The situation at first light
Two commanders were wrong about each other, and both stayed wrong for way too long.
Napoleon believed the army in front of him on the Landgrafenberg was the main Prussian army. It was not. Late on the 13th the larger part of Brunswick’s force some 60,000 men were marching away from Jena toward Auerstädt, and Napoleon was dictating orders on the assumption they were still to his front (Chandler, The Campaigns of Napoleon, p. 640). What he could actually see burning on the plateau was Hohenlohe’s command alone: Holtzendorff’s 5,000 out toward Dornburg, Tauentzien’s 8,000 around Closewitz and Lützeroda, and about 25,000 more south and west of Vierzehnheiligen, with Rüchel’s 15,000 still sitting at Weimar (Chandler p. 640).
Hohenlohe was wrong in the opposite way. He had no suspicion he faced the mass of the Grande Armée, “believing that the French troops on the Landgrafenberg only comprised a flank guard” covering a French advance north toward Leipzig (Chandler p. 640). He had been given a rearguard’s job and he read the noise in front of him as a rearguard’s problem.
The night before had been spent getting an army up a goat track. One narrow road led from Jena to the plateau and it had jammed solid with more than a hundred transport wagons & carriages. Napoleon took a lantern and supervised the widening himself. Marbot, then a young officer on Augereau’s staff, describes the method: four thousand pioneering tools sent for from the engineer and artillery wagons, every battalion working an hour in turn and then forming up silently on the heights while the next took its place, the torchlight hidden from Prussian eyes “by the blaze of Jena beyond” (quoted in Chandler p. 637). By dawn 25,000 men and 42 guns were on the summit.
They were packed in tight enough that Marbot thought it should have been noticed: “the breasts of the men of each regiment were practically touching the backs of those in front of them… although the enemy were only half a cannon shot off, they perceived nothing” (quoted in Chandler p. 642).
The fog is why. Every source reaches for it independently in Chandler he has it “considerably” hampering Napoleon’s final reconnaissance (p. 642); Fisher says he has men claiming they could not see thirty feet (The Napoleonic Wars, Osprey, p. 57); Van Creveld has Lannes’ troops simply “lost to the emperor’s view” until the mist lifted around nine (Command in War, p. 107).
What the opening scenario is about
The scenario opens at 0600 with Cospeda already in French hands, which surprised me until I checked. It is correct. The fighting on the 13th had driven Tauentzien’s advance elements back onto Lützeroda and Closewitz (Chandler p. 636), and Lannes’ corps assembled that morning on the sheltered plateau below Cospeda (Fisher p. 56). The village was part of the lodgement before the battle proper began. What Chandler describes on the 14th is Gazan attacking beyond Cospeda and being repulsed not a fight for the village itself.
So the first two and a half hours are not a battle for ground. They are a fight for room to deploy. Napoleon’s whole problem on the morning of the 14th was that he had an army arriving all day on one road and nowhere to put it. Chandler’s verdict on how he solved it is worth keeping in view, because it is not flattering to either side: the French got their deployment space “thanks largely to the enemy’s unnecessary withdrawals” (p. 644).
What historically happens in the window: Suchet takes Closewitz on the right and holds the wood by about eight. Gazan is repulsed on the left. Tauentzien, confused, gives up Closewitz, Lützeroda and Cospeda’s approaches, keeps skirmishers in the woods, and falls back on his Saxon reserves then he rallies five thousand men and counterattacks hard enough that the French “reeled back in confusion,” only one squadron standing, the rest split into two groups at Krippendorf and in the Isserstadt woods (Chandler p. 643). And then the Prussians decline to press it, turn about, and march away west.
The game
Vive l’Empereur, Volume IX: Trois Batailles en Allemagne. Regiment-level, with each counter carrying fire value, melee value, steps, morale and movement allowance, plus a range for those units that have one. The left-hand strip gives corps, division and unit so V-1-17 le reads as V Corps, 1st Division, 17th Légère.
That numbering matters more than it looks. Historically Suchet had V Corps’ 1st Division and Gazan the 2nd (Chandler p. 643), so the game’s divisional split is the Suchet–Gazan split, and which division you push where is the historical decision restated via the counters.
What Lannes has on the table:
The two légère regiments are the sharpest things in the corps fire power rating 4 and a range of 2, where the line regiments sit at 1 to 3. Worth noting that Fisher p. 58 names the 17th Légère specifically as the regiment Napoleon sent forward first that morning, and it advanced slightly oblique into flanking fire from Zezschwitz’s Saxons. The game gives us the same instrument history used, with the statistics that make you want to use it the same way.
