Bob Ward’s supporting FergusonMay 7, 2020 Tim WorstallEconomics36 CommentsTherefore, ineluctably, Ferguson is wrong. We need no further evidence than Bob. previousNot convinced this worksnextTre Professori on banking 36 thoughts on “Bob Ward’s supporting Ferguson” dearieme May 7, 2020 at 8:35 am “a scientist … with evidence”. Oh balls, the predictions of his model are not “evidence” in a scientific sense. In fact it’s not clear to me that he’s worth calling a scientist at all. The wonder of it all is that the govt should have chosen this berk as its crystal ball man: his record is dismal. Mad Cow, Foot and Mouth, flu: he’s buggered them all up. I can see, though, why he’d appeal to a Global Warmmonger. Steve May 7, 2020 at 8:39 am Ferguson has been under attack ever since his research team’s modelling suggested in mid-March that hundreds of thousands of deaths in the UK from Covid-19 were possible if stronger efforts were not made to curb the growing epidemic. Is Ward a trick cyclist, because that’s one smooth backpedal. Maybe the public would trust scientists more if they didn’t lie all the time? Pat May 7, 2020 at 9:05 am Ferguson never published his model. Hence it was entirely unchecked. (Alright I believe he has belatedly published a tidied up version, which is reckoned to be crap). Professor Jones never published data or calculations for his work, maintaining that he had lost them. Since they were produced at public expense, the public should have got their money back. Professor Mann only released his code under pressure, and that turned out to be crap. He still maintains the data to be his private property. Bob Ward approves all of the above people and promotes their work. Tim is right, if Bob Ward says it, it is untrue. Dave Ward May 7, 2020 at 9:12 am “Alright I believe he has belatedly published a tidied up version, which is reckoned to be crap” https://lockdownsceptics.org/code-review-of-fergusons-model/ Bloke in North Dorset May 7, 2020 at 9:36 am I really shouldn’t have wasted my time beyond this in the first paragraph: “Britain’s political-media complex”, but I did. The rest of it is a frightened defence of modelling and those who carry it out, with accompanying ad hominem attack on its critics, lest the public figure out that climate modelling is based on the same scam. Gamecock May 7, 2020 at 10:07 am They abuse the word ‘modeling.’ It’s simulations for entertainment purposes only. ‘a campaign to discredit him by those ideologically opposed to government interventions’ OMFG! We still have people opposed to government interventions? In the 21st century? ‘and who have used such tactics against scientists in other fields, particularly climate change.’ ‘Such tactics’ are the intellectual property of the Left. How dare anyone else use them! Ferguson predicted 2.2 million deaths in the U.S. He is one of those responsible for my gym being closed, my shooting range being closed. DocBud May 7, 2020 at 10:48 am As deariemie quite rightly points out, numerical modelling of complex systems with significant uncertainty and unknowns is not evidence. At best it is an educated guess. Complex situations can be modelled effectively where the interaction between the constituent parts is known and the properties are certain, e.g. a bridge (accepting that there are manufacturing tolerances but these can be taken into account). Pandemics cannot be accurately or even reliably modelled because a host of inputs, such as the the infection rate and human behaviour are uncertain. The same is true of other disciplines characterised by variability and uncertainty such as geotechnical engineering and the study of the climate. PJF May 7, 2020 at 11:05 am He is one of those responsible for . . . my shooting range being closed. Commercial opportunity for new target design; get ready for reopening. Rob May 7, 2020 at 11:15 am Expert epidemiologists think they are also expert programmers, now shown to be a fallacy. Amazing that one of the biggest policy decisions probably since 1939 was dependent upon a secret computer model, never reviewed by actual experts in the field. ANY model government policy depends on MUST be completely public, so that it can be verified to actually work and also all assumptions are clear. Bob’s worried – if this model can be shown to be an amateurish mess, what about all the climate ones? It’s unthinkable, so the call goes out – “Save Mr Ferguson!” JuliaM May 7, 2020 at 11:28 am @Steve: ’ Maybe the public would trust scientists more if they didn’t lie all the time?’ See also: journalists. Bloke in North Dorset May 7, 2020 at 12:09 pm They abuse the word ‘modeling.’ It’s simulations for entertainment purposes only. As that review says, its nothing more than Sim City without the graphics. As a society were getting so used to models telling us stuff and the MSM and government accepting them without question we’ll soon be at the to the point where if reality doesn’t match the model realty must be wrong and we must change. Matt May 7, 2020 at 12:46 pm Generally, scientists are a pretty honest bunch. Due to their honesty, they are rarely certain about anything because they are aware just how complex everything is and how many caveats and exceptions there are to any model. The problems are twofold: (1) non-scientists/engineers — i.e. the media and government — tend to want certainty rather than truth. Certainty fits in a tweet or a headline, the whole story doesn’t: as above, real scientists don’t do certainty (2) there are wannabe politicians pretending to be scientists who are either certain of stuff (because they’re too stupid to see the situations in which they might be wrong [Murphy]) or willing to pretend to be certain of it (because they want to have political power/be on TV [Ferguson]) The solutions are to have a better media/politician caste, or to refine the term ‘scientist’ to refer exclusively to somebody who studies the natural sciences (physics, chemistry, biology and the specialisations derived therefrom). “Social* science” is just a branch of the humanities and its adherents should be named accordingly. But we ain’t gonna get either of those, now, are we? *prepending “social” to anything appears to have the effect of inverting its meaning, see also: justice, security, care, democracy Jason Lynch May 7, 2020 at 1:11 pm I made a reasonable living for quite a while doing computer modelling (of “warheads on foreheads” rather than epidemics, but you get the point) and now make an even better living out of technical assurance (checking research to be sure that it’s answered the question and the evidence is to a high enough standard). There was a presentation by an incredibly talented, handsome, well-endowed gentleman at the International Symposium for Military Operational Research a couple of years ago, looking at the limits and risks of modelling and wargaming:- All models are wrong, some models are useful A career lesson I learned early and applied to others, is that just because your model has a hundred thousand lines of code, and a thousand programmer-years of effort went into building it… the results may still be garbage, they’re just really expensive garbage if your inputs are incorrect or inaccurate. For one real-world lesson from a while ago… a missile system was being developed that used an infrared camera to track the incoming target and the outgoing missile, so it could guide the missile to hit and kill. The system to detect “target” and “missile” in the infrared scene was built using a simulation of infrared imagery, one accepted to be very good based on a lot of observations taken in different places and conditions). And in its first real-world outing, it failed really badly. The modellers had got the environment, the sea, the sky and the sun all right. They’d allowed for cloud, rain, mist, snow, and other environmental conditions. They’d measured the smoke from the outgoing missile’s rocket motor, and the thermal “glare” it would produce. So what went wrong? Turns out, two problems. A burning rocket motor generates a lot of heat. And that mass of hot, turbulent air distorts and confuses the camera’s view: that hadn’t been considered in the model. Nor had an issue where the rocket burned out, and the last chunks of burning rocket fuel were ‘spat’ out of the exhaust: which the system would sometimes track, and try to issue guidance commands to (sending the missile off in wild directions). Both fixed fairly quickly, but later in the process than we’d have liked – and the original plan had been to do no firing trials, relying only on simulation and modelling for acceptance. (We had managed to crowbar a couple of trials in… and waited until nobody was looking before doing the “Told You So!” dance on our desks) jgh May 7, 2020 at 1:30 pm Fergusson hasn’t published his model, he’s only published a flawed implementation of his model. We still have had no sight of the model that the software is attempting to implement. Bloke in Wales May 7, 2020 at 1:50 pm the original plan had been to do no firing trials, relying only on simulation and modelling for acceptance. You really get people suggesting that missiles go into service having never flown before there is a real, incoming target coming