Napoleon's Last Battles: Whole is not the sum of the parts

Napoleon's Last Battles is one of the true classics of the SPI era, going through numerous printings and being resurrected in a Decision Games version that's still in print. It's also a popular game set offered by the online game-playing service Hexwar.com, which maintains statistics on the games, including totaled results for each scenario.
NLB is an unusual quadrigame because it includes four battles that were all part of the same campaign, as well as a campaign game. But the campaign game is not simply a linked-together version of the four component battles but a much-expanded standalone game that includes some significant differences. Foremost among those differences in a whole command-control subsystem based on leader counters, which play no role at all in the individual battles.
Because of this playing the campaign game feels much differently than playing the battle games, despite the fact that they use the same map and many of the same counters.
The individual battle games are fairly straightforward Napoleon at War-style quad games, the major difference from most of the other NAW series being that NLB is on the brigade level instead of divisions and therefor allows stacking.
While entertaining, Hexwar stats indicate that most of the individual battles are not especially well-balanced. Worse, the imbalance often runs against the historical winner.
The most popular scenario on Hexwar is La Belle Alliance, which depicts Waterloo. It's a tough challenge for the French, with just 590 wins as of Jan. 30, 2008, compared to 1,023 Anglo-Allied victories. Waterloo is a challenging battle for the French in most game depictions, but the special terrain status of Hougomont and La Haye Sainte create very formidable forts that break up the French attacks thoroughly to an extent rarely shown in other Waterloo titles. Adding a fatal blow to French hopes are the special rules dealing with the Guard, that inflict a severe morale penalty if they suffer an adverse combat result. This wouldn't be so bad except that an "Exchange" is deemed an adverse result for this rule. This means that there is no safe odds for a Guard attack. Even at maximum 6-1 odds there's a 1/3 chance for an exchange and a big morale hit. Indeed, the best odds for a Guards attack is exactly 4-1, but even then there is a 1/6 chance of an adverse result. This means that the French really can't use the Guard as part of their battle plan and must save them for use only if they are about to lose anyway.
This Guard problem is also largely to blame for the unhistorical and lopsided results seen in the Ligny scenario, which has resulted in a Prussian victory 676 times compared to just 365 French victories. In the actual case, the Prussians were defeated. In the game, however, the French can't really use their Guard troops for fear of a sudden loss.
The best balanced of the individual battles is Quatre Bras, which has 683 French wins and 548 Anglo-Allied. This is a low density and fast-playing scenario which can go either way and often can turn on a single die roll.
The least popular of the four games is Wavre. It also ends up being biased against the historical victor, with just 244 Prussian victories on Hexwar compared to 369 French.
None of the games ends up with many draws, only Quatre Bras has more than 1 and it has just 6, so the games are, at least, admirably decisive.
The character of the game changes immensely with the campaign game, including it's component parts. The Quatre Bras portion of the fight becomes considerably hard for the French as Ney is badly outgeneraled by Wellington, while action on the Ligny front is more promising for the French because the Guard rule has milder effects in the campaign rules and only applies to part of the Guard. This allows the Guard corps to play a much bigger role against Blucher.
The overall course of the campaign does not have to follow the historical sequence, although it will surprising often. There's still usually a showdown near La Belle Alliance near the end of the game unless the Allies have had greater-than-historical success at Ligny and/or Quatre Bras. There's much less chance of a Wavre-equivalent in the games I've seen. Most French players prefer to maximize their strength for the main effort rather than send a sizable force chasing the Prussians. A smaller pursuing force is usually enough to harry the Prussians and complicate their attempts to reorganize.
This is a much longer game to play, lasting as much as 36 turns, so it's not surprising that there's not as many completed playings yet, but it is popular. It's also a fairly even fight, with 339 French wins and 403 Prussian-Anglo-Allied victories, which seems well in line with the historical possibilities.
The campaign game is very engrossing and is highly recommended. The other battles are entertaining, but players should be aware of the scenario imbalances and plan accordingly.