I was intrigued with the following talk by former USN Commander Phil Pournelle discussing future fleet design, but drawing also on ideas from Science Fiction, games and literature. In particular work associated with the AI pioneer and Computer Scientist Doug Lenat who used a Science Fiction game (Traveller: The Trillion Credit Squadron) as part of evaluating his research tool, Eurisko. This was way back in the 1980's (see link below, note I have heard Commander Pournelle speak several time at the Connections UK Professional Wargaming Conferences, and he certainly knows his stuff and he is an enlightened professional who welcomed input from the recreational side of the hobby - at under 15 minutes, it is well worth a listen to):
https://www.youtube.com/watch?app=desktop&v=7Z4PMcgFIGE&t=12685s
The whole conference had a "what can we learn from Science Fiction" theme to and is available online (see link below, something for me to chew on over time):
https://cimsec.org/navycon-2020-navies-science-fiction-and-great-power-competition/
Phil confessed to having a long addiction to Science Fiction (his late father Jerry Pournelle is a Science Fiction author as he is himself), an interest which goes back to the 1980's where he played Traveller [and other game sci-fi games systems] but of particular interest to the above conference was Traveller, specifically - Traveller: The Trillion Credit squadron (aka big space fleet battles, using miniatures). Due homage was given to the late Doug Lenat who "computed" a winning TCS Competition Fleet [twice, Origins (San Mateo, California) in 1981 and 1982] using his post-doctoral computerised research engine (a sophisticated logical theorem and problem solving tool called Eurisko, an extension from his PhD work AM [Automated Mathematician], 1977). The Lenat/Eurisko duo produced a novel counterintuitive solution that went for "many smaller [75]" instead of the "few large [20]" space ships.
Footnote Addendum (see comments from Martin Rapier and my reply): Small is a relative term, Lenat/Eurisko fleet had as its backbone 75 "Eurisko Class" ships each 11,100 tonnes (but no jump drive which saved a lot of "tonnage") so not really small "patrol craft" in the scheme of things (?) discuss, but very much cheaper in cost than say to a Battleship (I must admit I am struggling for an appropriate word to call the "big space ships" so Battleship will do - but I know it is wrong). You can have almost four Euriskos for each bespoke Battleship (there I used the word again) and "Euriskos" carried the equivalence of one big weapon each.
So, instead of the few expensive "all singing, all dancing ships of the line" dreadnought star ships (Battleships?) and some accompanying cruisers/destroyers there was a "mosquito fleet" of heavily armoured, almost immobile bit gunboats (see below, a rather pristine looking copy of a Traveller book and Doug Lenat with his charismatic trademark smile - Lenat famously said the work was 40% Lenat and 60% Eurisko, he basically kept the machine from "going off the rails" adding additional heuristics as needed [he even showed his work (on Eurisko) to Richard Feynman, who was suitable impressed with its "almost creative power" at problem solving]):
Lenat's research took him away from TCS and he disappeared to a remote part of the Texas dessert for forty years (which sounds rather Biblical) doing research underwritten for the most part by the USN (emerging with a product called Cyc). Meanwhile, for Phil Pournelle there were other Science Fiction Game Systems of that era to play. they also showed similar "lots of little" beating "the big" meme - despite this not being the original Games Designer's intent. A game called O.G.R.E. getting a specific callout for the "tons of speedy hovercraft option" that so enraged the "big O.G.R.E. tankers" that the game's designer Steve Jackson created a subsequent errata to "try" and reduce, or rather moderate their "too successful" anti O.G.R.E. capability - which it has to be said, it only partially succeeded at). Where most players borrowed from the genre like Star Trek and focused on "the sexy big Enterprise" ships, Phil saw the ungodly potential of tooling up many small patrol ships with weapons to be able to take out squadrons of Star Ships when they came too "planet side close" (see below, Phil obviously enjoyed endless hours spent in intergalactic mayhem and destruction with the occasional model-making/painting thrown in - Note: examples include both tabletop and computer games - I suggest you hear it "on tape" through Phil's own words on the YouTube link - see first link in article):
