Blog Tools
Edit your Blog
Build a Blog
RSS Feed
View Profile
« April 2014 »
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 30
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
Monday, 27 June 2011
Historical Justification in Game Design
Topic: For The People Material
"FTP is an outstanding game that is very, very well balanced and has wonderfully inventive and absolutely elegant systems, but in game play will generally not yield a narrative that is familiar to anyone with any knowledge of the actual Civil War." 

I appreciate the kind words on the design, but you raise an interesting point about the narrative value of any wargame. I would first offer my Clio Corner #1 where I cover your point directly. 

The question that I always face when designing a game is how to capture the strategic situation with all of the considered strategic choices, while factoring in player hindsight. Unless you deeply study any particular historical set of events the narrative history, that which happened, is the standard for judging the historical path. Although this should be a starting point there is a basic view that the historical narrative is the baseline and the most probable path. In my mind that is a false assumption to make about any historical situation. All one can say is that this is what happened and the objective standard that any book must meet. 

Wargames in a way have to meet a higher standard. Once you consider 'what if' history you have to research deeply enough to understand what the historical participants thought was possible. One of the biggest 'historical' critiques of FTP is the unrealistic way in which forces can raid off of their supply lines. I have commented on this before and would ask anyone interested to find those posts, but just last night reading in the OR I found a letter from Lee to Davis during the Gettysburg campaign that he (Lee) intended to maintain his army in the North into the Autumn without a rail connection. Clearly Lee is not worried about the unrealistic tactic of maintaining his army on his opponents resources. Then of course there is the ultimate raid, Sherman's March to the Sea. 

My point being is the kindly worded comment reflects the writers tolerance for historical 'what if', which we all possess in greater or lesser degrees. All I can say is all parts of the FTP narrative have a basis in the historical narrative that did not happen, but was considered. 

FTP requires some skill to play well. I make no excuses for that since I believe that is why we are having this type of discussion over a decade after publication with an ongoing tournament adding to our collective understanding. The net result is in any unique game play situation if a player gambles or does not correctly anticipate how things might evolve the historical paths not taken can create for some interesting game play and for some the perception that the game has gone down an unreasonable path. Given my earlier comments I firmly believe that everything in the game follows a reasonable 'what if' but we all have a different tolerance for what that would be. 


Posted by markherman at 9:33 AM EDT
Share This Post Share This Post
Post Comment | Permalink
Saturday, 23 May 2009
FTP Recruitement
Topic: For The People Material
More or less...

I can gather it is a sociological inferation circa 1861-65, that the CSA Army cutting off Illinois, Indiana and Ohio would be a more economic negative substantial effect, as opposed to Union occupation of Georgia, Virginia and Tennesee in terms of SPs according to your data. 

To some degree this is an apples and oranges comparison. All is not always equal when it comes to production. The two sides use different recruitment models using the same factors.
The rail line discussion got me to go back into my notes (over a decade old) on this topic and see what data I had used for the recruitment model that generates SPs during the reinforcement phase. FTP is looking at frontline strength based on a soldiers on the rolls with some factors thrown in for frontline strength versus logistic troops. The overall analysis shows that once you get into mid 62 until almost the end of the war, the amount of men for each side hit a steady state that averaged around 48 (plus or minus 4) for the USA and 34 (plus or minus 3) for the CSA.
Remember, the historical model is based on the historical actions, so if the Union doesn't spend sufficient effort to shut down a blockade zone with amphibious forces or succeeds in accomplishing the capture of TN early in the war, etc. then the generalized model will produce different results.
When it comes to the overall model of what makes an SP it is a blend of Administration, People, and Ordnance. At all times there was some set of factors that limited output with the 1863 period being the one most in balance. So early in the war there was a surplus of people (recruitment enthusiasm), but a deficiency in administrative acumen and ordnance. Late in the war there was greater administrative efficiency and more available ordnance, but a deficiency in people (conscription difficulties due to going to war not being that attractive any longer). It is this deficiency that drove the Union to incorporate colored soldiers and the South to consider conscripting slaves into the army. When it comes to the Union, approximately 2/3rd of the SPs are being produced on the map and 1/3rd are being produced north of the map, so on a normal Union turn 12 SPs are produced on the map and 6 SPs are produced north of the map. I chose to consider Union State Capitals as key locations that impacted Union production as they were the administrative and production centers (ordnance and armories). My basic premise is if an administrative center is occupied it reduces the potential SPs available due to administrative and production losses. I viewed cutting railroads as disrupting production. I viewed DC as a major administrative center whose loss or isolation (loss of telegraphic communication) would create inefficiencies in SP production akin to when the Federal government shut down recruiting for a short time in 1863. There is also the Union Governor reaction effect whereby some amount of state troops would be held back akin to the National Guard of today.
Rolling back to a Confederate force operating across the three Western states would represent the disruption not only of state ordnance production, but occupation of the means to move people around. It also is severing the links to states like Michigan and Wisconsin (note that portions of the Great Lakes are just off the northern edge of the map). Perhaps more importantly is the Union states are not producing SPs equally. On the map IL, IN, OH, PA, and NJ are producing more than their share with NY, WI, MI, and MA the major contributors north of the map. So, in the aggregate the railline cutting rules are trying to capture an impact that taking out three major production centers and isolating them from the broader grid is taking out about 2/3rd of Union recruitment capability.
The southern model uses the same factors but behaves differently. The production centers are the resource spaces and the values represent their historical output normalized on 100 percent. New Orleans was always a problem as its potential production was never fully developed before it was captured, so I chose to model that blockade zone differently (takes only two to shut down) to account for this factor. The South was also dependent on finished ordnance and critical metals to maintain their production base, which is all tied into the blockade model. The issue with the south was the stronger state control over local production and recruitment. The Southern railroads could not support the Union industrial model making it far more local in nature. So although TN and VA were the major production centers the people were more distributed than they were in the north (New Orleans was the only major American city that went with the South). Consequently the Southern recruitment model is more resilient, but more vulnerable on the SW side if you were to lose both TN and VA (57 SW points).
So, different views of the same recruitment model. Now back to Riku's gambit. Quite frankly I put the rail cutting option into the game with a severe penalty with the knowledge that a Union player would do what it takes to stop it. The question on the table for me is whether this is a trick that cannot be stopped or just another historic vulnerability that has to be paid attention to. As far as whether it can or cannot be stopped, I submit that can be stopped without significant resources as per Riku's earlier comment. Columbus OH is the key coupled with effective use of strategic movement of reinforcements. Effectively if the CSA is threatening an invasion north of the Ohio, the Union has to have some rear area garrisons to limit Southern options.

