Master of Orion 1 unofficial patch

[kyrub]

Note Waste happens when AI is given free colony ships (ECO% is already calculated before maint suddenly rises) so I assume Psilons were given third free colony ship in 2361.

Free colony ships are given when both
a) there is less than three of them (if I remember correctly)
b) the d500 roll is not bigger than the reserves BC
This happens very early in the impossible games because the taxation is set higher. I have not seen a reduction of the reserves as if the colship is bought, but I may be wrong.
And yes, the reserves are especially used to help new planets develop.

negative pop tenths - a bug
reserve spending - insufficient and reserves reduced too much
taxation - a waste

- Negative pop tenths means we will have to call off the leisure time of children, I guess. Back to work, and it should be okay. As it seems, changing the structure of the Next_turn process isn't the way to go, hehe.
- the reserve thing may be actually a problem of the original game, when the AI is stuck on one planet, no? And let's not forget the transfer bug...
- I don't understand "taxation - a waste"

The major things are hopefully only one or two big errors in the AI tech/prod selection code. Sargon, do you think you could have a very quick look at the other races to see which problems they "share" with Sad Psilons? No big analysis, just a small look at their capital planet (like in 1 or 2 years).

[RefSteel]

Free colony ships are given when both
a) there is less than three of them (if I remember correctly)
b) the d500 roll is not bigger than the reserves BC

Well, a) is presumably "there are three or less of them" since four colships seems to be standard (at least in these recent games, where I was paying attention and could see OPEs' entire fleets).

And yes, the reserves are especially used to help new planets develop.
[...]
- I don't understand "taxation - a waste"

I believe what Sargon is saying here is that the tax slider is almost always a waste of resources - particularly so in this stage in the game, for a One Planet Empire with no prospect of new colonies! It costs 2 BC of production (which could be spent on factories/tech/defenses) for each 1 BC produced, plus additional wasted production due to rounding errors, which is bad enough to begin with. But with that tax money being salted away to spend on new colonies which have no prospect of being built now and probably never will be - thanks in no small part to the empire's wasting money on taxes and fleet maintenance instead of teching to e.g. R5/Dead - in effect they're actually getting zero BC of production from what they take in taxes ... and (per your rules for colship acquisition above) simultaneously increasing the liklihood that they'll be given "free" colony ships with nowhere to go but that cost them still more in maintenance!

If I were designing the code for the game, assuming the AI does not pay for its "free" colships from reserve (and it can't possibly be paying in a meaningful way under the rules you described above, since there's a chance of it getting a "free" colship with a reserve of 1 BC!) I might try something like this:
1) Tell the AI never to use the tax slider if Reserve > (race's_colship_count + race's_planet_count) x 50
2) Make colship acquisition rule b) above something like, "if the d500 roll plus 50 times the number of colony ships already controlled by the empire is not greater than the reserves BC plus 50."

An alternative - since taxes really are wasteful! - might look something like this:
1) AI doesn't take taxes, ever.
2) Colship acquisition rule b) becomes: "if d500 + (100 x race's_colship_count) >/= 10 x (race's_planet_count x difficulty_level)
Assuming simple = 1, easy = 2, average = 3, hard = 4, Impossible = 5.

Note that these ideas are just brainstorming and may well have holes in them.

[sargon0]

RefSteel: You are right about my comments on tax and reserves being examples of poor strategy, in the base game. I like your ideas for improving the AI's decisions although it may be difficult to implement and may have side-effects. It seems this crude old AI prepares reserves for future colonies and keeps supplying colony ships to settle those new colonies unless reserves are being depleted by feeding those new colonies. This can work well unless you are isolated (and cannot work out how to tech out of it) with reserves rising and more pointless colony ships being supplied but this is not necessarily as bad as the Psilons make out here, see below.

Kyrub: Other than the isolated Psilons the AI's are doing well. In 2362 all other AI homeworlds are max'd and have double digit MB's. Klackons with 6 planets should be winning, tech advancing well. Mrrshan tech is progressing but a little dissapointing for a 3PE but that's cats for you. The Meklars are isolated like the Psilons but have max'd planet and are still competing. They have similar ship maint (4xcol, 5xlarge) as well as high MB maint (61) but their fully developped robot hard level production still pumps a decent profit. They have at least one tech in each field and in construction already have RW80, II8 & II7! They have R4 and are focusing on Dead with a Dead rich which looks 4 parsecs away! Interestingly they only have 182 BC in reserve (0.45% tax). Other reserves Mrrshan 917 (tax 0.3%), Klackon 234 (tax 0.4%). Hope this helps.

[kyrub]

Ahhhh, discussion about AI improvements. Nice!

particularly so in this stage in the game, for a One Planet Empire with no prospect of new colonies!

