Naval Academy Question

[Nick Thomadis]

Here is a question that could help us improve Naval Academy missions.

Random AI builds aim to increase replayability of missions and the uncertainty of what kind of enemy you will face (if you know your enemy 100%, then you find the optimal design once and beat the missions easily and consistently until you get bored of them).

Would you like better to have more constant AI enemies? Enemies that you can more easily predict their strength?

[HusariuS]

@Nick Thomadis 

Two questions:

1. Is it possible to give AI only one design for specific missions like in "Prove your Might" where we are supposed to fight against Bismarck and Prinz Eugen (EDIT: at least according to mission description it WAS looking like they are Bismarck and Prinz Eugen.) ?

2. If yes, then is it also possible to give players an option for those kinda missions to choose either "Historical design" or "Nonhistorical design" for AI warships?

Edited May 29, 2020 by HusariuS

[johnson smith]

the naval mission is fine. I just don't want to chase a BB with 35 + knot or DD with 45 + knot when the game win condition is knock out 100% of the enemy forces 

[Cptbarney]

@HusariuS makes a good point which i would like to echo, with a different mode type for the missions, historical could follow more closely to historical designs if the missions is based upon an actual event, or towards actual ships in general, while non-historical will go for what we have now, but still closer to functional designs (think of blueprints, what ifs but sensible ones i guess).

This would offer a nice balance of consistency and unpreditability since yes the ships might use the same hulls and probs super structure, but the other stats could maybe tweaked somewhat higher or lower indifferent areas for historical, while non-historical would still feature mostly reasonable, but the occasional (u wot m8) moment ships.

 

[fsp]

I like the inconsitency, not knowing what you are up against, as otherwise it is really to easy to make tailor-made ships for certain missions. However, sometimes this also results in opponents that are easily beat, e.g. an enemy BB with 13" guns while you are already sporting 16".

 

[Reaper Jack]

2 hours ago, fsp said:

I like the inconsitency, not knowing what you are up against, as otherwise it is really to easy to make tailor-made ships for certain missions. However, sometimes this also results in opponents that are easily beat, e.g. an enemy BB with 13" guns while you are already sporting 16".

 

The opposite is also true however, and I've found that for most BB missions you cannot afford to not have the biggest guns as even something like 381mm weapons are just inadequate to do any real damage to AI ships that have 35 knot speed and 450+mm of armor. 

[SiWi]

personally I can deal with random enemies better then random allied ships...

 

I mean if the light cruiser escort is design without torps, I can't see the point of them...

[Reaper Jack]

5 hours ago, SiWi said:

personally I can deal with random enemies better then random allied ships...

 

I mean if the light cruiser escort is design without torps, I can't see the point of them...

This too, I made a fleet of 2 BC, 2 CA, 2 CL in custom battles yesterday, trying to simulate two German 'Raiding Squadrons' that what still on their way to the Atlantic, together so far, being intercepted by a British Patrol Fleet (3 CA, 3 CL, 6 DD) and the BC I designed myself (better Scharnhorsts basically) and they could pull 34 knots, but the CA's could do 26 knots only, had no torps, 8 guns but also had ludicrous armor (270mm belt) while the 2 CL could only make 24 knots, had no torpedoes and additionally had 70mm armor at the most, and were destroyed almost instantly because of it. 

The biggest problem with friendly AI designs I've found so far is they are either very slow (and therefore easy to hit) or their engine efficiency is...abysmal. Wounded Beast gave me a BC that could pull 40 knots...but had one funnel of the oldest, worst type only. It took it 45 minutes to hit top speed. 

[TAKTCOM]

"Completely random" sound extra fun and sometimes it really is.  But much more often it becomes extremely frustrating: too fast ships, incredible (sometimes incredibly low) firepower, phenomenal survivability, and so on. Too often you have to restart the mission again and again, just to have ships that allow the player to win. This works for both enemy ships and player ships, built by auto design.

So I choose the last option.

[admiralsnackbar]

In my mind the answer to this question depends on what purpose the naval academy missions are supposed to serve. 

If they're designed to function as tutorials, teaching you about certain build/tactical techniques to win engagements, having randomization of enemy capabilities defeats that goal unless the purpose of the mission explicitly revolves around dealing with uncertainty. 

But in the worst cases it turns into a frustrating situation where it's not a question of whether 'did you build a ship that is well suited to the mission as we describe it' and instead 'did you build a ship that just so happens to be good against the randomly designed scenario' 

Requiring the player to design a well rounded ship through randomness might be a good idea, but then the random ships they face can't be so strong that only a ship heavily optimized for fighting said foe could win -- because only chance will allow such a ship to match the scenario. 



