Saturday, January 22, 2022

Review of Adjutant Introuvable and Test Game

Adjutant Introuvable by Nic Birt is billed as an "Auto Strategy System for Miniature Wargames". 

Adjustant Introuvable (AI) aims to provide a strategy (the overall plan) to the absent opponent when engaging in solo wargames. AI attempts to maintain a reactive and dynamic plan throughout the battle through some general tactical guidance.

Strategy is fairly generic across historical military periods and therefore this system is appropriate for most ages from ancient to modern. However, the tactics have changed … and for this reason the tactical level of AI is provided in simple generic terms that will require interpretation to suite (sic) particular armies.

So now we understand what the rules are trying to accomplish, let's review what I think a good programmed opponent provide. Generally speaking, it must answer these four questions:

  • Which side gets to act next?
  • Which unit gets to act next?
  • Which actions will the acting unit take?
  • How will the acting unit execute those actions?

As previously noted, so rules answer these questions automatically, without the use of special "solo gaming" mechanisms. For example, if you are using rules that are IGO-UGO the answer to the question "Which side gets to act next?" is taken care of. Whatever side is the active side gets to act next until their turn is complete.

For games where the turn sequence consists of phases where all units perform the same actions at once, in any order the player likes (most IGO-UGO rules), the question of "Which unit gets to act next?" is also generally moot. This usually only comes up when two units are trying to move through the same physical space, in which the player generally moves the unit that can move the farthest first, so its movement is not blocked by the slower unit.

For rules where a unit can move, fire, and conduct close combat all in a single turn, the question "Which actions will the acting unit take?" tends to be less critical of a decision than in rules where, say, a unit can only move or fire. Nonetheless, this is where the programmed opponent starts to come into play as you need to know what the objective the unit is trying to accomplish.

Finally, few solo game systems answer the last question. In order to specify "How will the acting unit execute those actions?" they likely need to be written specifically for the rules you are using and possibly even the scenario and side you are playing. (The new Undaunted: Reinforcements, which I hope to be reviewing and testing here soon, would be an example.)

Before we take a look at how AI handles these questions, we note that "AI has been designed to work with wargame rule systems that operate with an alternating turn sequence (one side moves, then the opponent moves, and repeat). It is assumed the average game length will be between 6 and 12 of these double turns. For an effective use of the strategy plans it is best to have at least 10 units that are deployed by the AI side." For this reason I will be using the rules One-Hour Wargames (OHW), which I have reviewed here, only modified for using a 6" square grid.

Setup

The first task is to divide the board into a three-by-three grid. Given that out board is already nicely divided into a six-by-six grid, the AI grid cell will contain four cells of the board. Note that this makes counting terrain values (see later) much easier.

Set up the battlefield as normal and deploy the player's forces while also providing orders for the first turn. If the player wishes to simply react to the programmed opponent then the player may not make any moves for the first two turns.

In this test game I have decided to play the defending side in a battle similar to Guilford Courthouse, of the American War of Independence. The Patriots are therefore initially passive, firing at the advancing British, so counting them as passive seems the better option.

AI provides nine strategy cards, which outline the plan for how the programmed opponent will conduct battlefield operations (at a high level). The rules direct you to draw three randomly and then perform a calculation for each strategy to determine which of the three to use. The strategies are named: Equal Attack; Center Attack; Left Flank Attack; Right Flank Attack; Envelop From Left; Envelop From Right; Oblique Left Attack; Oblique Right Attack; and Ambush. I chose randomly and drew Envelop From Left, Envelop From Right; and Center Attack.

The strategy calculation consists of looking at the aggressiveness of the strategy on the left, center, and right and comparing it to the terrain in the left, center, and right. Here is what the board looks like, set up.

As shown in the image above, the map contains a grid for easy reference to terrain, units, movement, and combat.

  • E1 contains a Town. (In this scale, it is a Courthouse and outbuildings.)
  • B2 contains the West side of a Hill and a Patriot Infantry unit with 15 hits remaining.
  • C2 contains the East side of a Hill and a Patriot Infantry unit with 15 hits remaining.
  • D2 contains a Patriot Artillery unit with 15 hits remaining.
  • E2 contains the North side of a Plowed Field.
  • F2 contains a Patriot Cavalry unit with 15 hits remaining.
  • E3 contains the South side of a Plowed Field.
  • A4 contains the West side of a Woods.
  • B4 contains the East side of a Woods and a Patriot Skirmisher (Rifle) unit with 10 hits remaining.
  • C4 contains a Stone Wall and a Patriot Infantry (Militia) unit with 10 hits remaining.
  • D4 contains a Stone Wall and a Patriot Infantry (Militia) unit with 10 hits remaining.
  • E4 contains a Woods and a Patriot Skirmisher unit with 10 hits remaining.