Was there any "earth-bound" evidence of historical equivalent in "the many beating the few" .. Phil showed a USN WWII ship production data table for all naval construction. There was a trend away from the production of the sexier end [battleships, aircraft carriers and cruisers. even destroyers], to patrol -craft . These categories had a huge increase in numbers (but not necessary greater tonnage) . It was the sheer "amount" of smaller ships and "other" types, with a parallel to the many Shermans tanks beating the few but better Panther tanks on land (see below, hmm, I can see the point Phil is trying to make - but you are excluding aircraft from this assessment and other in the Pacific Theatre of Operations .. and perhaps the European Theatre of Operations working under different dynamics - the land example also has its problem, Eisenhower called the Sherman an "artillery tank" not a "main battle tank", most of the time tanks did not fight tanks but supported infantry who appreciated a nice bit of 75mm HE direct fire support to knock out that German machine gun nest that was being so troublesome):
Interestingly Phil brings in one of his significant mentors from the USN Naval War College .. Captain Wayne Hughs .. and focuses on his salvo equations ("first effective salvo" winning a battle - ambush style in particular) and the relationship between the potential damage delivered from a small ship, especially in ambush prone littoral settings, where as Phil puts it "most people live" (see below, a footnote being appropriate force composition is essential, you don't want to have the wrong stuff in the wrong place defended by the wrong things .. and by definition leave it vulnerable to "mosquito fleets" .. yikes):
This chart Phil used still worries me very deeply (see the picture above two and repeated below), because it is conflating all "operations of war" under one banner (akin to just computing the greatest Lanchestrian "fighting power" and saying that is "job done"). It is the interactions between the types of ships that matter too, critically so. Even in Traveller's Trillion Credit Squadron it was not just the Lanchestrian equation of force at play ["fighting power squared"] that won one Lenat/Eurisko its battles and two TCS titles. There is a hidden but very important lesson to be learned in the first Tournament Fleet Battle Final. True Lenat's Eurisko fleet fought and won the final .. but Lenat was worried because .. it faced off against a very different style of opponent. On a superficial examination, it looked like a near identical fleet [one that came from a fertile mind of a teenager without the aid of vast PDP computer time from a university - so true respect to him (who he was and what became of him I know not)]. Yet .. Lenat sighed in relief .. the opponent's fleet was not exactly the same as his. On the surface in one aspect it actually looked better, its Lanchestrian Fighting Power was higher, as it had more or slightly more powerful little fighting ships - its composition is sadly lost in the deeps of time, we only know of Eurisko's fleet listing). What am I getting at?
Lenat/Eurisko did not win by a random chance (rolling good dice) despite starting teh battle at a slight disadvantage in Lanchastrian Fighting Power strength. Yes, read that again. My conjecture is that Lenat/Eurisko's fleet actually looked a weaker fleet! We can imply that the other fleet was in fact stronger because, Lenat has spent "points/tonnage" on non-Eurisko class ships. The question is "why"? Wasting points like that would lose battles by reducing your fleet's Lanchestrian Fighting Value? [If this is incorrect it is my bad, but I think it holds water!] But it had teh opposite effect.
Lenat's Eurisko Fleet won because it had other "minor part" players that critically turned the tide of battle for him. The special ship in the first final was a "life boat" or "defensive shield" which if deployed correctly shielded the fleet from further damage. So, when the fleet was being beaten (yes "when", not "if" - assume during the course of the battle your fleet can be "placed in a losing situation" bu gos play from teh other side). This "shield" allowed the Lenat/Eurisko fleet to retire behind a shield and rebuild a badly damaged fleet in "game-battle time" (not campaign time) and then be ready to "go again". Yes, Lenat/Eurisko had a "repair ship" capability too. The nameless teenager's fleet was based on one good idea and over optimised for it - but Lenat had revised and refined his ideas further, by a proces that was effectively "Red Teaming". Eevry potential fleet solution was tested against "all other potential fleets he could think of" - rigorous and diverse testing.
He analysed his losses as much as his victories. When Eurisko was fighting Eurisko in a multiple simulation this was a hard thing "not to do" - you see the reason for your loses as well as your wins).
