The most recent issue of The Courier opens with a substantial piece by the late George Jeffrey, edited by Ned Zuparko. Over the years Jeffrey became an interesting sort of wargamer celebrity: a man who was well-known and well-regarded for something that never happened. Everybody talked about his "Variable-Length Bound" game, everybody praised it, and nobody played it. That was mainly because - although very few people ever admitted as much - it simply doesn't work. My own encounter with this concept began about twelve years ago. In those days Dan Beattie owned a hobby shop called "The Standard Bearer," and knowing of my fondness for avant-garde game systems, he handed me a little blue five-dollar booklet (Xeroxed from typewriters), and said I should try it. It was an incarnation of the VLB system, circa 1989. As with all VLB games, George Jeffrey was listed as the inspiration, but somebody else had actually written out the rules. On the off-chance that there's somebody out there who still hasn't heard about it by now, I'll try to explain VLB in a (very small) nutshell: Rather than having turns which represent fixed increments of time, with fixed divisions into phases and so on, VLB supposes that we can fast forward time until we need to pause to react to some "Change of Situation. " (COS) Instead of expressing unit capabilities in terms of what they can do in one turn, we can express them in terms of effectiveness per minute (how far they can march, how many cannonballs they can shoot, etc.) The opposing sides "dialogue" with each other to determine when COSs occur, and therefore when something has to be resolved. There are eight possible COSs, such as seeing a new enemy, or being threatened by an existing enemy, or getting new orders from a higher echelon, or so on. Only when a COS occurs may you change the orders of units under your command. The hoped for outcome is that a game compresses all that time it doesn't need, such as units moving when they're not close to the enemy, and thus also reduces player control by only allowing for changes in plan when some sort of COS warrants a change. Those changes (the sending of new orders) have to travel up and down the command chain by courier, and thus also take time. Brilliant So I thought: "This is brilliant!" I immediately tried to put together a game using the rules. It sort of worked for about an hour. Over the course of the next year I tried VLB again and again. People in my gaming group had become so familiar with the VLB concepts that they were actually calling a "Change of Situation" a COS (pronounced "koss.") We tried tinkering with this and that, we tried zooming in and out, scaling the game differently, inventing all sorts of limitations on commander omniscience, but no matter what we did, no matter how familiar we became with the principles of VLB, the game always broke down for the same reasons. They were: 1. As soon as the two armies get well engaged at multiple places along the front, there are so many Changes-of-Situation (COss) being generated, that it becomes virtually impossible to sort them all out. The game's "bounds" get shorter and shorter, as we have to respond to a zillion different COss. 2. Confusion mounts as the simultaneous movement crashes against the variable length of the bounds, and the result is seven or eight little games, each happening in various sectors of the board, each on its own clock. And since battles are messy by nature, some of those little battles will crash into others (like a cavalry charge breaking through somewhere), causing yet more consternation. 3. With so many COSs being generated, and the VLB system requiring real-time transmission of reports of them (and then, orders in response to them), a million little slips of paper are flying around, each with the futile purpose of "telling" me something I can plainly see for myself, but which I can't react to, until I get this little slip of paper telling me about it. As a result, one loses track of what he does and doesn't "know" in the game-world. After something like a dozen VLB games we gave up, never to return. The group consensus was: "This would be great if it could be run by a computer that would keep track of everything, and do all the updating for us." But nobody, to my knowledge, has ever tried to make a computer-based VLB game. The basic premise of VLB is tantalizing. Who could argue with the idea that we should get rid of fixed-length turns and try to compress time and thus "feel" the battle like a real general would. But over the years I've noticed that most people who talk about VLB rarely get past that initial bout of praise for the noble goals. In actual fact, aside from the glaring technical problems with its application, VLB also suffers from some basic theoretical problems: 1. Hmm... I Sure Feel Like my Situation has Changed... Jeffreys wrote: "...the player cannot react to the information for any of his commanders for whom the information doesn't cause a change of situation." But this is simply not true. All players stand over the table. All can see what is happening to all other players. You might have this game mechanism that prevents them from saying they're reacting to something they shouldn't know about, but what's to stop a player from moving his command figure to a hilltop, just coincidentally enabling him now to see that exposed enemy flank. And now that he can see it: voila, there's a COS. You could only prevent something like this from happening by having a straitjacket rule, which basically says that corps and divisional commanders can't move, can't turn their heads, can't do jack, until they receive a message of some kind. But is that realistic? If you do that, you have the absurd problem of a commander frozen in place. And even that's not workable. Let's say that Marshal Ney is advancing his corps toward the town of Schmuck. Over on his