Posted by markherman at 9:39 AM EDT
Updated: Saturday, 23 May 2009 9:59 AM EDT
Share This Post Share This Post
Post Comment | Permalink
Saturday, 17 January 2009
For The People Riverine Q&A?
Topic: For The People Material

The one aspect of For The People that continues to create problems is the riverine rules. It is a function of a non-intuitive concept (non-bilateral relationship) and the need to read the rule and apply it literally. To aid this process the final version of the rules included extensive diagrams, which apparently most people do not want to look at. Anyway, here was a good exchange that captures all of the issues that seem to crop up with this rule and my responses. I felt that I needed to capture this as a future resource.


 Windopaene wrote:

To Mark:


What makes them confusing for me is a couple of things.


Upriver/downriver - there are a bunch of those rivers that don't run the way I think they should, (all heading south like the big M), so when reading the rules, I get confused about what is being described in the examples.


"Crossing a river" - There are one or two examples that talk about crossing a river when leaving a space, even though there are rules before that say that crossing a river only impacts crossing a blue line when entering a space that contains the blue bar. I may still have this confused.


And that's just the basic stuff. Doesn't even start to consider Riverine movement, and the like. And the coastal forts/inland associated port stuff.


I think I understand the rules, but don't feel like I am at all comfortable with them, and any attempt to apply them will result in having to go over them step by step as I try to take a move, to be sure I'm not screwing them up...


And Empire of the Sun is to For the People as For the People is to We The People... 



On the up/down river question, remember in the real world all rivers run to the sea, which is New Orleans in this case, so north-south has no relevance to rivers. However, we did put the arrows on the map to show you which way the river is running. I would use the arrows... as an aside, if you ever get lost in the woods and see a river, follow it downstream, eventually you will get somewhere... :-)


Crossing the River: this is the one that continues to confuse people, but mostly folks are not reading the rule, although the graphic chosen doesn't help. You are only crossing a river when you enter or exit a riverine port space. The part that gets missed is the last part, "...riverine port space." A riverine port space is any port space that has a blue bar. So entering any space that is NOT a port space regardless of whether you exited a Riverine port space to get there is NEVER crossing a river. 


As far as the rest of the riverine rules they are easy to summarize, but hard to write unless you use alot of words. Basically assume that the Union navy has river control unless there is a Confederate fort in the riverine port space, not adjacent, to the port space (see crossing the river above). There are really only two situation you have to consider. 


Situation 1: If you have a CSA fort (or ironclad), the CSA can cross the river. If not assume you cannot. Neither side can cross a river into a port that contains a fort. This is the corollary to why the CSA can cross from a port space where they have a fort. The simple concept is neither side can cross a river into a port space where there is an enemy fort and the CSA can never cross a river unless they have a fort in the space. This is by far the major case you have to remember regarding river crossing. 


Situation 2: There is another significant case when the CSA has a network of forts, such as in the opening, where they bookend a major portion of the river. (see the diagram on pg 28). Here, the combination of Fort Philip-Jackson, Columbus KY, and Dover TN creates a large Mississippi zone where the Union does not have naval control (deny Union naval control=DUNC). As soon as any of these forts fall, you more or less find yourself in the first case. 


