How Long does a Fight Take?

[Mark G]

Sabin has some bits on this, the short of which is that as the punic wars went on, the battles go longer.

earlier wars - especially the better documented greek wars, seem to be very quick indeed, as do most but not all of those involving gauls and germans.

and the short of all that is that its not down to predicting individual behaviour, as much as it is about the changing nature of the armies and the changing nature of the battle outcomes - subjugation, extermination, enslavement or hegemony - which we have overlooked a lot so far.

[Patrick Waterson]

I suspect it was more like a medieval mastermason than a modern civil engineer.  He knew his materials and their properties by experience, not by scientific experiment. 

While I can see and appreciate the thinking, I am intrigued by the assumed distinction between experiment and experience: what would be the practical difference in applicable knowledge, if any?

Interesting question.  How you get to proficiency would be different - theoretical learning, academic as against hands on.  And you could become proficient quicker.  Also, I suspect that innovation would be easier/ more likely.  I leave it to you to apply that to a battlefield, though I would draw the parallel to the staff college model developed in the 19th century against what came before.

And the 19th century staff college concept seems to have arisen because the pace of warfare - both mobilisation and campaigning - had stepped up to the point where traditional methods of troop concentration, movement control and supply were becoming overwhelmed.

Hannibal seems to have been that comparative rarity: a thinking general whose aspirations and proficiency reached beyond his experience, or perhaps more accurately enabled him to combine facets of his experience in order to develop novel concepts on the battlefield.  While there is nothing new in the concept of the ambush, there was in the way he lured Sempronius into one, and encirclement was not in itself a novel concept, although there was novelty and ingenuity in the way Hannibal set Varro up to take the fall at Cannae.

[Erpingham]

Hannibal was able to execute manoeuvres with precision and to assess how long it would take for certain troop types to overcome other troop types. 

Not wanting to dwell but the original reference was to precision of timing, not of manoeuver.  I think the ability of generals to estimate how long troops could hold for is one of the basic skills for generals.  It was down to experience, intelligence (as both thinking power and knowledge of the troops and conditions) and skill.  Which makes it a tricky element to build into any refight, of course. 

[aligern]

We are dealing with battles where, mostly, the general can see everything. hence I have great doubt that they would attempt to calculate how long a melee will last and long redoplyment will the for skirmishers and mostly operate on the 'master mason' principle of when x happens I order the cavalry to move forward, when Y happens the skirmishers are to pull back through the line. Whilst the Ancients could estimate time and distance I just don't see them making a plan on that basis because there is too much friction in war. Indeed Hannibal might have had a reserve behind the Celts and Iberians to feed in and keep to his plan or he might just have reasoned on them holding long enough. Until the 18th century march and deployment rates are not particularly scientific.
Roy

[Patrick Waterson]

Sabin has some bits on this, the short of which is that as the punic wars went on, the battles go longer.

earlier wars - especially the better documented greek wars, seem to be very quick indeed, as do most but not all of those involving gauls and germans.

and the short of all that is that its not down to predicting individual behaviour, as much as it is about the changing nature of the armies and the changing nature of the battle outcomes - subjugation, extermination, enslavement or hegemony - which we have overlooked a lot so far.

And which deserve further study, or at least attention.

The Punic Wars do show an imbalance of forces in many of the engagements described: the actions tend however to be decided by an imbalance in command as much as anything else.  In the First Punic War, we have several actions in Sicily that centre around sieges, and two in North Africa (Adys and Bagradas) which involve an army attacking an army.  We can look at Bagradas below.  Then, once Regulus has been removed from the scene the war shifts back to Sicily, with Hamilcar Barca commanding the Carthaginian army, and he conducts NO battles!

In the Second Punic War, the Trebia lasts the better part of a day, Trasimene somewhat less, Cannae the better part of an afternoon and Zama a good part of the day.  The battle at Herdonea in 210 BC (Livy XXVII.1) seems to be over in fairly short order, with the Romans enveloped and suffering a spreading rout along the line after the 6th Legion was charged in the rear.  After this defeat, Claudius Marcellus and his army turn up to fight Hannibal and the battle lasts "from early morning to nightfall" (Livy XXVII.2) with neither side prevailing.