With the terrain set up and the player's units deployed, it is time to determine which strategy AI will use. You start by calculating the strength of each strategy versus the terrain it is facing. Each strategy card shows whether the left, center, and right sectors (indicated by the red grid) are moving rapidly, advancing, ambushing, or holding. You use a table that compares that strategy to the terrain found in that sector to derive a score, which is then combined to determine how effective the strategy is to this scenario. The highest score of the three selected strategies is the strategy that the programmed opponent will use.

Without going into the details of the formulas (that is what the $5 price is for), the Envelop From Left strategy calls for a rapid advance on the left, holding in the center, and an advance on the right. The Envelop From Right is the mirror to that. Finally, Center Attack is an advance on the left and right and a rapid advance in the center. The significance of the movement (hold, advance, etc.) is how far through the board the program intends to advance. Hold only considers the programmed opponent's baseline ares (rows 5 and 6 in this case). Advance considers up through no man's land (rows 3 through 6), and rapid advance considers all rows in the sector. Let's compare the scores.

Envelop From Left

The Rapid Advance on the left starts (reading from the opponent's baseline) with Open terrain (0 points), then moves through the Woods (-1 point), then ends in Hill terrain (-1 point) for a total of -2 points.

The Hold action in the center only considers the Open terrain on its baseline, which gives it a score of -1.

The Advance on the right scores 0 points for the Open terrain and -1 for the Woods in no man's land, for a total of -1.

After scoring each sector a random factor of 0-3 points is added for a total. This strategy earned a -3 (after rolling +1 point for the random factor).

Envelope From Right

I will condense these down to Left -1, Center -1, and Right -2, Random 0 for a total of -4.

Center Attack

Left -1, Center -2, Right -1, Random +1 for a total of -3.

Because Envelop From Left and Center Attack are tied, there would normally be a roll off. However, I have decided to go with a Center Attack because Woods in OHW are essentially impassable terrain to five of the six units that the British have, thus attacking troops would have to flow through to the center, making this a Center Attack of sorts anyway.

With the strategy decided we now need to deploy the units to either the left flank, center, right flank, or the reserves. Troops are categorized as either Regular, Mobile, or Support. This is cross-referenced with the mission (Advance, Rapid Advance, Hold, or Ambush) and the target terrain (the terrain in the grid indicated by the mission). This yields a die roll to indicate if the unit is deployed in that sector or not.

Essentially you are rolling a die for each unit to determine if it is assigned to a specific mission, rolling for them in the order of Ambush, Rapid Advance, Advance, and Hold. Any units not assigned to those missions are placed in the Reserve.

So I have four Infantry, one Cavalry, and one Skirmisher. Starting with the Infantry (Regular) and rolling to assign to the Rapid Advance in the Center, with the target terrain of the Hill, I would need a 5+ to have a unit assigned. I roll 2, 2, 3, 5 so one Infantry unit is assigned to the Center. Next rolling for the Cavalry and Skirmisher (Mobile), still for the Center, I would need 4+ to have a unit assigned. I roll 2, 3 so neither is assigned.

Now I do the same for the Left, which is an Advance to the Woods. Infantry needs a 4+ and Mobile needs a 6. Because one Infantry is already assigned to the Center, I only have three Infantry remaining to assign. I roll 1, 1, 6 so one Infantry is assigned. I roll 5, 5 so no Mobile is assigned to the Left.

Finally I do the Right, which is an Advance to the Woods. Only two Infantry to assign and I roll 3, 6 so one gets assigned to the Right. I roll 2, 6 for the Mobile and the Cavalry gets assigned to the Right. (I rolled off to see whether it was the Skirmisher or the Cavalry.)

The remaining Infantry and Skirmisher unit gets assigned to the Reserve, which according to the strategy card, is behind the Center.

So Now What?

Now, it is time to start fighting the battle. When it is their time to act each mission force (sector) will roll a die, cross reference it to their mission, add or subtract 1 based on how well the mission is going, and get a Tactic to use for that turn. (Troops in the Reserve will deploy on a 5+.) The tactics are as follows.

  • Charge - move rapidly towards objective, initiate combat whenever a win is possible, and bypass strong opposition.
  • Engage - move steadily toward objective, initiate combat only when favorable results are probable.
  • Probe - move cautiously, initiate limited disruptive combat when favorable result is most likely.
  • React - hold ground, initiate combat only against weakened or breaking opponent.
  • Relieve - retire to advantage, make defensive formations and positions.

These tactics are listed on each strategy card, along with the die rolls needed for each action.

Some of these will be hard to interpret, especially where in this scenario the programmed opponent starts within musket range of my troops. Here is the final setup.

As you can see, two of the British Infantry are Elite, and thus have 20 hits remaining, while all other units have 15 hits remaining.

