BoB - German bomber kill ratio

[simovitch]

Update: Turn 63 and nearly 1000 of my German bombers shot down so far, and only 1 kill... by a Do17.

[harley]

Ron and I have been discussing this offline for some days. We don't have a solution, but it seems the lower critical hits are the cause of the problem. The pea-squirters in the BOB bombers don't have a high chance of a CH, and BOB airfields are relatively close...

Potential solutions are:

1. Lower fighter Dur across the board
2. Raise pea-squirt effectiveness.
3. modify the CH for 1940/41 - work on the assumption that technology advanced by '43 with better critical system armor and redundancies
4. increase the crash-chance for 1940/41 for the same reason.

Part of the trouble is lack of information on what actually happened. We don't know how many fighters were shot down by defensive fire vs dogfights.

Happy to hear suggestions - right now I'm leaning to 3, and maybe 4 as well, but both will result in more dogfight losses, especially for the germans with the longer return journey.

[Hard Sarge]

I am still looking though my info, I have the data that Harley wants, it is just finding it, and finding it in a way it can be used

[Nicholas Bell]

What about a simple increase in the accuracy & effectiveness value of the bomber guns in the data?  I realize that is more a "design for effect" solution, but there's a lot of that around here.  [:)]

HS - Check your copy of Price's "The Hardest Day" (I'm at work so I can't) as it has a breakout of the number of planes damaged and shot down by bomber return fire on Aug 18.  It's more than I would have expected - but then it probably is on the high side of statistical normal.

Nick

[Hard Sarge]

thanks Nick

I put in a order for a used copy of it (I think I got it, but where to find it

makes me mad, as I can see the data in my mind, so I know I have seen the whole break down of the battle, but those two dead computers have really ruined a lot of my reseach sites I had saved up

plus most of my doc's don't open with this computer (yet)

[Nicholas Bell]

August 18

2 Hurricane shot down and destroyed by Do-17s
1 Spitfire hit by Ju-87 return fire, damaged beyond repair after ditching
2 Hurricanes damaged by Do-17 and force landed (but not written-off)
1 Hurricane damaged by He-111 and force landed (but not written-off)
1 Hurricane damaged by Ju-87 and crash landed on airfield (but not written-off)
1 Hurricane damaged by collision with He-111 but landed okay
1 Spitfire damaged by He-111.

[Hard Sarge]

Thanks Nick

from some of claim hunting I found 4 Stuka kill claims and one from KG51 during Aug

working on some ideas (but keep getting side tracked)

(when I seen those 3 claims on the same day, I remembered the story about how a Cane Squadron got hit HARD when they thought they had a easy target)

[Nicholas Bell]

ORIGINAL: harley
Ron and I have been discussing this offline for some days. We don't have a solution, but it seems the lower critical hits are the cause of the problem. The pea-squirters in the BOB bombers don't have a high chance of a CH, and BOB airfields are relatively close...

Potential solutions are:

1. Lower fighter Dur across the board
This will increase German losses and they seem correct as is. So I think this would be bad.
2. Raise pea-squirt effectiveness.
Or accuracy? I would go this route first, although "logically" how can a single pea shooter be as effective as a .50 cal?
3. modify the CH for 1940/41 - work on the assumption that technology advanced by '43 with better critical system armor and redundancies
I don't think one needs to justify "why" as long as the results are good. The 1.02 patch gives excellent results for BTR, but the conditions in 1940 are just very different!
4. increase the crash-chance for 1940/41 for the same reason.
Again this would only increase the German fighter and bomber losses to higher than realistic. Having played quite a bit of both 1940 and 1941 from both sides lately, the German losses seem "right".
Part of the trouble is lack of information on what actually happened. We don't know how many fighters were shot down by defensive fire vs dogfights.
See results for August 18, 1940

Happy to hear suggestions - right now I'm leaning to 3, and maybe 4 as well, but both will result in more dogfight losses, especially for the germans with the longer return journey.
As noted above, I do not think this is the right way to go.

