|Products||| Napoleon’s Triumph||| Design Diary||| The Sound of Silence|
It has been over 4 months since the last design diary entry; rather a long delay given that the plan was for one or two entries a week. In that time, there have been numerous inquiries about the status of the design, and those asking have received very little in the way of a response. The reason for this isn’t complicated: the design was stuck and there was nothing at all happening with it. This shouldn’t be understood as meaning that I wasn’t trying to make progress, but rather that my efforts were yielding little or nothing of any use.
The problems with the design were numerous, and in fact were very difficult to even sort out into any kind of structured form, and few problems can be solved when they cannot even be articulated. Within the last few weeks, however, I’ve been able to make some headway, and would like to share the problems as I currently see them. Hopefully, later diary entries will deal with proposed solutions, but for now an enumeration of the major problem areas must suffice. They are as follows:
I am Clausewitzian enough to make the first object of almost any wargame design the destruction (physical or moral) of the enemy army, but that as an object is in and itself is generally insufficient for a wargame. This is partly for the prosaic reason that mapboards have physical limits and these limits require the action to be kept within the board (years ago, SPI did an Austerlitz game in which the Allies could pretty much guarantee themselves a marginal victory by running away; refusing battle was not historically an unwise Allied strategy, but it certainly didn’t make for much of a game). To resolve this problem, it is generally strongly desirable to attach additional victory conditions tied to territory on the mapboard. The main justification for this from a game design perspective is to keep the action from drifting out of frame, but there is a historical justification for it as well: traditionally, the victor was considered to be the side that held the battlefield at the end, so victory conditions that correlate to possession of the field do not have to be ahistorical.
There is however a larger reason for the inadequacy of the destruction of the enemy as the sole determinant of victory, and that is that that objective in and of itself does not place the burden of attack on either side. If we agree with Clausewitz that the defense is the stronger form of war, the failure to assign the burden of attack becomes a critical failing: neither side has been given an adequate reason for risking an attack. Of course, it is possible to simply state that the result of a game in which the two sides do nothing but stare at each other from a distance as a draw, but draws are the most unlovable of outcomes for games, and to have a design that actually penalizes the first side to try to avoid that outcome is to have a design with a severe defect at its heart. One way to avoid this is to put territorial objectives into the game such that force one side to do something or lose the game by default, and it is this imperative for one side to act that is the engine that drives the game.
This sound general principal, applied to Austerlitz, however, yields something of a puzzle: where should the burden of attack fall? If we look at how the armies behaved historically, the Allies attacked first, but it was the French who had the more urgent need for a battle: there has been a broad historical consensus that the Allied attack was a mistake: time was on their side. On the other hand, a game in which the burden of attack is on the French simply doesn’t feel like Austerlitz, and if the Allies were not to attack, what were they doing there at all? This ambiguous historical background fails to yield a clear answer to the problem. Further, the armies were (apart from leadership) fairly closely matched, and the French leadership advantage is part of history, but cannot be part of the game. (Who would want to play the Allies if they were forbidden by the rules from managing their affairs better than Alexander did in the historical battle?) Thus, each army would have its hands full attacking the other, creating an imbalance for either side if it has the burden of attack placed upon it.
I was of course aware of this problem very early on in the design, but my original ideas to solve it simply didn’t work. I played around with secret victory conditions, random victory conditions, purchasing of victory conditions (basically using morale points as currency), but none of these seemed satisfactory. One of the reasons I had chosen Marengo as a subject for my last game was that the burden of attack so clearly rested on one side. Structuring the victory conditions was easy conceptually; the only problems were those of fine-tuning them for balance. In the Austerlitz design, however, even after months of thinking, it remained unclear how this critical design element was to work.
Most players would, I suspect be puzzled by what is even meant by this “problem”. If, after all, the subject of the game is the battle of Austerlitz, then surely the scope of the game is geographically defined as the dimensions of the battlefield while the time-frame would be the day of the battle. Apart from some “edge conditions” (pardon the pun) surrounding the exact definition of what constitutes the physical limits of the battlefield, it might well seem that there really wasn’t much to say about it.