Now Bagradas: the Roman army under Regulus had previously defeated a Carthaginian army at the Adys: the one change the Carthaginians made was the commander, and all else flowed from that.  The armies were the same as before (except that the Carthaginians had lost a number of Celtic mercenaries at the Adys and had them partly replaced by the mercenaries - apparently Greek - Xanthippus arrived with) but the Carthaginian commander was not, and he proceeded to find a good battlefield to maximise the effectiveness of Carthaginian cavalry and elephants, something his predecessor had conspicuously failed to do.  He also drew up his forces in an effective manner in which they could synergise their strengths - we call it 'combined arms'.

The result was a one-sided massacre.  The Roman cavalry, outnumbered 8:1, fled at the first charge (and appear to have been the 500 men picked up with Regulus after the battle), whereupon the Carthaginian cavalry, assisted by the 'most agile' of Xanthippus' foot, surrounded the Roman infantry who found themselves being gradually stamped out of existence by the solid line of elephants backed by Carthaginian spearmen and shot to pieces by the Carthaginian cavalry.  Carthage's spearmen only had to kebab a few groups of Romans who forced their way between the elephants.  Polybius does not tell us how long the action took, but it was over soon enough for the Carthaginian to catch Regulus and his cavalry before the day was over, so given that the armies would have been deployed and closing each other by, say, mid-day, we can surmise that the serious fighting, or rather killing, took perhaps three hours, as the Carthaginians had time to strip the dead and get back to camp before the close of day (in addition to picking up Regulus).

This is similar to the duration of Cannae, which we know began shortly before mid-day (the Volturnus got up soon afterward) and was over in time for the Carthaginians to rescue their own camp, take the Roman camp and round up practically every escapee from the battlefield.

Zama, which involved much speechmaking along with unusual deployments, could hardly have got underway before noon, and thereafter had a prolonged struggle between the Roman first line and the first two Carthaginian lines followed by an even longer one between the Carthaginian third line and the whole Roman infantry force.  We do not seem to have any reliable time indicators for this battle, except that it would have begun around noon or very late in the morning and was over before dusk.  Livy has Scipio plunder Hannibal's camp 'immediately' (confestim = straight away, without delay, forthwith) after the battle and take the booty to Lentulus' fleet at Utica, but one presumes the carriage of booty to Utica, a straight-line distance of 100 miles, was not accomplished the same day.  We might estimate a duration of perhaps five hours for Zama, which represents how long an outfoxed Hannibal was able to stave off defeat.

[Patrick Waterson]

Hannibal was able to execute manoeuvres with precision and to assess how long it would take for certain troop types to overcome other troop types. 

Not wanting to dwell but the original reference was to precision of timing, not of manoeuver.  I think the ability of generals to estimate how long troops could hold for is one of the basic skills for generals.  It was down to experience, intelligence (as both thinking power and knowledge of the troops and conditions) and skill.  Which makes it a tricky element to build into any refight, of course.

Both precision of timing and precision of manoeuvre are important, though as you say building the accumulated wisdom and experience of generals into a refight is decidedly challenging - substituting players for the original generals is already a major change from the historical prototype.  Paradoxically, even the most dedicated players rarely if ever seem to have the same grasp of deployment and manoeuvre timings as real period generals, and how long a troop type can hold, or how long it takes to overcome particular opponents, is very much at the mercy of wargame rules abstraction and interpretation.

What I hope to establish is that there are cases where resisting power, and hence duration, can be quantified and shown to be (reasonably) consistent, which would make it possible for designers validly to use mechanisms along the lines of: troop type A is facing troop type B; troop qualities and experience are similar and there is no leadership advantage, so troop type A will overwhelm troop type B in X turns.  The owner of troop type B had better have some idea of how he is going to either relieve them or otherwise relieve the situation, but he knows he has about X turns before the first cracks appear.  This kind of approach seems to reflect the pattern of many classical period infantry battles and could be a great time-saver on the tabletop without losing essential realism.

How useful it will be for other periods and armies I am less sure.  Less disciplined and more volatile armies may not be entirely anemable to this approach.

