Reexamining the First Barbary War through a solo play of The Shores of Tripoli

American interception.

One of my favorite things with historical games is when they get you to really think about what did happen, and why, and what might have happened instead. Back in 2020, I looked at the Battle of Chancellorsville through its Battle Cry scenario. Now, in a similar vein, here’s a look at one particular solo session of The Shores of Tripoli, which I played Sunday morning.

First, a little background. The Shores of Tripoli was first published in 2020, by designer Kevin Bertram through his own Fort Circle Games. It’s a one- to two-player game that takes about 45-60 minutes, as per its BGG page, and that timeframe seems accurate based on my plays. It focuses on the First Barbary War, largely a conflict between the United States and Tripolitania (with some additional involvement from Sweden, Morocco, Algeria, and Tunisia) that historically ran from 1801 to 1805 (although the game can run through 1806).

This is an asymmetric card-driven game, with both sides having unique decks. It takes place over up to six years from 1801-1806, and each year has four turns for each player, divided by seasons. On each turn, the active player will play a card, either for the listed event or as a discard for a basic always-possible action. The U.S. player always goes first. The Americans win either by successfully playing “Treaty of Peace and Amity,” which requires three separate conditions, or by successfully invading Tripoli itself. The Tripolitanians win by gaining 12 gold through pirate raids and card play, sinking four U.S. frigates, or eliminating a U.S. land army.  If neither side hits its conditions by the end of 1806, the game is a draw.

I quite like The Shores of Tripoli, especially for the different stories it tells with each play. The event cards, and when you happen to draw them or how their use goes, do a nice job of illustrating the various possible permutations of this conflict, as I discussed in my segment on this game (and other plays from the January 2021 Armchair Dragoons Digital Convention) on Whatcha Been Playing Wednesday #3).  The components are beautiful, evocative, and functional, and the game has lots of strategy without lots of rules overhead or a long playtime. It also uses “buckets of dice” to good effect in smoothing out probability curves (something discussed by Kevin in a recent Connections Online panel, and discussed by fellow panelist Amabel Holland in a follow-up Twitter thread), which makes reasonable outcomes probable but not certain.

The Shores of Tripoli is available for free online play at Rally The Troops!, and there’s currently a tournament ongoing there. In fact, preparing for my next tournament games (go Team Shadow Dawn, where I’m paired with excellent game designers Sobhi Youssef, Joe Schmidt, and David McDonough) motivated me to dig out my physical copy and take on the T-bot to see if I could improve my U.S. strategies. (The standard two-player game sees one player controlling the Americans and one controlling Tripolitania, but the solo version sees a human U.S. player facing a bot-controlled Tripolitania.) Here’s a look at how that went through photos (an approach inspired by Hiew’s Boardgame Blog):

Shores of Tripoli setup.

In the two-player game, the board starts with four Tripolitanian troops in each of Tripoli, Benghazi, and Derne (the three red circles left to right above), four Tripolitanian corsairs in the harbor of Tripoli, two Tripolitanian corsairs in the harbour of Gibraltar, and three American frigates in the harbour of Gibraltar. The solo setup changes this slightly: the bot gets an extra corsair in Tripoli and two extra infantry in Benghazi and Derne.

The 12 cards at the top left are the bot’s event card line and battle card line; the three at the bottom right are the American core cards. The bottom left has the Tripoli deck, with the U.S. hand face down next to it; the U.S. deck and discard pile are at the top right. This photo comes after the first U.S. turn, which saw me play “Naval Movement” to send two frigates to the Tripoli patrol zone and one to the Gibraltar patrol zone. (Normally, the U.S. options are to play a card for its effect, discard a card to move two frigates, or discard a card to build a gunboat in Malta. “Naval Movement” is a better version of discard to move two frigates, as it lets you move up to four frigates.)

Swedish interception.

One of the key early decisions the U.S. has to make is whether to play the core card of “Swedish Frigates Arrive” or not. This brings the two Swedish frigates (yellow) into the Tripoli patrol zone, where they’ll take part in interceptions of pirate raids, but not battles.

The upside is that this helps provide some early interception before the U.S. gets more frigates out there (the U.S. starts with only three, with one more set to arrive in each of 1802, 1803, and 1804, and two others available from the “Congress Authorizes Action” card). The downside is that this allows Tripoli to play “Sweden Pays Tribute” in 1803 or later, removing the two Swedish frigates and gaining two gold.

