streakeagle

Keep in mind, SFP1/WoX series had absolutely horrible AI for dogfighting from the release of WoV to the release of WoI, that's quite a few years! They barely engaged in useful maneuvers, and when they did they almost never fired, and if they fired at all, they only hit anything on accident. For the duration of those years I had only two options: fly online (which I did as much as possible given the small online crowd and my own schedule) or use an old SFP1 SP2a install for single player, especially for the Korea mod. WoI brought back good AI, but they crashed into the ground quite a bit. The latest version of the AI is by far the best implementation, especially considering weapons employment from missiles to rockets to guns, but they still like to hit the ground a bit more often than I would prefer. The only glaring limitation I see in the present AI besides ground collision risk is a general preference to fly the 2-d turn fight. Fortunately, my enemies are usually angles fighters anyway, so it doesn't hurt them. But watching heavy energy fighters try to turn fight with agile fighters is quite painful. The flight modeling seems too generous in available lift, stability, and usable angles of attack, as it seems like a can turn fight against MiG-17s with heavy fighters like F-100s, F-104s, F-105s, and F-4s when I shouldn't have too much of a prayer without using vertical tactics or lucky "Maverick" style overshoot snapshots. But overall, the balance and net results appear to be far more realistic than competing sims. Before missile accuracy/reliability was souped up to make the game "more fun" during SF2NA patching, the results were exceptionally close to reality in AI vs AI missile fights. Gun lethality has always been substantially higher than other games, both in % of rounds hitting and the damage caused by a single hit. I am ok with 20mm/23mm doing this to fighters, especially the very vulnerable MiG-21 where a single hit could result in a catastrophic loss. But 0.50 cal/12.7 mm and 0.30 cal/7.62 mm are typically way too effective as well. DCS swings in the other direction with multiple 20 mm hits on a MiG-29 doing little more than causing fuel leaks and possibly an engine fire. Against a MiG-29 flying straight and level, I can empty quite a few rounds into it until I light both engines on fire. With the wing planform exposed by a hard turn, I can easily saw a wing off for a very quick kill with minimum rounds. A stern approach requires me to do one of two things: multiple bursts until I light both engines on fire or a single burst with some rudder kicked in to saw across both engines. The nice part about air combat in DCS is that I can get into steady state gunsight tracking to verify the accuracy of the gunsight and observe the strike pattern. Unless a target is focused on going home, SF2 targets always know you are there and jink like mad if they have the energy to do it. So SF2 is more realistic, but not as good for practice and learning the physics of the bullet stream. Back in the day when I played SFP1/WoX online almost every night, I got extremely proficient at hitting with US 20mm whether it was a 6-barrel or four separate guns in the belly or nose like the F-100 and F-8. But I don't play enough to have that instinctive judgement and have to rely on the LCOSS too much, not effective against SF2 flip-flopping fish AI. To be fair to DCS, real pilots don't magically maintain situational awareness and flop all over continuously in a pattern that is perfect for breaking a gunsight track. Top Gun and Red Flag gun camera footage from the 70s and 80s clearly show opponents caught in steady state turns/loops allowing the attacker to momentarily sustain classic steady gunsight tracking. What I would prefer is to fight AI that flies historically correct with the tactics and skill levels appropriate for the nation and time frame. SF2 is the only sim that I know of that permits modding the AI (per aircraft type!) to get the desired results. Playing online isn't really good for historical fights: guys who fly online all the time typically become above average honchos and the talented ones are almost unbeatable no matter what aircraft they fly against any reasonable numerical odds, which is not very realistic and certainly not fun for pilots who can't fly enough to be above average.