I submit that this latter case is very visual if you just look at it logically, but beyond the opening set up, it is hard, approaching I have never seen it in actual play, where the CSA can do much better than their opening set up. Assuming that the CSA controls Fort Philip-Jackson OR has a fort in New Orleans, the CSA needs a fort in Columbus KY or Memphis TN or Vicksurg MS to hold any section of the Mississippi. Once New Orleans (and its associated fort) are Union, basically there are only two crossing points at Memphis and Vicksburg if there is a CSA fort present (situation 1 again). 


The only other portion of the river that ever gets in a similar state is the fort at Dover allows the CSA to cross at Clarksville and Nashville TN. Once Dover falls you find yourself back in Situation 1. There are other possibilities, but in actual play they just do not occur, so just do not try and consider them.


Hopefully that helps, but I would also not worry about all of the details, especially when you are learning the game. Basically deal out cards and move armies and conduct battles. The riverine rules tend to sort themselves out if you just look at all situations through the Situation 1 lens.


As far as the EoTS to FTP comment... that is a personal taste issue. What I find is folks never read or follow the sequence of play in EoTS (rule 6.2). I then get a whole host of questions that indicate that no one is following the sequence of play. In a hobby of rules lawyers I find this the most curious situation of all. I get beat up all of the time for not being precise in my wording, but the corollary is most people are not reading what is written, so we are even.


My biggest piece of advice is, stop reading the rules per se, but set up the game and play out, move by move using the 1861 turns at the front of the rules. We spent hundreds of hours creating and checking this for accuracy, but no one ever seems to use the resource. Once you get through the third turn in the example, just continue playing into 1862. If you do this you will experience just about every rule in the game through the example. I hope that helps your endeavors... 




Eric Brosius wrote:

MarkHerman wrote:

So entering any space that is NOT a port space regardless of whether you exited a Riverine port space to get there is NEVER crossing a river.



I think you are implying that the CSA can move from Louisville to Bloomington, or from Cincinnati to Falmouth, at any time regardless of the UNC situation, without a fort or ironclad in Louisville or Cincinnati.


I didn't think this was legal.


In rule 6.2 it says "A force is crossing a river the instant it enters or exits a riverine port space by crossing a blue bar which is part of the port space." (Italics supplied by me.)


To me this implies that moving from Louisville to Bloomington is crossing a river (and hence requires DUNC for the CSA,) while your quote seems to contradict this.


I recognize that you're the designer, Mark, so I may very well be missing something.



You have put your finger on the thing that continues to trouble folks, so you are not the first one to get this wrong. However, right in the rule that you cite, there is a diagram (bottom right of pg 24) that explicitly demonstrates the point that I am about to make. By the way the diagram is the Louisville-Bloomington situation that you mention. However, if you read my earlier note, sans a fort, the CSA cannot cross any rivers except when their fort network has protected a large segment of the river as in the opening set up. 


I cannot re-iterate this enough, practically speaking, unless you are crossing the Mississippi in the deep south with the original fort configuration, it is easier to assume that the CSA can never cross a river except if they have a fort in the space (note: ironclads allow CSA river crossing like forts, but are vulnerable to removal ala the USS Monitor unlike forts).


As to the citation you mention, you need to read the next sentence. It is important not to take just one sentence out of a rule, but apply the entire rule.


Rule 6.2...Forts affect entry into a riverine port space when crossing the blue bar that is part of the port space..."


The connection to spaces is not bilateral, in that what is permitted in one direction is not permitted in the other direction when a fort and a river are concerned (see the really informative diagram on page 24 that covers this exact situation). The question I always get asked is can a Union unit move from Washington to Manassas if there is a Confederate fort in Manassas. The answer is yes, because Manassas is not a port, so the river rules are not relevant. The opposite situation is not true, you cannot cross from Manassas into Washington if there is a fort in Washington because then you are effected by the river rules.


Before trying to apply the river rules ask yourself is the space a port, if the answer is no, do not look at the river rules. The closest wargame metaphor is when a game portrays rivers within a hex as opposed to sitting on a hexside. In FTP the rivers are effectively running through the hex.


Besides the written rule and a diagram, not sure what else I can do. People who have tried this game and gotten frustrated, like the nice guy who started this thread, continue to tell me how much the river rules confuse them. I hear their pain, but the rules here are written very precisely and each concept has a diagram, beyond that all I can do is dialog with folks when they ask for help. The key is if the space is not a port then the riverine rules that continue to bother some just do not apply. I hope that helps,



Posted by markherman at 9:59 AM EST
Share This Post Share This Post
Post Comment | Permalink
Monday, 31 March 2008
CRT reply
Topic: For The People Material
That was a good post and worth an answer. There are many situations in the world which are not monotonic. Something happens in the ACW casualty statistics that is not a linear progression, but has a step function in it. It has something to do with the density of targets, but more defenders does not always equate to better results. This is why I did not go for a straight odds CRT, 2-1 at one size battle behaved differently than 2-1 in a medium battle. In fact this was an earlier controversy with the CRT, the results were insensitive to force ratios until they got near 3-1. Some thought that was highly incorrect and illogical, but the numbers are what they are.

