RE: Utter BS (Full Version)

All Forums >> [Current Games From Matrix.] >> [World War II] >> War In The Pacific - Struggle Against Japan 1941 - 1945



Message


Sneer -> RE: Utter BS (6/24/2006 12:21:53 AM)

good
my contact with statistics instruments ended few years ago when i got promoted from analyst position...
but remember as it was yesterday from my university that if sb want to run short research about sth it is advisable to use it if number of runs is < 30 and as long as it is close to that value t-student probability distribution should be used to describe population.
normal distribution is not valid unless you want to run >30 or so :-)
certainly there will be sb here with math degree who can take care with results


BTW Student was a nickname from Gosset -




Charles2222 -> RE: Utter BS (6/24/2006 12:25:09 PM)


quote:

ORIGINAL: ecwgcx

Man, I must be doing something wrong. I've sent six BB's and assorted CA's against targets on Honshu that recon says has almost 100 a/c and 70+ LCU's and there were less than 10 a/c destroyed and maybe a few hundred casualties. NO DAMAGE to port or airfield[X(]. The only tac-nuke I've had was the first bombardment of Truk. On all turns after (I'm talking for weeks after) there was "realistic" damage: a few a/c and some inf and guns. Doesn't matter how many BB's I put in the TF.


Any chance you were using the same TF, IOW, the same commander?




Feinder -> RE: Utter BS (6/24/2006 4:09:26 PM)

I'm glad you think it's such a fabulous idea sneer. As you pointed out, objectivity and repetition are crucial to a tests validity. Since I'm going to create the scenario (no small feat if you've ever actually created a scenario), I'm sure you'd like to volunteer to run the first round of twenty tests.

He're what I'll do.

Create a scenario with 4 bases. I might be able to use the Saipan scenario as a starting point but if the terrain types don't match Akyab, I'll just use Burma. No no biggie. But these four bases will have the same terrain type as Akayab (once again, we're trying to limit the amount of variables to be able to focus the test, but I'm sure you knew that!).

I can name the bases appropirately, so that when the combat comes up, it'll refect which base (and test) is being run.

Base #1 = A9 = Allied base with fort Level 9
Base #2 = A0 = Allied base with fort Level 0
Base #3 = J9
Base #4 = J0

Setting them up in this way will (hopfully) allow us to test the following:
1. Do fort levels have any reducing effect on bombardments? (I believe not, but that's for you to prove!).
2. Is there any bias for Japan or Allies in the bombardment routines? (Again, I believe nukes are -not- biased toward either side in execution).
3. Establish if available ops points (thru range) has any effect on the damage inflicted.
4. Attempt to establish a baseline thru repetition, and see if we can obvserve what triggers the massive "nuke" result. We can do this by...

The IJN bombardment TFs will each have
CL Sendai
CL Kuma
CA Suzuya
CA Atago
CA Myoko
BB Kongo
BB Kirishima

(matches the conditions of the bombardment mentioned previously, we're trying to duplicate the result after all).

There will be 10 IJN task forces, with the above ships listed in each. Set to bombard each of the F(0) and F(9) bases, there will be 5 TFs, one at range 1, one at range 2, and so forth, out to range 5.

I'll change the names of each ship so that it looks like :

CL Sendai - F0 - R1
CL Kuma - F0 - R1
CA Suzuya - F0 - R1
CA Atago - F0 - R1
CA Myoko - F0 - R1
BB Kongo - F0 - R1
BB Kirishima - F0 - R1

CL Sendai - F0 - R2
CL Kuma - F0 - R2
CA Suzuya - F0 - R2
CA Atago - F0 - R2
CA Myoko - F0 - R2
BB Kongo - F0 - R2
BB Kirishima - F0 - R2

Reflecting the base to bombardded, and the range that the group came from. What you'd end up seeing in the combat report is something like

---

Naval bombardment of A9


Allied aircraft
no flights


Allied aircraft losses
P-40E Warhawk: 17 destroyed
T.IVa: 4 destroyed
Swordfish: 3 destroyed
Spitfire Vb: 10 destroyed
Catalina I: 2 destroyed
Wirraway: 1 destroyed
Hurricane II: 1 destroyed

Japanese Ships
CL Sendai - F0 - R1
CL Kuma - F0 - R1
CA Suzuya - F0 - R1
CA Atago - F0 - R1
CA Myoko - F0 - R1
BB Kongo - F0 - R1
BB Kirishima - F0 - R1



Allied ground losses:
1800 casualties reported
Guns lost 29
Vehicles lost 7

Airbase hits 15
Airbase supply hits 18
Runway hits 142

---

See how that makes the job of analysis easier? I just knew you'd think it was pretty slick. But it's the communtity that will do the "analysis", not you or I, because objectivity rules!

Only one bombardment at each of the four bases per turn (obvioulsy, or you'd stack the damage).