If you can overlook the "easy" FMs and AI cheats, SF2 single player dogfight AI is almost the best there is. In my opinion, only Battle of Britain 2 Wings of Victory has more challenging/realistic dogfight AI. I have been playing tons of DCS, and now that the F-15C has a very detailed flight model, I have been spending most of my time flying the Eagle in DACT with guns only. It doesn't matter whether I fly 1 vs 1, 1 vs 2, or 1 vs 4 against the MiG-23, MiG-29, or Su-27. The MiG-29s are by far the most aggressive of the three, but I can still smoke all 4 with relative ease. I can flip it the other way and fly a MiG-29 or Su-27 vs 4 x F-4E, F-15, or F-16 and still end up with the same results. If I don't go for a head on shot at the merge, I can easily avoid taking any fire at all, pick a target, focus exclusively on my target, then look for another target until all four are dead. But in SF2, things are quite a bit different. Using 4 x F-16A Netz to represent the MiG-29, I am in danger of taking fire at the merge whether I take a headon shot or not. In fact, there is a fair chance of being shot down altogether. The AI is aggressive and accurate. Combine that with the much higher lethality of gunfire in SF2, and you have a real challenge. Unlike DCS, the other AI pilots in the same flight will aggressively pursue and shoot even without regard for what the leader is doing. If I focus on one target and ignore the others, one or more of the other will rapidly position themselves for a good shot and possibly kill me. My target may try to evade, extend, force an overshoot, and/or drag me past his friends. In DCS, I can casually and carefully line up the gunsight with the target, then fire to hit specific parts: engine, wing, cockpit, etc. Not so in SF2, that baby jinks and jinks well, frequently spoiling my aim and certainly preventing shots aimed at anything more specific than center of mass. In SF2, it is rare for me to kill more than one target before getting sliced and diced from the rear or side. Of course, the AI in SF2 specifically focus on the player and know where you are at all times regardless of whether you get below and behind them. Perfect situational awareness or not, SF2 pilots with max skill level are most certainly capable of killing me especially with a 4 to 1 numerical advantage. DCS AI pilots only kill me under very specific circumstances and only rarely when in those circumstances. I doubt DCS is going to improve their AI anytime soon. At least the AI in DCS don't form a "conga line" like they do in Falcon 4 where I can catch up to the tail and munch on some aircraft that were all trying to follow me in one big circle.

There is an entire 3rd party WW2 game based on the ArmA2 engine, but free mods include Invasion:1944: http://www.moddb.com/mods/invasion-19441 If you don't mind paying you can try this ArmA 2 WW2 mod: Iron Front: http://ironfront.deepsilver.com/en/

A great game that can't be beat for moddability/fun is the old Operation Flashpoint GOTY which has been re-released as ARMA Cold War Assault on Steam. This will run on just about any PC given its age and is a great shooter with fun arcade action in tanks, helos, and planes. My son loves driving around in an M1A1 or M60A3 and decimating infantry. Of course, if you have modern, strong gaming PCs, then ArmA 2 or ArmA 3 might be what you are looking for. All of them have tons of free mods.

Did FC3 sell well because the aircraft were easier to fly? or because there are a heck of a lot more people interested in flying F-15C/MiG-29/Su-27/Su-33 than the A-10C/P-51D/Ka-50? Who doesn't want an Su-27 with a PFM that accurately models the Flanker's unique high AoA controlability? Some people may prefer easier flying, which is already available anyway even with full DCS level aircraft, but it is the type of aircraft that people are more concerned about. Throw out an F-14, F-16, or F/A-18 or for those more interested in stealth/tech, F-22 or F-35 and the dollars will roll in. Obviously, the newer aircraft cannot be modeled very accurately at all, but people still want to fly them. Casual or not, jet sims have never sold as well as WW2 prop sims. I am willing to bet that despite being full DCS complexity that the WW2 modules will kill the post WW2 thru modern jets in sales, including FC3 if the AI doesn't screw up the experience. I don't understand why Korea has not been covered my more companies. Many of the same principles as WW2 even some of the same aircraft plus early jets with minimal radar and no guided missiles. Of course the era I prefer is both the least modeled, the least profitable, and probably the most complex to model in any detail: 1960s/1970s.

