Ok...call me a cheater...but the tactic of shooting and running works pretty well...

I know the theory of bases automatically defending themselves...however...after several years of playing harpoon..as a layman who loves the hypothetical conflict scenario but doesnt overanalyse it..I have always worked on the basis that if you leave it to the automatic defence reaction to defend your base you are buggered!

I quite like it, scramble the crap out of aircraft to intercept attacking missiles...or overload (eg a hell of a lot of SEAD in the attacking force against enemy bases...does this make me a Harpoon heathen? ( I am kind of serious about that question)

Ok...call me a cheater...but the tactic of shooting and running works pretty well ... I know the theory of bases automatically defending themselves...however...after several years of playing harpoon..as a layman who loves the hypothetical conflict scenario but doesnt overanalyse it..I have always worked on the basis that if you leave it to the automatic defence reaction to defend your base you are buggered! I quite like it, scramble the crap out of aircraft to intercept attacking missiles...or overload (eg a hell of a lot of SEAD in the attacking force against enemy bases...does this make me a Harpoon heathen? ( I am kind of serious about that question)

 

Play in whatever fashion makes you happy.

A possible issue specific to my machine is memory shortage. I typically have a number of apps running concurrently with HCE, and there tends to be a lot of swapping at times when HCE is in the background. Supposedly, this shouldn't screw anything up, but...???

 

I'm no computer expert but I would tend to think that if some sort of memory leak was affecting the die rolls, it would be happening randomly, i.e. just as often against the AI as it would in favor of the AI. Is your PC a Cyberdyne Systems model by any chance? <g> Seriously, though, wouldn't you expect it to screw with the numbers generally and not in favor of one side or another?

I quite like it, scramble the crap out of aircraft to intercept attacking missiles...or overload (eg a hell of a lot of SEAD in the attacking force against enemy bases...does this make me a Harpoon heathen? ( I am kind of serious about that question) <_>

 

You aren't the only proposed heathen. I don't like fair fights and I play the game to have fun. For me that pretty much means big scenarios I'll definitely win involving F-14 and Tu-22. Realism isn't key to me as long as I'm having fun. If you are a Harpoon heathen then so am I.

 

That said I do care about realism when making changes to the game and strive to increase realism as we make changes to the game mechanics.

SHUK, do you get demolished every time you play HCE?

 

Short answer is no...

A possible issue specific to my machine is memory shortage. I typically have a number of apps running concurrently with HCE, and there tends to be a lot of swapping at times when HCE is in the background. Supposedly, this shouldn't screw anything up, but...???

 

I'm no computer expert but I would tend to think that if some sort of memory leak was affecting the die rolls, it would be happening randomly, i.e. just as often against the AI as it would in favor of the AI. Is your PC a Cyberdyne Systems model by any chance? <g> Seriously, though, wouldn't you expect it to screw with the numbers generally and not in favor of one side or another?

 

Nope, no Cyberdynes... Just assorted Toshibas and Acers/eMachines... (but all instances of this PH issue has pretty much been seen on my ancient Toshiba Satellite).

 

Frankly, I don't really know what affects memory leaks could have within this app... but it seems plausible that they could bung up data that's in memory, and only in certain spots that would have limited scope of effect. Also, without knowing what the data structures look like, and are they static or dynamic, etc. etc., all I can say is that it's in the realm of possibility, albeit perhaps not particularly probable. Anyway, I'd lean more towards the issue stemming from a selective corruption of some data that is a common factor in the PH calculation, yet still specific to one side's calculation... or else a corruption that gets replicated into each new group when new groups on one side or the other are subsequently created.

