A2A and Nik Mod questions...

[treespider]

Please try to keep your answers to these question short and direct...

I’ve started looking at Nik mod and have some observations/questions...

His intent was to make A2A less bloody…

To achieve this it appears that he has increased Aircraft Durability by 50% and cut A2A weapon accuracy by 50%.

However this change apparently affected AA so he had to increase AA/DP device effects by 50%.


Question #1 - Wouldn’t the change to AA and DP gun device effects also affect the ‘effect’ of DP guns in ground combat and/or naval combat?

Question #2 - Instead of increasing aircraft durability by 50%, what would the effect be if A2A weapons device effect were decreased by 25-33%? (IE the .50 cal would go from effect 3 to effect 2, 2’s to 1, 1’s remain the same, 4’s to 3’s etc).

Question #3 - What would the effect be of cutting A2A weapon ranges by 25-33%?

Question #4 - What would the effect be of combing #2 and #3?

[el cid again]

Question #1 - Wouldn’t the change to AA and DP gun device effects also affect the ‘effect’ of DP guns in ground combat and/or naval combat?

Question #2 - Instead of increasing aircraft durability by 50%, what would the effect be if A2A weapons device effect were decreased by 25-33%? (IE the .50 cal would go from effect 3 to effect 2, 2’s to 1, 1’s remain the same, 4’s to 3’s etc).

1) Yes.

2) I HAVE cut the value of a .50 cal to 2, based on the actual value of the round - which is closer to 2 times that of a .30. However, instead of increasing durability, I radically reduced it, by an average of divide by 6, with NO impact on operational attrition, but some impact on AAA attrition.
I also reduced the range of air-air weapons - most are range = 1 - because the game system is completely fictional in this regard. Reducing the range of weapons and the punch of .50 cals overcompensates for the drop in durability - and my system is much less lethal. At the same time, AAA is more lethal. To address that, I restricted AAA weapons to effective ceilings (rather than maximum ones) and to effective ranges (which equal effective ceilings) - with the unpleasant side effect on dual purpose guns of reducing maximum range in naval battle (a price I find worth paying since the chance of a hit at such ranges is not that great).
To insure it is not too easy to overfly the AAA I reduce aircraft ceilings to a value halfway between service ceilning and optimum operating altitude - which is 80% of service ceiling for normal engines, 90% for supercharged engines and 95% for jets/rockets (on the average). [The latter matters since I allow jets and rockets]. Note, however, that canon values need to be higher (If we are simulating) - not lower - only MGs are cut - .30s from 2 to 1 and .50s from 3 to 2 (in most cases - where they were 2 and 3 to begin with). And durability needs to be calculated on some basis - not just "divide" or "multiply" current values - since relative durability is apparently "seat of the pants" estimated - and you have some clearly incorrect relative values. I elected to leave heavy bomber values the same and correct everything else by calculating their relative durability on that scale (which turns out, on the average, to be greatly overstated for smaller planes). It turns out (testing shows) that LESS durability means LESS damage - a fighter cannot survive 29 rounds of air combat to keep dishing it out (with the present unlimited ammo and high durability it can - higher durability only exaggerates that capability). It also turns out that if even a cannon must close to ranges where the enemy can shoot back there are fewer "free shots" - so every pass takes a chance of the lower durability plane getting hurt. Overall, it means the aggressor fighters don't get to be as aggressive - no free shots - and the return fire matters more.

[treespider]

It turns out (testing shows) that LESS durability means LESS damage - a fighter cannot survive 29 rounds of air combat to keep dishing it out (with the present unlimited ammo and high durability it can - higher durability only exaggerates that capability).

Less Durability means Less Damage?

[el cid again]

Less Durability means Less Damage?

Things are not always intuitive.

In our special case, we have a strange routine that has UNLIMITED ammo for air air combat (ironically at player request - wish Matrix had not listened to us on that).

Also, in our special case, fighter planes with cannons OUTRANGE the normal defensive mg on targeted enemy planes - in particular on light fighters (like Ki-43) and bombers.

What that means is that fighters have UNLIMITED FREE SHOTS with NO ONE SHOOTING BACK. ONLY if range closes to mg range is there any risk at all. In that case, our durability rating for fighters is much too high - about six times too high - relative to say a B-17. So the chances a fighter gets badly hurt are very low EVEN IF you are able to shoot it with your mg.

My system changes all that:

1) First the range of effect is almost always 1: if your fighter can shoot so can its target - unless the target has no guns at all;

2) Second, if the target can shoot, the fighter is much more likely to get hurt - it has on average much less durability

THAT means, on the average, IF the routine ATTEMPTS to let the fighter go round after round on the attack (with its unlimited ammo) it does not SURVIVE long enough to score many kills.

Statistically, it may survive 3 rounds (about right by the way) - so IF it scores every time, and itself then becomes a victim, that is 4 kills. A whole lot less than if it scores 16 times. Statistically, the total number of kills goes way down. But only for large air combats.

More important to me, the relative chance a fighter is killed vs a heavy bomber is more correct. And a Glen is not tougher than a much bigger plane!

[timtom]

Sid, have you saved any testing data? Without it you're precenting us with theory, not fact, as I'm sure you're aware [:)]

[el cid again]

Sid, have you saved any testing data? Without it you're precenting us with theory, not fact, as I'm sure you're aware

A number of people have been testing and sometimes I get reports.
Regardless, I myself am testing continuously. I save every test for a while - but typically I overwrite older data. I see no point in saving data for a long time - and it takes storage to do it. Even a save is not very useful - most of the data about specific encounters is not preserved. On the other hand, I value the statistical totals for a campaign - which tracks planes by type - highly- the more so the longer the campaign.