Basically, I'll set up similar TFs for the Brits to bombard the Japanese bases. I was thinking of putting a typical, and relatively like-gunned, group together:

CL Dragon - F0 - R2
CL Durban - F0 - R2
CA Norfolk - F0 - R2
CA Dorsetshire - F0 - R2
CA Sussex - F0 - R2
BB Resolution - F0 - R2
BB Ramilles - F0 - R2

But you can pick the Allied ships, if you want, no big deal to me.

For each turn, there will be four bombardments, on against each base. You'll just select a TF from range 1, and send it to its destination, and likewise for the other three (4x tests at once). The combat results themselves will display all the relevent information for each test case.

You run the test 20x for range 1. 20x for range 2, and so forth until 20x for range 5. I completely agree with you, 20 tests will give us an excellent sampling!

===

And of course, anyone who doubts -your- obectivity (I have complete faith in you by the way), can run the tests themselves (I'll post the scenario for all to dowload). If somebody thinks your testing is questionable, all you have to do is tell 'em to dowload the test scenario, and run it for themselves.

It'll be "Put up, or shut up" time [;)].

=

Some things about the above bombardment that I can't include in the tests:
The minefield. In game, it's fairly large, but you can't place mines with the editor. So we'll have to do without.

There is a PT boat squadron with 4 PTs that triggered a combat before the bombardment. I can place them if you want, but frnakly I don't think it's absolutely nessary (since they obviously didn't mitigate the bombardment at all anyway). You call if you want them placed or not).

===

More to come, feedback early on our test plan is welcome. Natureally, sneer has to sign off on the test plan before I code the thing, but community input (objectivity and relevence) is requested.

Toddler needs breakfast, but I'll check back in shortly...

-F-




Sneer -> RE: Utter BS (6/24/2006 4:32:26 PM)

how about randomiser - as far as i know it is fixed within save file so result may go same - similar ????




ADavidB -> RE: Utter BS (6/24/2006 4:58:19 PM)

quote:

Setting them up in this way will (hopfully) allow us to test the following:
1. Do fort levels have any reducing effect on bombardments? (I believe not, but that's for you to prove!).
2. Is there any bias for Japan or Allies in the bombardment routines? (Again, I believe nukes are -not- biased toward either side in execution).
3. Establish if available ops points (thru range) has any effect on the damage inflicted.
4. Attempt to establish a baseline thru repetition, and see if we can obvserve what triggers the massive "nuke" result. We can do this by...


The other variable that should be kept as consistant as possible is TF commander. Can you identify opposing leaders whose command characteristics are closely alike?

I have experimented against the AI on the effect of changing commanders on naval battles and I convinced myself that differences in commanders make a difference in battle performance. (I've seen the same thing for air commanders and land commanders.)

And no, I have no intention of doing any sort of statistical analaysis - I'll leave it to you professionals...[;)]

Dave Baranyi




worr -> RE: Utter BS (6/24/2006 5:15:07 PM)


quote:

ORIGINAL: Feinder


And yet, who knew Japan had nukes in July of '42...



I'd say this is the expection.

My experience with IJN BBs on a bombardment mission is you are luck to do any damage. Once in a great while you get a good hit in but the norm seems to just dink the shore.

Worr, out




Nomad -> RE: Utter BS (6/24/2006 5:47:20 PM)

Feinder, when you get the test senario built, sent it and instructions and I will run a number of tests for you. It it time to run the tests and publish the results. without controlled tests arguement is worthless. Let's start with the facts and then analize whether the game is right or not. [:)]




Terminus -> RE: Utter BS (6/24/2006 5:58:01 PM)

"Analize"? Does that involve some sort of probe?




Speedysteve -> RE: Utter BS (6/24/2006 6:26:36 PM)

LOL[:D]




Feinder -> RE: Utter BS (6/24/2006 6:27:50 PM)

Randomizer - Actually, what you'll be doing is loading it as a new scenario as head-to-head (that way you can view damage to both sides).  Then issue the orders for each of the 4x TFs to bombard.  Check the damage.  Then fire it up again.  You can't run it successively, because the damage to the bases will need to be reset.  But I've done fairly extensive testing (on other things), and it'll create the new random seed with each restart.  Won't be a problem.

TF commanders - good point.  I'll set them all to the same stats.  I'll ask Kurt will send me the name of the TF commander in his group, and I'll make clones of him for each TF.

-F-




Nomad -> RE: Utter BS (6/24/2006 7:55:03 PM)


quote:

ORIGINAL: Terminus

"Analize"? Does that involve some sort of probe?


Yes, please bend over. [X(][8|][:D] [sm=sterb011.gif]




dtravel -> RE: Utter BS (6/24/2006 8:02:46 PM)


quote:

ORIGINAL: Nomad


quote:

ORIGINAL: Terminus

"Analize"? Does that involve some sort of probe?


