Aircraft database available to test.

[el cid again]

It appears most aircraft in the WITP & CHS databases have speed data in mph and endurance data calculated from that. The manual (and SOME peope at Matrix) say the data should be in nautical miles. Clearly SHIPS move in knots.

There is a theory that fighters are "too effective." One factor in air combat is relative speed advantage. As little as 10 mph matters, and for planes moving at 250-350 knots, a 14% difference is much more than that! So MAYBE defining the data as the engine intended will matter. Maybe not. Anyone wanting to test can have a database converted for speed and endurance (if speed is in knots, the endurance in minutes changes, increases actually).

Now my database has fixed ALL Japanese aircraft errors data wise. But it has ONLY changed speed and endurance data for Allied aircraft, except for F4F, F4U and B17 models, which also corrected ceiling and ROC entries. This database is only for testing purposes at this time. IF you want to play with it, I will send it to you.

[DFalcon]

I ran a test using the speed in knots information provided by Sid. I left all every other data variable the same. The test I ran is a duplicate of the F4F-4 vs A6M2 test I had run earlier and ran only the new data through the tests. The comparison information is ripped from my first test of the standard data.

The change reduced the range of the Wildcat and the Dauntless to 3/4 and the Zero to 6/9. This put the allied planes at extended range in the test but it did not seem to effect the results too much.

Test Result #1

Day Air attack on Rabaul , at 61,88

Japanese aircraft
A6M2 Zero x 42

Allied aircraft
F4F-4 Wildcat x 33
SBD Dauntless x 58

Japanese aircraft losses
A6M2 Zero: 8 destroyed, 2 damaged
D3A Val: 9 destroyed

Allied aircraft losses
F4F-4 Wildcat: 7 destroyed, 3 damaged
SBD Dauntless: 9 destroyed, 13 damaged

Japanese ground losses:
80 casualties reported

Airbase hits 9
Airbase supply hits 1
Runway hits 19

I ran this type of battle 10 times with the following settings;

Defenders
72 A6M2 80% Cap
72 D3A targets

Attackers
36 F4F-4 10% Escort
72 Dauntless

Average losses per dog fight (Standard / Mod)

A6M2 4.1 / 5.3
F4F 9.4 / 9.0
Dauntless 6.9 / 7.1
Airbase hits 15.6 / 17.6

[DFalcon]

This is an inverse of the first test with the A6M2 attacking the F4F CAP. With the data in knots the Val also has a range of 4/3. The results are turning out to be the same as the standard data.

Test Result #2

Day Air attack on Torokina , at 63,92

Japanese aircraft
A6M2 Zero x 33
D3A Val x 57

Allied aircraft
F4F-4 Wildcat x 42

Japanese aircraft losses
A6M2 Zero: 3 destroyed
D3A Val: 3 destroyed, 7 damaged

Allied aircraft losses
F4F-4 Wildcat: 10 destroyed, 4 damaged
SBD Dauntless: 7 destroyed


Allied ground losses:
7 casualties reported

Airbase hits 6
Airbase supply hits 1
Runway hits 14

I ran this type of battle 10 times with the following settings;

Defenders
72 F4F 80% Cap
72 Dauntless targets

Attackers
36 A6M 10% Escort
72 D3A

Average losses per dog fight (Standard / Mod)

F4F 9.5 / 9.3
A6M2 5.0 / 4.6
D3A 6.4 / 6.6
Airbase hits 21.4 / 18.1

[el cid again]

The results are turning out to be the same as the standard data

Boy is that different from my reading. I see a significant decline in fighter casualties. With less fighters taken out by the escort, it is logical more Vals will get hurt. With fewer Vals hitting the target, it is logical the target gets hit less. These proportions are in the range I expected - at the low end - but in the range. I do not expect anything dramatic from this. And I wanted to be sure the speed change does not hurt the problem. Clearly it doesn't.

[el cid again]

At first I did not like the Forward Test. For one thing, it is invalid:

You asked me to "modify a couple of planes" and I did - but Dauntless was not one of them - so the ROC and other fields were not reviewed - only the speed was changed to knots. It may well be too fast - most of the planes I did review were overrated.

But (a) we can fix that and I will send you a reviewed Dauntless and
(b) IF the Dauntless data is close I have a new view of the test:

1) The Zero is less lethal. It kills fewer F4Fs because it has less of a speed advantage and its improved ROC does not make up for it;

2) The F4F, taking fewer kills, gets more kills, and that may make sense;

3) The fewer surviving Zeros kill fewer Dauntlesses - SO

4) The greater number of surviving bombers score more hits.

And it indicates the change to knots does nothing radical - it just shifts marginally in a direction we want to go in anyway (IF we want less lethal air battles).

But I want to see this with a real Dauntless data set - not just convert some unknown number to knots - but look up the number.

[DFalcon]

If my guess is right only the difference in speed seems to make a difference for air combat.

The main difference the Knot data seems to make is in the range of the planes.

[el cid again]

If my guess is right only the difference in speed seems to make a difference for air combat.

The main difference the Knot data seems to make is in the range of the planes.

My theory was it would affect air combat slightly.

The range effect is also real - or 14% of it is. But some planes had the wrong data period. All Japanese planes were corrected for this - but so were the planes you were to run for your test.

For Dauntless, turns out there is a real error in range - but in the opposite direction. The plane CHS and WITP use must be SBD5 (which is wrong) - but it outranges the Kate (for recon anyway). In realty, not game data.

SBD3 should be used. It has 20% less power. But BOTH are actually faster than the game version. And BOTH carry 650 pounds more bombs than the game version! AND BOTH significantly outclimb the game version. So EITHER is going to do better in a test than the one you ran. Dauntless is a fabuluous plane - better than the game one - in spite of the fact it has bad data!

[el cid again]

I now have a version of the database in which all aircraft have a special modification in the "range" field NOT visible in the editor. This correction allows for use of speeds in knots, true endurance values, but nevertheless range calculated in terms of the hexes on Andrew's map. I can modify this field for an operational reserve, if anyone likes that idea.
Typically non-combat missions have a 5% emergency reserve, and combat missions as much as a 10% emergency reserve.

In addition to all Japanese aircraft, this mod also has validated data for F4F, F4U, B-17C, and SBD (which is now SBD-3 - a separate SBD-5 entry was added but unless you edit or upgrade to it, it won't get used).