It’s not always worth it for the U.S. player to bring in the Swedes, with that decision often depending on your starting hand. In this case, it seemed worth it to me, and it certainly was in this Winter 1801 interception pictured above.

This took place on Tripoli’s turn, with the bot launching a corsair raid. A raid (one of the basic possible actions for both human Tripolitanian players and the bot) sees all corsairs from Tripoli move to the naval patrol zone. All frigates (U.S. and Swedish) then roll two interception dice each, hitting only on a six. Any sixes sink one corsair each, so my above roll (I had four U.S. frigates there thanks to grabbing my 1802 frigate via “Early Deployment” and moving the Gibraltar frigate to Tripoli with a discard after the corsairs there broke out) gave me four hits on 12 dice. That meant only three corsairs survived to raid.

American interception.

The game progressed from there, with the U.S. building up forces (including the three gunboats in Malta, which are useful in later bombardments and invasions) and Tripoli doing its own buildup and raids. The next critical moment came in the above photo, a Tripolitanian raid in Winter 1803. Here, the U.S. frigates rolled extremely well, sinking the entire Tripolitanian fleet.

Battle for Derne.

The interception success meant that I didn’t have to worry too much about Tripolitania reaching 12 gold, especially as they hadn’t had a ton of success before that. So the next key thing for me was focusing on my own victory conditions. The U.S. has two potential paths to victory, either successfully playing “Treaty of Peace and Amity” or successfully assaulting Tripoli, and it’s helpful to pick which one of those you’re going to focus on relatively early.

In this case, I targeted the treaty, so I needed to assault Derne. To do so, I moved a frigate to Alexandria in Spring 1804 (it’s ideal to do this late in 1803, but I couldn’t pull that off) so I could play “Hamet’s Army Created” in Summer 1804, then attacked Derne in Fall 1804 with the “General Eaton Attacks Derne” card (which I’d been saving since drawing it at the start of the game). Aided by frigate and gunboat bombardment, I took the city, but at significant losses (the two extra infantry in Derne the T-bot starts with relative to a human player adds challenge). And that further confirmed my plan of focusing on a treaty win, as I didn’t have enough troops left to slog through Benghazi and then Tripoli. (You can attack Tripoli without taking Benghazi by playing “Send In The Marines,” but it’s tough.)

Tunisia declares war.

After taking Derne, there are two big issues for the U.S. to focus on in order to get a treaty win: taking Tripolitanian allies out of the fight and destroying the Tripolitanian frigate they get from “Constantinople Sends Aid.” This photo shows the Spring 1805 situation, with Tripoli playing “Tunis Declares War” right before that. I cancelled this with “A Show Of Force,” moving three frigates to Tunis and taking their allied corsairs out. This left not much blocking raids from Tripoli, but the bot didn’t have enough gold or corsairs for that to be a real problem. So that left the frigate.

Burn The Philadelphia.

There are a few potential ways to get rid of a Tripolitanian frigate without the full Assault On Tripoli. One is simply moving frigates into Tripoli’s harbour for a one-round naval battle. But this carries problems from play of “The Guns Of Tripoli” (and/or the Fortress Tripoli variant, which is being used in the tournament and which I’m now using in my regular plays (multiplayer and solo) as well), and also is a challenge if Tripoli has enough corsairs left to soak damage there.

The other path is from an event card. Both “Launch The Intrepid” (destroys one Tripolitanian frigate on a 5-6) and “Burn The Philadelphia” (damages one Tripolitanian frigate to the next year on a 3-4, destroys it on a 5-6) are options here, with “Burn The Philadelphia” being slightly better for these purposes (“Launch The Intrepid” makes up for that with its potential use against corsairs as well). “The Daring Stephen Decatur” enhances either of these by letting you roll two dice and choose one. I used “Burn The Philadelphia” with “The Daring Stephen Decatur” in Summer 1805 and rolled a 6, destroying the frigate.

Final situation.

This would have cleared me to play “Treaty of Peace and Amity” in Fall 1805, the first possible option. However, the bot pulled one of the only cards that could stop me in Summer 1805, “Morocco Declares War.” Fortunately, though, I still had “Tribute Paid” in hand, and the bot wasn’t close to its 12-gold win. So I played that in the fall, and then the bot got a card that didn’t screw things up on its turn, paving the way for my treaty (seen on the board west of Tripoli) in Winter 1805.

