[Java MOO] Feedback on bugs, cheats & exploits

[Ray F]

This forum is a valuable resource for me, and I would appreciate any feedback on this post: http://www.java-moo.com/?p=530

In particular, are there any I missed?

[Zeraan]

This forum is a valuable resource for me, and I would appreciate any feedback on this post: http://www.java-moo.com/?p=530

In particular, are there any I missed?

While digging around in MoO 1 OSG and observing the gameplay mechanics for my MoO 1 clone, Beyond Beyaan, I discovered several bugs that was never reported anywhere else. I will link threads where there's discussion on them, and summarize each bug:

Pollution bug - The more factories you have, the less it pollute actually, until it reaches a point where it doesn't pollute anymore! Also, Sakkras lose 2x as much population from pollution as other races, so their growth modifier may have an effect on the pollution formula. The reason why this isn't really noticed is that the ECO slider automatically resets to at least "CLEAN" each turn. So if you want to exploit this, you'd have to drag the ECO slider for every planet for every turn.

Gaia Discount - Discussion on Gaia and Fertile improvements led to discovery that if a planet was naturally fertile, it does not discount the Gaia cost (should be 150 BC for each upgrade, totalling 300 BC, but if a planet is naturally fertile, you still pay 300 BC instead of 150 as it should have)

Factory Refit - I researched on how MoO 1 does refitting exactly. The manual and OSG both state that it should cost 10 BC for first 2 factories per pop, then refit them at 5 BC cost each to upgrade them to tier 3, then build tier 3 factories at cost of 15 BC each for each pop, and so forth. So you should see "FACTORY 3.4/turn" -> "REFIT" -> "FACTORY 2.1/turn" -> "REFIT" and so forth. In reality however, when you get a new RC tech, all the new factories change to that new cost. So what "REFIT" does actually is just waste production until the number of existing factories * the new cost is reached (it stores some kind of total cost paid somewhere) before building the new factories. While this part may be common knowledge, this next bit is new: You can immediately move population out of a planet as it immediately applies the new RC level. For example, with homeworld pop of 100 and 200 factories, when you research RC III, you can send 33 population off planet immediately, and the remaining 67 pop will utilize ALL of the factories (67 * 3 = 201). So this can be an exploit.

Granted, those are minor bugs, but could potentially be used as exploits (the pollution and refit parts that is) by players with deep knowledge of the game.

[RefSteel]

On factory refit: There is actually NO "refit" cost for factories. (Sargon found a cost of 1BC TOTAL for the planet, but whether that exists or not is really splitting hairs.) So no cost is actually wasted. The planet screen says "refit" but actually all the "refit" production is going into new (higher-cost) factories. It is also by no means a minor bug. It severely affects optimal tech tree paths and timing, to the extent that drawing RC3 at an ART world in the early turns effectively adds multiple difficulty levels to the game. (Of course, II8, IER, Range 6, or Controlled Dead can reduce the effective difficulty severely).

Other bugs ... I'd have to rack my memory. Off the top of my head, some weird ones:

- Negative Fleet Bug. AI fleets occasionally appear with a negative number of ships (which later "corrects" to >32000) for unknown reasons.

- Comet and Pirate mitigation: This is inexplicably based strictly on the number and size of the ships. The most consistent way to damage/eliminate a Comet or Pirate in a given system is to throw a bunch of Small ships at it, regardless of whether they're armed. (Huge ships do random damage between 1 and some insufficiently large number, whereas small ships always do 1 damage each).

- Going Diplo: The OSG claims that the diplo event severely reduces relations, and may even lead to war. If fact, the Diplo event leads to a war declaration every single time, without exception, even for allies at maximum Harmonious relations. In fact, it results in a war declaration message even when you are already at war!

- In the base game, an AI with 4 colony ships will not initiate any colony missions (if I'm remembering kyrub's analysis correctly). If this happens with no colships already en route (or if they are prevented from colonizing their targets but not destroyed) the AI will be unable to peacefully expand again until it loses a colship in battle, scraps its colship design (for a newer one) or colonizes a planet accidentally (i.e. by sending a fleet that happens to include a colship to a world where, upon the arrival of the fleet, no colony is present and the environment is suitable to its technology.

[RefSteel]

On the list you've compiled so far:

(Spoilered for length. By which I mean for lennnnnnnnnnnnnngth.)