Something that I have learned about designing games is some things are true, but not entirely explanable, and counter intuitive. I can either ignore it and let logic apply or I can actually be an historian and try to make sure it doesn't get left on the cutting room floor because it is inconvenient. So, you can sit back and go, that doesn't make any sense, but at the macro level some things are true even if they are counter intuitive, which is why they are called counter intuitive. All of my data comes directly out of a wealth of Operations Research papers that proliferated during the mid-60s, which were and remain the major sources for my ACW combat research.

1. Example #1: the Union attacks Richmond with 4 SP, the CSA defends with a fort and 1 SP. There is 0% chance of victory. Add a second CSA SP and the Union can win if they kill 2 CSA SP to 1 Union loss. Twice the defense = worse defense. The results of going from a small battle to a medium battle.

FTP has three basic defensive situations, unfortified, forts, and fortified lines (resource and capital spaces). So this example is a fort/ fortified lines situation. I should note that if there is no fort in Richmond then the Union can win a small battle, but this is a fort/fortified lines example which makes a difference. There were seventeen instances of assaults on fortified lines during the Civil War (based on Livermore's calculations). In all small battle equivalent situations on fortified lines there were no successes. At the medium battle level there were three successes. A great question is why? I have some theories, but it appears that less is more in this particular case, at least based on the historical record. The statistics are what they are... so in this example you do pick up a 1 in 18 (as close as I could get it) chance of winning a medium battle in a fortified line situation.

2. Example #2: The much discussed 1SP attack of a large force(5SP) or more. This force will always cause a 1SP loss and if can roll a 6 will cause 2 SP loss. Only risking 1SP. If you can get modifiers by leaders or defender OOS this can be lucartive killing 2 SP 50% of the time. But look at what happens if you are only attacking 4 SP. Suddenly you can only kill 1 SP 50% of the time with no chance to kill 2SP even with DRMs. Is this is not gamey, or unrealistic? A small force attacks a weaker force and has less than 50% chance of matching the results if they had attacked a larger force? News FLASH! : This is a game. Do we really want more rules?

Again this phenomena is again what it very small battles and very large battles there is little variation in losses for both sides, in essence more or less equal. However, in medium density battles there is a slight bias whereby the defender lost a bit more than the attacker. I believe that this was based on a number of factors such as some magic density of targets to maneuver room relationships and historically it was due to the CSAs ability in the aggregate of being able to create local advantage even though they were outnumbered, but the Union had a few successes when they pulled this off. Again that is what the numbers show, so this is how I captured it. It should be noted that the probability of the 1SP in a small battle being eliminated is 84% vice 100% in a medium battle, so there is an attacker disadvantage when going from a small to a medium battle (without modifiers).

3, Example #3 The CSA has a dream team with an Army with DRMs of +9. If they make a medium attack against 15 Yankee SP they can not lose unless attacking a resource space. This would require them to attack with 4 SP. However if they attack with a stronger force say 5 SP they could lose if the Union can roll a 10. They would also lose their Army and 10 SW. All for making the attacking force stronger.

This situation is taking advantage of historical hindsight to some degree as I pointed out in the earlier example concerning the medium battle bias. However, in FTP terms this example has other costs that are not being evaluated. First off, the ability for the Union to roll a 10 prior to Grant with a 15 SP army is in itself very small. You need to have McClellan or Meade with Pleasanton in a fort, or Lee has to be OOS for this to occur. I would also note that leaving the leader casualty rules aside, the Union in a medium battle is very likely to give as good as it gets due to the lopsided nature of the force ratio giving the Union a +2, since its an army the lowest drm for the Union is probably a +2, giving the Union a 50% chance to kill 3 CSA SPs, which was my statistical point that the larger the fight the smaller the variability in losses. I should also note that the Union doesn't have to roll a 10 to eliminate a 5SP army in a large battle, so the bigger message is small armies may get the edge in one transaction (battle), but payback is hell (when the 10 SP army counterattacks against 1 or 2 surviving SPs).

There is no question that there are a few arbitrage situations where players can gain some benefit from the numberical relationships of a medium battle over either a small or large battle. However, that is the point, historically that bias was there. You cannot win a game working the seams of this historical bias, but as an historian I try to make the system as faithful to the reality as I can. Hence the way things I said the CRT is not flawed unless my interpretation is flawed and I have looked at this in great detail, so it is what it is for a reason, not a mistake or something I had to fudge to make it work as a game.

Great discussion,


Posted by markherman at 4:38 PM EDT
Updated: Monday, 31 March 2008 5:08 PM EDT
Share This Post Share This Post
Post Comment | Permalink
Sunday, 30 March 2008
FTP 'suicide' attacks and OOS
Topic: For The People Material

I was at a conference over the last couple of days, so I could not enter into this discussion in any detail as my fingers get tired on a blackberry.