your way? And then discovering all the “Oops, didn’t think of that one!” moments that all complex systems have? Sounds like it’s time to reinstate the Roman practice of forcing engineers to sleep under the bridges they just built – or in this case put them on the ship with an exocet heading their way. Jim May 7, 2020 at 4:04 pm “You really get people suggesting that missiles go into service having never flown before there is a real, incoming target coming your way?” You mean like the US Navy did with the Mk14 torpedo in the 1930s? Which if they’d tested properly and ironed out the bugs would maybe have knocked months if not a year off the war in the Pacific, given the number of Japanese ships they would have sunk in 1942 and 43 if the goddamn torpedoes had actually exploded on impact, or run at the correct depth, or not veered off in random directions? dearieme May 7, 2020 at 4:58 pm @Jim: it was a conscious decision not to test them, on the grounds that each torpedo you tested reduced your inventory by one. If this weren’t Worstall’s blog it might be tempting to refer to the bone-headedness of admirals, generals, and the like. Tractor Gent May 7, 2020 at 5:02 pm BiW: Perhaps the bean-counters ought to join them! Kevin B May 7, 2020 at 6:16 pm The situation is a lot worse in the climate panic industry; there they refer to the outputs of models as data. “The data proves the world is warming!!!” they say. “What data?” say I. “The average monthly global temperature has gone up over 1deg C since 1850. Ha! Ha! So there!” “But that’s not data, it’s the output of a model.” “No it’s data, denier!!!” “Show me on the globe the location of the orifice that I can stick my thermometer in and measure the average monthly global temperature. Or better yet, publish the code that calculates the monthly average global temperature and the data it uses to calculate it. Oh and the fiddles and fudges it uses to ‘get rid of the forties bulge.’ and the ‘hiatus’ and all the other inconvenient challenges to the narrative.” Perhaps, as some opine, the current mess might cause our betters to question the outputs of models and the experts who peddle them as real, but I have my doubts. They seem to be suckers for a good sales pitch and a few flashy graphs, perhaps because they use those techniques so much themselves. Pcar May 7, 2020 at 6:47 pm @jgh Fergusson hasn’t published his model, he’s only published a flawed implementation of his model. We still have had no sight of the model that the software is attempting to implement Correct: In March Prof Neil Ferguson told Sunday Telegraph, on Sat, he would release his model “Next Week” for public scrutiny He lied and still refuses to release full model code to Gov or public Show Your Work Ferguson Sex Scandal Prompts Calls for UK Govt to Release Coronavirus Models “Conservative MP David Davis said that “a bigger issue than Professor Ferguson’s private life is the accuracy of his model. When applied to the Swedish policy, it forecast 40,000 deaths by now, over 15 times the reality.” “We need the whole model, its assumptions and working in the public domain. We can no longer run our strategy on secret advice and potentially flawed calculations” Hugely irresponsible & asinine of Govs to accept his results at face value, despite his previous massively failed models. Even more so given the Lockdown’s huge cost to UK: £2.4 Billion per day Pcar May 7, 2020 at 6:48 pm What led to this absurd worship of academic models? Models are what children and child-adults build and play with because they can’t do it for real @BiND I remember when I switched from questioning to full sceptic on Global Warming: – Met Office CEO speech “…we must ignore empirical evidence and trust what models tell us…” Ummmm May 7, 2020 at 8:35 pm Given that the model has a mortality rate of 1.2% that’s going to account for big chunk of the difference. That was a figure quoted by the Canadian govt in its models with a note that the source was the Imperial model Raffles May 7, 2020 at 9:36 pm Didn’t I read the the Royal Navy’s new anti-aircraft (planes or misses) missile system supplied by a French company has never been tested properly? From memory it failed its subsonic test, but was accepted anyway even though it was never tried against supersonic targets – its primary purpose. Apparently the simulation results were good enough for acceptance into service. All just from memory, so I may be off target.. Jason Lynch May 7, 2020 at 10:04 pm Raffles, If you mean Sea Viper (on the Type 45s) it’s been tested on everything up to a GQM-163 Coyote (Mach 3 seaskimmer) successfully, which