Blog Tools
Edit your Blog
Build a Blog
RSS Feed
View Profile
« February 2012 »
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29
You are not logged in. Log in
Entries by Topic
All topics  «
Churchill Design
Design Diary
Empire of the Sun
For The People Material
Wargame Design Musings
Wargaming For Leaders
Washington's War
CIO Insight Reference Links
Must-read Books
Leadership Blog
Buy/Order Wargaming for Leaders
Barnes & Noble
Wargaming for Leaders
Book Site
Mark Herman's Wargaming Blog
Tuesday, 14 February 2012
FTP CRT Debate continues
Topic: For The People Material
The flaws of the CRT are not merely a personal opinion, but are supported by history. Cold Harbor was not a 5/3 result. It was closer to a 5/2. Confederate losses did not exceed 5,000 for the entire battle. A 5/3 result would put rebel losses at something close to 7,000. Is this splitting hairs? Certainly, but when playing the game I never had a battle that seemed like Fredericksburg, Franklin, Cold Harbor, or Ezra Church, where the losses were of such a a wide disparity. In my experience the CRT works wonderfully for pre-1864 engagements, but does not model the power of entrenchments with the same effectiveness.

I could of course make an esoteric argument that all of it is personal opinion, since figures are rarely so accurate. Flaws abound in each game. I prefer the Victory Games classic, but it does occur in a political void, which makes it imperfect. My point here is that my view of the CRT is not just a cavalier personal opinion. Regardless, my problems with For the People have more to do with the way the game plays in totality than it does with the CRT or its simulation value. I believe on the later point is succeeds quite well.

This initial comment is beneath you and not in keeping with a adult conversation, but typical of the internet. I will also predict that no matter what I say or demonstrate it will not be sufficient to change your mind and as I will tire long before you do, so you will get the last word for whatever that is worth.

My comment on the popularity of the game was in reference to your negative review comment. I do not care that you do not like the game and like others. I have lived that experience for over 35 years, so get in line. Those who like the game give the same reasons for liking it as you do for not liking it. My point was not your review which I stated was well considered and thoughtful, but your one line dismissal of a great deal of work and analysis.

Eric Lee's game is awesome and as one of the lead playtesters who had real input, good news. I would also note that Eric Lee in his own words rates FTP as one of the best games ever designed on the topic, kind words from an old friend. 

Now to my main bone of contention. First off the remarks smack of Wikipedia. There is a range of historical facts on the losses at Cold Harbor that differ significantly by approximately a factor of 2, so from the analysts do not agree by a significant margin. At the end of the day I had to state my view that differs from yours, but stating it is an historical fact is not credible as even the experts are off by a factor of two using the same data but using different assumptions. Like all of these authors I have done my own research and come up with my own view, clearly not in line with yours.

The low end Union losses are shown to be in the 6-7000 range and the high end is close to 15,000. The same goes for the CSA losses that range from 2,500 to 5,000, just for June 3rd, but even though June 3rd is the main focus there are a series of engagements across a broad area over a week period. At the FTP scale one card does not represent one day, but at a minimum a major offensive muscle movement. Bottomline at the scale of FTP a card play represents a significant number of engagements over a period of time that is imbedded in a four month period. You state a flawed CRT based on the notion that you calculate the losses for Cold Harbor as a 2 and I come up with 3, so at the end of the day we disagree by one SP which is the granularity of this design and most others at this scale. So all I can surmise is your standard of whether a crt is flawed is a rounding error based on a narrower view of what a card play represents. 

I have looked at the same data and applied additional tests that constitute losses as a percentage of forces employed in addition to relative losses. However at the end of the day you are stuck with some atomic level of granularity, so no matter what happens, the same is true for The Civil War and the rest that use the 5-6k, one division SP that as Eric Lee states was a Dunnigan view that I agree with. At the end of the day you cannot go below the atomic level set by the smallest unit in any design be it one man or one division. What people generally like is linear monotonic CRTs, basically more is better with some minor variation. Unfortunately most combat outcomes are non-linear as well stated by a range of Ops Research papers produced around the Centenial. I used these and other very analyses and drew several conclusions, but the main one is linearity is not part of the actual facts.

My view at the end of the day was not raw casualties, but divisions made ineffective. In FTP it takes approximately 2400 casualties to make a division ineffective (40%, but some division became ineffective at less and greater percentages, with the mean around this value), which puts the CSA losses at 2-3 sps for the May 31st-June 3rd actions. The USA losses are between 4-6 sps for the same period. At a fundamental level I feel that Rhea and Trudeau who did more recent and extensive number counts would tend to round up not round down the CSA losses and the CRT will produce the 4-6 USA losses. The other issue is CSA end strength counts for this period are notoriously inflated as desertions and detachments are poorly accounted for and units were encouraged to inflate their end strength to draw extra rations. This puts emphasis on rounding up. In the end I have the CSA forces on the line as about 25% less than the approximately 57k to 62k reported by the various roll calls. The Union numbers are also inflated, but by a lesser factor. In the end if you have studied these numbers as long as I have, there is no agreement and at the end of the day you have to pick a place. In your estimation we disagree by one CSA SP constituting a flawed CRT. in my view we are using different calculations. What most sources do agree on is that Lee lost a greater proportion of his force than Grant did even given the disparity in losses in this one short set of engagements.

My conclusion is that the historical Cold Harbor is either 6-3 or 5-3 not 6-2. No combination of sources comes up with a 3-1 loss ratio and all sit somewhere around 2-1, which the FTP CRT will produce although my personal view is it is less than that as I think that Rhea has done a deeper analysis of the numbers and the returns. However that is an opinion not a fact as Rhea and myself were not there and the records even with heroic analysis cannot be fully reconciled.

As far as your post that you never saw or felt a battle that felt like Fredericksburg I can only ask why? Playing any game a few times does not constitute a valid sample but the CRT makes this all visible. The large battle CRT and the CRT in general is built around the notion of a 6-1 dice outcome to handle the historical outliers. So the CRT can and has on many documented tournament outcomes a 1/6 SP loss exchange and everything in between assuming that one side has no drms versus one that has maximum drms. Note that a Burnside led army probably has +0 modifiers and a Lee-Stuart-Jackson-Longstreet combo almost always generates a 6sp loss. In addition the design has compelling reasons why the UsA might willingly accept this type of battle although my hedge is there is a card, again witness real world tournament activities as demonstration that this happens in most games at least once. So, if you have not seen or felt this type of outcome it is what it is, but the CRT math is quite visible in this example.

What should be obvious is the CRT behaves in a non-linear fashion based on the drms. For example in a medium battle with no drms, say Fremont versus Polk the medium CRT is defense dominant. Put Sherman versus Polk and the attacker is slightly advantaged, put Sherman versus J Johnston and the CRT becomes defense dominant again. Put the AoNVa at its height versus Burnside, Hooker etc. and it is once again attacker advantaged. My point is the non-linear nature of the CRT is designed to handle a wide variety of situations to include everyone that is recorded. 

Anyway, post your next message where you disagree, you win. As I said I have never seen an internet conversation change anyones starting positions. In any case it has been fun trading perspectives.


Posted by markherman at 10:38 PM EST
Share This Post Share This Post
Post Comment | Permalink

View Latest Entries