left, but just out of sight, the Russian cavalry has broken through Lannes' corps and now threatens his flank. According to the VLB system, this will be a COS for Ney only after: 1) Lannes reports it to Napoleon, and 2) Napoleon issues new orders to Ney, saying, "Watch the left flank!" But hey, Ney is moving, right? So what's to stop him from deciding that he needs to move up on that little hillock over there, and all of a sudden he can now see the trouble for himself. Had the disaster in Lannes' corps not occurred, the guy playing Ney would never have seen the need to get on top of that hill and take a look. Are you going to try to invent a rule that would forbid him from doing it? Good luck. So clearly, Jeffreys was wrong: what players can see always affects what their character/generals can "see." They will find a way to react, like it or not. You can't shoot down the Command Helicopter. 2. I Have Decided to Take the Peach Orchard Jeffrey wrote: "Whilst a commander is free within the limitations of his control orders... he is most definitely not free to `do his own thing.' ...A commander is always acting under orders." The VLB system places total faith in the obedience and cooperativeness of subordinate generals. If Ney gets his orders to move to Schmuck, then he'll move. If he gets orders to stop, he'll stop. The only exception allowed is to act in defense against an immediate threat. And even after that, Jeffreys writes, "As soon as the threat was removed, however, the brigade commander would automatically resume operating under the divisional commander's orders, return his command group to his previous position, and reform his line." History is replete with examples of subordinates "doing their own thing," sometimes brilliantly, sometimes disastrously. If we played Jeffrey's system for a Gettysburg game, then Sickles would never advance into the peach orchard. Murat would never envelop the Allied left at Dresden. Steinmetz would never ignore Moltke's orders at Gravelotte, the Union assault at Chattanooga would never have happened, and so on, and so on.... In short, the decisive events in about half the battles of history can not be replicated using the VLB system, because there is no randomization, no provision for insubordination or confusion. VLB defenders might retort by saying, "No, see, one of the things that causes a COS is the receipt of a message with some new intelligence, and that message doesn't necessarily have to come from a higher-ranking commander. So perhaps the Dan Sickles player gets a report from one of his brigadiers saying that the Rebs are falling back, and this can prompt him to act in a new way." That's an example of: 1) rationalizing as a gamer in order to cheat a bad rule system (see my observation about Ney getting on top of the hill, above), and 2) one of the problematic contradictions of VLB. According to one of the core VLB principles, Sickles as a corps commander can not change his own orders. Only Meade, his boss, can do that. But a commander is allowed to change the orders of units under him, upon receipt of a COS. So... is Sickles allowed to advance into the Peach Orchard or not? How would changing the orders of your subordinates not be the same thing as changing your own orders? (Perhaps Sickles can send one of his divisions into the orchard, but not both? Who knows.) 3. Good God, are all these My Hats? The reason our gaming group finally decided that VLB needed a computer was that there was such a blizzard of information to keep track of. Playing VLB made us long for fixed-length turns because at least in a fixed-turn system we could know when this message got to that general, and when this gun actually fired on that target, and so on. In a VLB game you have to be the master of dozens of little generals, each reacting to different things by sending reports of COSs back up the chain of command, and each in turn waiting for new orders to come back down. Sure, this is probably how it really worked, but is it a good game? It doesn't take much imagination to realize that, with a corps of about 12 brigades, plus artillery, fully engaged with an enemy corps, there are going to arise all sorts of complicated situations that desperately need some logical order to resolve, because the outcome of each bears upon the outcomes of others. That artillery unit fired on you for 5 minutes, but then you passed by it, exposing your flank, which means its fire was more effective for another 6 minutes. Oh, but wait, three minutes into that last six, it was attacked by your cavalry and forced to fall back. No, wait - that other enemy battery over there was able to fire at your cavalry before it hit the first enemy battery, meaning it was shaken going into the charge and thus didn't rout the gunners... Oh hang on a second - this is all pointless because my first infantry unit would have had to stop after four minutes because of the enemy cavalry on the left... Oh, no, I forgot, they'll be falling back at that time, because they would have become shaken two minutes earlier, due to This is the curse of simultaneous movement systems. The invite a swamp of arguments about who did what to whom, in what order. VLB makes it even worse by getting rid of turns and phases. Jeffrey may have thought he could speed up wargaming by compressing time, but in reality what he did was create a game with an unlimited number of one-minute turns. Wargaming is full of rules that work perfectly as long as players use them in exactly the way they're illustrated in the designer's examples. ("My combat system is flawless, as long as you attack units one-on-one, in only one terrain type, only from the front..." etc.) But wargamers will never play your game the way you think they will. It will take about five minutes for them to generate every possible contingency under which your system doesn't work. So good rules have