If the avionics are little better than SF2 and the SFM is absolutely inferior to SF2, why would I ever fly DCS? At a bare minimum, I expect the AFM of the Su-25T, but having enjoyed the PFM, I really don't want to buy anything less. Having thoroughly enjoyed the UH-1H and P-51D more than any other aircraft I have ever flown in all of my flight sim history, I am more than patient of enough to wait for more aircraft of their equal. I truly hope the MiG-21bis has an FM that is as good or better than anything ED or Belsimtek has done to date. Anything less can be had in countless other sims from Jane's FA to Jane's USAF to MS FSX to SF to Falcon BMS. DCS's unique market niche IS DCS level aircraft. Falcon BMS and SF2 are essentially free at this point and MS FSX/P3D appears to be as strong as it ever was if you don't need combat. If the only thing added to DCS between the P-51D and the imminent future had been FC3 as a pure import of FC2 into DCS and more aircraft of that quality, I wouldn't have spent a dollar on DCS. The announcement of the MiG-21bis followed by the release of the UH-1H and the promise of DCS F-15C, FA-18C, F-86F, and AH-1G are what got me to buy everything they have released and are planning to release. You would think that It certainly wouldn't hurt if some 3rd parties would flood the the market with FC3 aircraft with the promise of upgrading them later. But all of the current 3rd parties had that option and decided it was financially in their best interest to do otherwise. The Hawk is a hybrid, initially being released with the advanced clickable systems and an SFM, while the F-15C is the inverse with a PFM but with mainly basic non-clickable systems. But the Hawk is supposed to end up with an EFM comparable to ED's PFM. The F-15C might ultimately get upgraded to true DCS product with advanced clickable systems. I am sure Typhoon and F/A-18C fans will truly appreicate full-blown DCS releases over mere FC3 level modeling. Once the F-86/MiG-15 are paired up, I am hoping a 3rd party will provide a good historical match for the MiG-21bis. A "quick & dirty" solution is to develop a stock F-15A from the F-15C. The PFM would be the hard part since the older F100 engine would be somewhat different and far more tempermantal in behavior than the digitally controlled variant in the F-15C. I am not sure how much work would be involved providing the AIM-7F and AIM-9J missiles or the original, simpler APG-63 radar. But I would much prefer a more balanced historical opponent like the F-4E, F-5E, or Mirage III.

The MiG-21bis should finally be coming very soon, latest at the end of June if their latest release date discussion is accurate. It seems like I have been waiting for this forever and the false "end of March" announcement didn't help either. AFAIK, the F-104 is permanently out of the picture. The Hawk is imminent and is supposed to be out before the MiG-21, though the initial release only has SFM to be retrofitted later. I am not interested in aircraft with SFM, but will buy the Hawk to help guarantee it gets the FM upgrade. Razbam has no shortage of 3d models, but no releases. Not sure when or if any of their planes are ever going to come, especially complex aircraft like the F-15E and M2000, though they say the M2000 will be their first release despite all the effort put into the Navy T-2 trainer. Iris? Not really ever going to be in the picture as far as I can see. From what I have read about them, I don't know if I really want their business model/honesty/integrity delivering any products I want/like. Not sure about the Super Hornet. If it ever gets released, I will wait for reviews/debug before I consider buying it. I am far more interested in the ED DCS F/A-18C, closer to my preferred generation of aircraft. Belsimtek is my hero: UH-1H and Mi-8 that truly rock and an F-15C flight model that makes it far more interesting to FLY the F-15C. While I prefer true DCS aircraft with both clickable fully modeled systems and the PFM, if I can only have one, gladly take the PFM. I can't wait for the AH-1G and F-86F. If the F-86F comes out alright and a comparable MiG-15bis becomes available, I might actually be drawn back into online multiplayer :)

<S>

<S> May the next generation serve as well as the past ones.

A very important milestone!!! How long after the last treatment before you will know if it was effective or not?

Sounds like you found a real gem. Just read a bunch of reviews on this. As long as the game doesn't benefit from more than 2 cores, this processor won't slow you down much if at all when paired up with a strong gpu. It also comes with another benefit beyond the lower price: lower power usage/less heat generation.