 

[The Fundamentalist]

I think it would be best if the Naval Academy was designed to help the player learn the mechanics of the shipyard, battlefield and campaign. The game relies heavily on a basic understanding of naval tactics and how contemporary navies formed battle lines, so it would be nice to shallow out that learning curve. 

Right now it's kind of singleplayer mission set 

[Skeksis]

Following a “certain rule of design” should put more focus back into designer tool ‘designs’ and less focus in missions of victory by random chance.

If/when you discover or predict enemy strengths, then it becomes a ‘practical’ choice of 'designs' to defeat the enemy, this is very appealing and I think it's the general bases of the game, so anything along those lines is got to be good.  

Edited June 15, 2020 by Skeksis

[Shaftoe]

We should assume that naval intelligence exists, so some tactical information should be available, such as top speed, max gun size/number of guns of different types, and some key armor values. Basically, the stuff that could be learned from open sources/observation at the time. Other weaknesses should be known, too - like poor turn rate/acceleration, etc.

[shieldy44]

i guess my thing with the completely random thing is you end up sometimes with small ship designs that just seem completely unrealistic, i have shot up small boats that have maximum bulkheads and just seem to last forever there is even one with supply ships and those things could sometimes be ridiculous taking damage like it had serious armor and bulkheads. So i think some guidelines would be nice

[Christoffer]

The best would be all of the options above.
for each mission select
"Historical" :Same ship every time and you can examine all ships ahead of time
"Variant"  Resonabely same ship  and you can examine allied fleets ahead of time
"wild"  you have no idea. (maby option to examine allied fleets ahead of time.)

[Zak MacKay]

I would rather they be known for the most part. Missions are meant to be more a test of skill and learning right? If I wanted random missions theirs the custom scenario for that. I feel like missions should be more crafted events where restarting for a new seed is not a big factor. Having a crafted ship may give players hints on what a good ship looks like. Instead of something vomited out. With random ships it may be reset and try again. Which is just a RNG check not skill. Where if you lose against a pre determined ship you just need to learn and improve.

I feel like hard missions could then be much harder. Because theirs no randomness to make the mission easier or harder. So it would be a achievement to do it without some seeds being impossible to win. Yes you can prepare exactly for it but the challenge does not have to be randomness it can just be a unfair fight you need to win with superior crafting ability.

Maybe their should be a second type of mission which are more random and intended for fun. But I don't think their really needed.

[Cpt.Hissy]

I am normally supporting randomness everywhere in games, as it makes things interesting... BUT with this here thing something's wrong.

Mission briefing creates some background of when and why, describes for us numbers and classes of enemy ships, and sometimes additional info, and expects us to utilize our (gamey?) naval warfare knowledge to deal with them.
For this, we apparently should use the common understanding of what defines each ship class in given time frame, to interpret what we're going to deal with, right?
Now, whatever game throws at me almost never is the thing i'd expect by it's designation, especially for bigger ships.

I'm promised an old dreadnought? I expect something like Orion class. Slow-ish, small-ish uniform battery battleship. But i get a 55000 ton abomination with at least 3 different calibers of main guns and wide spectre of secondaries from 8" down to 2" and torpedoes, running at 30+ knots.

Generation is great, but generation absolutely must follow the rules. If thing's called battlecruiser, or dreadnought, or whatever, it must actually be one, at the very least.
Also I'm not sure what purpose the Academy is intended to serve, be it tutorial section, or just there to play with while actual game isn't existing yet, but if it IS tutorial, then following the rules is even more important. This teaches you what you'll meet in campaign later. And i really hope to never see these 30 knots 40 guns "dreadnoughts".

[Steeltrap]

I made some comments quite some time ago about what the Academy could become, and others have echoed some of it here which makes me suspect we're thinking similarly.

I think a REALLY important idea is teaching people HOW to design ships. There are SO many interlocking aspects that it's not necessarily easy for everyone to get to grips with it, especially as a new player.

I know I got a lot of benefit from asking others when I first arrived, and there's been a steady drip of posts about "I can't win, this scenario is BS" (I should know, I wrote at least one myself, lol).