One of the constants in our hobby is when something occurs in a game that doesn't feel right the first reaction is to declare such an occurance 'gamey' (my personal favorite), 'flawed', or 'broken' (without moving parts this is an interesting metaphor). I have no trouble with the expression of personal views, its why people do or do not play a particular title. Where I get cogniative dissonance around is the fact that the statement is made as an ascertain of fact without any facts.

I would like to take a moment to describe the 'suicide' one SP tactic in FTP in historical terms. Playtesting back in the 90's revealed the tactic, so I had to understand it in terms of the war or I would have added rules to eliminate it. In all wargames there is some atomic level unit or said another way every game has a smallest increment of strength below which the game cannot go. Technically this is the level of simulation granularity resident within the design. In FTP the smallest increment is actually the forts with zero SP (representing 2500 gunners), but practically it is one SP (about a division: 6000 men, although this number changed over the course of the war). An issue in any design is how to handle the collision of two atoms in a perverse version of simulation quantum mechanics. Most of these interactions are covered in 7.33 and 7.34, although the rule on General casualties also accomodates this notion.

I checked this once with Dr. John Hatcher (National Park Service Superintendent of the Gettysburg Battlefield) and he tells me that they have not discovered any reason to alter the official view on casualties for this battle. My point being is the factual data on battlefield statistics has not changed in a very long time. As I am working on a new ACW game at this time, I am deeply steeped in the statistics of the war and nothing has changed that would cause me to change the FTP CRT in any manner. If it is flawed it is my mathematical interpretation of the facts. Basically in the ACW both sides in almost every engagement lost about the same number of soldiers. There is variation and some notable exceptions (e.g., Fredericksburg), but this basic fact remains.

It is also true that units rarely fought to the last man, so the CRT is based around an SP being eliminated when losing around 40% of its strength, which is calculated into the CRT in synchronization with the reinforcement and attrition rules to get to a reasonable approximation at any point during a game of how many effective divisions you have on the map. All this was preamble to get to the issue of the how I view an allowed tactic of sequential assaults on a major campaign card.

There are several historical models for this, but I think the 1864 campaign (as noted in an earlier post) is a good instantiation of my point. Now one of the downsides of a manual simulation is simultaneous movement is difficult to portray especially with three activations unless you have three hands. Grant's constructed coordinated advances for numerous forces, but in the East he envisioned an advance in the Valley, from Fort Monroe, and the main force advancing directly on Richmond. The flanking forces were small 'armies', but this is what a major campaign is simulating. And historically each of these forces although advancing more or less in a coordinated fashion were dealt with sequentially by the South in three seperate actions that permanently halted the flank activity and led to a protracted series of flank movements that were successfullly blocked until the siege of Petersburg began. So from my perspective handling a major campaign card as three seperate actions makes reasonable historical sense.

Then there is the issue of being OOS. What does it really mean in this period? First off, 19th century armies are not 20th century armies. At the operational level warfare is non-linear and there is no POL requirement, or artillery shell needs, that ties armies rigidly to lines of communication. There are lines of communication to be certain, but they have a very different impact as the forces do not require regular delivery of supplies to function at full effectiveness (which is why I treat OOS as a positive for the attacker and not a negative for the defender). Again note the fact that Lee at Gettysburg has a LOC (so in FTP terms he is in supply), but there is no rail connection or supply columns moving toward him from some distant base to keep the AoNVa in the field. It should be noted that Richmond and Petersburg fell to the Union not because they were successfully assaulted, but their LOC was about to be cut due to Five Forks and Lee withdrew. That is how 19th century generals dealt with this situation. I would also note that the more aggressive generals in FTP were willing to forgo their LOC on occasion, hence their one rating, which occurred numerous times during the war. Many have commented on the fact that CSA raids, a staple of FTP tactics, would not have occurred because they would be out of supply. Of course the Vicksburg campaign and four CSA invasions of Union territory would seem to dispute this view.

In FTP when a force is OOS, although it is still portrayed as being in a space, in actuality (the spaces are very large areas) the army is dispersed, not necessarily concentrated. So the normal dynamic of a combat whereby both sides lose the same amount of casualties can be altered as the raiding SP in our case is running into a more dispersed force and it is conceivable in this circumstance that they will give more than they get. Again the one SP suicide force is not a bunch of Union soldiers arranged like kamikaze pilots, but a small raiding column that is attacking dispersed foraging columns and catches them in an unconcentrated manner. Once the raiding column loses around 40% of its strength, the column withdraws (in FTP it is removed from the map) to be reconstituted (achieved through the reinforcement process). If you look at the CRT for this interaction on the small battle CRT (the quantum side of a 1sp vs 1sp or such encounter), the losses are usually 1-1 with no losses and the attacker being repulsed. Again a low density fight.

However, in a medium battle, the 1sp force is running into a higher density of enemy forces and a more intense fight is inevitable with the attacking SP being eliminated in every case, although usually only taking an equal number with them. The reason you get a different dynamic is the attacking force is not afforded the luxury of pulling out and a cornered force is much more dangerous than one with a way out, so the defender losses are more substantial. The attacker can get lucky with a six, but overall the attacker without drms loses 5 out of 6 battles, demonstrating the defense dominated nature of ACW combat. As you add drms to the medium CRT you get an array of results based on the ability of one side to shape the battle a bit more (picking and using terrain in a superior manner or catching the defender dispersed in the OOS situation).

