In the previous post I presented the currently accepted understanding of the diekplous and discussed the two main “schools of thought” attempting to explain how the diekplous was practically performed. Was it by single ships or by squadrons and, in the latter case, by squadrons in line ahead or in line abreast? I lamented that both “schools” relied on mostly philological arguments, which was not really helpful in understanding the specifics of the manoeuvre. I concluded that I was able to find only one study which focused on how triereis would practically interact with each other at sea.
The study is by Andrew Taylor and is published as a chapter of the “Olympias Final Report” with the title “Battle manoeuvres for fast triremes”. Using a mathematical model based on the Olympias’ performance at sea, Andrew was able to assess which kind of movements would such a ship be theorically able to do. On this basis, he replicated the movements of triereis in battle and had them perform manoeuvres and adopt tactics - including the diekplous - in a simulated environment. Thus he was able to define precisely how a galley could successfully break through an enemy line of battle.
In this post I provide a brief sum-up of Andrew’s findings on the diekplous, with some short stop-motion clips to illustrate the several interesting points he makes. The clips are basically animated versions of some of the key tables of the study. I strongly suggest you to read the original version of Andrew’s article as well, also because it discusses other manoeuvres, including the periplous, and provides an overall paradigm for fast trireme tactics.
One last clarification. The distances between ships shown in the clips are also taken from Andrew's data and then scaled down to 1/2400 (the scale of Alkedo ships, where 1 real life meter equals 0,4mm). The relative positions and distances of ships that you will see in the clips, therefore, are exactly the same as they were in the real world. They really would get so close to each other!
Let’s start with ship performance, because this is the “raw data” of Andrew's model. According to the Olympias trials and Andrew’s calculation, a trieres can get from zero to full speed, around 9.5 kts., in around 40 seconds. It can reverse from full speed to full stop in under 20 seconds. It can move backwards at around 80% of full speed – rowers would take the oars out of the water, turn around on their seats and start rowing the oar in the other direction. A trieres can perform a full 180° turn at full speed, a “fast turn”, with a diameter of only 145m. By slowing down to some 6-7 kts, it can do a 180° “tight turn” with a diameter as small as 60m. That is, slightly more than a ship lenght!
These performances have important implications for ships tactics. In fact, they mean that a fast trieres can approach a stationary enemy line up to a very short distance and still be able to turn and run away without being caught: an anastrophe. Provided, of course, that its captain and pilot are quick enough and that its crew is fast in responding to orders.
In the first clip, two triereis demonstrate a 180° anastrophe in the face of the enemy. The Petomene (“the flying one”, the red ship in the background) does a fast turn at full speed of around 9 kts. The diameter of the turn is 145 meters. With such turning strategy, the ship is able to close to 250m from the enemy before being forced to turn if it wants to get away with it. It seems a long distance, but it is only six ship lenghts. The other attacking trieres performs a “tight turn” instead, slowing down to 6,5 kts. This gives it a turn diameter of only 80 meters and the possibility of arriving at 180m from the enemy, or little more than four ships lenght, before having to turn away
In the second clip the Petomene demonstrates how an attacking trieres, by choosing a more oblique angle to approach the enemy line, can come much closer to the enemy and still sneak out if necessary. This is because closing at a less sharp angle allows the attacking ship to turn a smaller turn than 180° while mantaining high speed, thus allowing it to turn away much faster. In this way, Petomene is able to close to 75 meters – less than two ships lenght!
Now, let's assume that no one from the stationary enemy fleet reacts, and Petomene can row along it at short distance. Maybe the stationary fleet is afraid to break ranks because they are green crews, or maybe they are in a purely defensive posture, like the Peloponnesians at Naupactus. Petomene is now rowing along the stationary enemy line at a distance of only 60 meters, which is one and a half ship lenghts. And it can still theorically disengage if attacked, provided it is able to start turning the instant an enemy ship starts rowing!
The tactical implication of the extreme anastrophe capability of a trieres is that an attacking column could probe a stationary enemy line almost up to melee distance, in search of a gap for a breakthrought. If it was attacked, or found no gap in the enemy line, it could anastrophe and reform away from the enemy. If it did find a gap, it could try to perform a diekplous. Of course no admiral would voluntarily leave an open gap in his line. However, a gap could form by itself along a line of maybe 60 or 80 or 120 ships, because of mistakes, currents, winds, etc...
Now the question is, how big of a gap would an attacking squadron need to perform a diekplous? Andrew’s calculations shows that a 150 meters gap would suffice, as shown in the next clip.
The defending trireme A is inside Petomene's turning angle and might be able to ram her, but as she does she exposes her own flank to trieres 2, following Petomene in the column. I guess trieres 2 could then, in turn, be rammed by defending ship B, which then would be rammed by the third ship in the column, and so on... This would definitely be a failed diekplous! However, it would also be suicidal for everyone involved and I doubt this would happen in the real world. The other defending trieres, at the other side of the 150 m. wide gap, is not able to turn around fast enough to ram Petomene unless it somehow moves before Petomene starts turning. But it can hardly do so, because it would mean breaking rank and maybe put itself out as a ramming target if Petomene does not turn and simply stays the course.
Andrew shows that the very same diekplous, with a more direct angle approach, could also work with a shorter gap – 130 meters! That is, a gap a little larger than three ships lenghts!
So as you can see, a clear paradigm of diekplous emerges from Andrew's study. As I understand it, it goes like this.
The attacking squadron approaches in column and comes as close as possible to the stationary enemy line. It is looking for a gap to sneak into.
If it is attacked by one of the defending ships, or if it can't find a gap large enough to be exploited, it can try to anastrophe and then safely decide what to do next when it is well away from the enemy. Of course, if the defending enemy has faster reflexes and is able to catch the column off balance, the attackers are in trouble (ask the Imperial fleet at Artemisium…).
If the column is not attacked and finds a suitable gap in the enemy line, it can try to break throught – the diekplous. Again, it needs to be quicker than the enemy: if not, the gap will probably be closed by the ramming manoeuvre of a defending vessel against one of the attackers. In this case, the rest of the column would have no gap to pass through, and the whole thing would probably degenerate in a bloody and messy melee. But, if defenders are indecisive and are not able to react in time to the diekplous, the attackers will successfully sneak in the gap.
I think this model make sense, is supported by solid data and is fully compatible with what ancient sources say. I therefore adopted it as my version of “the Truth”. The house rules I am developing for “He hemetera talassa”, my to-go ancient naval rules, are among other things also an attempt to incorporate this model into it. What I plan to do is not allowing a break through as a possible outcome of a melee (i.e. I am taking it out of the melee table). I am making the diekplous a different manouver, the success of which is based upon a separate opposed die roll. This can result in a success, a partial success (which would force the attacker to anastrophe) or a failure (which would immediately force a melee, with the attacker being OOF). I will of course share it here as soon as it gets more playtest.