Despite a small loss in the optimal fighting strength, Lenat deliberately incorporated other ships types in the Lenat/Eurisko for special vital roles", for example such as a "lifeboat killer" that were never used in most competition play to my knowledge, but needed to be included for "completeness". That is to say to be called upon if needed - their role was a contingency against a certain type of fleet turning up (see below, look again at the chart, the "chart" hides many interactions that run deep about the historical period. Be careful what you wish for based purely on a surface reading of the chart - you need to understand the history (and I know Phil certainly does, he has the sea legs and been tutored in teh ways of the US Naval War College); his "good enough and in sufficient quantity" meme is very valid. In fact it sounds like the USN C-in-C Admiral King said in 1942, when he wanted an "offensive spirit" to emerge from teh USN in 1942, now and not next year. True there were plenty of ships destined to come online 43/44 but they could all be lost if they came onto a chess board set up for failure [to their collective credit Admirals "King-Nimitz-Spruance-Fletcher-Halsey" got the victories of Midway and Guadalcanal under their belt, despite the "Europe First" soldier friendly camp in the Supreme Allied High Command holding sway]):
Coming back to ship specifics, your ability as a ship to survive damage only slowly increases with size/tonnage, but the ability to inflict damage is not so constrained .. firing more missiles is cheaper. Phil highlighted an interesting dilemma for those living and working in ships in the "missile age" (see below, the bigger the ship the easier you are to be found and hit .. hmm .. what is the optimum size to be, or "how many small but expendable ships is best" - it circles back to Lenat and the question he was trying to answer, what is the best fleet composition to have?):
What does this mean for shipbuilders and navies? The Death Star dilemma ... which implies a navy must have enough ships as to not get too precious about losing one of them, because if that is not the case and you are too precious about your ships, your follow-on actions, albeit well intended, are naturally going to try and make the ships "better defended". This will actually make them even more precious to you, so paradoxically you cannot risk them where they are needed. They will be a bigger resource grab on your budget and more paradoxically a better target for hostile powers to sink and really hurt you. Care for your ships but they are to be used. Tonnage however is subject to the law of decreasing returns on protection levels (see below, in the end a "Death Star" is created and we all know how this ends, as it becomes something the Empire simply cannot afford to lose - but will. Footnote: "Use the Force Luke" .. but remember Luke has to hit a very small spot and got one shot at it - not easy but not impossible):
Alternatives? There is the "Shell Game". Can you find the target like finding a pea under a walnut shell, you are never quite sure what is under each shell so to be sure you have to hit everything which is beyond your offensive (first strike) resource capability This benefits lots of small craft over few of the large (see below, just like the proposed MX ICBM deliver system of the 1980's - it impossible to hit the American ICBM missile with 100% certainty [and therefore stop a retaliatory attack]. All by just adding a few more possible launching sites - the MX ICBM is listed as a reason why the (First) Cold War ended, as economically the Soviets could not afford the logical counter to it, which was a lot more expensive offensive first strike missiles):
But are .. you putting a best case scenario forward for the "little guys". I mean a "heavy sea" can swamp them right? But as time goes on they are getting better. You can cram them with missiles and when the conditions are favourable they are devasting, like Landsknecht Doppelsoldners they have "their day", a gad-fly brief summer day, whose life is gone in a flicker of the eye (see below, life would be certainly exciting to say the least on one of these in a combat zone):
But how can these "little ones" travel and victual? (see below, there is always a larger "mother" to hand", which can cross the Pacific and Atlantic Oceans, dropping off their charges and stay around to supply):
There is also stuff that is just living on the drawing board or virtual reality simulation today (see below, the student projects of today, are by graduation tomorrow's naval craft living and breathing in the water [for example The Minuteman Missile Corvette Concept Ship]):
Sometime the "little guys" are .. not so little (see below, an Ambassador Class Missile Corvette is not a speedboat or experimental catamaran - it can cruise quite long distances and packs a punch):
Therefore we are now back to the Trillion Credit Squadron (TCS) situation again, as a Lenat/Eurisko produced a "tug carried fleet". Eurisko Class ships got to places by using drop tanks. Once there in the battle they were immobile gun and missile platforms until the battle finished, rotating and tactical thrusts to position line of sight was all that was needed. See what happens when you tap into the genre of science fiction gaming, played in a way you it "dares you to think in the art of the possible, even if that is highly unconventional" and then see "if it is plausible within the rules of the game". There was an interesting interlude when a man called Doug came into town with an interesting horse called Eurisko that could "run fast". Lenat was prepared to follow the answer to the end with an unconventional answer to the TCS Fleet Problem (risking mockery from his academics peers, by playing games). The fleet problem is there, waiting for more innovative answers. However I think the last word should be left with Mr Jerry Pornelle, Phil's late father who obviously encouraged and inspired his son to be curious and imaginative, an author of works of Science Fiction in there own right (see below, "total respect" and I am going to look forward to reading them!):