If I can find the post I made, I might paste it here, too. In essence it divided missions into "principles" and gave the player very specific objectives. As an example, you could set some pretty technical things (not unlike requirements the genuine design bureaus faced) such as armour, armament, speed, total displacement AND COST etc. This would hopefully force players to get to grips with things that will matter when playing the campaign, and also set forth some challenges to come up with designs where they may need to compromise on some things in order to satisfy others (which is very accurate).

Some of the initial missions are clearly intended to do this sort of thing, and it's obvious why it wasn't continued in great detail (resource constraints and priorities).

As a different part of the Academy could be those that are more like what we have now, specific scenarios against certain opponents. As has been suggested, ideally players might like to be able to choose, but that's more work for your dev team. I don't mind variety of opponents, but I DO mind if the gap can be large enough that you win easily in one battle then really struggle the next time all due to the massive differences between the AI designed ships from one to another.

I suppose I'd say the AI ought not have the ability to produce ships that make you question the briefing you're reading, which is what I felt was a big problem on the "German Raiders" mission as it initially appeared. I had one where I cleaned up easily, and several others where I got wiped. Unsurprisingly, the degree to which the TR were armed AND how many bulkheads they had were rather central to the difference, lol, but then so was the weather.

(speaking of which, any chance we can get an indication of the prevailing visibility in battle? I have some where I'm blind as a bat, others where I can see much further, and nothing in the interface I've found gives me any specific idea. If I can only see 6km where my 'standard' visibility is 8km, that's kind of important AND something the fleet would know in general terms)

Edited September 3, 2020 by Steeltrap

[Nick Thomadis]

36 minutes ago, Steeltrap said:

As a different part of the Academy could be those that are more like what we have now, specific scenarios against certain opponents. As has been suggested, ideally players might like to be able to choose, but that's more work for your dev team. I don't mind variety of opponents, but I DO mind if the gap can be large enough that you win easily in one battle then really struggle the next time all due to the massive differences between the AI designed ships from one to another.

It is a work in progress to offer difficulty modes in Naval Academy, so that AI opponent can be defined in a more strict way.

36 minutes ago, Steeltrap said:

(speaking of which, any chance we can get an indication of the prevailing visibility in battle? I have some where I'm blind as a bat, others where I can see much further, and nothing in the interface I've found gives me any specific idea. If I can only see 6km where my 'standard' visibility is 8km, that's kind of important AND something the fleet would know in general terms)

Currently, weather is random and it is working in an initial WIP state, as weather visuals do not always match the modifiers.

Just a reminder, in the top left corner you can find the state of weather modifiers during battle. Regarding an extra visibility info, we will certainly allow that later, when we finalize the weather mechanics.
 

[Steeltrap]

Hi Nick, hope you're well.

Yeah, I know about the thermometer being where you can find current weather adjustments and that they don't necessarily match the conditions on screen. First thing I do is pause the game and check those, because they're important when I'm deciding if I want to max my "own cruising speed" bonus to compensate vs run at higher speeds to throw off the enemy if the weather is neutral or providing a net accuracy bonus.

I asked about the more general info because I wasn't sure if it was available and I'd not seen it or if it's not yet there but might be something for the "when more important stuff is done" list.

I think at some point we ought to be able to get some idea of our effective viewing ranges, but I also have questions about how easy that will be to do given the way in which the visibility system has been designed (assuming I've understood it).

I suppose at the least there could be a circle showing the distance out to which you technically can spot things given applicable tower and tech, or two showing "normal visibility for this time of day" v "current visibility given prevailing weather conditions and any damage to towers/tech".

As I said, clearly a "nice to have" feature as opposed to being particularly impactful on game play.

Cheers

 

[draconins]

I voted for "Something in between, enemies should follow a certain rule of design that does not variate so much. "

Don't make it completely random, that would make your the mission description useless, which is what already happen. Some randomness needed not only for replayability but also give some challenge, however complete random make it hard for many player to design ship matching the situation. Theoretically you can make "fixed" design historical mission in easy/normal mode, while some randomness in harder mode (current setting is already hard mode).

Current randomness can make some mission extremely hard if you are unlucky, simply because of RNG giving say 20 km torps CA/DD escorting 16+inch guns BB with anti 16+inch plating (belt/deck), with extreme accuracy, with player ship practical design limit of say 12-14 incher. This is while at other time your enemy is 14incher, thin armor, with escort DD with 5km torps. In fact the easiest way to differentiate between easy, normal, and hard probably range of design allowed with enemy ship.

Additionally, you may be able to even add mode where the enemy is picking from "design library", this may even useful for historical mode.

Edited September 5, 2020 by draconins