to be simple and universal in their applicability. Whenever George Jeffrey gave examples of how VLB ought to work, he cited one or two brigades in action, sending back reports to a commander, who then sent them up the chain of command to another commander, and so on. That's a nice image, until you realize that no wargamers on earth are going to sit still and do nothing while only a single brigade at a time gets into harm's way. Look at any convention game, for example, and see the whole army surging forward at once, and tell me that a game system based on simultaneous movement and written reports of every COS is going to work? Napoleon's army at Waterloo had something like 40 brigades. Can you imagine keeping track of the orders for 40 brigades, when each of them is constantly generating reports where either it or the enemy has done something to change the situation? The result is that old wargame standby: after a while everybody cheats and just does whatever needs to be done at the local level, since he's sure that there are enough COSs to warrant making whatever changes need to be made. That's how our VLB games always ended: they degenerated into no command system at all. With this recent Courier article (which closed with a promise to try to publish some kind of VLB game someday soon) the Variable-Length Bound seems poised to become wargaming's Holy Grail: the unattainable goal which everybody thinks is excellent mainly because nobody's been able to do it and therefore prove that it isn't. Since I have already got a reputation as the "critical" guy, then let me be the first to say it: We need a better Grail. Objective-Horizon Games I was reading MWAN117 a few months ago and came across Jim Getz's piece on his new Piquet module. I generally like Piquet, although it has a few too many gadgets for my taste: clocks, cards, calipers, and all those different dice. But Getz really caught my attention when he casually mentioned that his variant uses "Objective-Horizon Movement." He elaborated in a lengthy footnote, and mid-way through his explanation I was sitting straight up on the sofa thinking: "This is brilliant!" The idea here is that the player designates an objective for his unit's move, and then he places a little "picket" (a single figure in a smaller scale) on that objective. He calculates how long it will take the unit to get there. When that time has expired, the unit is moved all the way there, in one wollop. You compress time, in the sense that instead of moving four 9" increments, you simply move one 36" increment, but only after four "turns." (Since he's playing Piquet, the turn system isn't really the same - he's waiting for the right card to allow him to do what he wants to do.) He also achieves the nifty result of forcing a player to commit to a move in its entirety, and not having four turns to think it over, react to changing circumstances, and so on. I immediately set it up and began experimenting. And instantly there were frustrating problems: 1. If the line between your unit and your picket lies across any enemy's path or possible path, you have all sorts of potential conflicts. Imagine, for instance, a field like the face of a clock. Let's say your unit is at 6 o'clock, and you're moving to attack an enemy who's sitting at 12 o'clock. It will take you four turns to get there. But there's another enemy unit at 2 o'clock. Now what? When, exactly, does he fire on you? And what if he's moving, but he started at a slightly later time, and is moving further, and not at a perfectly perpendicular angle to your move... It's like a naval game; you're trying to calculate when exactly your courses cross, and then what happens as a result. Otherwise, you have the weird illusion of teleporting suddenly from one part of the field to another. 2. This system doesn't go very well with most notions of command control and unit integrity. Let's say you've got some units in motion, some not, and of those in motion, some are moving to different objectives than others. At any given moment, do you know where all your units are? The ones that are "in motion" aren't actually moving on the table - they're sitting right there with the ones that are, well... sitting right there. Mating this concept to the Piquet system could potentially make the problem even worse. Because Piquet allows for wild swings of luck on that all-important D20, it is not only possible, but commonplace, to find units zipping all around each other without consequences. What does this have to do with command control? Well, it depends on whether or not you want your commanders to keep their units in some sort of order and relative formation. In order to do that, you have to know where they are. I think the Objective-Horizon concept has a lot going for it. I especially like the notion of committing a unit to an objective, regardless of how many moves it takes to get there. But here again, this is something that would a lot better on a computer screen, where a constant flow of time can be modeled, than on a game table, where little lead men have to be picked up and put down with our hands. The issue is one of events in real time: picking up a stand of figures and putting it down somewhere else is an event. It is done at a certain moment. It is not a process. And what Getz seems to want is a model of a process - the process of moving, rather than the event of moving. I agree: that would be nice. But you can't really use a discrete event to model anything other than a discrete event. The rest is imagination, and that's where we get into trouble. Now What? Back to MWAN # 121 Table of Contents Back to MWAN List of Issues Back to MagWeb Magazine List © Copyright 2003 Hal Thinglum This article appears in MagWeb (Magazine Web) on the Internet World Wide Web. Other articles from military history and related magazines are available at http://www.magweb.com