Most designers, however, might be more sympathetic to my citing this as a problem. What may seem obvious to the player is not always obvious to the designer. The first scope problem was to what extent should the game presume the historical deployment. Does the scope of the game include that time prior to the start of the game clock when the armies were deploying? The amount of freedom permitted in set-up had numerous significant effects on the game. The most important of these was replay value; the more constrained the set-up, the more players would be locked into certain lines of play. On the other hand, start-up time was also important: in general, the freer the deployment, the longer it would take, and I have always viewed start-up time as a decided negative that should be minimized as much as possible. The next effect was on historicity: free set-up tends to create a puzzle as to how the army got there (and what the other army was doing at the time). Finally, set-up affected limited intelligence: the less free the set-up, the more intelligence the other side gained from their knowledge of the historical battle. All of these effects made the set-up problem difficult, but of these, the problem of lines of play was the one I thought most critical: it simply wasn’t reasonable to expect people to pay much for a game that could only be played a few times before they would feel that they had done pretty much everything in it that could be done.
One approach to resolving the set-up problem was to open up the scope of the game in terms of time. The battle may have been on December 2, but the game did not necessarily have to begin on that day. By moving the start of the game earlier, players would have fewer constraints upon their lines of play. There were two negatives associated with this idea. The first was that starting the game earlier meant increasing the number of turns. Increasing the number of turns meant increasing the playing time: all things being equal, more turns means more playing time, and I was firmly committed to the idea that Austerlitz was to be a game that could be played in a single session. The second negative was the potential for a dull opening game in which the Allies deployed for turn after turn while the French sat on their hands and waited.
The scope problem was difficult, and was made still more so by its strong interaction with the game objectives problem. As a result, I tended to address the two problems together, hoping that some magic with the game objectives could solve the problems of the game scope: my particular hope was that I would be able to go with a December 2 start-time and a largely historical set-up (with just enough freedom to make limited intelligence work) and have the victory conditions introduce enough variability into the game to open up multiple lines of play and hence give the game substantial replay value. Numerous efforts and ideas, however, failed to yield the desired results, leaving the scope problem unsolved.
Bonaparte at Marengo had a highly abstracted command system. While this system had few virtues from a simulation perspective, it was cheap in terms of complexity, and was a great positive in terms of the game experience. Without in any way feeling that I had made a mistake in the Marengo game, I wanted to enlarge the complexity budget for command in Napoleon’s Triumph, and see if I could come up with something that was had some real claims to being a simulation of Napoleonic command, and which also added to the enjoyability of the game. I came to characterize this as the two-command-level problem: really the player of the game is not strictly speaking taking the part of the army commanders: their scope of decision making is too broad; it really comprises the army commander as well as the corps-level commanders at the same time. This two-level command problem meant that really two distinct sets of command problems had to be simulated: the highly localized problems faced by the commanders of the corps, and the broad battlefield management problems faced by the army commanders.
My initial attack on this problem was to simply put corps designations on the playing pieces and promote keeping the corps together through rewards and penalties in combat. (Rewards and penalties in movement might have appeared the more obvious idea, but limited intelligence caused trust issues for such rules: if the players couldn’t see which enemy units were being moved, how could they be sure that any command rules on movement were being obeyed?) A number of problems, however, nagged at me, and left me increasingly uncomfortable with the solution. The first of these was a concern about increased production costs from increasing the number of different types of pieces in the game. The second was a closely related concern about manufacturing quality control: in Bonaparte at Marengo, a defective, damaged or lost piece was of little consequence: extras were available for all the types of pieces needed for play; however, putting corps designations on the pieces made the pieces harder to replace, which would substantially increase the percentage of customers with problems in their pieces inventory. The third was that it slowed set-up time: players had to match not only the types and strengths of pieces during set-up, but corps designations as well. The fourth was that it badly compromised limited intelligence: once players knew where one unit of a corps was, this gave them a great deal of information about what other pieces were likely to be in the area, through the application of historical hindsight. The fifth was that it restricted the lines of play for both sides, since it dictated to an extent what corps would be capable of what missions. Finally, and most fatally, it just didn’t work very well in doing what it was intended to do: it did little to introduce the sense of a two-level command system.
Road movement is critical in Bonaparte at Marengo, but the rules transferred literally from that game worked very poorly in Austerlitz. The first problem was the usage levels of local roads. At Marengo they were used very little, whereas at Austerlitz a number of major corps movements in the battle were by local roads, moves that would not be possible under the Marengo rules. An additional concern was that the Marengo rules, although they provided a generally realistic depiction of how road columns worked, were mechanically awkward, especially when the unit and road densities were high, conditions that very much prevailed at Austerlitz.