 

Because I've also seen occasional indications that things get screwed up under certain rare conditions that involve the destruction of a unit or group - for example, when part of an air group gets shot down while the group is in landing mode, or when the readying status of landed aircraft gets changed while a group of the same a/c type is in the process of landing, or when the last ship of a group gets sunk while there are still attached air units, or when sometimes an enemy base can no longer be attacked after one of its attached AD units gets destroyed, etc. - this leads me to suspect there may be some bug somewhere in the "unit-destruct" code segments that doesn't properly "clean-up" all of the related data structures, and thus leaves latent problems laying around. This tends to fit with the observation that the trouble usually showns up only under conditions where there are either a whole lot of groups in the game, and/or there has been a lot of preceeding "action" - where lots of units and groups have bitten the dust. Either condition might "stress" the memory situation enough that stray memory accesses would actually start hitting something of importance. Obviously, though, it couldn't be "random" memory hits or it would likely show up in assorted ways, not just in terms of PH. Yet, it certainly appears that the more memory that has been used in these ways, then the more likely it is for this PH issue to crop up.

 

Now, there is an aspect to this whole situation that could be relatively unique to the way I typically play, and which might tend to tax any memory issues even more than perhaps is the case when others play. That is, I typically use small groups - especially with air groups, where I generally send out one- or two-unit groups, but at the same time, typically have almost all of my air assets in flight at once, at least at some times during the game. When adding the AI air groups that may launch to counter all of my groups, I'd estimate it would not to be unusual to exceed 100 individual air groups going all at once, plus whatever other types of groups may exist in the scenario... which themselves may be numerous in the WestPac demo scenarios that I've been playing, due to all the land and Mobile AD groups that exist in those scenarios. And I typically play with the "wipe out" mentality (where my "mission" is to eliminate the enemy, to the fullest extent possible that does not violate the orders... ) This tends to pile numerous missile groups on the heap, too. So, with all that stuff involved, and with so much unit/group destruction going on, if there is any "residual" stuff that doesn't get fully cleaned up when a group is destroyed (or it is simply removed due to returning to base, etc.), then this playing mode may lend itself to developing the issue - more so than a "less aggressive" style of play might.

 

The bottom line is that there very likely is a correlation between large numbers of groups and heavy action and the probability that the issue will show up in any particular instance. This, in turn, suggests that there may be some "houskeeping" issue involving the construction and/or destruction of the data structures for the groups or units, that in some way eventually "feeds" bogus data into the subsequent PH calculations. But, as you say, it's all just anecdotal "evidence", so who knows?

I also sent game saves that consistently showed the same results when I ran them... but you said they worked fine when you tried them.

 

Those savegames are still available for download here if folks want to take em for a spin.

Joe K, this is what stands out from what you've written:

 

1. You claim your "AI always hits is reproduceable" from a saved game. Even better reproduceable after shutting down the game and restarting.

2. You have not provided said saved games for the current 2009.050 demo, nor any other.

3. Now is a really good time since I actually have 2009.050 available for debugging unlike the usual situation where I've moved beyond the current demo (I have in this case but maintain 2009.050 as a 'stable' version).

4. None of your issue reports here at HG deal with the AI always hitting so please make an issue report using one or more of those indicative saves just before missiles hit (or even just before they are launched).

5. If it isn't machine specific and I expect it isn't then I'll finally be able to reproduce the issue and do something productive about it.

Joe K, this is what stands out from what you've written:

 

1. You claim your "AI always hits is reproduceable" from a saved game. Even better reproduceable after shutting down the game and restarting.

2. You have not provided said saved games for the current 2009.050 demo, nor any other.

3. Now is a really good time since I actually have 2009.050 available for debugging unlike the usual situation where I've moved beyond the current demo (I have in this case but maintain 2009.050 as a 'stable' version).

4. None of your issue reports here at HG deal with the AI always hitting so please make an issue report using one or more of those indicative saves just before missiles hit (or even just before they are launched).

5. If it isn't machine specific and I expect it isn't then I'll finally be able to reproduce the issue and do something productive about it.

