Bizarre and just plain unfathomable....

Matrix Games Forums

Forums Register Login Photo Gallery Member List Search Calendars FAQ

My Profile Inbox Address Book My Subscription My Forums Log Out

Bizarre and just plain unfathomable....

View related threads: (in this forum | in all forums)

Logged in as: Guest

Users viewing this topic: none

Printable Version

All Forums >> [Current Games From Matrix.] >> [World War II] >> Uncommon Valor - Campaign for the South Pacific >> Tech Support >> Bizarre and just plain unfathomable....

Page: [1]

Message

<< Older Topic Newer Topic >>

Bizarre and just plain unfathomable.... - 9/10/2002 11:04:39 AM

wmtiz

Posts: 30
Joined: 6/4/2002
From: Belton, Texas
Status: offline

Let me first preface this by stating that I have been playing UV since it was released. I have played through all of the patches and after going through the Coral Sea scenario twice, winning once decisively and once having my head handed to me, I have played scenario #17 exclusively since. I have had many "start overs" as the many patches have come out in order to take full use of the rule changes, OOB adjustments etc.

As you can see from my post count, I do a lot more playing than posting, not that posting a bunch is a bad thing mind you, it's just a matter of making the most out of my limited amount of "free" time.

I have seen some very bizarre happenings in the various versions/patches of UV. Some/most were bugs of one kind or another. Most squashed quickly, some have survived like cockroaches, others have leaped into existance as new patches have been released. Writing program code is one of the things I do for a living and I'm well aware how code changes can have unforseen "side-effects" in other areas of code in a program that was operating perfectly previous to the change.

I have witnessed teleporting and disappearing units from time to time. The most recent I will describe shortly. But the most predominant, ruinous problem I have encountered is the fact that the TF setting "Do Not React to the Enemy" does not mean that at all. Before another round of "use cautious commanders instead of aggressive commanders" starts up again, let me put it this way; You are the Theater Commander in this game. Although you can micromanage some aspects of the game, others you have only general control over. You are the one in charge of seeing the Big Picture. It is your strategic skill (or lack) that is being put to the test here. You set the Rules of Engagement for your TF commanders and it is up to them to carry those directives out.

In that sense, the directive "Do Not React..." would be analguous to telling the TF commander to not seek out hostile action, but leaving them free to respond to attacks (ie counter-punching). The setting "React to Enemy.." would be giving the green light for the TF commander to seek out and engage any sighted enemy they deem worth attacking. I have seen many times where even a cautious commander has reacted to enemy ships, always advancing towards the enemy to shorten the range, perhaps to allow the TBDs to carry torpedoes instead of bombs or whatever. My gripe with this arrangement is does not "feel right" to have to assign a cautious commander to lead the TF from point A to point B and then docking at a base in order to change over to an aggressive commander when you want to go on the offensive.

Here is a case that happened in my most recent Scn#17 that I simply cannot resolve the logic of the TF commander AI coding routines.

It's mid-September 1942, the Allies have been in control of Lunga and Tulagi since the beginning of the scenario (the Japs invaded Buna and Gili Gili instead of going after Lunga at the outset). Their carriers were caught and hammered, but not sunk, by Allied LBA from PM as they were supporting the twin invasions. There was a period of about 40 days where there was no Jap CV sightings in the area. This time was used to reinforce both Lunga and Tulagi. Lunga stands as level 5 Airfield, Level 2 Port, Tulagi at level 1 Airfield, level 6 Port. Lunga has one squadron of F4F-3 and one of F4F-4 available for CAP (survivors of the carriers Enterprise and Yorktown which were damaged in the area a few weeks before and could not sustain flight ops) Tulagi has one squadron of P-39D located there.

Subs patrolling the slot and PBYs on naval search spot multiple Jap TFs coming down the slot from the Shortlands. I have two CV TF available, containing the Hornet and commanded by Spruance, the other containing the Lexington and commanded by Fitch. If you picture a triangle of hexes with Tulagi being the top apex, Lunga the bottom left, the two TF are sitting in the bottom right hex of the triangle. Both Hornet and Lex have cap set to 80%, in addition, the F4F-4 squadron are flying LRCAP over the hex the two CVs are sitting in. Both TF are set to "Do not React..."