As mentioned off the top, one of the things I quite like about The Shores of Tripoli is the way each play tells its own story. That’s been noticeable throughout my tournament games, many of which have followed dramatically different arcs depending on card draw, player actions, and particular die rolls. Of course, it’s not quite as wide-open as some grand strategic scale whole-day games, but there are still six different outcomes here (the three Tripolitanian win conditions, the two U.S. ones, and a draw), and even many games hitting the same outcome have significant differences in how they got there. (Also, this game is appreciated for its full “Historical Supplement and Designer’s Notes” booklet in addition to the rules: this is well worth a read, and both the historical essay from Dr. Abby Mullen and the designer’s notes from Bertram add a lot to the game.)

And, impressively, this is true with solo plays against the T-bot as well. The bot has some predictability compared to a human player (which is offset by the extra resources it gets and its constant availability of its main events and battle cards), but it still has some uncertainty based off its card deck and based off how rolls go. It’s a worthy opponent, and one with impressively low overhead to operate. And it’s a fun way to play this game when you don’t have a human opponent around.


Exploring Chancellorsville through Battle Cry

The Battle Cry box.

The two most crucial days of the Battle of Chancellorsville took place on May 2 and 3, 1863, so this weekend marked an excellent opportunity to recreate it. This battle’s long fascinated me, especially with the divide-and-conquer flanking march through The Wilderness Jackson pulled off on May 2. And given that I don’t have a battle-specific game for this one at the moment, the Chancellorsville scenario in Battle Cry (the 1999 Richard Borg design that kicked off the Commands and Colors system, since seen in the likes of Memoir ’44, Commands and Colors: AncientsBattlelore and more) seemed like a good way to recreate it. Here’s a look at how the battle played out this time, with me playing both sides. For background info on Chancellorsville, the American Battlefields Trust page and the Wikipedia page are quite good; there was also a very funny thread from Matt Palmquist (@CivilWarHumor) this weekend, starting with a good map overview.

And for those unfamiliar with Battle Cry, it’s a low-complexity wargame where each side takes a turn playing one command card. Those cards usually activate one or more units in one of the three regions of the map. All moves are done before battles, and battles can be resolved in any order the activating player chooses. Infantry and cavalry can move and battle, artillery can move or battle, and generals can only move (but they can attach to other formations to give them an extra die). Battle involves just the attacking player rolling a number of dice; that number starts at five for artillery attacking an adjacent hex, four for infantry attacking an adjacent hex, and drops by one for each further hex away. Some terrain also reduces the number of dice the attacker gets; notably for this map, forests and hills drop the attacker’s dice by one, fieldworks drop it by two. All units always attack at their full strength even if they’ve suffered losses.

The Battle Cry box.

The six-sided dice have two infantry results, with the other sides being cavalry, artillery, sword and flag results. To record a hit, you need to roll the type of unit you’re attacking. Swords are wild (hits on anything), and flags force retreats. A hit removes one figure; infantry have three basic figures and a flagbearer, cavalry have two and one, artillery have one and one, and generals only have a flagbearer (but can only be killed if they’re attacked when they’re not attached to a unit and if a swords result is rolled). The goal is to eliminate six enemy units’ flags, and the flagbearer is always removed last.

And now, on to the game report in photos (a technique I particularly like thanks to Hiew’s excellent blog).

