MANIPULATING ZAMA

Using the wargamer-friendly information about the 202 BC contest between Hannibal and Scipio found on page 121 of Warfare in the Classical World (the main diagram includes a map scale), I estimated the length of the Roman battle line at around 2,900 yards. The left wing, commanded by Laelius and including the Numidians of Dacamas, represented approximately 500 yards of this frontage. The four legions along with the Numidian infantry supplied by Masinissa, had a footprint of about 1,400 yards. The Numidian cavalry on the right wing, also under the command of Masinissa, extended for something like 1,000 yards. Employing these same amateur skills, it appeared that the Carthaginian cavalry on Hannibal’s right had a footprint (or hoof print) of about 300-350 yards. The first line of Hannibal’s deployment consisted of Mago’s troops and matched, approximately, the combined frontage of the Roman legions, Allied legions, and Numidians. The loyal Numidian cavalry on the left wing formed a line stretching for about 800 yards. 

The diagram of this famous battle shows Hannibal’s 80 elephants screening Mago’s infantry. In fact, the symbols representing the pachyderms are in front of the Carthaginian light troops or skirmishers, so it appears that the diagram is following the narrative accounts provided by Polybius and Livy, as opposed to the description written by Appian. Now then, if I am close with my estimates regarding frontages of the various portions of the opposing lines of battle, simple division informs that there was an elephant about every 17 yards along Hannibal’s first line. 

On another wargamer-friendly page in Warfare, informative illustrations and text are found about “The Post-Camillan Roman Army.” According to one paragraph, “The depth of a legion in open order from the front line of velites to the rear rank of the triarii was c 100 yards (91 m), and its frontage, depending on depth, would average 200-250 yds (182-230 m).” Taking the larger number and multiplying by four gives me exactly 1,000 yards. This would leave just 400 yards, approximately, in which to squeeze 6,000 Numidian foot soldiers. (Estimating a yard frontage per man, this block of Numidians would have to be 400 men wide and 15 men deep.) Upon review, perhaps it is better to take the shorter legion frontage. This would give the legionary infantry 800 yards, and 600 yards (a little more room, the additional 200 yards being the equivalent of a couple of American football fields) for the infantry of Masinissa. 

Returning to the previously established elephant estimate, it appears that there would be nearly a dozen of the nellies lined up against each legion formation. So, if I suddenly became interested in refighting just a portion of the engagement between the armies of Hannibal and Scipio, it seems that it would be quite possible to recreate the fighting that involved just one of the Roman legions and the enemy formations or units it faced on the fateful day in 202 BC.

Building a Functional, Inexpensive, and Unattractive Model of a Roman Legion

After further study of the telescopic illustration and accompanying text on page 112 of Warfare, I decided to establish a “figure scale” of 1:1 and a “working ground scale” of .25 inch represented 3 actual feet or 1 yard. With this “foundation,” I established the strength of my centuries at 72 men each. A maniple of hastati would have 144 legionaries as would a maniple of principes. The triarii would have just 72 men in each maniple, but these would be tough veterans, armed with spears. The attached velites of the cohort would also number 144. These young men would be charged with screening the heavy infantry of the legion and with harassing the enemy. The total strength of my “model” legion would be 5,040 men. [1] Deployed on my tabletop for pretend battle against the Carthaginians, my legion had a frontage of approximately 60 inches or 240 yards by the “working ground scale” previously established. Dividing this estimated frontage by the previous estimate of 17 yards per pachyderm resulted in 14 animals in a line with their trunks and tusks pointed at the shields and standards of the Roman infantry. 

Preparing the Opposition

Instead of fabricating all three lines of infantry in Hannibal’s army, I made formations and units that would model the first two lines of Carthaginian foot. After reviewing the various ancient narratives, I decided to go with Appian’s description. For my first “miniature” line then, I built 20 elephant stands (due to space constraints and overall look, 4 of these stands were not used), 3 units of Moorish archers, 2 units of Balearic slingers, 7 units of Celts (warbands), and 3 units of Ligurian light infantry. The second line consisted of 9 units of Libyan levy and 6 units of Carthaginian citizen militia.

The additional increase in the number of nellies was the result of reading “Amazed am I ere I made Zama,” an excellent article penned by the late and still greatly missed Patrick Waterson, wherein he posited an elephant frontage of approximately 10.5 feet per animal. (Please see the January 2009 issue of Slingshot, The Journal of The Society of Ancients, pages 11-16.) 