During the air phase of the turn, there are three Jap CV TF spotted within air range for the US CVs. The Lex group stands its ground, but Spruance reacts one hex......directly EAST. Not only reducing his cap from the CVs by roughly one-half but also willingly moving out from under the LRCAP from Lunga, completely exposing his TF leaving the CAP from Hornet as his only Fighter coverage. The air attacks that followed had predictable results, the small raid sent towards Lunga was engaged by the F4F-3 and only a small number of bombers got through, damage was minimal, the larger raid sent at the Lex TF was hammered by the Lex's fighters and the LRCAP from Lunga with substantial losses, no damage to the TF. The medium size attack hit the Hornet group, some were splashed, but three bombs and four torpedoes were more than enough to blow the Hornet out of the water. Interestingly enough Spruance did not go down with the ship.

I find this completely unfathomable because I cannot see how any coding routines would have evaluated the options available to Spruance, and would select a one hex move away from the enemy and reduce his effective CAP by two-thirds as an acceptable solution.

If the "Do Not React.." setting is truly not going to remove the TF commanders from reacting, then at least the reaction should make some sort of tactical sense.

As the smoke cleared on this debacle, I checked Lunga and found no trace of the F4F-4 squadron that had just been on LRCAP there. Checking at Noumea, I found the Enterprise was now down below 50% system damage (ie could operate aircraft again) and, yep, the F4F-4 squadron and magically teleported back to their home CV from Lunga to Noumea.

I know this was long and thank you for reading it. I would really appreciate any feedback on this, especially anyone very familiar with the coding routines that handle how the AI evaluates its situation (CAP strength, having or not having LRCAP coverage etc) without giving away any restricted information of course.

_____________________________

Know your enemy and know yourself and you will always be victorious -- Sun Tzu

Post #: 1

- 9/10/2002 12:14:40 PM

pasternakski

Posts: 6565
Joined: 6/29/2002
Status: offline

Please take these comments as analytical, not personal. And respond where you think I'm full of s*** (which I often am).

I don't think the problems you encountered were the fault of the coding or anything else about the game system.

First, having two carrier TFs with one carrier in each is an invitation to disaster. At Coral Sea, this was one of the reasons Lexington was lost and Yorktown badly hurt (the Yorktown and Lexington groups had little practice in operating together. When the Japanese strike force showed up, the Yorktown and her escorts veered one way, the Lexington and her group the other, leaving both carriers vulnerable, particularly to torpedo attack). In early posts after this game was published, I noted a few players advocating this as a viable distribution of strength. I tried it. It was a loser. Your game is an example of why. Maybe when you're "flush" with carriers, multiple air combat TGs of two or so CVs each, one following the other as a unified TF, works. In the early game, and while you have fewer carriers than the IJN (most especially at PBEM), your naval forces are like a big, happy family: they like to go everywhere together.

Second, putting the aggressive Spruance in charge of one with the p**** Fitch in charge of the other is asking for trouble. I don't know if you had one TF following the other, but you should have. Even worse, why were you seeking a carrier battle with three IJN CV TFs with two CVs after having had two CVs knocked out in earlier action? I don't know what your earlier success might have been, but I suspect that you were facing the "Midway" carriers at this time, and they are a serious load in sc17.

I find your preparations for combat and operational execution of your plan seriously flawed on the basis of the information you give. "Do not react" orders are not imperative on an aggressive commander who tries to interpret them in such a way as to give himself the greatest latitude of action in seizing a perceived tactical advantage, and they may have had no effect here, if, as you say, Spruance's group was moving away from the IJN carriers. Perhaps the fault was with Fitch not following Spruance. Perhaps another fault was that the TFs were not placed sufficiently within the range of LR CAP to allow for limited reaction to the approach of significant enemy forces. If Spruance were moving singlehandedly toward the enemy in a foolhardy manner, maybe a court martial should be in his future. This does not seem to be the case, however.