First, on the whole I like the modifications you suggest in the linked post. I strongly disagree with your definition of "cheats" - a word which carries moral connotations at the least when we're just talking about different players' preferences on how to enjoy a video game played against a machine - but that's semantics, and in nearly all those cases, I think it's desirable to modify or eliminate the loop-holes in question anyway, no matter what we call them. Some questions and suggestions then:

- "Swing Votes": The more I think about this, the more I like your solution. The one caveat is that random early wins or losses (due either to sheer luck or random/periodic/permanent AI alliances) are the least fun part of the game, so if those could be mitigated in some way (regardless of player vote) that would be neat.

- Canceling Retreats: Will players be allowed to redirect retreating ships to planets other than the one from which they just retreated? Or are they considered "already in transit" so they can't be redirected at all? (The latter seems cleaner, but the former will be less tedious in some late-game situations and more tense/dangerous/exciting when the AI makes use of it, as I presume you would "teach" it to do.)

- Dodging missiles: Forcing missiles to move last would modify the dynamics of combat significantly. As I demonstrated in OSG-26, you don't need to use the "Wait" button to dodge slow missiles with fast ships - you just outrun the things! - but this doesn't work the same way when missiles target a ship's final position instead of "following" it as it moves. I think ultimately I like the dynamic you propose, but be aware that it opens the door for other exploits. For instance, in some circumstances larger ships with "maneuvering missiles" (weapons that can be "turned off" to allow a ship to move after firing its other weapons) would become extremely powerful. (More so than it is already I mean. That's another trick/bug/whatever to add to your list of problems to fix...)

- Instant Fleet Construction: I actually love the way the base game does this, as it rewards planning ahead and makes intuitive sense.. It's also not a really huge advantage compared to your proposal as I understand it. I assume by "moved to the colony reserve" you mean it's added to the colony (at 1:1) as though it had been "transferred" there from the treasury on the Planets screen. The problem is, this means instead of pre-building a placeholder ship for a different, newer fleet, players will (still do that, but with slightly slower effect, but also...) prebild a placeholder spaceship for terraforming, factory construction, and planetary shields! If you convert the reserves at worse than 1:1, it would mean optimal play requires micromanaging ships to complete to the exact BC (even at 2:1, it's better to build IND for reserves usable empire-wide at the same ratio than to allow overflow from a ship). If empire reserves are fed this way instead of just the planet's, even at 2:1, it lets you build reserves from backlines worlds without max factories and still demands micromanagement of ship construction at front-line colonies where you want to spend your treasury instead of generating it. If you allow 1:1 but it only applies on the same planet, and only to shipbuilding ... actually I think that would work, as long as it either combines with normal reserve spending (so you can get up to 3x normal production toward ships each turn if you spend full reserves and have overflow from an unbuilt ship) or at least doesn't "get in the way" of normal reserve spending if you suddenly switch from a half-complete 7-turn Huge to a few small defensive fighters and all the missile bases (or terraforming or factories or...) your reserve-boosted planet can give you. The only trouble here (and with this solution in general) is that it's essentially adding a new mechanic to the game. (Also note: The conversion from ship to reserves should happen when the player ends turn with a different ship in the queue, not just when a player mis-clicks the ship button or clicks it on purpose to examine other possibilities....)

- UR Reserves: Your solution significantly changes the value of the planetary reserve and especially of "special" worlds. It definitely reduces micromanagement however, and the special worlds are significant enough even without reserve spending that ... I think I like this. Wait, does this mean Poor and UP worlds don't apply their penalties to reserve spending either? ... Actually, I think I like that even more.

Kudos on thinking all this stuff through; hopefully the ideas above will be of use as well!

[Ray F]

My reply to Zeraan:

Pollution bug - The more factories you have, the less it pollute actually, until it reaches a point where it doesn't pollute anymore! Also, Sakkras lose 2x as much population from pollution as other races, so their growth modifier may have an effect on the pollution formula. The reason why this isn't really noticed is that the ECO slider automatically resets to at least "CLEAN" each turn. So if you want to exploit this, you'd have to drag the ECO slider for every planet for every turn.

This factory pollution bug will not occur in Java MOO. Also, the Sakkra do not suffer any special pollution penalties. Are they supposed to?