Chancellorsville scenario setup
The Chancellorsville scenario setup. Jackson’s forces are to the left (west), Lee’s to the right (east). Hooker’s troops are in the middle around the town.
First attack
The Confederates go first here, as you’d expect. The hand I randomly drew for their side started with an assault (activate all troops) card for Jackson’s wing, which was a nice note to history. The first assault wiped out the infantry closest to Jackson’s lines (probably part of Howard’s corps) and pushed the other forces near the south of the map back towards town.
Further success
The Union was unable to do much with their first couple of turns (seems historically accurate!), with one turn even being a discard and draw thanks to only holding cards that couldn’t activate anything. Some further Confederate assaults led to this situation. The CSA has already knocked out two Union flags and weakened some other units without losing much themselves.
The Union army has had some rough card draws so far (and they’re hurt by only having three cards to the CSA’s five in this scenario), so they haven’t been able to do much attacking. But they did draw a fieldworks card, allowing the construction of two fieldworks around Chancellorsville. That gives those two units more defense (reducing the dice rolled against them by two). Unlike Hooker’s historical order for Sickles to abandon the earthworks in Hazel Grove, these units won’t be leaving these fieldworks any time soon.
All-Out Offensive
The All-Out Offensive card is the most powerful in the game, letting you activate all your units. The CSA drew it and used it to solid advantage, attacking with both Lee and Jackson’s forces in a move that feels like what historically happened in the morning of May 3. (Granted, at that point, Stuart was in command of Jackson’s troops after Jackson was hit by friendly fire the previous night.) You can see the mounting Union casualties at the top of the image, including four of the six flags necessary for victory; meanwhile, the CSA has only lost three figures and no flags. But this offensive did leave their troops more exposed outside of the woods.
Finally, some luck for the Union. They drew a Counterattack card, which copies the opponent’s last order. And All-Out Offensive is the best possible thing to copy, especially with several CSA troops now out of the woods and into the open. This turn sees the Union eliminate seven CSA figures, including one flag.
A couple of turns later, the Confederates have kept up their attacks, and they’re now one flag away from victory. They’ve also pushed the Union off both of the key hills in the middle, and forced the artillery to retreat to the buildings in Chancellorsville (and you can’t fire from a building, so that’s a useful result; it actually happened twice during this battle, with the artillery being pushed back there, moving back to the hill, and then being pushed back again). But the CSA has taken some heavy losses along the way, and the Union has done well to bring its northernmost artillery and infantry (likely Reynolds’ troops) into the fight. The flag score is now CSA 5, Union 3.
The Union has run out of useful attack cards, so they try to recreate the historical result of picking off Jackson (although that was by friendly fire rather than sharpshooters and happened earlier in the battle). But they fail their die roll, and he survives.
The Confederates press for the victory with some much-reduced troops from Jackson’s command, and they get it, recording their sixth flag. But they did take heavy losses. Final score, CSA 6, Union 3.

Overall, Battle Cry is far from the most comprehensive simulation, especially when it comes to what units you can activate being limited by your cards and to wounded units still attacking at full strength. And as an overall war title rather than one specifically focused on this battle, it obviously doesn’t have as much particular chrome or orders of battle as you might find in a more narrowly-focused title. But I love this game for what it’s able to do with a rules-light, easy-to-teach-and-play approach, and the scenarios are very well-designed, making strong use of terrain and of differing hand sizes to give you some similarities to what actually happened in a wide variety of different battles. Playing out the Chancellorsville scenario this way was a lot of fun, and a great way to re-explore that part of history.

The Play’s The Thing: Orange rush, Cinderella crush, code hush

Welcome to a new Board and Game segment, The Play’s The Thing. (Yes, inspired by Hamlet; hopefully it ends with less death and destruction!) These will be reports of different gaming sessions I’ve participated in. Here’s the first one, from a gaming session with friends Tuesday:

California Gold

Published: 2015

Designer: Patrick Stevens

Publisher: Numbskull Games

3-5 players

120 minutes

California Gold is a pretty intense economic game, focusing on farming oranges in California. Each player is running their own co-op and trying to add orange groves to it; groves pop up in the different counties of California, and you can only collect them by building packing houses there and spending an action to take the card. The more orange groves you have, the more money you’ll collect; Valencia oranges are worth one dollar per card, while Washington Navel oranges are worth two dollars per card, but are more vulnerable to weather that can keep you from collecting from them.

The grove cards also come in sets that help you gain influence with different railroads, bringing in even more money. You’ll get money for each train token you’ve placed (which requires three grove cards with that railroad’s symbol), and more money if you have more trains with that railroad than anyone else. You’ll need more packing houses to get these cards, though; you start with one in one county, and to get a card, you need to have a packing house with capacity (each can hold three cards) in the county the card is from. Each packing house you build both costs money itself and makes it more expensive to keep your workers happy (something you have to pay for every round, or risk being affected by strikes), so you have to carefully consider your timing in expansion, and also which rivals you’ll be competing with in a particular county.

Money and actions are tight in this game, especially early on. You have to build a significant engine to make your collect dues actions efficient, but to get there, you’re going to have to carefully grab cards early on and also monitor what your opponents are doing. Later in the game, money becomes less of a problem, but there are more things you have to invest in, such as advertising to boost your revenues, additional buildings such as nurseries and agencies to give you free actions, and political lobbying and infrastructure improvements to protect yourself from attacks.