SF2 still allowed the 1-pixel distance until somewhere around one of the expansion packs when the new ground-object fade-in feature was introduced. Aside from the loss of multiplayer, the view distance limits, some previously open files being hidden/locked, and cloud altitude being limited... SF2 is way too much better for me to go back to SFP1. RWR/ECM, dogfight AI, SF2NA naval features, etc. make the game way better. I was never happy with the preferred 083006 patch level that dominated for so long until WoI was released. WoI with Expansion Pack 1 almost gives you the best of both worlds, but it had several bugs that TK never patched. It was also a lot of work to port over SFP1/WoV/WoE/SF2 content into WoI Exp Pk 1.

His main argument is number of objects being drawn. He should have left it moddable/optional, then those with lower performance could use stock distance and those with higher performance could have the old "1-pixel range".

Pixel size has not been the only limit since the terrain fade-in was added. B-52s do no appear until they are quite large. This problem affects medium and larger aircraft. Only smaller aircraft render at the correct 1-pixel distances since this falls within the new artificial limit.

Congratulations on your return home! If CombatAce and Strike Fighters have been around long enough for an 8th grader to grow up, go off to war, and come home, then I must be getting pretty old.

Supply, demand, and the quality of the firearm. You can get 1911's in all flavors, but if you are going to carry it concealed as I did, you want it to be as reliable and accurate as possible. I bought my base Springfield Armory 1911A1 for about $530 in San Diego way back in 1995 or 96. By the time I was done upgrading it, I had spent a total of $2,000 with the labor of a reputable local gunsmith being a part of that cost. My Colt Sporter H-Bar was only $900, I added a scope and a bipod for less than $200. The rifle uses far more material and is inherently the better weapon both in terms of accuracy and overall effectiveness. But I can carry the pistol anywhere and be fairly sure it won't jam (throated feed, hard recoil spring), won't accidently pinch me (beaver tail grip safety), hits where I want it to (has snag free sturdy sights and a guiderod for consistent repeatable slide motion), and smooth trigger pull riding the line between too light and just right. Unfortunately for me, I had found out about the brand new Kimber 1911's just a little too late: everything I had paid $2,000 for only $600. I could have had two Kimbers and some money to party hardy for what I put into my one Springfield.

Gaining proficiency in beating the 2xSu-27s with only AIM-7Ms and AIM-9Ps. My flying/evasion/weapons/countermeasure employment hasn't gotten much if any better. But I am learning how to read the AI's intentions and exploit his moves. There are a couple of alternatives depending on how the AI reacts to my first AIM-7M shot. He usually turns away and loses his lock on me before his missile gets me, then either he turns back towards me and allows a nearly optimum short range AIM-7 shot or he runs away trying to drag me for his wingman. If I get him quick with an AIM-7, I struggle to find the wingman before the wingman gets me. If he runs, I try to shift targets and kill the wingman with an optimum short range AIM-7 shot. But then I have to struggle to figure out where the leader is before he shoots me. If they get an AA-11 off before I can threaten them with an AIM-7 shot, I almost always die. But when I have managed to maintain situational awareness and/or just gotten a lucky auto acquisition while pulling my nose around, I can kill the wingman quick for no damage. In SF2, I would simply go low, give them my beam, and watch them waste their missiles trying to hit a nearly impossible target. When they got close enough, I would enter a turn fight and patiently wait for a good tail aspect shot. But the DCS AI is really good at handling this: they get in close, use IR and/or helmet sights to make AA-11 shots to kill me quick. So, when it comes to missile combat, I can see the DCS AI using tactics good enough to beat me, particularly when they outnumber me 2:1. If the DCS AI could manage at least a 50% win rate in 1 vs 1 BVR fights with the above loadouts, I would call it pretty good since the Su-27 and its AA-10C/AA-11 missiles are a bit better than the F-15C with the AIM-7M and AIM-9P. It is only when I close to gun range that I see the DCS AI lose almost every time, even when they outnumber 4:1. Of course, SF2 is the other way around. Hard to get the AI to use missiles correctly, but great in a close-in gunfight.