 

Ok, let me try to clarify the individual points here:

 

1. Are you refering here to my previously-submitted game saves, or to situations in general? If to my submitted game-saves, then:

a. As I recall, the saved games that I provided were addressing the issue where the human side's groups were unable to kill any of the AI group despite repeated volleys, while the AI group was able to acheive a near 100% hit/kill. This is probably due to the same issue that we're discussing, but is a slightly more extreme manifestation. (In any case, please refer to the info submitted along with those saved games for specifics).

b. Reproducability: When I ran those saved games at the time, they produced the effect each time. Offhand, I don't recall how many times I tried them, but I'd guess it was at least three-five times each. As far as the shutdown/restart not clearing up the problem, that was unrelated to the submitted game-saves. Instead, that was merely observations that I've made along the line - usually when things got so bad that there was no point in continuing the game, and I was trying to "salvage" the game via desperate measures.

If you were instead referring to the general situation, then:

a. & b. On the occasions when I have tried re-loading to clear up problems, yes, in each such case, the re-load did not help the situation. I don't know that that constitutes "reproducability", though... because it could have been just my bad luck (as far as not trying it in more instances, I mean... after a few failures, I just didn't bother trying it anymore).

 

2. Are you referring here to the game-saves that I submitted back when? If so, those saved games were from a demo version - 042, IIRC. I have not submitted any from newer demos because they weren't requested, to the best of my knowledge.

 

3. If I can find my original game-saves, I'll re-test them within demo 050 (assuming that's what you meant?) If you want a new game that executed entirely under 050, then I'll have to see whether I can get one that will be comparably consistent. (Note: I said "if", because the hard disk suffered an impact crash sometime after all that game-saving occurred, and I don't know whether those original save-files are backed up anywhere).

 

4. Using the old save-files, or new ones? Again, my recollection of the issue involved in the game-saves was more that it related to the inability of any of the player's missiles to achieve hits, although, as I said, I think the AI's near-100% hit rate existed in that situation as well. Are you wanting to tackle both of those aspects, or just one or the other for now?

 

5. That would be great!

Joe K,

 

1. In this thread you made the statement of reproduceable across those methods. I looked at the issue reports you have submitted and none of them are for the AI missiles hitting 100%. They are also not for the Player missiles missing 100%. So I would read that to mean a new save and instructions would be handy. We haven't deleted any threads or issue reports so if you can find them already in place, great, I could not (I scoured all of the issue reports created by you, I did not scour issue reports from others or forum posts).

 

2. If you have some from 2009.042 that show the same behavior when loaded in 2009.050 then that'll do, otherwise new ones would be best.

 

3 & 4. They don't need to be new if the results are the same in 2009.050. New would be better but old is fine for now.

 

Thanks

Ok...since this is one of the most fired up forum topics i have seen in awhile...im going to throw my two buoys into the water..

 

First of all i rarely post here, i do most of my talking on the IRC chat, which i hope all of you try to get into, some really good discussions go on there, harp related and otherwise.

 

Secondly, a little background on me. I was in the US navy for 10 years, all of that time was as an sonar technician/submarines, and i spent the most time as an ST-S on the USS Topeka. I have been a student of naval warfare and tactics pretty much all of my life...having read everything i could get my hands on and even some of the rare stuff i shouldnt had gotten my hands on. ive been playing harp a LONG time, and seeing the best of it in Tony's hands with HCE.

 

Now to the main question of if the AI is too good or has an unfair advantage. I would have to say no, BUT it does things so damn well, our human mistakes make it look otherwise. Ive been playing HCE more than usual lately due to this thread and i cannot come up with anyone single reason why the AI would have a behind the scenes advantage. Ive shot missiles at AI targets, hes shot back and tried to shoot down said missiles...some were knocked down, some got through and killed the target. My Air to Air engagements pretty much go the same way. And in the real theater, the underwater fight, i can get to a target group and kill with torps and or missiles and get away. i use various ways to achieve wins in battles. One of the things the AI absolutely hates is when you attack from different directions at the same time, as do targets in real life. hit em where they aint looking. coming from just one direction guarantees that all the weapons will be pointed at you. One of the posts in this thread described an attack on Keflavik with f-4G's with HARMS. Id like to know if the base was radiating at the time. hard for HARMS to do much damage if the base was being sneaky with radar. plus your description made it sound as if the attack was from one direction, therefore everything was pointed at that threat axis.