Gaia Discount - Discussion on Gaia and Fertile improvements led to discovery that if a planet was naturally fertile, it does not discount the Gaia cost (should be 150 BC for each upgrade, totalling 300 BC, but if a planet is naturally fertile, you still pay 300 BC instead of 150 as it should have)

Gaia upgrade costs are only 150 BC from any Fertile planet.

Factory Refit - I researched on how MoO 1 does refitting exactly. The manual and OSG both state that it should cost 10 BC for first 2 factories per pop, then refit them at 5 BC cost each to upgrade them to tier 3, then build tier 3 factories at cost of 15 BC each for each pop, and so forth. So you should see "FACTORY 3.4/turn" -> "REFIT" -> "FACTORY 2.1/turn" -> "REFIT" and so forth. In reality however, when you get a new RC tech, all the new factories change to that new cost. So what "REFIT" does actually is just waste production until the number of existing factories * the new cost is reached (it stores some kind of total cost paid somewhere) before building the new factories. While this part may be common knowledge, this next bit is new: You can immediately move population out of a planet as it immediately applies the new RC level. For example, with homeworld pop of 100 and 200 factories, when you research RC III, you can send 33 population off planet immediately, and the remaining 67 pop will utilize ALL of the factories (67 * 3 = 201). So this can be an exploit.

Java MOO does not allow this exploit since the robotic controls level for a colony does not increase until all refit costs are paid. (I had to check my code to be sure...

[Ray F]

My reply to RefSteel:

- Negative Fleet Bug. AI fleets occasionally appear with a negative number of ships (which later "corrects" to >32000) for unknown reasons.

Integers in Java are signed 32 bits instead of the signed 16 bits in the original MOO (which is what caused the overflow at 32767). This allows you to build over 2 billion ships in a stack before hitting this problem.

- Comet and Pirate mitigation: This is inexplicably based strictly on the number and size of the ships. The most consistent way to damage/eliminate a Comet or Pirate in a given system is to throw a bunch of Small ships at it, regardless of whether they're armed. (Huge ships do random damage between 1 and some insufficiently large number, whereas small ships always do 1 damage each).

Java MOO bases this one the estimated firepower of the ships. Building a giant, empty hull would be completely ineffective against comets or pirates.

- Going Diplo: The OSG claims that the diplo event severely reduces relations, and may even lead to war. If fact, the Diplo event leads to a war declaration every single time, without exception, even for allies at maximum Harmonious relations. In fact, it results in a war declaration message even when you are already at war!

I'm in the camp that this event should always trigger war anyway. The whole purpose of these events is to shake things up by throwing you (or an AI) an occasional curveball.

- In the base game, an AI with 4 colony ships will not initiate any colony missions (if I'm remembering kyrub's analysis correctly). If this happens with no colships already en route (or if they are prevented from colonizing their targets but not destroyed) the AI will be unable to peacefully expand again until it loses a colship in battle, scraps its colship design (for a newer one) or colonizes a planet accidentally (i.e. by sending a fleet that happens to include a colship to a world where, upon the arrival of the fleet, no colony is present and the environment is suitable to its technology.

Sounds like a code bug in the original game. The AI in Java MOO builds all of its own ships (no freebies) and colonizes aggressively.

- "Swing Votes": The more I think about this, the more I like your solution. The one caveat is that random early wins or losses (due either to sheer luck or random/periodic/permanent AI alliances) are the least fun part of the game, so if those could be mitigated in some way (regardless of player vote) that would be neat.

In Java MOO, the council only forms after one race has contacted all others. I expect that this may push back the starting time.

- Canceling Retreats: Will players be allowed to redirect retreating ships to planets other than the one from which they just retreated? Or are they considered "already in transit" so they can't be redirected at all? (The latter seems cleaner, but the former will be less tedious in some late-game situations and more tense/dangerous/exciting when the AI makes use of it, as I presume you would "teach" it to do.)

Retreating fleets are allowed to redirect to any system except the system they are leaving. Which means if you retreat a stack but win the battle, you still have to retreat (no protecting your unarmed colony ships from stray missiles by retreating)

- Dodging missiles: Forcing missiles to move last would modify the dynamics of combat significantly. As I demonstrated in OSG-26, you don't need to use the "Wait" button to dodge slow missiles with fast ships - you just outrun the things! -

I want to correct how I described this fix. Since the ship combat isn't done yet, that was the "plan". But I have since recalled that in my original iteration of Java MOO that I solved this problem by having missiles move when their target moves. Since missiles more directly towards their target, they will catch you if they are faster.