We are dealing with battles where, mostly, the general can see everything. hence I have great doubt that they would attempt to calculate how long a melee will last and long redoplyment will the for skirmishers and mostly operate on the 'master mason' principle of when x happens I order the cavalry to move forward, when Y happens the skirmishers are to pull back through the line. Whilst the Ancients could estimate time and distance I just don't see them making a plan on that basis because there is too much friction in war. Indeed Hannibal might have had a reserve behind the Celts and Iberians to feed in and keep to his plan or he might just have reasoned on them holding long enough. Until the 18th century march and deployment rates are not particularly scientific.
Roy

Scientific, no, empirical, yes, as Anthony pointed out.  One can be just as predictively accurate empirically as scientifically, perhaps even more so as the intrusive and occasionally distorting layer of theory is not imposed.  However I would point out that Greek and Roman commanders had march and deployment rates down to a fine art thanks to regular drill and practice and the institution of the parade ground, Field of Mars etc. and had accepted conventions about such movement rates (and look at the planning and execution of, for example, Claudius Nero's march north in 207 BC, Livy XXVII.43 and 45).

I disagree that 'there is too much friction in war': this would only apply if one's army is not a well-oiled machine.    We might also note that a cooperative opponent goes a long way towards making effective prediction possible.

While Hannibal might have employed a reserve to feed in to support his Gallo-Spanish line, our sources indicate that he did not, but let it collapse, which had the effect of sucking the Romans into the trap.  His remaining Gauls (c.16,000) are not accounted for in his visible lineup, the obvious place to put them being in a solid second line behind the variegated and attenuated first.  In that situation they would be well placed to stop a somewhat disordered Roman pursuit of the first line and stop it cold long enough for the trap to close.  This Hannibal evidently deemed a superior approach to piecemeal reinforcement, especially as it had the effect of sucking the Romans into the trap rather than expending force to keep them out of it.

Cannae depends upon an accurate estimation of how long Gauls and Spaniards can resist Romans (and for that matter on how long Roman cavalry can resist Gallic and Spanish cavalry) so that the triggered actions will be triggered in the correct relationship.

[Erpingham]

Just to expand on our frame of reference, I dug out Rollason & prestwich The battle of Nevilles Cross 1346 because I remembered they'd given some thought to the duration of the action and had compared the sources.  Translated into modern timings, this gives the following

6.30 am - a clash between a Scottish foraging party and the English van alerts the Scots to the approach of the English
9.00 am - the English have assembled on the battlefield and hear speeches
9-00 - 2.00pm - the English and Scots look at each other across the battlefield, banners displayed
2.00pm - 5.30pm - the battle is fought.  There are at least two breaks for rest (this is one of the clearest accounts of a medieval battle having periods of mutual separation of armies).  Scots are broken by degrees, with parts of the army running but part attempting ( but ultimately failing) to make a fighting withdrawal.  Pursuit probably limited by nightfall.

So, to the participants, this is a dawn to dusk battle.  If we were refighting though, we'd probably take the 3 1/2 hours of action.  This includes the initial archery skirmishing through to the final collapse of the Scots and the capture of the king.  So we'd probably call the battle rather than play the rout and last stand phases, which might have taken an hour (one source credits the Scots with falling back two miles before they were finally brought to bay).

I'd have to do more research to find out the length of other battles but I don't think this is untypical.  But some were very short.  The battle of Bulgneville in 1431 was famously short, for example, with modern estimates from 15 minutes to an hour.