I am primarily a fly-boy, so I don't spend too much time on SB, but it is worth the price... and I haven't even done the best part yet: multiplayer. I like it both for hard-core first-person tanking and as a modern substitute for tactical board games / miniatures rules I used to play in the 1980s/1990s where you are a company or battalion commander trying to stop the Soviet hordes coming through the Fulda gap. Just playing the gunnery practice missions is good fun and the better score you get the better your AI troops will perform (their training level reflects your training level). Instant action with a lone M1A1 trying to stop an advancing Soviet column is great for the feel of playing the old Battlezone arcade game, but with realistic physics and graphics. But playing a mission where you have pre-program a battle plan, then adjust that battle plan in real time to adapt to the current situation is fantastic... and you can still jump in tanks/apcs to take part in the shooting/driving. I will say that it takes time to get proficient with battle planning, managing all the platoons, and making the artillery calls. I am still not all that good at timing my artillery fire. I don't play enough to compete against humans, but I always play on the hardest settings, which apparently makes the AI way better than real troops would normally be. I love the challenge. If the M60A3 ever becomes a fully modeled manned vehicle, or better yet the M60A1 is added and modeled to that level, it will get a lot more of my time. As a commander, I love to try to win a scenario with M60A3s that is incredibly difficult to win using M1A1s. I can almost do it. If I could man the M60A3s and shoot during critical moments like I do with M1A1s, I think I might even finally get a victory. Every now and then, I spend a week or two on Steel Beasts. The high initial purchase price is a one-time fee, and the paid upgrades are few and far between and easily worth the typically $25 cost. But OFP/ArmA is a lot more accessible for the not-so-hard core types and allows everyone to do what they like: snipe, run-n-gun, tank. fly helos, fly planes, etc. But the flight sims have gotten so good and my spare time has gotten so much smaller that I play Steel Beasts a lot more than the OFP/ArmA series. I really love the hard core sims and will gladly split my limited solo PC time between flying DCS P-51D, UH-1H, and FC3 F-15C, MiG-29, Su-27 with some SF2 F-4, F-105, MiG-21, and MiG-23 time thrown in for good measure. As the FC3 F-15C isn't modeled all that well yet and is easy to employ in combat, I mainly fly the old DACT missions I imported from LOMAC/FC2 which I modified to permit MiG-29 and Su-27 DACT missions, too. Whereas, the P-51D and UH-1H get a mix of cold start training and combat missions. I can spend hours just tooling around in the UH-1H to see where I can land without crashing. I see ED realized the problem in calling it the DCS F-15C and have reverted to the FC label. They really should have created a new category since it is still FC level avionics but has a max detail DCS level flight model. I am quite sure the level of effort they have put into the flight model and associated systems modeling easily justifies charging FC customers for the upgrade. But since it is going to remain freebie delivered with FC3, it is fair to give it an FC label.

As currently implemented in SF2, I like them a lot. I don't believe there is enough accurate information to say how good they actually were/are. Beyond the small warhead, which always is a problem, the fact that they are rather short ranged puts them at a big disadvantage when facing enemies with AIM-9L/M or better weapons. But, their maneuverability gives them an excellent engagement envelope if you do manage to get in close enough to fire. Firing at short ranges also gives the target less time see the shot and deploy flares. But small size might not only limit range and warhead, it may mean a simpler seeker that is more susceptible to countermeasures and/or less able to discriminate the target against background heat such as clouds/ground. It would be cool if a government with access to both AIM-9s and AA-8s could afford to do comparative analysis with real-world shots against QF-4s or similar. I would love to know the truth rather than the propaganda always touted by both sides.