validated the outstanding simulation results. One batch of missiles had a hardware problem that was found in a trial and needed fixing, which was done. While it’s really expensive, it has been tested – you don’t want to find a problem in wartime when it’s too late to fix. (Jim’s not the only person with memories of systems that didn’t need proper testing and got paper passes into service…) True dit: about ten years ago, we had HMS Portland firing a Harpoon antiship missile at a barge target off Benbecula, and HMS Diamond was in the area. A certain analyst suggested we could put Diamond in line with part of the missile’s flight plan (it was making some turns en route) so the missile would be running right at her, so we could test everything up to the “whoosh! bang!” part of an anti-missile engagement – detection, classification, weapon allocation, system setting up the shot and asking “please give me permission to delete this target?” and we just wouldn’t turn the key to FIRE. Misses testing the weapon, but ensures that the radars and combat system react properly to a real missile. Of course, said analyst was strongly invited to observe the trial from the Ops Room on Diamond… but if you’re asking others to trust their lives to the kit you shouldn’t be shy of joining in. Worked brilliantly, the only issue being that the missile (despite having a telemetry package instead of a warhead) set the target barge on fire. So, no second shot. The CO asked me^H^H the analyst to work out whether Portland could fire the second Harpoon at us, and we’d just shoot it down. Navs and I looked at each other, thought “well, WTF?” and started planning the shot. Apparently the scream of “NOOOOOO!” from Range Control could be heard in low earth orbit… but it was a fun idea. (We were probably never going to get to do it but why not try? A destroyer CO has to find his amusement where he can…) dearieme May 7, 2020 at 10:52 pm “What led to this absurd worship of academic models?” If they are as good as some of mine a bit of genuflection would be entirely in order. That’s why we could commercialise them: they were good. You’d want some software engineers to give them a GUI, make them easier for a non-expert to use, and even give the code an MOT. But any notion that a bunch of computer monkeys could have conceived and designed the models and algorithms is ludicrous. If you’re going to apply science you’d better know some, and a bit of maths as well. And even then it would be mad to fall in love with a model. Even if it can demonstrate lots of practical success, who knows what problem might defeat it? Mind you it was not my practice to guess at parameter values; the huge thermodynamics literature usually gave me what I needed (though sometimes after a bit of tidying up). What beats me is that anyone with any background in science or engineering could take models in the Ferguson style seriously. Judging by the comments of “Sue Denim” the buggers used ‘stochastic’ to mean haphazard. Bogus to his boots, Ferguson is. Snake oil merchant. So who in government/civil service decided he was the fellow to use? Pcar May 7, 2020 at 11:37 pm @Jason Lynch May 7, 2020 at 1:11 pm +1 That IR missile rang a bell: Similar to when Virgin & two other new teams joined F1 and were to be low-budget CFD only – after a while they had to start begging FOM to rent time in other teams wind tunnels Bloke in North Dorset May 8, 2020 at 8:50 am All this talk of missiles stirred the old memory pot. Given what we know about the memory these two may well be less than perfect (bollocks in the vernacular, but I still enjoy telling the tales). Just over 38 years ago I and 3 colleagues were whisked out of Germany to bolster a nascent Royal Marines electronic warfare organisation, Yankee Troop. Whilst at Lympston waiting to fly out to Ascension Island to join the rest of the task force we were given lots of briefings about the Royal Marines and the Royal Navy. One of those was about the Sea Wolf anti missile system and how good it was and how it would protect us against Exocets. As Wikipedia drily puts it: The Type 42/22 combination achieved some degree of success during the Falklands Conflict, but the Sea Wolf armed ships were unable to prevent the loss of HMS Coventry and heavy damage to HMS Glasgow Most Navy commenters at the time called it fucking useless. The other (less reliable) memory is being in Port San Carlos and watching either a Rapier or Blowpipe missile decide that the Royal Marine Snow Cat vehicle heading up the hill on the other side of the water was a much more interesting target than the Argentinian