- Instant Fleet Construction: I actually love the way the base game does this, as it rewards planning ahead and makes intuitive sense.. It's also not a really huge advantage compared to your proposal as I understand it. I assume by "moved to the colony reserve" you mean it's added to the colony (at 1:1) as though it had been "transferred" there from the treasury on the Planets screen. The problem is, this means instead of pre-building a placeholder ship for a different, newer fleet, players will (still do that, but with slightly slower effect, but also...) prebild a placeholder spaceship for terraforming, factory construction, and planetary shields!

Let me expound on this. As you build ships, each turn a "Ship BC" bucket is filled with that turn's production until the ship is built. If you change designs, all of that BC is moved to a "Ship Reserve BC" bucket (1:1 ratio). Each turn thereafter, any new BC dedicated to building ships is matched by funds from the "Ship Reserve BC" bucket, effectively allowing you to double your ship production. If there is no "Ship Reserve BC", then "Colony Reserve BC" can be used instead (which is filled from the Empire reserve). The "Ship Reserve BC" cannot be used for non-ship construction like factories, bases, etc.

- UR Reserves: Your solution significantly changes the value of the planetary reserve and especially of "special" worlds. It definitely reduces micromanagement however, and the special worlds are significant enough even without reserve spending that ... I think I like this. Wait, does this mean Poor and UP worlds don't apply their penalties to reserve spending either? ... Actually, I think I like that even more.

The bonus or penalty from special worlds only gets applied once, when BC is sent to the reserve. Reserve funneled back to the planet has no penalty or bonus. This will help poor and ultra-poor planets get up to speed faster while preventing the ultra-rich exploit.

[Ray F]

My reply to Thrawn2:

BUG: Invading Spored Planets
To make sure I understand correctly: if AI 1 owns a planet and then AI2 captures it, then on the next turn I move in and capture it from AI 2 before they've had a chance to refit any factories - I'll still have a 2% chance per factory to get a tech, but they will be from the techs that AI1 owns and that I don't, and it doesn't matter what techs AI2 has?

This is a nice touch but has a problem of its own. When I invade a planet of an AI, I get from the techs that they currently have. This way of handling it means that I can get from AI1's techs 10 turns after they were kicked out of the planet and possibly discovered a new shiny tech.

Overall it's a very nice touch, and if you keep track of the techs of previous owner at the time the planet was taken from them I don't see any problem.

It doesn't handle that particular case now, but I should be able to do it pretty easily.

BUG: Infinitely recharging specials
This is solved in the 1.4m patch. I don't know if you have looked at the full fix list but it would certainly be a good idea to. Also there was some 'embryonic' patch that turned out to be bugged and kyrub discontinued it but some of the fixed it made weren't transferred over, so you can check it as well.

Keep in mind that I am not inheriting many of the bugs in the original game since the Java MOO code was written from scratch.

CHEAT: Galactic council swing vote
The problem seems to be with the player getting the AI's goodwill by voting for them. If we assume that the AI players are smart enough to see past what the player is trying to do by voting for them, then removing the benefit will solve the problem. The penalties can be kept though
Regarding the order of voting - is there anything wrong with implementing blind voting? That way the player will always vote first and then see what the others voted.

Voting against an AI will have negative consequences, especially if you were a swing vote. Voting for an AI will not have positive consequences unless they are elected and, in that case, the game is over anyway.

Voting in order of population gives the great possible influence to otherwise insignificant empires. Blind voting would take that away.

CHEAT: Empty threats
The problem here seems to be the incorrect evaluation of the AI of the strength of the human player, as well as if the player will actually attack or not, and the farce of the peace treaty which it currently is.

That evaluation is what I specifically hope to improve.

I suggest that peace treaties are enforced by the game. This means that fleets will not engage as if there is an alliance, the bombardment option will not be available, and if there are incoming troops to a planet they will get disbanded on arrival rather than starting to attack the planet. The players should be able to violate a peace treaty in the way that they can break an alliance (but it shouldn't require contact with the diplomat in case he is unavailable) and this will be remembered as oath breaking with the appropriate penalties.