With regard to unit frontages and numbers, the Moors, Balearic Islanders, Celts and Ligurians all had stands measuring 3 inches across, so they matched the Roman heavy infantry. The depth for most of these units was 2 inches, which would translate to 8 ranks. The warbands of Celts were 3 inches deep, so these groups of irregular and fierce warriors were 12 ranks deep—twice the depth of the Roman heavy infantry. Adding up the various strengths of the various formations, there were 288 archers, 192 slingers, 288 Ligurians, and 1,008 Celts. The units in the second line had uniform frontages of 4 inches and uniform depths of 3 inches. Therefore, each stand represented 192 men. There were 1,728 Libyan levy present, accompanied by 1,152 citizen militia armed with spears. 

A comparison of the total numbers revealed that the “model” Roman legion had the advantage. There were 5,040 “miniature” Romans on the tabletop, facing 4,656 mercenaries and other types in Hannibal’s army. The Carthaginians did have the double-edged sword advantage of the elephants, however. 

Rules Review and Selection

Ideally, I should have liked to employ the Armati 2nd Edition rules for this exercise. Being able to note fatigue as well as the “step loss” of unit strength was very appealing as well as familiar. However, figuring out how to equate a full frontal fighting value to a single elephant as well as a single century of hastati proved rather problematic, at least in my estimation. (Note: I have a fairly long history of solo gaming with Armati. I was also inspired by Mark Fry’s engaging and entertaining “Zama with Armati” Battle Day report which appeared in the July 2010 issue of Slingshot.) Next, I scanned the pages and related army lists of the Hail Caesar rules. I liked the idea of using an initial clash value and then a sustained value for combat. I also liked the variety of “special characteristics” that could be added to each unit. However, the amount of dice involved (rolling for hits, rolling for saves, re-rolling because of “specials”) proved a little off-putting. Proceeding alphabetically, I briefly considered IMPETVS as this set of rules also has mechanics which allow for the degradation of a unit’s strength as well as organization. There were concerns about accurately representing the “traffic lanes” created by Scipio, however. I also looked at L’Art de la Guerre (3rd Edition, not the 4th). Simon Watson’s (a long-standing member of The Society of Ancients) preferred and thoroughly play-tested set of rules, TACTICA II, appeared to tick most of the boxes, as the saying goes, but I was worried about the number of dice rolls as well as about the number of amendments that might be needed. Finally, I took a look at Simon Miller’s To The Strongest! rules. This innovative set appealed because there would be no dice or rulers involved. The rules included a provision for rampaging elephants. I also liked the “supplement” Simon drafted about modeling and playing with a Polybian Roman Army. (Please see https://aventineminiatures.co.uk//wp/wp-content/uploads/2016/02/Polybian-Romans.pdf.)

After going back and forth on this question for a couple of days, I decided on TACTICA II. However, the spiral-bound rules would not be used “as written.” There would be a fair number of amendments. Most of these would be taken from the Armati 2nd Edition rules, written or at least approved by the same author. For just three examples, I would be substituting the Armati game move sequence for the TACTICA II game move sequence. I would also be using the movement rates but not missile ranges provided in Armati 2nd Edition. Finding the unit breakpoints and quality modifiers a bit unrealistic or at least limiting, I decided to tinker with these to suit my purposes. 

Have you “heard” the one about the solo wargamer who drafted a “trunk” full of Special Rules for Elephants?

On page 47 of TACTICA II, in Section 10.7.2, the equivalent figure “value” of African elephants is listed as 4. Thinking this too small a number, especially when the pachyderms would be facing attachments of velites and centuries of hastati containing 72 men, I increased this “value” to 8. The FV (fighting value, meaning the number on a d6 that had to be rolled by an attacking enemy unit) of elephants remained 5-6, but the melee ability of the pachyderms was directly linked to their current strength or stamina. An unwounded elephant received 8 melee dice. If there were two hits against the animal, then it would only receive 6 dice. 

Polybius describes the opening of this famous engagement in the following manner: “Hannibal ordered the drivers of the elephants to charge the enemy. When the trumpets and bugles sounded shrilly from all sides, some of the animals took fright and at once turned tail and rushed back upon the Numidians who had come up to help . . .” How could I model this on my tabletop?

Arbitrarily, I determined that the “range” of the Roman trumpets and bugles would be 10 inches or approximately 40 yards. Once it reached or crossed this “barrier,” an elephant stand would have to test for its reaction by rolling a d6. On a roll of 1, the animal would panic. A result of 2 would result in the elephant veering. If a 3 were rolled, the elephant would balk, meaning it would stop at that point and be agitated or confused by the cacophony of music, shouts, and banging of spears or swords on shields. Rolls of 4-6 would mean no effect. 