I bought Combined Arms, but I haven't even tried it. When I want to drive tanks, I play Steel Beasts Pro PE. If I ever get around to actually playing DCS World online, I like the idea of being able to man SAMs. I would love it if the DCS World SAMs had the option to be up to the same level of detail and realism as the free SAM simulator. As it stands now, I have more sims to play than I have time to play them. I have primarily been cycling between DCS P-51D, UH-1H, and F-15C with some F-4 and F-105 time in SF2, a bit of offline time in Aces High, and a touch of Steel Beasts. I almost haven't touched ArmA2 and OFP GOTY at all.

Great details on what is under the hood of the 1.2.8 beta F-15C AFM is here: http://www.digitalcombatsimulator.com/en/products/eagle/ I truly understand the work that has gone into this. I don't bother with DCS beta testing... that is for people with time to kill and don't mind being frustrated by major glitches. I can't wait until the major bugs are ironed out and I get to try the release version. I hope that a full DCS: F-15C is in the near future. In the absence of the MiG-21bis and no plans for an F-4, the F-15C is going to get some time from me if I like the flight model as much as I like the UH-1H, Mi-8MTV2, Ka-50, P-51D, and A-10C. I am thrilled by the possibility of flying an Su-27 with a reasonable flight model with an unscripted accurate Cobra maneuver! I hope Edge comes out soon, too. I want the performance gains and new maps. Red Flag with F-14s, F-15s, F-16s, and F/A-18s flying against each other sounds fun to me: a beautiful and historically correct map combined with the aircraft that dominated a generation of combat aviation.

I know a lot of aircraft are constrained by unmodeled systems: radar and various avionics not present on any existing DCS level aircraft. The F-15C should at least have educated ED if not 3rd parties on transonic/supersonic aerodynamics, particularly the transonic region where published data is minimal and linear mathematical modeling is inaccurate. Of course, I want the F-4 and many others want the F-14, and that is further complicated by the 2-seat issue. Multiplayer backseat support as full DCS level would be nice to have, but ED support game play when a backseater is unavailable. Of course, the SF2 approach for allowing the pilot to easily manipulate the radar/weapon/navigation systems as if he was giving simple commands to a skilled back seater is an easy way to go that avoids the need to develop complex AI. I don't expect to see a proper two-seater with a RIO/WSO any time soon... if ever. The progress on the UH-1H co-pilot/gunners has been impressive, if somewhat limited compared to a RIO/WSO in a supersonic interceptor with radar. It was interesting to see that A2A first postponed its F-4 to develop an F-104, then sidelined both with the present plan to attempt a T-33 when time permits. I am guessing that modeling jet engines and transonic/supersonic flight at the level A2A likes to achieve is presently far beyond their time/manpower/knowledge level. Whereas their target audience is pretty much just as happy getting Piper Cubs, Cessna 172s, and unarmed WW2 aircraft. I can't blame them for focusing on readily achievable goals that will be highly profitable. The complexity of the F-4 will only appeal to a very small niche. Why must I be that small niche? If it wasn't for TK, I would be stuck flying Jane's Fighters Anthology, Jane's USAF, or flying an F-16 reskinned as an F-4 in Falcon BMS. There is a reason SF2 was the number one sim on my hard drive for a decade and even now gets 25-50% of my PC flying time.

Even with the limitations of the current DCS AI, using a single F-15C with 4xAIM-7M/4xAIM-9P against 2xSu-27 with R-27ET/R-27ER/AA-11 pushes my limits. Even exploiting the AI response to the AIM-7M launch on the leader, sometimes I get shot in the face before I can get even one kill. I usually kill one of them, but die looking for the other one. But I finally got them as good as I get the MiG-23 and MiG-29: only one shot at me with no hits on me, clean kills on them: 3 x AIM-7M at the leader, then came around and found the wingman before he fired on me and used the last AIM-7M to get him to turn and run. Closed while he was climbing vertically to reverse back into me, fired an AIM-9P. Shot him with guns just before the AIM-9P hit... he couldn't be more dead. In SF2, I can notch my opponents radar and make them miss at will with radar missiles. Having a lot harder time dodging missiles once I go on the defensive in DCS. Once they get in close enough for them to unleash AA-11s, I pretty much die after burning up my E and losing my SA dodging AA-10s. In Vietnam/Israel era combat the missiles aren't nearly as likely to hit as long as you don't fly straight and their smoke trails make them much easier to spot/evade. So you have more options to disengage, re-engage, or simply press the attack with very little regard for the missiles. Flying a gunless brick against an agile gunfighter makes for some of the most challenging and fun dogfights, the sooner DCS gets some early to mid Cold War Era planes in the mix, the happier I will be. Really sad that the F-104G project was abandoned and it doesn't sound like Leatherneck is going to do the MiG-23MF anytime soon. So for now, I am eagerly awaiting the final release of the F-15C AFM and look forward to the F-86.