What’s in front of him: Tauentzien’s detachment: 7th Fusiliers, 45th, 5th and 12th Line, a grenadier regiment, 7th Hussars and 3rd Chevau-légers all reinforced by the 3rd and 4th Grenadiers and one battery from von Prittwitz’s Third Division. Ten units holding the triangle of Closewitz, Lützeroda and the Dornberg. They are shallower than the French: five or six steps against seven to ten, and morale down to 7 where every French line regiment is a 9.
Note who isn’t there. Von Zezschwitz’s Saxon division is on the battle’s full order of battle but not in this scenario, so the flanking fire Fisher describes hitting the 17th Légère on its oblique advance (p. 58) has no counterpart on the table. The scenario is a tighter thing than the morning it depicts.
The scenario: “Opening moves in the fog”
Six turns, 06:00 to the end of 08:30, half an hour to the turn. No reinforcements on either side.
The French must hold all three villages at the end of turn 6. Take Closewitz and Lützeroda, keep Cospeda. Fail on any one of them and it is a Prussian victory. That is a harder standard than the history sets: historically Tauentzien gave up the forward villages and Napoleon got his deployment room, which in the scenario’s terms is a French win, but the game asks you to do it inside six turns against a defender who cannot be dislodged by manoeuvre alone.
The Prussians are permanently out of command for the whole scenario, and no leader is present. In place of command they get a cohesion rule: each good-order unit must stay within two hexes of another. So they cannot advance, cannot reposition, cannot exploit — but they can fire, melee, change formation and facing, counter-charge, and advance after combat. A kind of “Not 1 Step Back” moment!
That is Hohenlohe’s morning turned into a hard rule rather than a decision. Chandler has him with “no suspicion that he was faced by the mass of the Grande Armée” (p. 640); Fisher has him dismissing Tauentzien’s first reports outright (p. 58). The design says: you don’t get to play the Hohenlohe who woke up.
It costs something, as Tauentzien historically rallied five thousand men and counterattacked hard enough that the French “reeled back in confusion” ending up split between Krippendorf and the Isserstadt woods (Chandler p. 643). A formation that cannot voluntarily leave its hex cannot do that. The scenario has traded away the one Prussian action that genuinely hurt the French that morning. No doubt the other scenarios will build on this theme in later parts of the action.
And the fog is on the turn track. Turn 1 at 06:00 is marked Thick fog; turn 6 at 08:30 is marked Thin fog. The scenario is bounded by the weather at both ends — it begins in the murk and stops as it lifts. Fisher has the fog lessening around 8.30 (p. 58) and Van Creveld has the mist clearing near 0900 to reveal Lannes in possession of Closewitz (p. 107). The designer has set the clock to the visibility, not to an arbitrary number of turns or variable via die roll.
There is a special rule for it, and what it does is cut movement, ranges and firepower. Fog doesn’t hide an army from a player looking down at the map from above. Though there is an optional rule allowing units to all be covered by a marker. so instead of pretending to conceal, it takes away reach. You cannot see far enough to shoot far, you cannot shoot hard at what you can barely make out, and you cannot move confidently through ground you cannot read. The effect on play is the effect on the morning: everything happens closer, slower and more locally than either commander intended. Groovy enough.
The fog is also what let Lannes get onto the plateau unseen and Marbot’s forty thousand packed shoulder to back, half a cannon shot from the enemy, and “they perceived nothing” (quoted in Chandler p. 642). The same fog is what makes it hard to convert that position into three captured villages before the clock stops.
What a village is actually worth
Infantry in a town is forced into general order, and general order sets firepower to 1 a flat number to account for dispersal, regardless of whatever is printed on the counter. So a village does not project force. Put your best shooting regiment inside and you have thrown its shooting away.
What the village gives instead is this: morale checks called by combat results are ignored. Not modified ignored! A rule I often forget. A garrison cannot be routed by fire or by melee. It can only be dismantled step by step or disorganized. Which means morale, the thing you would normally garrison for, stops mattering the moment a unit is inside; your shakiest line regiment defends a village exactly as well as a grenadier. The only currency is steps.
And there is a ceiling on that currency: stacking is limited to ten steps to a hex. You cannot make a fortress. You can make, at most, two average regiments’ worth of resistance in any one place.