Trillion Credit backlink:
Coming back to ship specifics, your ability as a ship to survive damage only slowly increases with size/tonnage, but the ability to inflict damage is not so constrained .. firing more missiles is cheaper. Phil highlighted an interesting dilemma for those living and working in ships in the "missile age" (see below, the bigger the ship the easier you are to be found and hit .. hmm .. what is the optimum size to be, or "how many small but expendable ships is best" - it circles back to Lenat and the question he was trying to answer, what is the best fleet composition to have?):
What does this mean for shipbuilders and navies? The Death Star dilemma ... which implies a navy must have enough ships as to not get too precious about losing one of them, because if that is not the case and you are too precious about your ships, your follow-on actions, albeit well intended, are naturally going to try and make the ships "better defended". This will actually make them even more precious to you, so paradoxically you cannot risk them where they are needed. They will be a bigger resource grab on your budget and more paradoxically a better target for hostile powers to sink and really hurt you. Care for your ships but they are to be used. Tonnage however is subject to the law of decreasing returns on protection levels (see below, in the end a "Death Star" is created and we all know how this ends, as it becomes something the Empire simply cannot afford to lose - but will. Footnote: "Use the Force Luke" .. but remember Luke has to hit a very small spot and got one shot at it - not easy but not impossible):
Alternatives? There is the "Shell Game". Can you find the target like finding a pea under a walnut shell, you are never quite sure what is under each shell so to be sure you have to hit everything which is beyond your offensive (first strike) resource capability This benefits lots of small craft over few of the large (see below, just like the proposed MX ICBM deliver system of the 1980's - it impossible to hit the American ICBM missile with 100% certainty [and therefore stop a retaliatory attack]. All by just adding a few more possible launching sites - the MX ICBM is listed as a reason why the (First) Cold War ended, as economically the Soviets could not afford the logical counter to it, which was a lot more expensive offensive first strike missiles):
But are .. you putting a best case scenario forward for the "little guys". I mean a "heavy sea" can swamp them right? But as time goes on they are getting better. You can cram them with missiles and when the conditions are favourable they are devasting, like Landsknecht Doppelsoldners they have "their day", a gad-fly brief summer day, whose life is gone in a flicker of the eye (see below, life would be certainly exciting to say the least on one of these in a combat zone):
But how can these "little ones" travel and victual? (see below, there is always a larger "mother" to hand", which can cross the Pacific and Atlantic Oceans, dropping off their charges and stay around to supply):
There is also stuff that is just living on the drawing board or virtual reality simulation today (see below, the student projects of today, are by graduation tomorrow's naval craft living and breathing in the water [for example The Minuteman Missile Corvette Concept Ship]):
Sometime the "little guys" are .. not so little (see below, an Ambassador Class Missile Corvette is not a speedboat or experimental catamaran - it can cruise quite long distances and packs a punch):
But if needs must [Minuteman Corvettes] can be transported piggy-back style (see below, that is "one heavy ass" lifter of a ship - something of that type of 'thing' brought back a Type 42 Royal Navy Destroyer all the way home (UK) from an embarrassing "holed below the waterline" incident in the southern hemisphere):
Therefore we are now back to the Trillion Credit Squadron (TCS) situation again, as a Lenat/Eurisko produced a "tug carried fleet". Eurisko Class ships got to places by using drop tanks. Once there in the battle they were immobile gun and missile platforms until the battle finished, rotating and tactical thrusts to position line of sight was all that was needed. See what happens when you tap into the genre of science fiction gaming, played in a way you it "dares you to think in the art of the possible, even if that is highly unconventional" and then see "if it is plausible within the rules of the game". There was an interesting interlude when a man called Doug came into town with an interesting horse called Eurisko that could "run fast". Lenat was prepared to follow the answer to the end with an unconventional answer to the TCS Fleet Problem (risking mockery from his academics peers, by playing games). The fleet problem is there, waiting for more innovative answers. However I think the last word should be left with Mr Jerry Pornelle, Phil's late father who obviously encouraged and inspired his son to be curious and imaginative, an author of works of Science Fiction in there own right (see below, "total respect" and I am going to look forward to reading them!):
Afterall who does not want to hear a good story, best wishes to you all and thanks to you for reading if you have got this far!
Trillion Credit backlink:
Footnote:
Legend has it, as written in his own words, Douglas Lenat was asked not to enter the the Origiins TCS Tournament in 1983. If he did organisers said they would just cancel the event. So Eurisko didn't enter, it retired undefeated, with the honorary rank of Admiral. Sadly I think we were all the lesser for that as who knows, Lenat never said a fleet devised by Eurisko could not be defeated, after all in its training - Eurisko was always trying to defeat itself as well as everybody else it knew of.
12 comments:
Geordie -
A very interesting and thought provoking article! One has a feeling 'small and lots' versus 'big and few' might well have implications in land and air warfare as well.