Did you have these TFs set on "do not retire" or "retirement allowed?" This factors into the TF commanders' decisions, as well. The most aggressive setting is the combination "react/no retirement." The most passive setting is "do not react/retirement allowed." I have worked and worked in my play against the AI to understand the mechanics of the various combinations, particularly as they affect TF commanders of various characteristics, and I am getting better, but the picture is still rather muddy. I find that PBEM opponents I have discussed this with are just as confused, if not more so, than I am.

There is no mystery in your carrier-based F4F squadron returning to carrier duty. That's where they are supposed to be. Your local commander just exercised the initiative granted him in sending them back to their assigned duty.

One last thing. Spruance could not be expected, in my estimation, to act in reliance on what a LR CAP might do. Was his TF retiring toward his home base? In any event, a carrier force commander who has to rely on LBA to protect his ships is, first of all, unduly restricted in his choices of action, and, second, unduly dependent on forces over which he has no control.

I am trying, through trial and error and accumulation of little bits of knowledge, to learn this game well enough to play it well. So far, I suck big, red, wet donkey dicks. But I'm getting better, and many things that used to frustrate me I now see as things I just simply didn't understand sufficiently before.

Ain't no bugs here that I can see.

Uncommon Valor. It's not just a job. It's an adventure.

I hope the local looney asylum has a rubber room reserved for me for when WITP comes out. I can tell already I'll NEVER get that one sorted out.

_____________________________

Put my faith in the people
And the people let me down.
So, I turned the other way,
And I carry on anyhow.

(in reply to wmtiz)

Post #: 2

Yes, all that, but - 9/10/2002 1:20:28 PM

Black Cat

Posts: 615
Joined: 7/4/2002
Status: offline

I think his concern is that the " Do Not Retire" routines are not working as well as they might.

(in reply to wmtiz)

Post #: 3

- 9/10/2002 8:26:51 PM

wmtiz

Posts: 30
Joined: 6/4/2002
From: Belton, Texas
Status: offline

pasternakski...

I appreciate the time you took to read my lengthy post and offer some insight on what you thought of my tactics at the time. Let me expand a bit into the tactical picture.

On Lunga, in addition to the two Wildcat squadrons, there were two dive bomber squadrons, one an SBD squadron (also from the mauled Enterprise) and a squadron of A-24s ferried over by the CVE Long Island. Additionally, a squadron of Hudsons and a squadrons of TBDs (survivors from Yorktown). Thus, Lunga was a big fat and Unsinkable carrier. The Hornet group had been sitting in that right apex of the triangle of hexes for about a week under the alternating LRCAP of F4Fs from Lunga and had been very successful in blasting over a dozen Jap APs out of the water as the Japanese initiated their first serious push to take Lunga. Also, the Japanese had been sending bombardment groups trying to knock out Lunga, they consistantly ran into a strong CA/CL DD surface force commanded by Ching Lee positioned in the Lunga hex. The course of events would be; Jap bombardment force enters night surface action with Lee's group, Lee bloodies them and turns them away, next morning any stragglers found by PBYs are blown out of the water by Spruance's CV TF. Other stragglers limping home had to run the gauntlet of 4 SS and some were picked off by those subs. Some poor ships got to face Lee in the night surface battle, then Spruance in the morning, subs during the day, and Spruance again in the second air phase. So, to sum it all up, the tactics were pretty effective to that point.

The Lexington group had just arrived to bolster the number of CVs in the area. Like you I have tried the tactics of separating the CV TF in the same hex and combining them into a large CV TF. I have had better results than you have at keeping them separate, but in the same hex than you have it seems. Thanks for your feedback.

_____________________________

Know your enemy and know yourself and you will always be victorious -- Sun Tzu

(in reply to wmtiz)

Post #: 4

- 9/10/2002 8:45:39 PM

Apollo11

Posts: 24082
Joined: 6/7/2001
From: Zagreb, Croatia
Status: offline

Hi all,

Please visit most excellent Spooky's Uncommon Valor Fansite:

http://uncommon-valor.chez.tiscali.fr/

There is "Strategy guide" and in it there is line about Air TFs
reacting in all cases.

My Air Combat TF moved towards an opposing TF even when I have them set to �Not to React�.