[Patrick Waterson]

Classical armies tended to measure a battle by the duration of fighting: first there was the deployment, then the sacrifices, then the speechmaking, then the skirmishing and then the action.  The troops were also unusually fit by the standards of other times, the Lycurgan Spartans and, as noted by Josephus, the Imperial Romans in particular placing great emphasis on regular and intensive training:

" ... they do not begin to use their weapons first in time of war, nor do they then put their hands first into motion, while they avoided so to do in times of peace; but, as if their weapons did always cling to them, they have never any truce from warlike exercises; nor do they stay till times of war admonish them to use them; for their military exercises differ not at all from the real use of their arms, but every soldier is every day exercised, and that with great diligence, as if it were in time of war, which is the reason why they bear the fatigue of battles so easily; for neither can any disorder remove them from their usual regularity, nor can fear affright them out of it, nor can labor tire them; which firmness of conduct makes them always to overcome those that have not the same firmness; nor would he be mistaken that should call those their exercises unbloody battles, and their battles bloody exercises." - Josephus, Jewish War, III.72-75

Romans were thus well fitted for longer battles, and an all-day fight (albeit with reliefs noted as occurring at least once) is perhaps not too surprising.

But let us look at another mediaeval action.

Roche-Derrien in 1347 is interesting as being mainly a night battle.  Sir Thomas Dagworth left Beghard, nine miles away, at about midnight and ploughed into the sleeping French camp at "a quarter before dawn".  Between then and dawn his men had a fine time and fended off three improvised French counterattacks, then at daybreak the French began to get their act together and things started to look sticky for the English when Totsham and the garrison of Roche-Derrien, seeing what was happening, organised a sortie into the French rear and broke them.

Dagworth's '300 men-at-arms and 400 archers' had sustained a fight against several thousand French troops (Froissart has 1,600 men-at-arms and 12,000 foot; Dagworth himself reckoned 1,800 men-at-arms, 600 archers, 2,000 crossbowmen and an unknown number of 'commune' (presumably levied foot).  Depending upon what is meant by 'a quarter before dawn', Sir Thomas' force may have sustained the action for twenty to thirty minutes or two to three hours.

The Earl of Derby had achieved a similar success at Auberoche in 1345, in daylight, but time indicators for this one are much harder to come by.  Even at the time nobody seems to have hazarded a guess.

[Patrick Waterson]

And one of our few classical era engagements with a definite timespan: Hispania, 193 BC.

" Nor is there any doubt that all Spain would have taken courage to rebel had not the other praetor, Publius Cornelius Scipio, the son of Gnaeus, fought many successful battles beyond the Ebro and so intimidated the natives that not less than fifty towns surrendered to him.  These were Scipio's achievements as praetor; when he was propraetor he fell upon the Lusitani as they were returning home after plundering the farther province, laden with much spoil, while they were still on the march, and from the third hour of the day to the eighth maintained an indecisive action.  He was unequal in number of troops, superior in all else; for with his troops in a compact body he had clashed with a column long drawn out and hindered by the great number of its pack-animals, and he fought with fresh troops against an enemy worn out by a long march.  For they had set out during the third watch; three daylight hours had been added to their night march, and the battle had followed at once upon the labour of the journey, with no time given for repose. Accordingly, only at the outset of the fight did they retain some energy of mind and body, and at first they had thrown the Romans into confusion; later the battle became gradually more even.  At this crisis the propraetor vowed games to Jupiter if he should rout and slaughter the enemy.  At length the Romans pressed on with greater vigour and the Lusitani gave way and finally fled; and while the victors pursued the fleeing foe, about twelve thousand of the enemy were killed, five hundred and forty were taken prisoners, almost all cavalry, and one hundred thirty-four standards were captured. From the Roman army seventy-three were lost." - Livy XXXV.1.3-10

This action has five hours of fighting before the battle began to shift the Romans' way: not, according to Livy, because of superior Roman stamina or technique, but because the proconsul had vowed games to Jupiter.  Could be worth a try if a tabletop battle is not going one's way ... if one knows Jupiter's favourite boardgame. 

[Patrick Waterson]

And another: the battle of Carystum, 173 BC.