There is a tendency in rules to make militia and elite troops overly weak or strong, in comparison to regular troops. For example, I have seen them being given modifiers to hit, i.e. the ability to inflict casualties, while simultaneously giving them morale modifiers, i.e. the ability to withstand casualties. Although many would say this is correct, I believe it is not in that the accuracy of the musket is horrendous (especially after fouling from a few shots) and thus the difference in the three types to inflict casualties is negligible. To me, the best way to represent quality of troops is to rate their ability to stay combat effective, which in terms of OHW, it the number of hits it can sustain.

Turn 1

The British start first. Looking at my strategy card the Center will Charge on a 4+, Engage on a 1-3, and Probe on a 0-. This is a D6 roll with a -1 because it is one British (Elite) Infantry against two Patriot (Militia) Infantry behind cover. I roll a 1 making it 0, thus the British decide to stand and fire.

On the left and right flanks they are advancing, so the cards says to Engage on a 5+, Probe on a 2-4, and React on a 1-. Both the left and right flanks get a Probe result. The Infantry will stand and fire while the Cavalry will hold in place.

Because this is the first turn, there is no roll for reinforcements.

As stated earlier, the best solo systems answer all four questions in some way. What action the units will take has been answered (firing instead of moving, for all three cases), but the answer to the last question, which is how to execute the action, i.e. which unit do I fire at, may not be clear. The rules do state that the combat should have a "favorable result". So, what is that?

In OHW I often analyze combat in terms of Average Turns to Eliminate (ATE), or the number of turns, on average, it takes to eliminate the enemy unit. In this case the British Infantry roll 1D6 to determine the number of hits inflicted on the enemy, so 3.5 hits per turn. Units in cover, such as in Woods or behind a Stone Wall, take 1/2 the hits, rounded up. If the target has 15 hits then Infantry's ATE is 5 if the target is not in cover and 8 if the target is.

In the case of the left flank Infantry firing, both targets in range are in cover so both are equally valid. One could next look at the hits remaining (essentially the ATE) to determine the weakest target, and thus the most vulnerable and therefore the higher chance of success. Again, both targets have the same number of hits, so the ATE is tied. At this point either target is equally valid and the rules are thus silent on selection, so presumably a die roll would be in order. However, one additional difference to not is which target can inflict more damage (thus making them more of a threat), i.e. what is the target's ATE against you. In this case the Patriot (Militia) Infantry is lower and therefore the greatest threat. That said, I am going to shoot straight forward and into the sector where my mission is located.

On the right flank the case is largely the same: fire at a Skirmisher in the Woods or an Infantry behind a Stone Wall? The difference is that if the Infantry fires on the enemy Infantry, it is potentially doubling up its fire against that target.

Here is the result of the British first turn.

Remember, because I did not write order before the British set up, I am hampered from making any moves for turn 1 and 2, which is acceptable as I am simply going to fire at the attacking British. Here is the end of turn 1.

While the Skirmishers on the left muffed their shot, the right taught the enemy Infantry to ignore it at its own peril. The two Militia Infantry in the center absolutely pounded the center British Infantry. Had it not been an elite unit (starting with 20 hits), it would now be within one good die roll of being eliminated.

Turn 2

As with the first turn, each sector needs to roll its tactics, and reinforcements can be rolled for.

Left - Probe; Center - Charge; Right - Engage; Reserve - no reinforcements.

The Center charges as the British realize that they cannot afford to slug it out with the two Militia units. Because it moves forward, the British Infantry moving to D5 now blocks the Infantry in E6 from firing on the Militia in D4.

On the Right the Cavalry moves forward. If it moved to F4 it risks being fired upon by the Artillery and charged by the enemy Cavalry, so it only moves to F5.

This is the last turn in which the Patriots cannot react, so all they do is fire.

The British Grenadiers in D5 are looking pretty bad. With it being 'outnumbered', it is unlikely to be able to roll a 'Charge' result. This is one thing I have a criticism about with chance-oriented programs. It sometimes feels like 'morale' is baked into the results. Shouldn't the core rules cover morale, rather than the solo mechanics?

Turn 3

As with last turn, each sector needs to roll its tactics, and reinforcements can be rolled for.

Left - Engage; Center - Charge; Right - React; Reserve - no reinforcements.

The British on the left and right flanks continue to fire, but the cavalry stands. The British Grenadiers in the center rolled well and can continue the charge. Had they rolled less, they would have stood and fired, which would have been a bad result.

This was a good turn for the British. Both the Infantry units on the left and right flanks scored a '6', resulting in 3 hits each on the Patriot Skirmish units. The British Grenadiers charged, inflicting a single hit and forcing the Patriot Militia to retreat.