First, I forgot to mention there is a time delay before the reserve filling process begins (I believe it to be 2330). Second, as far as I know, there's no One-planet emergency plan in the AI conception. That is a pity, and we could surely think of a bunch of places where a sub conditio could be inserted, along with some strategy.

On the other hand, I still have what I believe to be a much better AI expansion routine (or rather a series of small-but-important tweaks in different areas of the game) on my harddisk. With it, I hope to make the OPEs considerably less frequent. If we can add some emergency OPE point or two, that's all the better.

The Meklars are isolated like the Psilons .... in construction already have RW80, II8 & II7! They have R4 and are focusing on Dead with a Dead rich which looks 4 parsecs away!

Exactly - nice to hear this. Let's say that Meklars (patched) are a "role model" for the new expansion routine (but it will conserve a unique objective accent for every AI player). There may still be OPEs, but they will try to dig themselves out of the hole.

I believe what Sargon is saying here is that the tax slider is almost always a waste of resources

The equations cannot be inserted in this place. Both fonctions (tax slider and colony ships building) are small, distant slumps of code.

When I analysed the tax code, my first idea was to erase or decrease it considerably as well. But, having seen the AI reserve usage later, I now appreciate the coder's work. The reserves play an important role in the way AI manages its planets... and some other things. And it's coded shortly and simply (important for the Moo game). Surely, there are big flaws (they did calibrate the system for a normal, not impossible level - and then they made the economical bonuses bigger in the latest patch), so it needs some patient tweaking and toning down. But I find the whole concept neat and fonctional.

BTW, speaking of a resources' waste, have you ever noticed how much does an impossible lvl AI spend on the internal security slider, when you have 'espionage' on? That cannot be even called a waste, it's self ruining.

Other reserves Mrrshan 917 (tax 0.3%)

Interesting stat, comparatively.
Thanks for your work. You may not believe it, Sargon, but reading your last report and seeing the unicity of the Psilon OPE problem, I suddenly got a small personal enlightment about the possible whereabouts of the bug.

[Waterd]

When I read in the other thread about cutting in half the ship maintenance and increase the maintenance of bases I got pretty excited. I thought it was an objective to give some alternative to the common flow of master of orion, instead of just making a more powerfull AI.

With the time as I understood better the game, my strategy to win got more polarized. What's more important I find the fact that the best strategy is not to have a fleet for a long period of the game, wich is less exciting of what it can be.

My point is. Bases are too powerfull. Bases are defensive. Defensive is fine but is less exciting than attacking.

It's no surprise than fighting races are far inferior to diplomatic and production races. (Btw I disagree with what most players seems to think that Darloks are weaker than Alkaris, even if darloks have generally inferior diplomatic relationships).

Anyway is not my point to argue about Moo strategy since i think most people will agree with me.
Ships are only usefull early on to abuse AI flaws and fights the planets for early colonization. Once planets are colonized, bases rule the day for a long period of time and generally you don't build ships to attack unless the game is already won (what i mean with attack is going to dominate a planet, just light harrassing , distraction and abuse of the AI doesn't count) It would be nice , as least for variety and as an option, to have a balance were building ships early on is actually a good. That may even make Mrrshans more likeable in large/huge maps, wich i find now very very hard to win without luck on those maps. Generally when I play Mrrshans i feel I have only drawbacks when playing on large+ maps(I'm always talking about impossible difficulty). Part of it is that winning strategies tend to the "expand fast, mass bases, research, research, research, win by vote, if you want to kill, then build a fleet" IS in fact that 90% of games i won I don't even need a fleet, just win by diplomacy.

My point is, it would be interesting if in some games attacking would be actually a good option. Is likely to be a patch where you have this option? weaker bases stronger ships?

[RefSteel]

@kyrub: It's good to hear Sargon's data was able to suggest a solution! Let me know if you need any additional play-testing or confirmation of various theories at the moment....

@Waterd: First and most importantly, a belated welcome! It's always good to meet another player who enjoys this delightful old classic of a game!

On bases being too powerful - I think what makes them so is your playstyle; in other words, what's really overpowering is your ability to plan the strategic game! I think Sullla would argue that's one of Orion's strengths, as defensive games allow for a much greater range of strategic options, and I would add that I find defense more exciting than attacking, since my empire has so much more to lose in case a battle is lost! Nevertheless, if you prefer to play an offense-based game ... why not try playing one, even if it's not in line with your usual (or what you consider the best) strategy? I think what can make a game boring quickly is playing it the same way again and again, regardless of what that way may be; that's why we play with crazy variant rules so often here!