"There was a battle also in Liguria at the town of Carystum in the Statellate country. A large force of Ligurians had concentrated there.  After the consul M. Popilius reached the place they at first kept within their walls, but when they saw the Romans preparing to attack, they formed their line of battle in front of their gates.  This had been the consul's object in threatening an attack and he lost no time, therefore, in commencing the action. They fought for more than three hours without any certain prospect of victory on either side.  When the consul found that in no part of the field were the Ligurians giving way, he ordered the cavalry to mount and deliver as fierce a charge as possible on the front and flanks of the enemy's line.  A good many broke through the enemy's centre and got behind the fighting line.  This created a panic amongst the Ligurians; they broke and fled in all directions, very few reached the town, the cavalry mostly intercepting them. The obstinacy of the fighting proved costly to the Ligurians; 10,000 men are said to have been killed and more than 700 prisoners taken; 82 standards were carried off the field.  The victory was not a bloodless one for the Romans: they lost more than 3000 men; the loss fell mainly on the front ranks owing to both sides refusing to give ground." - Livy XL.7.3-10

[Duncan Head]

That's a good example that I hadn't previously noticed, of cavalry breaking frontally through "line" infantry (if not especially "heavy" ones).

[Jim Webster]

That's a good example that I hadn't previously noticed, of cavalry breaking frontally through "line" infantry (if not especially "heavy" ones).

But infantry that have 'fought for three hours'.
So you could 'soften' your enemy infantry up with prolonged 'combat' and then hit them with cavalry.

Jim

[Patrick Waterson]

So you could 'soften' your enemy infantry up with prolonged 'combat' and then hit them with cavalry.

The Romans did do this on occasion, e.g. in 203 BC:

"In the same summer in which these measures were taken at Rome and these operations carried on in Africa, Publius Quinctilius Varus, a praetor, and Marcus Cornelius, the proconsul, fought a pitched battle with Mago the Carthaginian in the territory of the Insubrian Gauls. The praetor's legions were in the first line. [2] Cornelius kept his legions in reserve while he himself rode up to the front. And from in front of the two wings praetor and proconsul kept urging the soldiers to advance their ranks against the enemy with all their strength. [3] When they failed to drive the enemy back, Quinctilius then said to Cornelius: "The battle is slowing down, as you see, and their unexpected resistance is hardening the enemy against fright, and the danger is that fear may turn into daring. We must rouse our cavalry to a sudden charge if we wish to confuse and dislodge them. [4] Accordingly, either do you in the front line keep up the fight, and I will lead the cavalry into the fray. Or I will command here at the front, and you shall send out the horse of four legions into the enemy." [5] As the proconsul was ready to accept whichever part of the task the praetor wished him to take, Quinctilius, the praetor, with his son, an active youth whose praenomen was Marcus, made his way to the cavalry and ordering them to mount suddenly sent them out against the enemy. [6] The confusion wrought by the cavalry was heightened also by the shouting of the legions, and the enemy's line would not have kept its position if Mago at the first movement of the cavalry had not at once led the elephants, which were kept in readiness, into battle. [7] Terrified by their roar and odour and by the sight of them the horses made the assistance of the cavalry useless. And although, so long as they were in the thick of the fight, where they could make use of the lance [cuspis] and, at close quarters, of the sword, the Roman horsemen were the stronger, still when they were carried to a distance by frightened horses, the Numidians were the more successful in hurling javelins from a longer range." - Livy XXX.18.1-7

Forgetting about the elephants rather spoilt the surprise in this instance, but the idea was there, although in this case desperation rather than calculation seems to have been behind what seemed like a good idea at the time.  The Romans eventually won by committing their reserve legions.

I seem to remember someone doing a Slingshot article on this particular battle. 

[Jim Webster]

Yes, at the time it was one of the battles which made me wonder if Roman cavalry would best be described as Irregular Knights (perhaps inferior?) who dismounted as Warband :-)

Jim

[Erpingham]

A few more medieval examples.  Because these tend to be less than unequivocal, I've put my sources

Agincourt : Battle starts with English advance c. 10 am, fighting to between 12.00 and 1.00 (Curry)
Courtrai : Battle starts with crossbow exchanges around 12.00, finished by 3.00 (Verbruggen)
Flodden : Fighting starts between 4.00 and 5.00, finished by 7.00 (nightfall) though pursuit and looting continues (EH battlefield survey)

These were just three I had to hand - I didn't just select them because they were the same length   Though obviously not a scientific sample, we seem to be finding quite a number of actions around the three hour mark.  What, if any, effect should this have on our gaming?  E.g. what is the average length of time represented by a move and does this affect how we view what happens in that time?