A new set of road movement rules had been devised that were (somewhat) less mechanically difficult than those used in Bonaparte at Marengo, but the mechanical improvements were not enough to offset the problems resulting from the larger armies and their (much) greater use of dense and interlocking local road networks. I felt that the road movement rules in Marengo were at the upper end of complexity that I was willing to live with, and that Austerlitz simply could not be more complex still: they had to be simplified.
Added to the complexity problem was the problem of simulating what a Napoleonic command system was capable of in terms of traffic control. The command rules in Marengo mandating each unit moving by local road to cost a full command may have been crude, but they did a good job in preventing units from using the local roads in a manner that simply would not have been possible historically, because such complex movements could simply never have been coordinated by couriers riding back and forth over the battlefield. The armies at Austerlitz had much greater knowledge of the local road network than the armies at Marengo, and made much greater use of it, but they still were gravely limited in the complexity of the movements they could try to undertake, and it was critical that these restrictions be simulated in some way, or the movements of the armies in the game could be wildly ahistorical. Added to the strict simulation problems, permitting such complex movements would drastically slow playing time, as players would have to invest considerable effort working out which pieces would be crossing which approaches at which times, as they made their various criss-crossing moves across the battlefield. Given the amount of time I had invested in the not-entirely-satisfactory road movement rules for Bonaparte at Marengo, the prospect of having to devise rules that would have to be an order of magnitude better in some ways was a daunting and disturing prospect.
The original assault rules for Bonaparte at Marengo were considerably more complicated than those that were ultimately published. The original rules included an elaborate counter-attack mechanism. This mechanism, which was intended to simulate my understanding of Napoleonic shock combat, proved overkill for Marengo: the complexity of the mechanism just didn’t provide enough bang-for-the-buck in that game and a much simpler mechanism proved adequate (the shock rules in Bonaparte at Marengo still retain some little-used features, such as defensive artillery fire and cavalry pursuit that arguably could have been dropped, but I retained them because the historical outcome could not have been simulated without them).
The simplified assault system, however, did not prove adequate in Austerlitz. The higher force densities exposed its limitations, both in simulation terms and in terms of player interest. The edge they gave to the defense increased as the force density went up, making approaches held by more than a few pieces all but impregnable within the time frame of the game: the losses required to weaken them to the point where they could be taken were insupportable, and the process was so slow that defensive reinforcments could be easily rushed to the spot in time. The rules tended to produce a WWI sort of feel, which was hardly what I wanted.
Quite a bit of work went into variations on this system, mostly having to do with restricting player choice as to leading units in an assault (this consisted in large part of making such choices “sticky” in some sense – such that units that led a defense in one turn would have to keep leading the defense in future turns) combined with a restored counter-attack system, in which each side could commit additional pieces to an assault, increasing the assault’s tempo and the likelyhood of a decisive, rather than attritional, result. At this point, I feel reasonably confident that the right choice exists somewhere in that mass of variations generated by this process, , but I am not yet sure exactly which one will prove optimum. Making the determination on this point is probably going to have to fall out of the playtesting process.
Recent weeks have been interesting. I’ve had a number of new ideas on all of the major problems, ideas which have some promise to address all of them. I’ve cleared off the game workspace again, and development has actively resumed (one symptom of which is the resumption of the design diary). It is my hope and intention that later diary entries will go into these new ideas, but it seemed appropriate to review what hasn’t been going on (and why it hasn’t) before going into these new directions. One thing is for certain, however: for a design that I had originally hoped would be a quick project, it has proven to be anything but. The only cheering thing for me is that my favorite recording artist, Kate Bush, (not that I expect that name to be familiar to most folks reading this) will release a new album in less than two weeks – after a mere twelve year interval since her last one. I now have a schedule goal that doesn’t depress me whenever I think of it: to complete Napoleon’s Triumph in less time than it has taken Kate Bush to complete Aerial. Of course, while that goal doesn’t depress me, I suspect it may be possible that it will depress you. Still, we can all hope for the best, and thanks to everyone who’s written for their interest and good wishes.