Artillery is excluded from the arrangement entirely in a town it can only be limbered or disorganized, so it cannot fire from inside a village at all. The guns have to stand outside, in the open, where they can be shot at.
Now put those pieces together and the scenario’s shape appears. A garrison shoots at 1 and cannot be made to run. So in clear weather the French answer is obvious and cheap: don’t assault at all. Stand off in line at full firepower, shoot the village to pieces at leisure, and never enter it. The fog is what takes that option away. With range and firepower cut, standing off stops paying, and Lannes has to close and melee and melee against general order means all six adjacent hexes are front hexes, so there is no flank to turn and no rear to find. He has to grind, frontally, on a clock.
That is what the six turns of fog are for. The weather is not atmosphere in this design; it is the mechanism that forces the assault, and is what happened historically.
The Prussian plan, written before it was tested
Pre-registered 2026-08-02, before play. Claude set the Prussians; I play Lannes against them. Recorded here so the after-action can be checked against a prediction rather than assembled to fit whatever happened.
This is the AI ‘speaking’:
Start AI comment// The reasoning, in one line: with no manoeuvre for six turns and a ten-step ceiling on any hex, the Prussian cannot defend in depth, so he must defend in time — and the village that costs Lannes the most movement points to reach is where time is cheapest to steal.
What a village turned out to be
The plan was rewritten three times, and each rewrite came from getting a rule wrong rather than from thinking harder.
The rulebook says general order is mandatory in “town, castle, brush or wood” and confines the no-morale-check exemption to “town and castle only.” The word village appears in neither list. The game’s own terrain chart, meanwhile, puts Village/Town on a single row and gives that row general-order-forced and no morale check, for both fire and melee. Neither term is defined anywhere in the rules.
Kev’s read, which is the right one: the omission is an editing failure, and a village is a town for these purposes. There’s a structural tell — the melee row reads “attacker −2, −1 from town to town,” a clause that only makes sense if some settlements span several hexes. Villages are one hex. The game distinguishes them physically while treating them identically for effects, which is exactly what the chart says and what the rulebook’s list forgot.
So a village is: general order forced, firepower 1, all six hexes front so no flanks, no morale checks, and an attacker who comes in at melee value −2 against a defender at normal value. Fire against the target is −2 as well.
Which makes Closewitz close to unshootable. Fog halves French artillery, drops skirmishers to strength 1 at two hexes, and the terrain takes another −2 off. Lannes cannot burn the garrison out. He has to walk in — and walking in is the worst thing available to him.
And the −2 cuts both ways, which is what killed the previous plan. I had built a counter-attack reserve behind the village on the theory that out-of-command units may still melee and advance after combat, so a lost village could be retaken. They can, and it can — at −2. A line regiment at melee 2 attacks a French-held Closewitz at 0; in column, at 1. Cavalry are value 1 in a village and 6 MP to charge in. There is no realistic recovery. The same terrain that protects me protects him the moment he is inside.
So the reserve was deleted. If I cannot retake the hex, troops held back to retake it are troops doing nothing.
Final dispositions
Closewitz — 3rd and 4th Grenadiers. Ten steps, melee 3, general order. This is the battle. Everything else exists to make Lannes take longer to reach it.
Lützeroda — 5th Line and Det-Gr-Wn. Eight steps. Not a second redoubt, a bill. The road brings him here cheaply, so make him stop and pay rather than drive through. When it falls it will have cost turns, and the survivors retreat west along the Weimar road.
45th and 12th Line — forward, on the ground he must attack Closewitz from. On the best defensive terrain the approach offers, where an attacker is already at −1 before he reaches the village. Ten more steps he has to work through, in the open, on his expensive flank.
Battery — unlimbered, behind the line regiments, covering the Closewitz approach. Not the Lützeroda road; Lützeroda is going to fall whatever I do. Artillery loses a point per hex of range and fog halves it, so it must sit close and accept being a target. It is the only fire in the army not reduced to 1.
Both cavalry — in the open, within two hexes of his forming-up ground. Two steps apiece, a two-hex charge, and a charge adds a step loss against infantry. Their job is to punish French regiments while they are deploying from column in the open, before they reach the village where cavalry are worthless.
7th Fusiliers — in the gap, skirmishing. Range 2 at strength 1 in fog. Two steps of nuisance holding the cohesion chain together.
No reserve, because there is nothing a reserve could do.