Anyway this is a long discussion to say that the use of a major campaign card to conduct sequential attacks on an enemy force to gain advantage is based on a real historical model and the results that occur from this action are supportable from the historical statistics. People do not have to agree with me as most internet conversations do not change minds too often, but I think ascertians need to be debated otherwise they become facts.


Posted by markherman at 4:23 PM EDT
Share This Post Share This Post
Post Comment | Permalink
What makes a good ACW game?
Topic: For The People Material

I find Don's recent addition to the discussion to be outstanding and I hope he develops it into an article and gets c3i or ATO or whoever to publish it. If he wants I would be happy to host it on my website.

I would like to take a very different tack on what makes a good historical simulation. Don has picked up on the physics side very well, which is very amenable to calculations and data. I would like to say that all of that is necessary but whoefully insufficient.

Decisionsmakers, in this case Lincoln and Davis' and their constituents perception of their physical reality is the dominant variable. What is perceived to be true is in fact true, regardless of the facts until changed by those facts. The reason is that decisions are based on perceived reality not physical reality (the two notions can converge, but the perceived reality always wins). For example how I treated DCs vulnerability in FTP makes this point. Lincoln thought it was vulnerable, but if the game system, as calculated by how the simulation treats the DC defenses, says that it is not vulnerable (and in retrospect it wasn't that vulnerable) then the players are given the luxury of ignoring a key historical cogniative variable. As the Eastern campaign was dominanated by this perception how can leaving it out make for a better historical simulation?

Going to Taylor's point about friction. FTP deals with the activation of leaders using a friction model. Each leader's initiative can be thought of as an amount of standing friction that must be overcome to begin movement. The three rated leaders are harder to move than the one rated leaders because the card deck is not entirely composed of three OC cards. That relationship as represented by card probability and leader rating is a statement of standing friction.

What the leader rating represents is how willing the leader in question was in taking risk, particularly as regards to their logistic preparation. The same real world physics pertained to a McClellan as it did to a Grant. What I mean by that is the needs of a Western soldier were not different than an Eastern soldier, a horse eats what a horse eats etc. (I am a huge fan of Van Creveld's book, even met and talked with him once). However, Grant's perception of what was possible (ability to forage in the deep South) was very different than McClellan's view. A strictly physical simulation, which account for all of the ones used in the DoD (with the exception of mine, search for Entropy Based Warfare) make no allowance for cogniative differences.

The question I would ask how good can a simulation be if it ignores cogniative factors such as the ones I mention (e.g., DC)?


PS: I broadly agree with Dockter on his point about the enduring value of personality in modern times, but I would note that the rise of the general staff system (just beginning in the ACW period) tends to mitigate the impact of one person on a very large organization. I would say that the impact of a modern leader on an organization is he sets the tone (how aggressive, how spit and polish) which in the aggregate does impact performance, but less so than a Lee sitting on Traveler during a battle. A political struggle due to its very nature, such as Dockter's excellent Triumph of Chaos, tends to elevate the importance of personal leadership as these personages ability to instill purpose into an ideological struggle is critical to its outcome. Witness the current struggle between America and Fundamentalism, its hard to say that icons do not matter in the 21st century.

Posted by markherman at 4:22 PM EDT
Share This Post Share This Post
Post Comment | Permalink
Sunday, 27 January 2008
Fortune of War
Topic: For The People Material

Change of Fortune

Dan, The idea for this mechanic comes out of Clausewitz and a graph in an Archer Jones book. The point of the mechanic is to show the added effect that occurs when a sides fortune rises and falls over time.

To understand the strategies behind this mechanic you need to graph the relative SW over the course of a game. What you will see is a sawtooth type of activity, like a stock market graph, where there will be general trend over time in an upward or downward vector as a general expression of how one side is performing over time.

What is being simulated is you as the player are the President (Lincoln or Davis) and one of the political levers you have at your disposal is the ability to time events. So, if you know that you need to do something unpleasant politically or if something unpleasant happens involuntarily, you have the ability to manipulate public opinion by having some 'good' news in your back pocket to reduce the impact of 'bad' news on the electorate. If you have paid any attention to the current presidential campaign you will see multiple examples of this on an almost weekly basis, witness the recent release of the jobs report and how both sides reacted to it.

The strategies associated with this mechanic are to time political or military events to occur in a sequence that maximizes their effect on the electorate (SW). The basic strategies depend on whether the marker is on the negative or positive side. When you are in a negative condition, you want to arrange your affairs so that you take bad news before good news. So for instance if you were the Union and you wanted to fire your Army commander and conduct an amphibious invasion to capture a CSA coastal fort, it would be best to fire the general and then capture the fort. From a Civil War perspective just as Lincoln's opponents are attacking him for 'firing' their guy, positive military news comes out that dampens its impact (overall +2 SW for good timing: no additional effect for bad and +2 for change of fortune). However, if you were to do it in the opposite sequence you would whipsaw public opinion and feel it in the polls (-1 SW for bad timing: +2 for change of fortune then a -3 for the rapid change in the other direction).