For instance? Check out my reports for Imperium 12 and Imperium 13! The first is a Hard game as Meklar, but we don't allow ourselves to invade any planets or acquire Planetology tech by any means! The second is an Impossible game as the Alkari (no variant for that one). As it turned out, I built a total of 6 missile bases throughout my entire empires in the course of those two games combined! Perhaps not coincidentally, those two are also great examples of games where I sent all-out attacks long before the game was won, in order to achieve a winning position. So, it can be a viable strategy....

In any case, if you're interested, you might try your hand at our current Imperium game. It's only on Hard, but we're only allowing ourselves to have research spending in a single field at any given time, resulting in (among other things) quite a number of interesting choices you wouldn't have to make in a normal game of Orion. If you want to try a low base-count or early-attack variant on top of it, you're certainly welcome to do so too! Just don't post any spoilers about the game or your results until closing day (September 7th) - it would be great to see a game report from a new face then (or the following day) win or lose! Oh, and if you do decide to play, don't feel obligated to write something as ridiculously long, in-character, or screenshot-ridden as my reports in the links above. Some people just post turn-logs, others much quicker summaries, and even I don't post reports quite the length of those early ones anymore. (Posting reports even of the length I still do in fact probably means I'm insane.)

Welcome again in any event; I hope - via kyrub's patch-to-be, variants, or both - you continue to enjoy the game!

[Waterd]

Thanks for the reply.

First, I read a lot of your reports and also Sirian reports. So my thought hasn't changed after reading that for some reason.

Second, Yea i will participate. Still i don't see how much can it change just researching a field at at time, except early on, where you can't see your early options for cheap. But later on the game I generally just focus on a research of a time (with just one dot in each field to not lose excesss of the previous research) So I fail to see how this variant can change the gameplay so much.
Or i'm not understanding the rules clearly? Maybe i'm missing some strategy by doing several researchs at the same time? It's unclear to me.

EDIT: Two hours later of this post i finished the save. I saw no difference on my strategy or how it affected going all in all the time in one research. Wich is basically what i do anyway. Maybe I lose some extra resources by not taking advantage of the %....but the strategy is basically the same

[kyrub]

Here is a sample for testing:

- it attempts to fix both bugs created by previous patch
- it conserves all other features

The transfer bug was "repaired" by removing the feature "after selecting the transfer reserve target, the target planet is centered in the PLANETs screen". I've tried other things to conserve the feature, but without any success.


Sad Psilons...

I found a small lapse I made in the AI_tech_prod_allocation routine. This bug has unknown consequences, it probably cuases some sort of memory leaks - or it does nothing at all.

The problem is I cannot continue my research before being sure THIS is / is not the bug. So any testing help with the sad_psilons savegame is welcome.


(rename the embryo_18.exe to starmap.exe)

[kyrub]

@Waterd
First things first. And then we will see if or what we should change in a (possible) second patch. That is my view of the process: I intend to repair the original first and then - afterwards, maybe, maybe not - to think about changing some game rules.


The MBs discussion is somewhat off-topic in this thread, but here's my view (being no expert in Moo):

I'm more or less on board with you about the defense being too strong:
- it leaves strategic plan of the game somewhat flat and repetitive (defend until you smash them)
- it leaves tactical combat too often relying on destroying the bases, there are few tactical options other then "MBs = first target"

On bases being too powerful - .....
- I think Sullla would argue that's one of Orion's strengths, as defensive games allow for a much greater range of strategic options, and I would add that I find defense more exciting than attacking, since my empire has so much more to lose in case a battle is lost! Nevertheless, if you prefer to play an offense-based game ...

I cannot speak for Waterd, but I think that it is not a case of playing offensively x playing defensively, but to be or not to be threatened by enemy offensive. Since the MBs (or should I say: the MBS + scatterpacks) are too strong, you do not have to fear really the enemy stacks in 98% of cases and do not have to try some counter-offensive tactic of your own. That is a pity, in my eyes.

The problem is, how to correct this issue without making the AI MBs' defensive significantly weaker - and the game significantly less challenging...

My thought are / were
- to learn AI to use WAIT fonction (that is next to impossible)
- to significantly weaken the to_hit_bonus of the scatterpacks - a small change that should not hurt the AI that much, but the human player who faces big stacks of AI ships will suffer. This will also promote the use of ECM on bombers.

[/offtopic]

[kyrub]

September started, I'm sorry I could not keep the schedule.

The Larval stage still not ready, but I am halfway through. I finally breached the combat code and more or less located the "Wait" issues. Don't know how to correct them, yet. Biobombing - found, corrected. Also, the double tech hit bug is confirmed (bravo to the spotters, I've played the game hundred times without noticing).

No progress on the Sad Psilons front, despite a significant effort. I am quite desperate about it (is this what a life of an IT is filled with, searching for one's own stupidities? I guess so).

I added some minor new issues to the 'to do' list.