 

And lastly you have to look at the scenario itself, some are unbalanced, some are pretty even. One of my favorite scenarios is from the MEDC battleset 6.0 Libyan Skirmish. (a bit mistitled, because Libya throws dozens of planes at you, and a couple of kitchen sinks) It makes for great practice on how to handle AAW for a CVBG. and about everytime i play it, i rarely lose one plane, Phoenix missiles are great to have but sooner or later it gets down to knife fights and even then the AI doesnt get much chance. And after youre done with the air battle, planning the strikes on the airbases takes a bit of work because of the SAM threat, i tend to use multiple packages and again, different directions to achieve my desired results.

 

Just before i started typing this out, i was playing a simple GIUK scen, and lost a RN Lynx on a ASUW mission with sea skua. I had it looking for two nanchuka's that i needed to find, and the nanchuka's downed her after she fired the skuas, and the skuas died because they didnt have radar illumination anymore. i went in dumb and paid for it. I did it mainly to test the theory of the AI.

 

The AI is spot on as far as im concerned, and while there are a few issues, the issues arent really the issue here.

Since I no longer have permission to post replies to the "I suck..." thread, I'm posting my reply here instead.

 

Joe K,

 

1. In this thread you made the statement of reproduceable across those methods. I looked at the issue reports you have submitted and none of them are for the AI missiles hitting 100%. They are also not for the Player missiles missing 100%. So I would read that to mean a new save and instructions would be handy. We haven't deleted any threads or issue reports so if you can find them already in place, great, I could not (I scoured all of the issue reports created by you, I did not scour issue reports from others or forum posts).

 

2. If you have some from 2009.042 that show the same behavior when loaded in 2009.050 then that'll do, otherwise new ones would be best.

 

3 & 4. They don't need to be new if the results are the same in 2009.050. New would be better but old is fine for now.

 

Thanks

 

1. OK, maybe that's what was confusing me... because I haven't ever submitted any bug report about AI missiles hitting 100%. But, they did include description and two game-saves relating to "Invicible Air Groups" - which is essentially a case of player's AAM's missing 100% of the time on one particular AI air group. Those were submitted at the spot referenced in CV32's post earlier in the "I suck..." thread. I don't have - and AFAIK, never have had - any saves demonstrating the other side of the coin - that is, where AI missiles hit 100%. But I see that latter situation so consistently that I assume no save is needed to "expose" that situation... but if nobody else is seeing it just as a matter of course like I do, then I can only assume it really does exist only in my particular installation. (If you wish to pursue this, and cannot see it in your own runs, then I'll have to capture some saves - although I think that effect can be seen in my previously-submitted saves, even though it was not documented).

 

2, 3, & 4. I recently re-played Middleweights (which was the scenario where these issues mostly showed up in 042 (or whatever it was then)), under 050, but I did not encounter the issues much at all that time. I also just located my original game-saves, and loaded those in 050 for some test runs. The results of that were a bit puzzling:

 

- The first three times that I ran the first of the two game-saves, I did not see any indication of the issue - that is, everything seemed very well-behaved.

- Figuring that game-save was a bust under 050, I tried the second game-save, and that one also ran normally (i.e. - no evidence of the anomaly).