Air Force whizzing past overhead. I suppose computer simulations weren’t as good then and the only real test is in battle. Gamecock May 8, 2020 at 1:27 pm U.S. Sidewinder and Sparrow air-to-air missiles were awful in the early years of the Viet Nam War. The brass was so enamored of them that they removed guns from U.S. fighters. “Who needs guns when you can shoot ’em down at 25 miles?” Oops. They were later added back in. It was a kluge for Phantom IIs, as they were designed from the beginning without a gun. It was EXACTLY like U.S. troubles with torpedoes in the early part of WWII, as commented above. Same sh|t, 30 years apart. rhoda klapp May 8, 2020 at 3:22 pm Sidewinder kill rate 12.5%, Sparrow about 8%. Didn’t change much during 8 years of air combat. Luckily the other side had a sidewinder copy.. Pcar May 8, 2020 at 6:41 pm @bis 1982 Falklands was our last war using equipment not much better than WWII Compare it to equipment used 8 years later in 1990 Gulf War I – Op Desert Shield The why: microprocessors – small, cheap, powerful @Gamecock MOD & RAF Chiefs here still insist guns not necessary, then have to try to retrofit Jason Lynch May 8, 2020 at 7:58 pm BiND, Sea Wolf worked really well when it had targets (three kills from five shots, outstanding by missile standards for 1982) – the problems were twofold. The only two Sea Wolf ships spent most of their time goalkeeping the carriers, since they were the only credible hardkill against Exocet; and when Broadsword was working with Coventry, the first raid exposed a software problem (see below) and in the second, Broadsword had fixed the problem but Coventry cut across the line of fire, aborting the engagement and winning the hits. (They used to have the 910 camera footage in the Sea Wolf shop at HMS Collingwood – painful stuff to watch as the RAT box locks up on the lead Skyhawk, then Coventry‘s upperworks block the view – those Skyhawks were coming in low) The software problem was one where the system would, in automatic mode, engage the most threatening target. The problem was that it was designed for straight-running missiles in the North Atlantic, not weaving aircraft popping up from over a hill; the aircraft kept changing places in the threat list and the system kept switching from one to the other. A quick fix by a Plessey technician aboard (the Mastermind rule, “I’ve started so I’ll finish”) sorted it out very quickly. Of course everything that would fire was turning live rounds into empty cases in the direction of the attacking aircraft as fast as feasible, but outside of CIWS systems like Phalanx or full radar fire control (HMS Avenger got a probable on a Skyhawk with her 4.5″ gun), 500-knot targets are only distracted by gunfire. Worth doing to maybe turn their attack from “hit” to “missed”, but don’t expect to shoot much down. Jason Lynch May 8, 2020 at 8:22 pm Gamecock, Bit more complicated than the version usually put about. First up, note that the USAF and USN both flew over Vietnam, by late in the war both using F-4 Phantoms as their main fighter aircraft (different versions, same basic airframe) and the Phantom carried Sparrow and Sidewinder missiles in both services. The USAF had serious troubles making missiles work in Vietnam, for a variety of reasons. Some were technical – these were weapons designed to swat Soviet bombers at high altitude, not agile fighters at low level. Others were down to the problem that on any given day there’d be 200-300 aircraft over Vietnam and maybe two of them would be NVAF, so visual ID was required (negating the range advantage of the AIM-7 Sparrow). Missiles were complex and not well maintained, and loaded, carried and unloaded often without being fired (air-to-air encounters were very rare) so they had a horrible failure rate. But the big problem was training: the 1960s USAF was dominated by the Strategic Air Command bomber generals, tactical fighters were a disliked afterthought, and the war wasn’t popular; so the USAF adopted a “no involuntary second tour” policy for aircrew. After all the F-105, F-4, and other “pointy nose fast mover” types had done one combat tour, the USAF were taking pilots from tankers, transports and bombers, putting them through a six-week course and certifying them as F-4 pilots. The course nominally included four (count’em, four) air combat sorties for training, but those were often skipped due to weather and safety concerns. So, many USAF pilots had no air-to-air combat training, and were carrying complex missiles they had received only the most hasty training on. Tactics, too, were basic in the extreme, using the