Panicked elephants would turn about 180 degrees and then flee for a distance of 1d10 inches. If the animal ran into any unit, that formation would immediately be marked as disordered. The roll of another d6 would determine if a melee took place, or if the contacted formation “made a way” for the panicked creature. A roll of 1-3 would mean combat; a roll of 4-6 would mean the elephant passed through the movement rate determined by the d10 result. 

If the result is “veer,” then the elephant is not really frightened by the noise, but is not all that fond of it. The animal changes direction in order to get away from the blaring trumpets and bugles. A d6 is rolled to determine the new direction and a second d6 is rolled for distance of movement. A result of 1 or 2 means the elephant wheels 90 degrees to the left. A roll of 3 results in a 45 degree turn to the left. Rolls of 4 produce a wheel to the right of 45 degrees, and a 5 or 6 produce a 90 degree turn to the right. 

Elephants that receive a “balk” marker halt in place. On the next turn, the driver of this animal must roll a 3-6 on a d6 in order to get the animal moving again. 

Wounded elephants must test each turn to see if they panic or go on a rampage. A d6 is rolled. On a 1, the elephant panics. If a 2 or 3 is rolled, the beast goes on a rampage. Panicked elephants will turn about and then roll a d6 to see what direction (the lower half of a clock face, where 9 o’clock is a 1, 8 o’clock is a 2, and so on) they flee in, and a d10 for the distance of that scampering away. Rampaging elephants will use a d6 and d10 as well, but the initial result represents the top half of a clock face - 9 o’clock is a 1, 10 o’clock is a 2, etc.

To be certain, these scenario or experiment-specific rules are a work in progress. They are subject to further adjustment or refinement as well as deletion. In addition to consulting previous efforts to reconstruct Zama in miniature, I also kept in mind the points raised in two well-written and thought-provoking pieces which appeared in the July/August 2021 issue of Slingshot. (Please see “Solo Wargaming,” by John Hastings, on pages 18-22, and “Game Mechanics and Realism,” by Anthony Clipsom, on pages 33-34.) 

Summary, Findings & Evaluation 

At the end of the eighth turn of play or experimentation, a halt was called to survey the status of the field and the opposing forces. Of the original 16 elephants on the Carthaginian side, only 3 had panicked at the clamor of the Roman trumpets, shouts, and banging of spears or swords on shields. Seven of the animals had been killed in action, whether by showers of pila or by velites and or hastati at close quarters. Five of the pachyderms were wounded, some more than others, and were currently rampaging about the field, not caring if they ran into friendly or enemy formations. One elephant had veered off course, and despite the efforts of its mahout, had not resumed its intended course. This lone animal had, however, managed to make contact with a unit of velites. The subsequent melee was ongoing. 

With regard to the first line of Hannibal’s deployment, in this specific sector of the larger field, most of the Moors, Balearics, and Ligurians had been sent packing. This was the product of being bested by the opposing velites or by being dispersed by friendly formations of heavier, formed infantry. The second line of foot, at least those units that had not been disordered and slowed by a few panicking elephants, had made a rather disjointed advance toward the Roman lines. The Celts struck first and caused some damage, but the Roman legionaries were stubborn as well as flexible. Even though some of the centuries were disordered and suffered losses fighting the impetuous warriors, the thinner grey line held, for the most part. Many more barbarians fell than Romans in this protracted contest. Indeed, on the far left of the Roman line, the legionaries were able to get around the flank of a warband and rout it. Two units of citizen spearmen were moved forward to threaten this local Roman advantage, but they were harassed and then halted by some daring velites of the Tenth Cohort as well as by the principes and triarii of the same formation. In the middle of the sector, hard fighting occurred between the hastati, the Celts, and finally, another phalanx of citizen infantry. Most of the Roman casualties were taken in this fighting, with Cohorts VI, VIII, and IX losing the prior centuries of hastati, while the velites of V, VI, VII, and VIII took even greater losses. On the Roman right, the Cohorts and their attached velites did rather well. They were able to withstand the enemy light troops and elephants; they were also able to dole out punishment to the Libyan levy troops when they advanced into contact. With the single exception of a century of principes belonging to the Sixth Cohort that had been “bumped into” by a rampaging elephant, none of the Roman second line saw any action. The triarii of this “isolated” legion spent all eight turns resting on their knees with shields and spears held on the ground. While the engagement was nowhere near over, it seemed that the Romans had or were rapidly gaining an advantage. The majority of the enemy elephants were out of action and the infantry battle was leaning in favor of the legionaries. Indeed, two-thirds of the Roman heavy infantry had not been committed to battle yet. The Carthaginians had six fresh units in reserve and a few more still sorting themselves out after being hit by panicked pachyderms, but the unit quality and die rolls certainly appeared to favor the Roman cause. 