- Getting a bit flustered over that, I re-opened the second game-save again (but without re-starting the app in this case). This time, I started seeing some evidence of the documented mis-behaviors - although not all of them. So, after eventually destroying the "invincible" group after several missile volleys, I re-opened that game-save and ran it again. That time, the behavior showed up pretty much exactly as documented. So, I repeated the process several more times. The overall results were less consistent than when I ran them at the time of submission, and gave these results for the runs:

1- invincible AI group: destroyed by first volley; player's group: unscathed

2- invincible AI group: destroyed by third volley; player's group: unscathed

3- invincible AI group: survived all three volleys of AIMs and two volleys of IRMs, and three volleys of assorted AAMs from a second player's group; player's group: both units destroyed by first volley of IR missiles

4- invincible AI group: survived all three volleys of AIMs and first volley of IRMs; destroyed by second volley of IRMs; player's group: unscathed

5 & 6- invincible AI group: survived all three volleys of AIMs and both volleys of IRMs; player's group: both units destroyed by first volley of IRMs

7- invincible AI group: survived all three volleys of AIMs and both volleys of IRMs; player's group: both units saved by manual intervention (i.e. - breaking off the attack)

8- invincible AI group: survived all three volleys of AIMs and first volley of IRMs, destroyed by second volley of IRMs; player's group: both units saved by breaking off and then re-initiating the attack

9- invincible AI group: survived all three volleys of AIMs and two volleys of IRMs, destroyed by two IRMs of third volley; player's group: both units destroyed by second volley of IRMs (allowed to continue the attack after firing their last volley of IRMs)

 

Note: AIMs = long-range missiles; IRMs = short-range, infrared missiles

 

After that, I tried the first save-file again a couple of times (without re-starting the app), and found that it generally followed the documented bahavior, but occasionally some of the player's missile did get kills.

 

So, overall, the game-saves were not nearly as consistent in result as I had seen at the time of their submission - especially not when I re-started the app each time - but overall, they demonstrated behavior that was closer to the documented anomalous behavior than to the expected normal behavior. This is unfortunate, because at the time I submitted them, I was seeing 100% conformance to the documented behaviors with each run (although, I believe I ran each only about five times because I was getting such repeatable results that I thought further testing was unnecessary). As it stands now, there isn't 100% repeatability in 050 - yet they still show the behavior more often than not... but I don't know how helpful that will be.

 

It would be interesting to see how the behavior of those game-saves now compares on your installations. Beyond that, what makes sense next? I was thinking I might run a large scenario, hopefully to the point that some issues start showing up, and then try opening the save-games and see what happens from that point... but that would pretty much throw any hope of repeatability out the window.

Since I no longer have permission to post replies to the "I suck..." thread ...

 

How's that? Are you getting some kind of error message?

 

Edit: I've merged the thread. JoeK, if you continue to have difficulties posting to this thread, please let us know before starting another one on the same subject matter.

Ok...since this is one of the most fired up forum topics i have seen in awhile...im going to throw my two buoys into the water..

 

...

 

Now to the main question of if the AI is too good or has an unfair advantage. ... plus your description made it sound as if the attack was from one direction, therefore everything was pointed at that threat axis.

 

The AI is spot on as far as im concerned, and while there are a few issues, the issues arent really the issue here.

 

IMO, the whole issue of "unfair advantage" for the AI has waxed and waned over the years, and - like Big Foot - it has been debunked and yet still holds a cadre of believers. For example, a friend of mine steadfastly insists that he once read somewhere that a vocal group of users insisted that the game was "too easy" and so the developers (whoever they were at that time) supposedly made some tweeks to the game in order to overcome some things that made it "too easy" for the player to win. Now, I've heard rumors to that effect myself, although never actually seen them in print anywhere as far as I recall... but, I can attest to my own observations of aspects which could easily be interpreted as "unfair advantage" in some light. Some of these observations apparently remain unconfirmed by others, but do exist - for whatever reason - in my own games. Now, it may be that I'm somehow mis-interpreting my observations, and they are in fact manifestations of some legitimate situation, but so far, I remain puzzled by them. These things include (but not limited to):

- The stuff we've been re-hashing in this thread, to wit, distinct and seemingly unreasonable mis-matches in the effectiveness of missiles when fired by the player's side versus the computer's side - regardless of the type of missile or launch platform. (This also includes torpedoes, although not being discussed in this thread).

- The possibly-related substantial discrepancy between the effectiveness of SAMs and AAA on the respective sides.