You can't bomb without breaking a peace treaty. Fleets will engage and will break the peace treaty if they engage. You will still have the option, however, of breaking the treaty and becoming an oath-breaker.

CHEAT: Canceling retreats
Since fleets can be contacted and given orders when they are located over neutral planets, it makes sense to me to be able to direct their retreat to a desired planet or even instruct them to pull out of a battle and stay where they are.
The AI doesn't do it, but maybe that's the problem - if the AI can take advantage of it too, then it's not an exploit.

A retreat is a retreat. If you choose to retreat, you must go to a non-hostile planet (no "retreating" to the next enemy target!).

This makes the decision to retreat a real decision and now some fluid game mechanic that can be exploited.

CHEAT: Dodging missiles with the 'wait' button
This is a cheat indeed and it's not always easy to avoid. Missiles move with the ship which seems fine, what can happen is that if a ship waits, the missiles don't move at that time, but move when the ship finally takes action - move or press 'done'.

That's the intended Java MOO behavior -- missiles moving when their target moves. I was incorrect in my earlier description.

Exploit: Instant fleet construction
Interesting idea. I think it has to be at the standard reserve penalty and move the funds to the general reserve.

This was a considered option and rejected because it would encourage players to tentatively build large ships instead of sending to reserve. The former allows you to build a ship and then change your mind with no penalty. As such, there would be no reason to specifically earmark BC to the Empire Reserve. The Java MOO solution requires this BC to be used only for ship construction on that colony, unlike general purpose reserve BC.

Also, there is one problem with it. If I start working on a huge ship expecting 2-3 techs to hit in the meantime, I redesign the ship slightly when they are done. For example something like a new battle computer can easily be fit at the end of the ship construction, so it is a valid tactic. With the change the ship will have be built to specs even if they are from 20 years ago with no chance for last minute upgrades.

Who says that new computers can be easily integrated into old designs? I think that is an unwarranted assumption. The new technologies also create miniaturization which can have ripple effects throughout the entire ship design ("oh look, I can squeeze in 25 ion cannons now instead of 23").

An alternative is to allow 'editing' a design of which there are no active ships yet - the editing can allow addition of new components at no costs, and the removal of existing components at a proportionate cost (for example 50% of the ship is constructed at a shipyard. a component that costs 10% of the ship's total cost is removed. this will leave the shipyard at 45% completion, and 2.5% will be added to the reserve). It’s workable if not perfect but it’s still a significant change from the original game.

The ability to edit ship designs on the fly removes one of the most underrated strategic considerations in MOO1 that exists in no other space 4X games... making the decision on when to commit to building a new ship design and not looking back. There's just no way I can remove that from the game.

OTHER NOTES:
Regarding the AI - if you can in some way make it moddable or programmable, fans will work on it and improve it ongoing beyond what can be accomplished with the original effort.

There is no need for me to make Java MOO moddable since it will be open-sourced upon completion.

Rounding! It impacts the game a lot and it shouldn't make any difference. If you make the game work with decimals or hundredths of a BC/damage point/etc in the background I believe it will improve the game greatly.

Accumulated things like population, factories, shield levels and missile bases are retained as 64-bit doubles. That's about 20 digits of precision :P

[Ray F]

In bombing and retreating fleets I agree that the player indeed has control. But transports that are on the way at least need addressing - maybe they can get erased the moment a treaty is signed.

In Java MOO, transports currently disband when they land on a friendly alien colony.

The main problem with the peace treaty is that the AI never breaks it whereas the player can - so it provides a convenient one sided immunity. The AI is also keen to sign a new threaty even if the player has violated a previous one, providing the one sided immunity again. In the very least the AI shouldn't offer any BC or tech for a peace treaty when dealing with a known violator.

The hope is that the AI will break treaties when it is to his advantage (after considering oath-breaker).

Exploit: Instant fleet construction
Interesting idea. I think it has to be at the standard reserve penalty and move the funds to the general reserve.

This was a considered option and rejected because it would encourage players to tentatively build large ships instead of sending to reserve. The former allows you to build a ship and then change your mind with no penalty. As such, there would be no reason to specifically earmark BC to the Empire Reserve. The Java MOO solution requires this BC to be used only for ship construction on that colony, unlike general purpose reserve BC.

