for my attack just south of Leningrad..."

Not really. Eastfront (and Eurofront) are not quite that bad, though you could roll up to 16 dies at once. But in Europe Engulfed from GMT, now there you could potentially roll over a hundred dice (as assault doubles the dies). They do provide resulttables (rolled with 3D6) though.

So Steve, if you like that, EE is for you. But maybe you need a second glasstable, just for the dice

.

Posted: **Thu May 07, 2009 3:15 pm**

ORIGINAL: Shannon V. OKeets

...It is a cumulative distribution. Here's the code:
... begin // 1943 + // Average = 284/94 = 3.02...

Steve,
Shoudn't it be 274/94 = 2.91 ?
Or am I missing something ?

Alain

Posted: **Thu May 07, 2009 3:45 pm**

ORIGINAL: Caquineur

ORIGINAL: Shannon V. OKeets

...It is a cumulative distribution. Here's the code:
... begin // 1943 + // Average = 284/94 = 3.02...

Steve,
Shoudn't it be 274/94 = 2.91 ?
Or am I missing something ?

Alain

Yes.

Paul and Nils and reviewing this distribution in detail, using several likely event calendars for the war and a simulation model to track the running US chit total(s). Their intent is to improve the distribution so it follows WIF FE experience more closely.

Posted: **Thu May 21, 2009 4:44 pm**

I am making a final push on finishing the Players Manual. One of the forms that I have been most concerned about is the setup tray. Here is how that form is described in the Players Manual (my final version, but the editor, Mike, will probably find ways to improve my writing).

First of 3 in the series.

Posted: **Thu May 21, 2009 4:51 pm**

2nd in the series of 3.

This form has undergone a least a couple of dozen of major incarnations. The beta testers had many suggestions (i.e., complaints). In essence, the form enables the player to perform an enormous number of functions. Designing the layout and the mouse & keyboard interface was not intuitively obvious.

The screen shots of the form you see here are for a screen width of 1280. The form can be resized from the left edge so only a half dozen units are visible in each list. Or it can be expanded to as wide as your screen(s) can show. I have double monitors and I could set it up to show 40 units in each row.

The vertical height was a major concern. We wanted the setup tray's footprint to be as small as possible vertically, so as much of the the detailed map (where the units are being placed) as possible is visible .

Posted: **Thu May 21, 2009 4:52 pm**

3rd and last in the series.

Posted: **Fri May 22, 2009 3:54 pm**

Crap. I am jonesing so bad for this game to come out!

[X(][&o]

Posted: **Tue Jun 02, 2009 5:39 am**

I haven't checked all 14 pages, so I hope I do not repeat anything, but I wonder if the LAIO specification is/will be available for download? [8|]

Posted: **Tue Jun 02, 2009 6:35 am**

ORIGINAL: Litsne

I haven't checked all 14 pages, so I hope I do not repeat anything, but I wonder if the LAIO specification is/will be available for download? [8|]

Welcome to the forum.[:)]

I don't believe I have posted that. If you are really interested, send me an email (SHokanson@HawaiianTel.net).

Matrix Games Forums

What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?

RE: What kind of manual(s) should we ask for?