Nemo121 -> RE: Allied fighters suck (12/20/2009 12:15:56 AM)
|
The Elf, Absolutely understand your position. In game development you have to prioritise or else you end up with a "Road to Moscow" vaporware project. I do wonder, however, if it mightn't be possible to, at minimal coding cost/time, move the model a bit more towards taking into account these important factors. The current model is a massive improvement over the previous A2A model, and congratulations for that, but I just think that it might be possible to swing it a bit more at minimal cost. What I'm thinking of could be described as follows: A. 3 Toggles. Toggle A toggles Energy vs Turn doctrine, Toggle B toggles Escort / Bomber targetting, Toggle C ( an only Japanese toggle ) turns on "Encourage Aerial Ramming / Discourage Aerial Ramming" Basically if the E vs T toggle is set to E then the plane will always attempt to attack from above with a single firing pass before extending away either horizontally or vertically. If T is selected then the plane will always attempt to engage in turning dogfights ( premium on its manoeuvrability ) Escort/Bomber targetting - would allow you to set planes to try to avoid enemy escorts and target enemy bombers or set them to target the escorts. Bombers, for this purpose, would be level, dive, torpedo or fighter-bombers. Ramming is self-explanatory. It would require some A2A coding but the guts of the current model could be left entirely intact. The only difference really would be adding in a slightly different engagement profile for E fighters. In the pre-engagement phase they would climb to altitude and take a risk of being spotted. When diving they could either bounce the enemy flight ( already in-game ), attack them but without the element of surprise albeit in a favourable position OR be spotted far enough out that the defenders turn into them and there's a head-on pass. The firing pass is unchanged but once the firing pass is made the E fighters would decide to extend either vertically or horizontally. A check would be made of the relative climbs of the E fighter and the defending fighter ( if the E fighter decides to climb out ) or of the E fighter and defending fighter's speeds if the E fighter tries to run away in the horizontal. Throw in a little randomiser for pilot error and poor decision-making and whether or not an E fighter gets attacked in the post-firing pass phase will depend more on the real characteristics it depended on than it currently does --- currrently it depends on relative manoeuvre which is the one thing it really didn't depend on unless the E fighter chose not to extend away. This would give an interplay of some powerful new factors for A2A combat in which planes would realistically seek to make use of their most effective characteristics and would win or lose based on those characteristics, luck and whether or not their pilot got suckered into the wrong type of fight. At present twin-engined fighters etc get reamed cause all fighters which go up into the air to intercept a raid seem drawn like moths to the flame of the enemy escorts. Having that toggle to prioritise what they intercepted by choice ( although obviously they often wouldn't be given that choice in real life ) would help improve the face validity of those types of fights also and introduce some tactical complexity. As to how the doctrine would be applied. A simple database of each plane model broken down by 6 month timeframes would work. E.g. a plane which might be a great E fighter for anti-fighter ops in 1941 might be EEN ( Energy, vs Escorts, No ramming ) and then by 1943 would be relatively slower than the Allied planes and become an TEN ( Turning fight, vs escorts no ramming ) before becoming an EBR ( Energy, vs bombers, ramming encouraged ) plane in 1944. That would allow realistic changes in role over time and make the AI seem reactive ---- which players then over-intuit to think that the AI is responding to them as opposed to a fixed algorithm which gets it to check a database and compare what the database says to the date [:D]. Instead of realising that players tend to go "Wow, when I started fielding more P38s which shredded the Zeros the AI started using Franks against the P38s and sending the Zeroes in against the bombers. This AI rocks." [8D] Obviously this would require some coding but since this approach should be able to leave all of the current code in whilst only adding a couple of new parallel branches to the A2A flowchart it shouldn't be as much work as something which requires much deeper integration into the code. Basically you only change what happens before and after the firing pass to determine whether or not planes get in and out cleanly. As I said the benefits which could accrue from face validity, immersion and apparent reactivity would justify the effort IMO. I've been on your side of this fence quite a few times in the past myself so I'll say this. My post is intended to be constructive and if you say that there just isn't the time to deal with the sort of minimal tinkering solution I outline above then that's fine. You have to draw the line somewhere. I do think though that with so many of the changes I see being made having such a small impact on gameplay I think that the argument for investing the time and effort into something which would significantly improve the veracity AND ( importantly to the yahoos on the forum whom I know have complained so vociferously... often with little justification ) face validity of the A2A model would be time well spent. Bottom line though: 1. Thanks for the open and honest contribution. I for one greatly appreciate it. 2. If you say the resources/time isn't there then that's good enough for me. Maybe not for others but I doubt those others have been on the same side of the line as you or I in the past ;-). P.s. Go, play and enjoy the hols and time with your family ( if you are not deployed obviously ).
|
|
|
|