I must also add that in 1941 the Bf 109 F-2 and F-4 are nearly worthless, as they usually only damage their target. While I acknowledge that the downgrade in armament in these types was looked on even then as a problem, nevertheless the LW scored disproportional losses against the RAF in Europe and Africa using these planes, as well as against the Red Airforce. This supports the possibility that the tweaks to the critical hit routines in 1.02 may be too harsh against lighter weapons (which are less prevalent in BTR, explaining why the issue has not been seen there)

[Hard Sarge]

for the F's, before the patch, they could more then hold there own, they would sit up there with the Spit V for kills vs losses

now, they do nothing

(right now, looks decent if I plot, looks pretty rotten if the AI does the plotting

[simovitch]

2. Raise pea-squirt effectiveness.
Or accuracy? I would go this route first, although "logically" how can a single pea shooter be as effective as a .50 cal?

The 7.92mm MG's had almost twice the ROF as the .50 cal... that should account for something?

[Hard Sarge]

I working on it, trying to get the fighters first, i could up the acc some more for the smaller guns

[Hard Sarge]

my latest run, I did well for a couple of runs, so then let the AI take over

had two decent runs in a row

then wanted to see how the AI did on a week long test

well, like the one guy says, you can't cure stupid

[Hard Sarge]

oddly, they did good damage to there targets, but what was the AI thinking ?

at least FOW is showing, losses were bad enough, but the RAF is overclaiming like a mad dog

[Hard Sarge]

overall, it is starting to look better, still pretty weak on bomber vs fighter, but I did get a kill (with 4 FB making claims)

also seen a Recon bird get a kill (but he may of rammed the plane, as he went down also)

but other then a few more little touch ups I want to do, I think I am about at the end of what I can with the data, other then totally drastic changes (I just can't go with the idea of making a 30 cal twice as strong as a 20 mm, or other stuff along those lines)

did have a good AI run

on the 7 day BoB41 battle (that is a HARD one for the AI to win, heck, I think it may be HARD one for the player to win now)

and other then a few brain farts, the AI did well

[Hard Sarge]

to be fair, that is a 6 day battle, the AI didn't bother flying on the last day

but still a good battle

[Hard Sarge]

just in case, this is a verison 1.02 AI battle before the changes

[Hard Sarge]

well, this is a combined, screen shot, showing how 3 different verisons of the game ran out

[Nicholas Bell]

Pretty bad for the LW!

Have you tried any tests with the tactics set to direct attack of bombers?  Usually this increases the number of interceptors hit.  Just wondering if it might skew the results too far (although I have no idea what you have done).

[otisabuser2]

Was thinking those split screens don't reveal much, except one game may have been randomly more bitterly fought than other.

No indications of what the LW strategy was, ie was it hitting the RAF planes on the ground in one, or flying against heavy flak targets in another.

Showing the LW Top Pilots List will show if any bombers scored hits and in what ratio ?

regards Otis

[Hard Sarge]

Nick

right now, I am going with AI vs AI, how the AI wants to attack, I am leaving it to it to decide

overall, I think I got the data twisted as far as I want to go, I think I got it pretty close to what I want, it is how the AI uses it that is the issue now, as I said before, you can't cure stupid, now, it is more a issue of tactic's, more so then data or stats

last run was close to what I wanted to see, almost 2 to 1 losses (1.5 to 2 to 1, should be good/close to what we are after)

last run, in 7 days, the bombers got 7 kills, I have had 7 runs of 1, a couple with 2, then started getting around 4 for the next couple

one a day, is close to the math I was working with (based on fighter claims and RAF losses (somebody had to get the other losses)

OB, those are just showing the overall numbers, I am watching the other stuff, and taking that into account, early on, the LW was losing over 100 FBs are run, it is now down into the 50's, kicking out some of the stupid stuff, and the numbers would look much better

what I really need is to get Harley to let me into the AI code

over all though, the LW AI is starting to get better, I would say it is winning about 7 out of 10 runs, and the 7 day 41 battle has been a very Ruff one for the AI

I need to see what a RAF player can do, vs the AI (any takers ???)

the RAF AI is getting pretty nasty !

time to hit the sack