From an historical perspective, one thinks back to the vast and expensive battle fleets of World War One that ended up not doing very much. During WW2, the Germans appear to have got more mileage from their small craft (e.g. E-boats, trawlers and suchlike), although their puny 'fleet' navy, just by 'being there', seems to have had an effect beyond their numbers.
After a bit of a flirtation with heavy cruisers, the British seem to have gone in more for lighter craft overall. For their purposes, the light cruisers were more 'cost effective' than heavier units would have been. Mind you, that might have been an economic measure than a military doctrinal one. Their destroyers, with their 4-4.7" guns, were also pretty effective against their larger, more heavily armed German adversaries.
Overall, there seems to be considerable historical support for the notions expressed here. Speaking of littoral operations, I have located my copy of Paddy Griffith's 'Fjord' game from his 'Sandhurst War Games' publication. Not sure how I am to adapt to my own games, though!
Cheers, and thanks for a great posting!
Ion
Thanks Archduke, I have edited the article a bit, so hopefully it will read better for that now. I intend to listen to the full conference having enjoyed that particular 15 minutes (over three hours still to see) - some interesting speakers lined up!
PS: I am still digesting what Phil said in the video - in the end I ordered the Wayne Hughs Fleet Tactics book - not sure if I will regret that or nor - I suspect incoming maths avalanche
That is very interesting, and chimes with the space combat conclusions in Ian M Banks 'Culture' series - lots of small/middle sized ships are way better than a few big ones. The same in Seastrike, although I'm surprised very small ships worked in TCS. In High Guard we tried fighter fleets and they were useless and cruisers ettc as they lacked the spinal mount weapons you needed to inflict serious damage. Perhaps the combat model in TCS is different to HG? The core of my High Guard battle fleet were jumpless battleships transported by jump carriers - the space saved on jump drives allowed more weapons, armour etc for a given tonnage, so a similar concept to the immobile ships mentioned.
The effect of the mix of ship types on the effective combat value, as opposed to the Lanchesterian combat value, is right out of Dupuy or Biddle, and is why I am dubious of Lanchester as a model of modern warfare.
Thanks for teh Science Fiction pointer, M Banks "Culture" series I will follow that up Martin
Very small I suppose is a relative term - cheap cost may be a better description, as the Eurisko class is 11.1k tons and armoured .. so you had to focus on it to destroy one, they were not light drones. 4 Eurisko to 1 Standard Line of Battle. Each Eurisko had drop tanks and did not have carriers, instead it got somewhere and then was stuck, until " something else" which was a fuel unit came up? I think Lenat exploited that this was not thought through in the first TCS Origins Tournament (1981). Drop tanks were prohibited in the second TCS Tournament (1982) .. teh fleet was modified .. I do not have those stats, so it would be interesting to "know" the differences.
Geordie -
One observation that made me prick up my ears was the 'delivery system' to bring small craft (whose operational range was more limited by their small size) into the battle zone (theatre of war, sharp end, happening area). These 'mother ships' seem to me, though, as a primary target for any enemy capable of reaching it with missiles, drones or strike aircraft. One forms the impression these days that it is very hard to hide large units, and I infer from the article and the talk that the 'mother ships' would be pretty sizeable.
On the other hand, judging by certain events in Eastern Europe and the Middle East, the drones themselves are beginning to look like the last word in small, and numerous; cheap and effective. The nature of warfare is changing very rapidly, methinks.
Cheers,
Ion
This post has made me think too, I have come back and edited a few (more than a few times) just to clear up my and ideas, After watching Phil Pournelles YouTube presentation I had to get the ideas out of my head and into the open. Your comments start more thought processes up. The mother ship itself becomes the Death Star if it is not a throwaway resource. Lenat's Eurisko ships had drop tanks .. to get to the battle but then were stuck? I will have to "go read" the rules a bit more. Your points about the Middle East and Ukraine are well noted! There is an Accountant somewhere calculating how much it costs to kill you (or Object A) and devising a KPI to say if it is worth it - on a real-time Dashboard!
I should have said .. "just to clear up my grammar and my ideas"
I see I am going to have to set-up and play a game of The Trillion Credit Squadron .. too interesting not to!
The specific Ian M Banks book I was thinking of is 'The Algebraist'. He wrote quite a few books, they aren't traditional space operas, one of them (Excession) is entirely from the pov of AIs.
Thanks! Cool, it sounds and a must read!
Post a Comment