- The ability of the AI to detect, intercept, and target player's groups when there is no apparent means for the AI to have even

detected

those player's assets... as contrasted to the extreme difficulty the player has in detecting, tracking, and engaging certain AI units, even when the player has employed substantial detection assets, and contrasted to the virtual impossibility for the player to successfully engage enemy assets without having a complete "fix" on them.

- The ability of the AI to successfully track and target player's groups without ever activating any radar -

especially

puzzling in the cases where the AI is firing radar-guided missiles.

- The relative difficulty in

killing

targets of the respective sides.

Now, all this stuff can be viewed as part of the "haze of battle", or whatever, but the aspect that bugs the you-know-what out of me is that, for whatever reason, all this stuff does not appear to apply equally to both sides. Not in the sense that everything has to be a fair fight, but rather that the laws of physics, or whatever, ought to apply without bias - unless there is some artificial skewing (like radar jamming, etc.). For me, the annoyance of it all is that there seems to be no rhyme or reason for the discrepancies in most cases.

 

I'm not looking for the AI to roll over and play dead, because figuring out legitimate ways to overcome legitimate advantages would be part of the fun and the challenge. It's just that from where I sit, these "advantages" don't seem to be legitimate, and that makes it difficult to assess the value of the chosen tactics.

 

I have to chuckle about the multi-axis attack tactic, because I've tried almost everything from surrounding the AI group with planes, ships, and subs, and loosing all the weapons timed as best I can to converge on the targets simultaneously, and in volumes hoped to swamp the defenses - yet the AI group dispatches them all as deftly as if I had sent in a massive single-axis group. I've also tried "waves" of missiles - where one volley follows fairly closely to a preceeding one... but with pretty much the same non-results. In fact, the only way that I've found to "crack" the nut, is just to keep pounding away with my unlimited supply of ASuW missiles until I deplete the magazines of the AI target... at which point, I gain success. But, that means expending hundreds of missiles - probably well beyond what would be available in reality. Maybe this is just another instance of mis-interpreted observations, but until the "reasons" become clarified, it still adds to my frustartion.

 

Bottom line is that although I have found ways to be (quite) successful in the game, I end up feeling like the tactics I've developed are largely "cheats", and so it leaves me wondering how well I might do in a "real" situation... but, then, I keep coming back for more...

IMO, the whole issue of "unfair advantage" for the AI has waxed and waned over the years, and - like Big Foot - it has been debunked and yet still holds a cadre of believers. For example, a friend of mine steadfastly insists that he once read somewhere that a vocal group of users insisted that the game was "too easy" and so the developers (whoever they were at that time) supposedly made some tweeks to the game in order to overcome some things that made it "too easy" for the player to win.

 

I don't know who the 'cadre of believers' are, other than you and your still mysterious friend, but they are welcome to speak up. Maybe if we can't get anywhere with your particular problem, we might be able to help them with theirs? As for the 'developers', two of them are right here in this thread. The game has seen plenty of 'tweaks' over the years, and is still being 'tweaked'. None of them ever created an invincible AI, however, or some method of the AI having a set of loaded dice.

 

I'm not looking for the AI to roll over and play dead, because figuring out legitimate ways to overcome legitimate advantages would be part of the fun and the challenge. It's just that from where I sit, these "advantages" don't seem to be legitimate, and that makes it difficult to assess the value of the chosen tactics.

 

Some folks think the AI is too easy as it is, right now.

 

Bottom line is that although I have found ways to be (quite) successful in the game, I end up feeling like the tactics I've developed are largely "cheats" ...

 

Well, since neither multi-axis attack or wave attack seems to work for you, except via brute force, I am curious about what does work. Can you be more specific? What methodologies do you employ?

 

Oh, and I'll ask again, what was the issue with 'permission' to post in this thread? (It seems to be okay now, I assume?)

 

Reminder: You're going to have to provide savegames again for us to get anywhere, ultimately, with figuring out your issues.