Yes, please bend over. [X(][8|][:D] [sm=sterb011.gif]


Isn't that supposed to be Nik's line? He is the one onboard the UFO after all.




Nomad -> RE: Utter BS (6/24/2006 8:06:50 PM)

Not if he criticises my spelling! [sm=00000036.gif][sm=00000018.gif]




Nikademus -> RE: Utter BS (6/24/2006 8:22:38 PM)


quote:

ORIGINAL: dtravel

Isn't that supposed to be Nik's line? He is the one onboard the UFO after all.


Come here....... <whiirrrrrrrrr>

http://www.matrixgames.com/forums/tm.asp?m=831932&mpage=1&key=Tactical%2Cnuke




Dino -> RE: Utter BS (6/25/2006 9:13:39 PM)

quote:

ORIGINAL: Feinder

Setting them up in this way will (hopfully) allow us to test the following:
1. Do fort levels have any reducing effect on bombardments? (I believe not, but that's for you to prove!).
2. Is there any bias for Japan or Allies in the bombardment routines? (Again, I believe nukes are -not- biased toward either side in execution).
3. Establish if available ops points (thru range) has any effect on the damage inflicted.
4. Attempt to establish a baseline thru repetition, and see if we can obvserve what triggers the massive "nuke" result. We can do this by...


I think you need to factor in the detection level of the base, but I'm not sure if this can be recreated in a test environment...




Nikademus -> RE: Utter BS (6/25/2006 9:16:00 PM)

lol....did anyone chk the link posted?




Dino -> RE: Utter BS (6/25/2006 9:46:24 PM)

quote:

ORIGINAL: Nikademus

lol....did anyone chk the link posted?



[:D]

I didn't even notice the link... I thought it was one of those hickup posts.[:o]




dtravel -> RE: Utter BS (6/25/2006 9:49:18 PM)

Don't go near the drunk fat bunny waving the chainsaw!  [X(]  It can't end well.


(The link is to a previous discussion of bombardments and the results of test that Nik ran more than a year ago.)




jwilkerson -> RE: Utter BS (6/25/2006 9:52:06 PM)


quote:

ORIGINAL: Nikademus

lol....did anyone chk the link posted?




I think this is my biggest nuke to date ... (this one's been posted several times by my opponent who was more "proud" of it than I was - he though he was on the receiving end ) ...

--------------------------------------------------------------------------------
Naval bombardment of Soerabaja, at 23,66


Allied aircraft
no flights


Allied aircraft losses
P-40E Warhawk: 3 destroyed
A-24 Dauntless: 5 destroyed
Kittyhawk I: 3 destroyed
Wirraway: 2 destroyed
P-38F Lightning: 2 destroyed
Beaufort V-IX: 3 destroyed
Hurricane IIb: 5 destroyed
Hawk 75A: 1 destroyed
OS2U-3 Kingfisher: 3 destroyed
CW-21B Demon: 3 destroyed
PBY Catalina: 1 destroyed

Japanese Ships
DD Oboro
DD Amagiri
DD Yugiri
DD Ayanami
DD Shikinami
DD Uranami
DD Hatsuyuki
DD Shirayuki
DD Isonami
DD Shirakumo
DD Usugumo
DD Shinonome
DD Fubuki
DD Natsushio
CL Jintsu
CA Suzuya
CA Myoko
CA Takao
BB Yamato
BB Kirishima
BB Hyuga
BB Ise
BB Hiei
BB Haruna
BB Kongo

Allied Ships
AR Castor, Shell hits 1, on fire
CL Mauritius, Shell hits 1


Allied ground losses:
14450 casualties reported
Guns lost 135
Vehicles lost 72

Airbase hits 32
Airbase supply hits 14
Runway hits 322
Port hits 3





Nikademus -> RE: Utter BS (6/25/2006 9:55:37 PM)

quote:

ORIGINAL: Dino

quote:

ORIGINAL: Nikademus

lol....did anyone chk the link posted?



[:D]

I didn't even notice the link... I thought it was one of those hickup posts.[:o]



I need one of those hand slaps forhead smilies..... [:D]


Three relevent points were:

1) Yes Virginia......Fortifications do reduce casualties/

2) Yes Virginia......Recon (i.e. Detection Level) is critical. It doubles bombardment effectiveness.

3) Yes Virginia.....there is an Easter Bunny....and he's kicking ass!





Terminus -> RE: Utter BS (6/25/2006 11:16:17 PM)

4) Yes, Virginia... There is an Easter Bunny... and he's quite mad, you know...[8|]




dtravel -> RE: Utter BS (6/26/2006 2:15:10 AM)


quote:

ORIGINAL: Nikademus

I need one of those hand slaps forhead smilies..... [:D]



[image]http://bestsmileys.com/doh/1.gif[/image]

(Sorry, don't have one with bunny ears.)




Page: <<   < prev  1 [2]

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
1.860352