I do not, however, understand your query/comment. I was not presenting theory at all - just analysis. I attempted to explain why the results are as they are with some theory. But I have not built a data trap to record sortee by sortee events. Whatever your intent may be - it is not important that you like my analysis: you are free to test this plane set any way you wish and come to your own conclusions - and record any data you like on your hard drive!

[timtom]

ORIGINAL: el cid again

I do not, however, understand your query/comment.

Whatever your intent may be - it is not important that you like my analysis.

No intent, Sid.

I'm just curious to see some of the data that leads you to draw the conclusions you do in your analysis. I'm not disparaging the validity or quality of your work. However if it was the other way round, I'm sure you'd expect and demand a quality of evidence better that "because I say so".

Fx:

tm.asp?m=993313&mpage=1&key=testing

tm.asp?m=763293&mpage=1&key=testing

But if you don't have such data to hand, I don't mind running them myself. Do you have any test scn I might use?

Regards,

Tom

[el cid again]

But if you don't have such data to hand, I don't mind running them myself. Do you have any test scn I might use?

I certainly have not maintained an archive of tests, nor done a statistical analysis of them. There is some data tracked in the game itself - and there is a way to get some of it dumped using utilities - both of which I find useful. But in general what I do is come up with a theory, put a box around it, measure whatever, and then change things, retesting until I
1) have total control of this or that factor
2) like the present range of outcomes.
Some things are too complex for this however. Some things won't be clear until we have vast numbers of turns - and until players do what players do - find things we never think of to try.

I am issueing a test set tomarrow - Saturday - April 29 - and you are welcome to test. I will change things tests indicate might benefit from changes prior to tentative formal release on Tuesday of RHSCVO.
I will then move on to do RHSBBO about two days later, RHSPPO about two days after that, and RHSESO about one week after RHSCVO.
The RHS scenario set titles mean:

RHSCVO = Carrier Oriented (you get ships as actually built)
RHSRAO = Identical to CVO with active Russians
RHSBBO = Battleship Oriented (you get ships as planned when the war began or shortly thereafter - including those cancelled soon after Pearl Harbor or Midway) - Russian active is the default case.
RHSRPO = Identical to BBO with passive Russians
RHSPPO = Based on RHSBBO but most units show up in home areas and you get extra political points to assign them to whatever front you wish.
RHSESO = Based on RHSCVO but the Japanese form a planning unit modeled on the one led by notorious Col Tsuji for Gen Yamashita's invasion of Malaya. This planning unit is strategic, not operational. Translation: What if the Japanese built the right ships, armed with the right weapons, to the extent these were available, as soon as they were available, etc.? They don't keep building Ki-27s until 1945. Stuff like that.

Note that in all versions the Japanese have the option to cancel - and should cancel - Shinano. But in CVO options she is building as a carrier while in BBO options she is building as a battleship. Same for Ibuki - but Ibuki should not be cancelled either as a CVL or as a CA - it is useful as either. Still - she was not given priority - and did not see service in the real war - so you may cancel her if you want to - and the resources saved will help escorts show up sooner (etc). There are more than a few other ships treated in a similar way - and a add to the list is two Japanese battlecruisers I thought were cancelled because the war began - in fact they were cancelled because the power of aircraft was demonstrated by Pearl Harbor - which may or may not happen in a game situation. They are not an option in CVO, but are in BBO - and the slips are therefore not available for carrier hulls.

[Sardaukar]

Reducing aircraft weapons range & firepower works just fine. I made own test mod by just cutting firepower to half using CHS1.06 as base and reducing ranges to 1. Worked fine to reduce A2A-losses. DFalcon did some testing about that too. Trouble is to balance flak, A2A and dual-purpose weapons. El cid and RHS seems to be good in that.

[el cid again]

Reducing aircraft weapons range & firepower works just fine. I made own test mod by just cutting firepower to half using CHS1.06 as base and reducing ranges to 1. Worked fine to reduce A2A-losses. DFalcon did some testing about that too. Trouble is to balance flak, A2A and dual-purpose weapons. El cid and RHS seems to be good in that.

In addition to dealing with things like flak - there were altitude issues and some guns that had no ceilings - there were problems of relative values for weapons. Cutting everything in half does not address that. While I did cut .30s in half (from 2 to 1), I cut .50s by 2/3 (from 3 to 2) and went the other way with cannon - a 20mm is a 7 for example (these are the normal cases - not the exceptions). The reason cannon are less deadly is they lack the range advantage of CHS and stock (except for the odd late war 57 mm and rockets). Cannon only work if you are in a situation where they can shoot back (if they have anything that shoots). In general mg raw power was too great, but cannon range was too great. The other problem was relative durabilities - in general fighters and odd planes - see Glen - were too great. This was so bad they patched it in code with what Joe calls a "knee" in the function - but I evaded the knee by staying below where it cuts in and came up with values that work properly relative to each other (at least bomber vs fighter) - and at the same time helps increase AA attrition. The hope it would increase operational attrition did not work as well as required - we needed 2000 times more attrition (!!) - and maybe we got 2 times - if we are lucky. Too soon to know for sure - operational attrition is a function of many things - and we don't know the variables - so it takes huge numbers in many different players hands to know we have real averages. But we sure didn't get anything like a thousand times more operational attrition.