WW2-vintage “fighting wing” formation where four aircraft flew in a tight formation, nominally to protect each other (it worked at 300 knots against enemies with machine guns) and only the flight lead was allowed to fire weapons offensively. This didn’t work well in the jet age. As a result, the USAF didn’t do well against the occasional NVAF MiGs. The US Navy had to fly off carriers on Yankee Station, so you had to be qualified to land a F-4 Phantom on a WW2-vintage carrier at night and/or in bad weather – no chance of drafting in large numbers of other pilots. This meant USN pilots knew their aircraft much better and had received more training (and by 1972 the USN had set up the Fighter Weapons School – the infamous “Top Gun” – to further improve its air-to-air skills). It did, of course, mean back-to-back Southeast Asia cruises for naval aviators with massive morale and retention issues… there’s no free lunch. As a result, in 1972 the USAF had “fixed its problem” with missiles by fitting a gun to their Phantoms (the F-4D had a gun pod under the fuselage, the F-4E had one built-in). During the Linebacker bombing campaigns of that year, they shot down 48 NVAF MiGs while losing 24 aircraft in air-to-air combat. The USN had shrugged off fitting a gun to the Phantom, and suffered terribly as a result: destroying 24 NVAF MiGs, while losing four aircraft air-to-air. The USN had better tactics, better training, and took more care of their weapons: the USAF, having declared the missiles flawed and faulty, continued to have very high failure rates (and yet, one point overlooked is that of those 72 US air-to-air kills in 1972, only seven were with gunfire: it was really hard and actively dangerous to try to get into position for a guns kill on an agile MiG-21, and at least one USAF aircraft was shot down by his would-be prey’s wingman). Air-to-air gun engagements had a very low kill probability even in cases where aircraft fired (under 10%) and the short range and very limited targeting envelope meant many more gun-armed aircraft never got close enough to fire. There was then, of course, a political backlash that It Was Only The Guns That Won It, because that allowed the USAF leadership to dismiss concerns about training, equipment, doctrine, and leadership and claim that “the problem was that we didn’t have gun-armed aircraft and we will never repeat that mistake”. Excuse the lengthy quote, but this is what a USAF pilot (who volunteered for a second tour in Phantoms after flying his first in F-105s, and so was pretty keen and grounded) had to say… You’ll seldom get a tactical aviator to confess that he didn’t have a clue, but I’ll admit it with regard to the F-4/AIM-7 system. (Of course, it’s just between you and me and if you tell anyone else, I’ll deny I ever said it–especially, don’t let Guy Alcala know!) I went to war in the F-4 after a “Cat IV” checkout which was 45 days and just under 30 hours flying time in F-4Cs at Luke. I arrived at Korat to fly E’s which had a considerably different weapons system–and within the first month the wing converted to TCTO-556 with yet a different cockpit configuration. I learned about Dive Toss fairly quickly and didn’t have much trouble taking advantage of the better radar, but that can be related to the fact that I’d done a lot of radar work in the 105. I don’t know how guys whose previous experience had been in the F-100 or 86 were able to get much out of the scope after a short course checkout. My understanding of the AIM-7E was pretty much that the WSO would lockup and I’d shoot when the dot was centered and the circle got big. In most instances I’d be swatting at the plastic tubing extension we all put on the weapons select switch (or after -556 flipping the pinky switch) to go to AIM-9 or guns because we couldn’t get a lock, we were inside parameters or the missiles didn’t “tune”–whatever that meant. I could understand “growl” and Sidewinder boresighting, and guns tracking was natural but things like English Bias and “interlocks” were all Greek to me. Training never improved much during the five years I spent flying Phantoms. The details of AIM-7E-2 employment were locked in the safes in the Wing Weapons shop and only doled out grudgingly by the “target arms.” The secrets of intercept geometry were readily available and with practice even an ol’ ground attack puke like me could figure out how to “hot up” an intercept to keep from tail chasing, but the details of even the bit checks were never well understood. (What was the significance of making the dot run around between the two circles anyway??) I’ve got great sympathy now for guys like you, Dweezil, who were trying to decipher what went wrong when some sloping forehead fighter pilot complained that the #$%#!