When your marker is on the positive side you have to be circumspect about voluntarily doing something that will cause a change in public opinion. A common one in the game is the circumstances by which Davis forms the AoNVa. If you form the army under anyone but ASJ, you will take a political penalty due to the turmoil that ASJ proponents will cause you for passing over their favored son. However, this may be a military necessity and if you were holding an SW event card you might want to first build the AoNVa and then deliver the good news (-1 SW). If you did it in the opposite sequence you would suffer for bad political timing (-3 SW).

Over the course of the game, failure to pay attention to change of fortune could cost you 10 or more SW points. When the South wins the long game (gt 13 victory), they usually have less than this many SW remaining, so failure to handle change of fortune could be the margin of defeat. In a close game Lincoln could lose an election by this many points.

As the old saying goes, "timing is everything."


Posted by markherman at 7:39 PM EST
Share This Post Share This Post
Post Comment | Permalink
Debate on For The People Play Balance
Topic: For The People Material

Over almost a decade the results from tournament play have had a more or less 50-50 win-loss ratio for each side. Here is an example of some responses to that continuing debate loved by gamers, play balance.


Baron, I think the fact that you state that the CSA has a measurable advantage and the post above it states the North is advantaged speaks volumes to me. There is no doubt in my mind that you are one of the better FTP players out there. As you know I speak from personal experience here.

The CSA can win early, but the Union has to win late and if it is not played precisely the Union can blow the win. My view is the CSA position is more forgiving of mistakes than the Union. I will note that in our second game, I made an avoidable mistake that gave you the game on the last card, which otherwise would have been a Union victory, but such is life (but a very enjoyable game).

My take on your CSA strategy is you put Lee in the West, dominate Kentucky and repeatedly raid the North. You support this with a strong fort position in West KY/TN. If you do not win by mid-1863, you use your interior lines to keep the Union out of the deep south as long as possible. The other important dimension is you try to keep close to sp parity by forgoing every offensive opportunity to spread out your sps to avoid attrition. All in all a very sound strategy.

I believe that there is strong Union counter play, and as always it depends on what cards you draw, but I believe that a prepared Union player has several strong counter play options. An aggressive Union amphibious strategy is key to create counterplay. The Union must create an sp advantage by reducing the CSA reinforcement rate. The maximum CSA sp production rate without cards is 15. If the South can maintain this rate for the majority of the game, barring unusual circumstances, they will win. The Union must conduct a game of economic warfare in order to reduce this rate. The best way is to take out the Transmississippi states and Florida, close down at least two blockade zones, and hopefully keep the South down to only KY until mid-1863. A reduction of 5-6 CSA sps gives the Union a 2-1 sp production advantage. If the Union plays an offensive game from mid-1863 through 1864, the South eventually runs out of soldiers. Its easier to state than it is to do, but that is the key to Union victory.

Also, I believe that wherever Lee is, is where McClellan has to be located. A McClellan/Pleasanton army on a fort is tough for even Lee with the dream team to beat in a large battle. All in all, I think personal play style determines which side someone is better with.

As far as you playing James in a long series, have at it. I do not think that it will prove what you think it will. Across the spectrum of games and players, the statistics are what they are. I take it from your statement that you think the CSA is a lock, you may get a chance to prove this in the next round. Good luck,




This is the kind of debate that I love to see. Thanks all...

I believe that Baron has put his finger on the key point. The sp ratio is a good metric on how the war is going. I agree that when the Union achieves a 3:2 or better ratio, then they are in the hunt for victory. This is why I advocate a strong Union naval game. Take a simplistic calculation, if the Union stops two blockade sps (combination of amphib and blockade drs), capture one CSA state (e.g., FL), and keep the CSA out of one border state, the South recieves 11sp versus the Union 18, which is one sp, Union favor, short of the 3:2 ratio. The CSA can expect on average 1sp per turn from cards over the course of the game. See below for why that is the case.

There are 12 CSA reinforcement cards of which they will on average get one every other turn, for an approximate average of 1sp per game turn over the course of the game. This gets the CSA to 12 (given the assumptions above) and the 3:2 ratio mentioned. The list of CSA cards is (Card#:No. of sps; ?=SW penalty for sps). #17:3, #21:3, #22:3, #23:2, #33:1, #37:3, #38:2, #41:2, #47:2, #79:1, #80:3?, #106:3

The Union gets 5 cards or two per game for an approximate total of 1sp every other turn.

