el cid again
Posts: 16922
Joined: 10/10/2005 Status: offline
|
quote:
ORIGINAL: turkey1 el cid 1) just looked at 5.03 and I note that the Japanese engine requirement for the Ishikawajima plane engine is 291,134 per month (all scenarios) REPLY: You have a problem. This is not the case in any game with source files in any scenario. Ishikawajima does end up needing over 1400 engines a month - prohibitively many - by 1943 - but it slowly climbs to that level. I cant see where these engines are being requested ie the japanese are producing anywhere close to that number of planes. 2) I also cannot figure out why you have reduced the durability of all planes so much. Wont this end up with a much higher casulaty rate in flak and aircombat REPLY: Two unrelated reasons - one of them now moot: there used to be a "knee" in the air combat functions above 40 - and we wanted to below that - so we had "honest" data processing - this I am told has gone away; the main reason - which remains - was that it would indeed result in higher casualty rates in ALL situations - including the one we strongly wanted more of: operational attrition. We also believed FLAK was far too ineffective - and not entirely because some guns had no ceilings, some had no detection gear, and so on - even properly set guns were not doing the job they should do. We then set the durability to the lowest possible level that distinguished between planes, measured the results with a statistical number of runs, and concluded we should apply a constant (K) factor of 2 to these values. This was done - and the new values have produced no complaints - but many very nice outcomes. I intend to run a more lengthy analysis after we get the new maneuverability values in place - to see if K is indeed right - but failing a measurable problem - it tentatively appears they are pretty good. 3) I also note the manoeuverability has been reduced across the board compared to vanilla witp . Interesting that late war aircriaft are getting higher numbers than early war aircraft. This seems incorrect as generally early war planes such as the Zero were more manouverable than later war craft which began to rely on higher speed eg the ME262a could not turn quickly but it was fast. Your stats have the P47 "JUG" turning quicker than early war zeros. !!!! The only thing a P47 could do was dive quickly and was otherwise noted as handling like a truck. I also note you have P40B's with the same manoeuvre ratings as the A6M2. Which seems to be incorrect. Sorry to seem to be having a go at your work but the air model you have adopted is a totally different approach to all the other mods and I was wondering why you have chosen these numbers. REPLY: Well - the approach was first of all systematic. The existing approaches were - whatever else can be said - inconsistently applied. We tried to understand the system, devise a similar one, and apply it consistently. At the same time, we WANTED it to be different. We wanted a lot less lethality in air combat - and we got it too. Above all we wanted what you seem to want - RELATIVE values to be right. Now that is hard to do. It is particularly hard to do with a too simple model like WITP uses. And ALL simulations must involve compromises - so we will never achieve perfection or escape a problem case that does not fit even if we made ideal compromises. Nevertheless - the maneuverability values were not entirely happy ones. Nor can they be for structural reasons: planes do NOT have the same vertical and horizontal maneuverability; planes do NOT have the same of either at all altitudes either - and we have only ONE value - so a plane good at this at one altitude may be good at that at a different altitude - and the two values must be averaged in a single field! Even so, we set out on a quest for a happier system, three times, failed twice - and finally succeeded. The only problem is - applying it to 249 types of planes takes a lot of research and calculation - and it has not yet quite completed. The ETA was today - and I hear it is almost there. Basically you should ignore the present values and look at the new ones. For samples go to the thread on the subject. To have pleased even the critics in this matter took a lot of work - and some technical help. Within the limits stated above - only one value cannot do this right - we may have done very well indeed. The release level for this new data is 5.10 Cobra - just downloaded the updated version of your map - Looks fantastic.
|