Note: British Infantry being able to charge into hand-to-hand combat and forcing Patriot Infantry and Skirmishers to retreat from hand-to-hand combat is not a normal rule for the OHW Horse and Musket rules, but is added by my AWI variant.

Now that the first two turns are up, I can finally order my Patriot units freely. At this point, however, I am going to leave the battle report. The idea was to test the AI system and see it in action.

Summary

Let me start by saying these rules are a mere $5, can be purchased online, and are delivered as a PDF. It is value for money. All of that said, to me it is an 'idea generator' rather than a full blown system of a programmed opponent. I see that these sort of systems can cover the following mechanics: overall battle plan; distribution of forces; unit deployment; unit mission; and unit tactics.

Overall Battle Plan

Basically the goal is to help the player come up with a basic strategy for the programmed opponent. AI does this providing you nine basic battle plans. Providing you a formula to calculate the best attack strategy based on terrain is very valuable. If fact, I think that is the singular best idea in these rules. For once a system takes terrain into account, giving it a value. Most points systems, for example, do not take the value of terrain into account, unless it is man-made like an entrenchment or a fortification. AI looks at the terrain within reach of your mission (how deep onto the battlefield you intend to take) and evaluates the difficulty of achieving it.

The only issue I take with the formula is that there are certain unstated value judgments built into that formula. For example, one element that kept coming up while using AI was that the rules I was using, OHW, does not allow most units to enter woods at all. Many rules take the approach that woods slow movement and provide cover, so the idea that you can attack a woods with any unit is somewhat universal. If you are using something like OHW where you can shoot at units defending a woods, but can never take it, the formulation should actually be harsher for the woods being present.

In the example above, if your mission is to rapidly advance on the left, i.e. advance to the enemy's row, the presence of the woods in row 2 makes this impossible if, say, your line infantry, cavalry, and artillery cannot pass through the woods and only your light infantry can. You could put infantry units in row 3 to shoot out the defenders, but at some point your infantry, cavalry, and artillery will have to bypass the woods through the center, while your light infantry moves in to occupy the woods, keeping the enemy light infantry out while your other forces pass by through the center.

So, is it perfect? No. It is naturally generic. That is why I say that these rules are great as a starting point, a template from which to build a more specific program that meet your scenario, forces, and rules needs and restrictions.

One final note: I am not sure why the author wants you to randomly select three possible battle plans, calculate the most effective one, and then use that, rather than having you calculate all of them. Given that you are playing solo, time is one factor that you have on your side. All of this work can be done before you actually lay out the terrain and troops on the board, so even if your time is limited in how long you can keep the game set up, this time calculating does not count against it and can even be done the night before.

Distribution of Forces

Again, my experience with this aspect may be colored by OHW not allowing certain unit types to move into Woods, but having a random roll to determine if a unit will be in a given sector feels … random. That said, I don't have a better way of changing the charts so that it makes more sense for OHW.

Unit Deployment

This was something that seemed missing. Although AI helped you determine which sector of the battlefield your units were going to be assigned to, no mention on how that unit would be deployed in relation to the other units also assigned to that sector. As I sit here and ponder that question, any such rules would probably be very complex and not work well for all rules, so it is probably just as well that it was not addressed.

Unit Mission

I felt like this aspect was something I had not considered, which is the mission of the unit. Granted, the same mission is given to all units in the sector, but the idea a unit would intentionally only advance so far (if at all) until the mission was interesting. Note that missions are always to occupy the terrain in the sector. Generally these might be the objectives in a scenario, so I could see reverse-engineering the overall battle plan based on scenario objectives and victory points.

Unit Tactics

This was the one area I felt worked the least because there was so much interpretation as to what each tactic – Charge, Engage, Probe, React, and Relieve – really means. Take a game like OHW. The basic decision is first "Do I move or fire?" This is something many rules avoid as they allow a unit to move and fire in a single turn. It seems like the tactics should be more aligned with how the unit fights. Does it primarily inflict casualties by fire or hand-to-hand combat? If the former, are you in range yet? Do you have range 'bands' (short, medium, long, etc.) which affect effectiveness? If so, which band are you in?

Considering these types of factors would lead more to rule-based decision making system, rather than one driven by chance elements, but I think it is something to consider.

As stated previously, there is also an element of morale sprinkled into the tactics table, with less aggressive tactics implying a partial failure of morale being the cause for the less aggressive approach.

Final Analysis

Overall, how can you say that this is not good value for money at $5? (I hope Nic does not raise his prices after this review! 😄) There are far better, well thought out ideas that are actually useful to gaming solo than many of the books I have purchased that promise to tell me how to wargame solo. Just consider that, unless the main rules you play fit well into the model published in these solo rules, you should consider this a template for how to convert this process to the rules you use most often; it is a starting point.