The list of Union cards is (Card#:No. of sps). #9:3, #10:3, #30:3, #40:5, #43:1

Of course every game is unique and is strongly effected by the cards you actually get, how those cards are played, losses due to combat, and how the two sides respectively handle attrition, which on average benefits the CSA. CSA players who use the last card play to launch a raid every turn are usually gambling that they can win the game before the Union reinforcement rate buries them. An alternate tactic is to spread some forces out to make better use of local supply and reduce attrition, which the South often did historically.

These are the reasons why I continue to point out the importance of the Union naval game and offensive actions against small CSA states (e.g., TX, AR, FL and LA). If the Union can create a long term 3:2 ratio or better on the board, then the Union chances for victory are good, less than that and barring significant SW advantage, the Union is probably losing.

Hopefully this analysis will help explain my earlier posts.



Posted by markherman at 7:37 PM EST
Share This Post Share This Post
Post Comment | Permalink
Campaign Card Distribution
Topic: For The People Material

This was in response to a thread around Campaign Card distribution…


Great thread unfortunately based on one of our brothers experiencing campaign distribution syndrome (CDS).

What I would like to contribute to this debate is how I believe multiple CCs give an advantage, but are not a show stopper in it of themselves. I have had to think deeply on the issue of multiple campaign cards and over the course of time I have come full circle back to where I began in the AH version of this game, no CC limits.

It is important to note that there is a basic assumption that a skewed distribution of CCs is a decisive advantage. There was also a good point made that the Southern railnet and logistic conditions would have prevented this during the real war.

FTPs military model runs off of two key factors: maneuver and logistics. The Union won the war by prosecuting a persisting logistic strategy. The CSA was trying to win by prosecuting a raiding combat strategy. FTP is built around this set of assumptions and concepts as interpreted by me from a generic model articulated by Archer Jones.

The maneuver rules are expressed through the play of cards in combination with general ratings and board position. How much maneuver a side can sustain is based on its logistic position over the course of the game.

One of my goals when I did this game was to design logistics into the core of the system, so the players would abstractly deal with it without any mechanics. I focused on SPs as the integrated logistic system for the game.

The relative reinforcement rates of the two sides is the basic logistic production model for the game. The Union has a large and relatively steady rate of SP production and the ability of the CSA to impact this rate reflects a successful application of their raiding combat strategy. The CSA SP production rate is represented at a more granular level. If the Union applies its historic persisting logistic strategy, the South will reflect the historical collapse. If the Union fails to steadily take the Southern logistic system apart (blockade, ports, state bureacracy and manpower), then the South does not collapse. This is not the historical outcome, but neither is the CSA player achieving a more than historical level of logistic cohesion.

The attrition rules reflect operational logistics by causing players to disperse their forces at the end of a turn if they do not have a good offensive option. This is exactly what Lee did several times during the war for logistic purposes. If you spread out attrition goes down (broader access to local logistics). You stay concentrated, disease and logistic considerations wears your force down faster. Since the CSA usually goes last, the Union has to remain concentrated more of the time, simulating, without overhead many of the war's logistic dynamics.

What should and does happen is the ability to score on your opponent with multiple CCs requires that you have sufficient SPs to expend in combat and maneuvers. Without sufficient SPs having multiple CCs will not yield decisive results, just a maneuver advantage. Having a 10-1 advantage in CCs is possible, but without considering the context of relative SPs and board position, it is very hard to evaluate how much of an advantage this really is. The fact that Dirion has not yet put the Union away by this point indicates that he hasn't been able to fully capitalize on Taylor's CDS problem.

As far as the point that the South couldn't accomplish this during the real war, the answer is the game supports this under certain circumstances. FTP is a model of the war, but players effect the model. If the South is allowed to have a surplus of SPs, this is due to the Union failing to reduce the CSA logistic infrastructure. In this case the historical result is no longer valid for that particular play through, because one of the key historical factors has been altered.

It is true that the Southern rail network was inefficient and poorly laid out for military operations, but somehow the South was always able to pull off concentrations when they wanted to during the war. The game only has two strategic concentration cards (Shiloh, Chickamauga), so statistically in a standard game, the CSA should get two of these per game matching the historical record. However this is a simulation (game) not an emulation (book) of the war, so variation is both intended and desirable.

The bottom line is if you pay attention to SP management, multiple CCs can be very useful, especially to the CSA raiding combat strategy. If the Union works to first gain logistic superiority via a persisting logistic strategy as happened in the historical model, it will not matter much how many CCs the South gets late in the war, because they will have insufficient SPs (as they did historically) to do much with them. This inability to do anything useful with the extra CCs simulates late war logistic problems (Richmond calls for action, but not much happens).

The opposite also applies, in the late war when the Union can do alot with multiple CCs, since their reinforcement rate (logistics) is large and steady under most circumstances.

I hope that helps. I am going to look over the game in question just to be empathetic, but FTP isn't chess and chance plays a role in the outcome. However, unlike chess where no matter how many times you play Kasparov you are guarranteed to lose, in FTP hope can spring eternal; at least once and a while. James, one of these days...


Posted by markherman at 7:36 PM EST
Share This Post Share This Post
Post Comment | Permalink

Newer | Latest | Older