The bet: Closewitz cannot be taken by fire, and an assault at −2 against grenadiers at 3 needs favourable columns Lannes will struggle to generate in six fog-shortened turns — especially if two or three of them go on clearing the forward line to get into position at all.
What kills it: he masks Lützeroda with one regiment, sends everything at Closewitz from turn 1, ignores the forward troops rather than clearing them, and wins a single melee. If buying time in the open does not buy time, there is nothing else in the position.
// END AI comment.
Lannes’ plan, also written before it was tested
Here is my plan to deal with an AI’s defensive scheme:
Suchet’s five regiments face Closewitz, Gazan’s three hold Cospeda and face Lützeroda, the cavalry are massed on the right by the wood, and the horse artillery sits centre with Lannes.
The French plan: shell Closewitz with artillery first, switch the assaulting regiments to line where their melee value comes up to 3 or better, then go in from three hexes at once. [EDIT: Column units gain a +1 to Melee, not Line.I will be attacking in COLUMN]
This math should work if I can knock off a step or two with combined arty unit Fire. 12 strength [two arty batteries firing together, halved to Six then -2 for village effect.
Then Melee.
Melee strength is value × steps, so with the grenadiers defending full strength at 3 with ten steps, the village is worth 30. A French regiment coming in at value 1 would need ninety steps adjacent to reach 3:1, and V Corps hasn’t got ninety steps. B
ut every step the guns strip off the garrison is worth three off that 30 , so we shell it down to six or seven steps and thirty-odd steps across three hexes is suddenly better odds
So, the assault is potentially the cheap half. The shelling is the expensive half, because it happens at −2 into a village with the fog halving French artillery, and it has to happen before the assault is worth making. Thus making combined arty very valuable, and why I move them to the extreme right out of shot of the enemy.
Dice matter here, and reminds me we need a discussion about CRTs one day.
The whole exercise potentially turns the scenario into a turn-budgeting problem. Lannes has to take Lützeroda as well. The 45th and 12th are ten steps at melee 2 strength 20, behind the same −2. Either Gazan does that alone with three regiments while Suchet works on Closewitz, or the two are sequenced and cost turns nobody has.
Six turns, one corps, two villages that both resist assault, and fog on the movement. Juicy!
//AI Comment: The Prussian counter-play is thin and entirely disruptive: the battery firing on French guns while they sit unlimbered, and the cavalry using a two-hex charge that adds a step loss against infantry, aimed at the assault stacks while they form up in the open rather than at the village they cannot enter. Costing Lannes one turn of re-forming may be the whole game.
The prediction, on the record: the assault works if it happens; the question is whether six fog-shortened turns are enough to close, unlimber, shell a village down, storm it, and do the same job at Lützeroda. My bet is that they are not. Kev’s is that they are — and he has said he has something planned for the Prussian battery, which is the one piece of the disruption I am relying on.
Three errors worth keeping in the record, because a piece about planning a battle you cannot see should show the planning failing. I read cavalry counters as eight steps when they carry two, and would have garrisoned the redoubt at four steps instead of ten. I read a rule about towns as though it covered villages, and built a plan on troops that could not run away. Then I built a recovery mechanism without checking that the modifier protecting my garrison would equally protect whoever replaced it. Every time, the information was on the table and I read it in the direction I wanted. That is the ordinary way plans fail, and it is a better story than a plan that worked. //End AI comment
My disposition
The thing worth recording before I get started: I set up without having read the battle, purely off the terrain and the enemy positions, and it came out as Lannes’ deployment.
My assessment here was built on purely on my estimate of timing to capture the villages and the high ground behind both villages. I want to bring arty to bear with no threat of being charged or fired upon, so as to quickly reduce Closewitz. Then push down the Prussian line and connect with Gazan who may require support. Each formation has a couple of high step count, good morale units that can melee like a boss. I want the Cav to be a threat to the rear.
But the claim I’d make is narrow. It isn’t that I thought like Napoleon. It’s that the ground and the enemy’s posture left one sensible approach, dictated by terrain and weather.
The counter-case of course is are we railroaded into this? Aa map can also railroad you, and a scenario that starts from the historical lodgement with the historical objectives has already narrowed the field before you touch a counter. There is no ‘appraoch to battle’ option. Whether this is fidelity or a rail is something the later turns answer, not the first two hours.
Full sourced background: BigBoard/Wiki/battles/jena-1806.md.