Taken from behind the right side of the deployed legion. This photo shows Cohorts I through IV. The “elephant lanes” are evident, even though they did not play any part, really, in the ensuing action. The attached velites are “stuffed” between the separated maniples. 

Taken from above the left flank of the Roman legion. This picture shows another view of the deployment and of the “avenues” for the pachyderms.

This picture shows a portion of the Carthaginian deployment. The first line formations have elephants between them. The second line consists of levy troops and citizen spearmen arranged in large phalanxes. 

Rolling dice to see what effect the trumpets and other noise made by the Romans has on the elephants. 

The velites have formed a screen for the heavy infantry of the legion. The various troops making up the first Carthaginian line have started to advance. 

The Celt warbands have launched an attack on the Roman line. 

A wider view of the state of the field shortly after the Celts struck home. At the top of the frame, a few rampaging elephants (identified with black markers) have produced disorder (larger yellow markers) in a number of formations (both friendly and enemy). At the bottom right, the red dice indicate the Roman “kills” against the Carthaginians, while the blue dice indicate the “kills” scored against the hastati.

The hastati of the Ninth and Tenth Cohorts are engaged in a desperate fight with a bunch of Celts. Some velites have managed to get around the flank of the barbarians. 

Four rampaging elephants caused quite a bit of confusion in the center of the legion. 

Over on the right side of the legion, the hastati have destroyed a unit of Libyans, and the principes of three cohorts are moving up in support. 

As this solo wargame or experiment was not played to an actual conclusion, I cannot state that history was repeated or reversed on my non-traditional tabletop. The limited scale of the action prevented this kind of resolution from being realized. That much admitted, what was the value of the exercise?

The recently completed wargame, experiment, or exercise had value (subjectively measured of course) in that it entertained and engaged me over the course of approximately eight days. It provided sufficient material for this article/blog post. Though the elephant rules remain very much a work in progress, I think that as they currently exist, these amendments or scenario rules provided for a certain degree of realism as well as fun. On reflection, I think I might need to shift around the sequence of elephant moves, at least for those animals that are panicked or rampaging. There is also a need for intervention by the mahout, so that panicked or rampaging elephants can be stopped (i.e., killed) by that spike driven into the base of the skull. The interaction between the elephants and supporting troops with the Roman velites was also fairly realistic and entertaining. The velites were sorely pressed at times and many of these cohort attachments paid a steep price, but they also proved quite capable in slowing if not stalling the Carthaginian advance as well as dealing with a number of the elephants. 

The simulation of the infantry combats, between the heavier units such as Celtic warbands and centuries of legionaries, was also fairly realistic in my opinion. It also tended to be rather labor intensive as well as protracted, as handfuls of dice were required to inflict losses and produce some kind of result. On reflection, there was not “line replacement” between the prior and posterior centuries, and there were no instances of principes moving forward to take the place of hard-pressed hastati. In this specific respect, my wargame or experiment could be criticized. Overall, however, I think the resiliency of the legionary cohorts was demonstrated.

Related to this counterpoint, I think my “model” of a Roman legion was very realistic. Admittedly, it was not much to look at, the visual impression was not very impressive, but the arrangement of its component parts was realistic and was, I would advance, to scale. Then again, such a scale limited me to the portrayal of a single legion on my tabletop. Ideally, I should have liked to deploy all four of the legions involved at Zama. 

As for problem areas or areas that could have used additional tweaking, I believe I have already mentioned a few. The elephant amendments need further work; the melee process needs refining so that the number of dice are reduced. (Perhaps I could assign a d6 to each group or rank of 12 “figures”?) I have also remarked on the atypical appearance of my Roman legion. The same lack of aesthetic appeal can be applied to their Carthaginian counterparts. I have no doubt that readers will produce mental lists of their own criticisms or perceived problem areas. I wonder, however, if these same individuals will take the time to post them in the comments section? 

Another “problem area” was the unit rosters. There were 19 pages of rosters prepared and used during the course of this limited engagement. As one might imagine, during turns where units were locked in combat from left to right, there was quite a bit of flipping through various pages and marking losses in addition to rolling handfuls of dice and tracking casualties. I do suppose that I could have decreased the size of the type used to produce these rosters and by doing so, reduced the number of pages required, but this still would have resulted in flipping through various rosters to find the involved unit or units and marking of losses. As I “worked” my way through this procedure turn after turn, I often thought about the Armati and IMPETVS processes whereby an immediate visual status marker or markers are placed on or behind the respective engaged units. 