Without something like this, huge ship lose the ability to get respeced while under construction which I think is one of their important functions. I wouldn't really start one a huge ship that takes 20 years to complete if I can't use technology that I expect to obtain in the meantime or have to start over even at double rate.

I respectfully disagree. At some point, a player should have to commit and say, "ok, this is my next Huge ship design" even though you KNOW that by the time the ship is built, there will be new technological breakthroughs. This is exactly the kind of difficult strategic decision that makes MOO1 a great game. "Should I start now, or should I wait 5 more turns?" Allowing you to bypass that decision, even for large ships, takes that away.

[Psillycyber]

Exploit: Instant fleet construction
Interesting idea. I think it has to be at the standard reserve penalty and move the funds to the general reserve.

The advantage of the current way ship production works under MoO (not penalizing accumulated BCs when changing ship production) is that the player never has to fiddle with the ship production slider on a turn-by-turn bases to make sure that JUST ENOUGH BCs are going into ship production to not cause any overflow, which might then be wasted if the player decides to change production to a different ship type next turn.

When producing small ships, the overflow is negligible, but when producing large or huge ships, the overflow can be quite substantial.

If you are going to penalize accumulated/overflow ship production in case of ship production changes, there would really need to be some interface button to help the player quickly round off the ship production slider to the next highest whole integer ship. That might be a way forward if we wanted to penalize this "wasted" accumulation/overflow of ship production by sending it to the general reserve at the standard 50% penalty.

RayF's proposal is another way of keeping the beauty of not having to worry about rounding while at the same time not allowing insta-fighter-swarms-o'-death. The only thing I'd worry about with his proposal is that having to take funds from a seperate ship reserve at each planet in addition to the common planetary reserve might mean more micromanagement for the player.

I do agree with RayF that it is an acceptable trade-off to have the problem of huge ships being slightly out-of-date from the moment they roll off the production line. Huge ships are already seriously OP. Consider how a huge ship fights at maximum effectiveness right up until the very moment that it drops below 1 hp. Consider how an equivalent swarm of fighters with an equivalent total hp have their effectiveness drop off exponentially in a typical battle long before they get to 1 total hp left between them. Consider all of the anti-small ship counter-measures that exist (pulsars, ion stream projectors, black hole generators, scatterpacks, tachyon beams, etc.) What counter-measures are there for huge ships? The death ray? And then when you throw auto repair and some of the other specials that huge ships can make particularly good use of, huge ships just become OP. One huge ship, if it can survive the initial battle with a consolidated AI fleet (if the AI exists in such an organized manner in the first place) can completely dominate the steady trickle of reinforcements that an AI will try to send at it for 50 years afterwards.

[Ray F]

Colony with base 100 BC production and 30% ship spending. I think the examples below should be accurate.

In all cases, Empire Reserve does not get the additional planet bonuses, and the Colony Ship Reserve will match the colony ship spending after planet bonuses


Normal planet, no reserve
100 BC * 30% = 30BC

Normal planet, 500BC reserve
100 BC * 30% = 30BC
Add 100BC reserve * 30% = +30BC = 60 BC

Normal planet, 500BC ship reserve
100 BC * 30% = 30BC
Add 30BC ship reserve = 60BC

Normal planet, 500 BC reserve, 500 BC ship reserve
100 BC * 30% = 30 BC
Add 30BC ship reserve = 60BC
Add 100BC reserve * 30% = +30BC = 90BC

Ultra-rich planet, no reserve
100 BC * 30% = 30BC
30BC * 3 UR bonus = 90BC

Ultra-rich planet, 500 BC reserve
100 BC * 30% = 30 BC
30BC * 3 UR bonus = 90BC
Add 100 BC reserve * 30% = +30 BC = 120BC

Ultra-rich planet, 500 BC ship reserve
100 BC * 30% = 30 BC
30BC * 3 UI bonus = 90BC
Add 90BC ship reserve = 180BC

Ultra-rich planet, 500 BC reserve, 500 BC ship reserve
100 BC * 30% = 30 BC
30BC * 3 UI bonus = 90BC
Add 90BC ship reserve = 180BC
Add 100BC reserve * 30% = +30BC = 210BC