*& missile didn’t work. I never did have a clue about what the Sparrow was supposed to do or how it did it. Ed Rasimus Fighter Pilot (ret) Apologies for the essay – this is one of those system engineering case studies where “what most people know, just ain’t so”… Jason Lynch May 8, 2020 at 8:34 pm Pcar, Note that when the Treasury tried to delete the BK27 cannon from Typhoon (then still Eurofighter 2000) in 2002 or so, the RAF nodded and tugged the forelock… then pointed out that to keep the aircraft certified to fly, an absolutely identical spaceweight replacement for the gun and its ammunition would have to be produced and signed off, which would cost more than just “fitting the gun and never using it”. Or, the RAF could pay for a new, bespoke “RAF Only” suite of Flight Control Software for the altered balance of the aircraft, spending a mere £2-3 billion pounds, in order to save some £5 million a year over twenty years. But, some savings could be largely realised by simply not training on or using the gun, and flying with inert drill rounds replacing the 27mm High Explosive Incendiary shells usually carried, so this option was taken (and the plan to reduce the Typhoon fleet size delayed with this saving credited against it). Of course within a couple of years we were bollocks-deep in Afghanistan, and there was pressure to send Typhoon out there to show it wasn’t a vanity purchase… and the urgent need was to get the 27mm cannon working for strafing, since at that point the Typhoon’s air-to-ground options were limited to “show of force” posturing or 1,000lb bombs and something smaller was desired for “troops in contact” situations, and since it works it might as well be fully operational… Since it was the same gun as fitted to the entire Tornado fleet, it was relatively trivial to restore it to full operating capability. It’s probably not necessary, but that wasn’t as clear in the early days of Typhoon design as it is now. Bluntly, when a missile cued by a helmet sight (pilot looks, missile locks, missile kills) can reliably hit a target 120° off your nose, trying to get into position for a guns shot is pure suicide. However, if the gun’s there you may as well keep it… There’s an argument over whether the F-35Bs we’re buying will get many or any of the external gun pod: some claim it’s vital, essential, et cetera, others note the cost and limited utility of bringing a low-observable aircraft within 400 metres of its target to try to kill it. rhoda klapp May 8, 2020 at 9:02 pm Jason, I remember Ed Rasimus from rec.aviation.military, twenty years ago. I’ve also read his bio of Robin Olds. IMHO the best discussion of the difficulties the USAF and USN had over North Vietnam is ‘Clashes’, by Marshal Michel. It echoes the problems you outine above but in book-legth detail. As for cannon, I was about eight when they first took the adens out of the Lightning, EE at the time, I suppose. I knew it was wrong then. You can do things with a cannon that you can’t do with a missile, especially in meeting unanticipated challenges. Like, in the Lightning case, three targets coming along. Gamecock May 9, 2020 at 12:08 am “trying to get into position for a guns shot is pure suicide” The American way is to sneak up behind them and shoot ’em in the back. USN doctrine in the Pacific in WWII was NO DOG FIGHTING. A pilot could get in deep trouble for it. You race up behind the enemy, blast them, then speed off. Don’t try to turn with them. Use your speed, armor, and armament to your advantage. Pcar May 9, 2020 at 7:23 pm @Jason Great post “agile MiG-21” wouldn’t have existed if Labour Gov’t Prime Minister Clement Attlee hadn’t ordered RR to sell USSR some RR Neme enginges despite RR & USA protests Sidewinder AIM-7 & 9: am I correct that it required IR lock on engine before firing and if IR lock lost after firing it wandered aimlessly until no fuel? “Typhoon’s air-to-ground options were limited to “show of force” posturing or 1,000lb bombs” Yep. In Libya they had to fly as “wing-man” to Tornados which did all the targeting – effectively Typhoon was a remote control weapon carrier drone Guns: always good to have a last line of attack/defence – as is using plane as weapon as a WWI pilot did & WWII Spitfire pilot did both over London Leave a Reply Cancel replyYour email address will not be published. Required fields are marked *Comment Name * Email * Website Save my name, email, and website in this browser for the next time I comment.