Air combat TF�s that have not aborted their mission and with at least 20 aircraft, may automatically move towards an enemy air combat TF after each air search phase. Allied air combat TF between 5 and 9 hexes while a Japanese air combat TF between either 8 or 9 hexes from the enemy will automatically move. (reported by Mark W. Carver)

Hope this helps...

Leo "Apollo11"

(in reply to wmtiz)

Post #: 5

- 9/10/2002 9:35:41 PM

Oleg Mastruko

Posts: 4921
Joined: 10/21/2000
Status: offline

wmtiz, while I agree with what pasternak said, I'd like to add few additional remarks...

First, and minor remark - you should have ordered one of your CV TFs to follow the other.

Second and more important: LR CAP over CV TFs is unrealistically effective, which I believe is a bug (see lenghty thread titled "Killer bug or killer CAP"). I sent the relevant files as to this bug to Matrix and hopefully they'll fix it. In the meantime I suggest a house rule "no land based LR CAP over CVs" to be used (which I believe is relistic anyway, as LR CAPs over CVs were not used in CV battles in Pacific).

MOST probably Jap strike against LRCAP-protected CV was massacred because of this bug. Otherwise I suspect Fitches carrier would suffer serious damage as well as Spruances did.

Considering all that is said in this thread (LR CAP bug, need to put TFs to follow one another if you want to have them in the same hex permanently, your risky tactic as pointed by pasternakski etc.) I don't think your AI-Spruance did such a horrible think at all...

O.

(in reply to wmtiz)

Post #: 6

Variability - 9/10/2002 11:28:16 PM

NorthStar

Posts: 219
Joined: 5/17/2002
From: New York, US
Status: offline

Another thing to keep in mind is that this game is DESIGNED to have some uncertainty in the implementation. AI Commanders do not always follow the doctrine settings to the letter. This is intended to simulate RL, and I think it is valid. Pick the battle, and I'm sure an example can be found of this type of event.

As far as the necessity of assigning the right Commander for the job, I'm not sure what the problem with that is. As a Theater Commander, you are "Upper Management". A good portion of success lies in making sure the right people are in charge in the right place. You're right in that changing Commanders like socks is a bit "gamey", but it goes right into the Computer Game vs. Real Life problem. In RL, you have to worry about a person's ego, so its impractical. In the game, you don't, so you can change more often. If it bothers you, take a RL tactic -- assign the correct Commander for the final job, and hope he doesn't screw up in the meantime.

(in reply to wmtiz)

Post #: 7

- 9/10/2002 11:52:29 PM

wmtiz

Posts: 30
Joined: 6/4/2002
From: Belton, Texas
Status: offline

Apollo11,Oleg,Northstar...
Thanks for all the constructive comments and analysis of the tactics employed in this example. I should stress that I have no problem with UV with respect to receiving a good thumping and losses when I implement an unsound tactic and take one on the chin, it's all part of the learning process and I certainly make no claim to be an ace player at this time :)

Apollo, in my example, part of my questioning as to this being very odd is that Spruance reacted East, away from the incoming Jap TF, not towards. I allowed for the fact that one or the other of my CV TF would react because I have seen it many times before. If Spruance or Fitch reacted towards the Jap CV, his choices was to move directly West, into Lunga's hex where he would be covered by the Hornet's CAP and the CAP over Lunga. Or, he could have moved NW into Tulagi's hex where he would still have had Hornet's CAP and Tulagi's CAP for protection. If he did nothing, he would have Hornet's CAP, Lexington's CAP and the LRCAP from Lunga. My grief is he moved to one of the only two hexes available that would reduce his available figher umbrella the most. That, to me, did not seem like a logical reaction.

Imagine you are the commander of a CV TF and you know a battle is imminent and enemy CVs have been spotted by patrols. You are sitting under the cover of nearly 90 fighters and you decide to move off 30 miles east thereby leaving yourself still in enemy attack range and by doing so you have automatically reduced your fighter coverage by 2/3 and your short range TBDs can only carry bombs instead of torpedoes in any strikes you launch at the enemy. This just did not make sense to me from a tactical perspective.