In summary, this project was educational, engaging, and interesting. While I have more information about what happens when an elephant collides with a formed body of men, I confess that I am still not exactly sure what happens and how best to reproduce this event on a wargames table. 

Sources & Acknowledgements

I was redirected to the ancient authors by Richard Lockwood, who penned the brief for the Zama Battle Day which appeared in the January 2010 issue of Slingshot. The narrative of Polybius can be found here: https://penelope.uchicago.edu/Thayer/e/roman/texts/polybius/15*.html. The account written by Livy can be read here: http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.02.0144%3Abook%3D30%3Achapter%3D31. The “asterisk marked” narrative from the stylus of Appian can be examined here: https://www.livius.org/sources/content/appian/appian-the-punic-wars/appian-the-punic-wars-8/#40. (Curiously, I noted that Dacamas had 1,600 cavalry with him when he joined the Roman side. In the Warfare orders of battle, the total number of Numidian cavalry on Scipio’s side of the field is given as 4,600. If the contingent under Dacamas is included, then this might suggest only 3,000 light cavalry on the Roman right. In Richard’s briefing, however, he supplies totals of 6,000 foot and 4,000 horse for the Numidians under Masinissa. It appears that there is a surplus or at least a transcriber’s error of 1,000 cavalry.) 

Like my library of rules, my actual library is also rather small, perhaps in the neighborhood of  75 volumes, the vast majority being about history. In addition to referencing the previously mentioned information on page 121 of Warfare in the Classical World, I read the relevant pages of Professor Lazenby’s book, HANNIBAL’S WAR, as well as pertinent sections of Professor Carey’s HANNIBAL’S LAST BATTLE: ZAMA AND THE FALL OF CARTHAGE. 

A number of authors, perhaps the equivalent of a century of triarii, have produced quite a bit of material about Zama for the pages of Slingshot. In addition to the dozen battle reports (Issues 271, 272), I mined the Index for other relevant articles. The analysis offered by Paul McDonnell Staff in Issue 142 still makes for engaging and excellent reading. Duncan Head’s focus on the elephants in Issue 145 proved very helpful in this amateur-by-comparison effort. A second acknowledgement has to be made to Patrick Waterson’s piece titled with a palindrome. I also need to make mention of Stewart Hey’s writing in Issues 273 and then again in 276. Both might be called reviews, but the former was oriented to the Battle Day proceedings while the latter was oriented to a re-examination of the historical battle. 

As one might expect, there are also numerous blog entries and other sites about “wargaming Zama” or the “battle of Zama” available online. I spent some time searching for and scanning what was found, but nothing new or of tremendous significance caught my eye. The narratives of the ancient authors, the analyses found in modern books, and the Slingshot articles more than sufficed, in my amateur opinion.  At the risk of appearing to self-promote or being accused of same, I dug out my old copies of the February and March 2010 issues of Miniature Wargames, wherein the editor kindly published my initial attempt to refight the Battle of Zama using the Armati rules. There were scenario-specific amendments generated, of course.  

Notes

1. If I were to pursue a traditional version of this project, a 6 mm Roman legion would cost me around 435 US dollars. This figure was calculated by finding the cost of a pack of Baccus Legionaries (96 per package) and using the exchange rate as of 26 September. Being completely inexperienced with preparing and painting metal figures, I would probably secure a couple of extra packs just to account for damaged or botched figures. My estimated cost does not factor in the price of brushes, paints, and other related materials. My guess is that a 6 mm legion at 1:1 scale would require another 100 US dollars, so rounding up, let us say 550 US dollars per legion. If I wanted to prepare the four legions present at Zama, then I would be looking at an investment of around 2,200 US dollars. If, on the other hand, I wanted to use 15 mm figures, provided by Essex, then I would require 630 packs. (Each pack of 15 mm Essex Roman infantry contains 8 figures.) At approximately 5.75 US dollars per pack of 8 (again, according to the exchange rate as of 26 September), a 15 mm scale Roman legion would require an investment of around 3,600 US dollars. Four legions would cost approximately 14,500 US dollars. The final price would be higher, when the cost of paints and other materials are factored in. Obviously, storage and set up for 15 mm figures would be greater than that needed for 6 mm figures. Sufficed to say, with any figure size, a 1:1 scale model of Zama would be quite an expensive undertaking.