California Gold, early in the game.
California Gold, early in the game.

Attacks? Yes, there are attacks. What separates this from a lot of economic simulators like, say, Power Grid, is that there’s a ton of direct interaction as well. Competing for cards is one thing, and one of the possible actions you can do is to wipe and refresh the deck of available cards (which can be very mean if the next person’s targeting some; however, you only get two actions per turn, so you can’t do this regularly), but what’s even stronger are the event cards. Each player has a hand of six of these, and the start player in each round gets to play one (and then draw back up to six). Most of them are nasty effects that go off under certain conditions; for example, a drought prevents those who haven’t taken the irrigation action from collecting dues this round. This is done in a cool way; the one-per-round mechanic keeps them from getting overpowered, the necessity of playing one (and not requiring an action to do so) means that players both can’t opt out of it and can’t go for an attack-focused strategy, and the choose-one-from-six gives players far more agency than a random event deck would have. Also, it’s nice that while most of these effects are damaging, the majority are about preventing players from doing something for a round than taking something away from them. Thus, they’re more delaying than destructive, providing interactivity without stealing a player’s hard-won cards (important, considering that the win condition is collecting a certain number of groves).

I had a lot of fun with California Gold, and not just because I wound up winning with 21 groves. There are plenty of interesting decisions in this one, and it’s a great underutilized theme with what looks like an impressive attention to historic detail (while still keeping it game-focused). Our play with three players (and the Northern Counties expansion) did take a little over the listed 120 minutes, and the others said it felt a bit draggy at the end and is better with more people (which doesn’t necessarily make it take longer, as the numbers of groves to collect for victory decrease), but I enjoyed it fine with three. If you like economic games, and also like more interaction than they often provide, this may be one to check out.

Too Many Cinderellas

Published: 2014

Designer: Nobutake Dogen, Nao Shimamura

Publisher: Grail Games

2-4 players

10 minutes

Too Many Cinderellas.
Too Many Cinderellas.

We played this fun little filler Tuesday after wrapping up California Gold. A fourth player had joined by this point, so we had a few quick rounds of this. In summary, Too Many Cinderellas features 18 possible Cinderellas, each with traits and rules. The traits, along with the Cinderella’s number, help determine if they are the real Cinderella if left in your hand; the rules come into effect if you play them to the table instead, and help restrict who the real Cinderella is. Perhaps she isn’t young, or doesn’t like ice cream, or doesn’t have glasses. Players get four cards and take turns playing one until everyone has two left in hand. After playing a card, there’s a vote to see if its rule will take effect; everyone else has the opportunity to reveal a “No,” which nullifies the rule, but which each player can only do once per game. This leaves everyone with two candidate cards, plus one revealed off the deck. The real Cinderella is the one numbered closest to 1 (unless the cat has been played, which reverses the numbers). A good quick game; there’s only some strategy, and a lot of chaos, but it has fun moments.


Published: 2015

Designer: Vlaada Chvátil

Publisher: Czech Games Edition

2-8 players

15 minutes


Codenames is a blast, and is quickly becoming one of my favourite party games. There are two- and three-player variants that I haven’t played yet, but the base game with 4+ is fantastic. You divide into red and blue teams, each with a spymaster, deal 25 clue cards out of a deck, have the spymasters look at a location card that shows which word card corresponds to which agent, and then have them proceed by giving their team one-word clues with a number of words the clue links to. (For example, “Imperial 2” might mean “tie” and “fighter”.) The team must then discuss among themselves (if there’s more than one team member guessing; with four players, each team has just one spymaster and one guesser) and point to the card they think is indicated. If they’re correct, an agent of their colour’s revealed, and they can take another guess. If they’re incorrect, it can be a bystander (neutral, but ends the turn), an enemy agent (helps the other team advance towards winning, ends the turn) or the one lone assassin (instantly ends the game in a loss for the team that picked it). The winning team is the team that reveals all of their agents first, nine if they received the first turn or eight if they didn’t.

Codenames is very simple and easy to learn, but has tons of replayability, and forces your brain to think hard and make clever connections. It may be somewhat better with groups that know each other at least somewhat and know what references they can make, but there are enough connections between these words you can get there without too much trouble most of the time. I’ve played this three or four times so far and had a blast each time. This is currently the top-ranked party game on BoardGameGeek, and it might be the best one I’ve played too.