Oleg, I read your comments about the uber results of LRCAP on the forums after I posted my initial comments on this issue. You are correct, the attack that went after the Lexington was the largest one launched by the Japanese that turn and it was chewed up extensively. I have no desire to use a bug exploit so I have refrained since from using LRCAP in that fashion.

This does not have any impact on whether or not I will continue to play UV, on the contrary I will continue to play UV and patiently await WiTP next year. All in all this is a superior game and well worth playing.

BTW the F4F-4s returning to their home carrier when it becomes operational again is understandable, except that their home carrier was quite a bit out of their flying range; Lunga to Noumea is quite a journey and unless these Wildcats were loaded with some very economical aviation fuel :) I don't buy that they made that trip without something else going on.

_____________________________

Know your enemy and know yourself and you will always be victorious -- Sun Tzu

(in reply to wmtiz)

Post #: 8

- 9/10/2002 11:57:14 PM

Yamamoto

Posts: 743
Joined: 11/21/2001
From: Miami, Fl. U.S.A.
Status: offline

I would like to see the "do not react" setting actually mean DO NOT REACT. I've been hurt by that 1 hex reaction move too.
If my forces HAVE to do that stupid reaction move then, as the Japanese, I would like to see them react AWAT from the American CVs. My planes have a longer range. Why the hell would I EVER want to move CLOSER to the enemy?

Idealy reaction should be able to be set somewhat like the old PacWar: 0-9. Or maybe 0-X,MAX, where MAX would be the task force's max move - their cruising speed move. That would take into account that some part of the day had already gone by.

Yamamoto

(in reply to wmtiz)

Post #: 9

- 9/11/2002 12:33:18 AM

Joel Billings

Posts: 32265
Joined: 9/20/2000
From: Santa Rosa, CA
Status: offline

It appears that the one hex auto move was triggered in one case and not in the other. I have seen this happen, although I don't know why this would happen with undamaged ready for action carriers. This auto move is not impacted by retirement or reaction orders, but is supposed to happen after all other movement has happened.

As for why your ships moved east, there may be something limiting movement of TF's into bases during this auto move, or the path may be away from shallow water. Again, I don't know. Gary would have to study his code (and then it's only 50/50 he could figure out why it happned this way). We are working on trying to fix the LR-CAP bug now, as LR-CAP was not intended to work well with carrier TF's. Wish I had a better answer for you. I agree that strange things routinely happen in UV (as they do in war), and we've worked to try to eliminate as many of them as we can. Gary created UV to be a fairly automated game system, and in the automation, many things happen that we haven't fully explained.

Joel

(in reply to wmtiz)

Post #: 10

- 9/11/2002 1:26:31 AM

wmtiz

Posts: 30
Joined: 6/4/2002
From: Belton, Texas
Status: offline

Joel,
Thanks for your input and insight on this issue. It is greatly appreicated and is another example of the great feedback and support this game and its players enjoy.

The not being able to react into a base hex is an interesting possibility. Of the four hexes available to move into, two were base hexes and these were the only two places that Spruance could have moved to and head towards the enemy. If he is not able through code to react into a base hex, then the only option would be to back away and set a course around the island and that coupled with a one hex only move may be what the program was attempting to do. I can buy that, but as a tactical move it was a unwise choice of action that cost a pefectively fine CV.

As I play this game I avoid the RPG-ish power gaming method of saving the game every turn to allow for a "Mulligan" (do over to the non-golfers out there) is something goes against my liking in a given turn. As stated above if I make a dumb move and get whacked for it, that's too bad for me. Learn from the mistake and press on. I don't want the temptation of having a way to go back and restart the turn when a battle goes against me etc. I'd rather learn as I go and adjust and recover when a tactic does not work out quite right as long as the results are not questionable like this instance.

Thanks again.

Wm

_____________________________

Know your enemy and know yourself and you will always be victorious -- Sun Tzu

(in reply to wmtiz)

Post #: 11

Page: [1]

All Forums >> [Current Games From Matrix.] >> [World War II] >> Uncommon Valor - Campaign for the South Pacific >> Tech Support >> Bizarre and just plain unfathomable....

Page: [1]