SF2 AI has a long and sordid history. For the longest time, the best AI dogfight logic was in SFP1 SP2a. From the release of WoV until the release of WoI (quite a few years), the AI could hit me with guns if I flew in front of it in a straight line. From WoI on to the present patch revision of SF2NA, it generally continued to improve, or at least remain as good as it was. In some areas, it is still too stupid. In some areas it is has been made too smart. I have never like the fact that the AI will automatically drop whatever it is doing to engage you if you padlock/target him. This was done to make the dogfights more interesting and challenging for the player instead of having most of the action take place between AI opponents. It makes the game more like a movie where you are the star/hero and get to hog all the glory. Additionally, the AI at some point lost visual arc restrictions. You can approach low and from his dead 6 when he is the only plane in the sky and he will react as if he can see you, even if he has no RWR or your radar is off. This was another fix to address the complaint about AI tending to fly straight and level when you attacked them from behind, never reacting to your shots or their exploding wingmen. But for all the flaws/cheats to make the game "better" for the player, the AI can't see through clouds and will use somewhat appropriate tactics and skill to engage you. I can beat an F-15A fairly consistently while flying a MiG-21bis or MiG-23MF, which shows that the AI can't fly the F-15A as well as it should when set to max pilot quality... but to be fair, I have to fly very low, be very patient, and carefully use doppler notch techniques just to get within dogfight range, then fly nearly perfect hi yo-yo and lo yo-yo maneuvers to slowly gain an angle advantage and try to work in a shot before running out of fuel. Saying I can win consistently is not the same as saying I can win easily! In DCS, when I fly the F-15C against MiG-23s, MiG-29s, and Su-27s or fly the MiG-29/Su-27 against F-4Es, F-15Cs, and F-16Cs, the AI flies in a way that seems more defensive. I don't always win, but I easily end up on the tail of my target and his wingman (wingmen when I do 1 vs 4 fights) usually leaves me alone until after I have killed the flight leader. When I die, it is because I got too focused on killing the flight leader and lost track of where the wingman/wingmen went. The AI is so predictable, that I can derive a formula for winning that will work every time as long as I don't make any big mistakes, especially not losing situational awareness while engaging/killing any one target. Basic rules for me to win in DCS with an F-15C: 1) Lock on and kill the leader first. He is the only one really trying to get you. 2) Engage at long range in a way that the target knows you have fired, i.e. use STT and an AIM-7M. The target will immediately turn to break the lock/evade the missile. He will lose lock on you and any radar guided missiles he has launched will go ballistic. 3) If you have the patience/balls to hold AIM-7 fire until it is within optimum launch range and successfully hit on the first shot, you will have plenty of time to go into vertical aquisition mode and roll left to lockup and shoot the lead's wingman. If you have done this quickly and correctly in a 1 vs 2 situation, you will never even get to dogfight. Both bogeys will get an AIM-7M in the face and you should be undamaged. I can replicate this over and over: fire AIM-7, watch bandit break and dive left, and wait for results, then roll left and hunt down survivors one at a time. Whereas if he had some balls, he would maintain lock and watch the missile he launched well ahead of my launch force me to break my lock on him to evade and possibly get killed. As long as he has exactly the same response to my actions, I can push him into my weapons. If I can get the job done quickly, the surviving wingman/wingmen will never get a shot off on me.