AI enhancement (test games)

[jmas]

Thank you for sharing those ideas--good ones!

You had (earlier) mentioned improving the interface somewhat. Is it possible to add a vital piece of information to the tech screen? Namely: "how much do I need to invest in a given tech field (Computers, Weapons, whatever) to get the maximum interest allowed on this turn"? This number could be expressed as a percentage of the current total research budget (which would make it easier to know exactly where to put the slider for that field) or as an absolute number of RPs.

I know that this number can be estimated in the game as it stands now, by playing with the slider and watching the lightbulb to get an estimate of the current investment, then multiplying by 7.5%. But it seems to me it would be a lot easier for the human if just a single number relevant to the calculation could appear on the screen. Even if all that is possible to program is making visible the precise number of RPs currently invested in a tech, it would make it feel easier for me.

?) to decrease the average damage done to POP, thus changing the ratio of damage dealt to POP/ind to about 1/5 or even 1/6 (now 1/4) - to make the conventional bombing effect clearly distinct from Bio bombing and to 'counterbalance' the increase for the missiles

Correcting missile damage will make bombing stronger, where missiles are in play (though I tend to use missiles only for bases until mid- to late-game when I can build good quality, medium-size missile boats), but from my perspective, when I'm on offense, I will go to great lengths to try to capture a planet with factories intact rather than bomb. I know it can be difficult when at a ground combat disadvantage, but the potential payoff in captured techs is just huge, in my eyes, at least for the first few invasions. So the biggest benefit for me that I see in the change you're proposing is that the bombing-happy AI would be more likely to wipe out the factories at one of my planets and therefore fail to capture any techs from me. Fine with me. Here's to you, AIs: :neenernee

EDIT: The above may not have made it very clear, but it does seem reasonable to me to stop the halving of missile damage. Changing the ratio, so that more factories are destroyed per unit of population eliminated, is what I am saying will make me even more averse to bombing. Meanwhile, as the game is currently coded, AIs will bomb until a planet's pop is less than the number of transports they have on the way--and they always send half the population of one of their planets to invade (and only from one planet at a time, I believe). So they often bomb quite a bit, and the change in the factory-to-pop destruction ratio will have them destroying even more factories. Good for me, but maybe not optimal for AIs. Not that I necessarily know what will work best--maybe this is a place to experiment to see what ratio seems to work best in the modified game? (EDIT end).

??) to halve the beam damage multiplier 30 > 15, at least
???) to alter the strange role of computers in the bombing result

Good idea on beams. About the computers, are you saying you found they DO affect bombardment outside of tactical combat?

[kyrub]

You had (earlier) mentioned improving the interface somewhat. Is it possible to add a vital piece of information to the tech screen? Namely: "how much do I need to invest in a given tech field (Computers, Weapons, whatever) to get the maximum interest allowed on this turn"?

Great, I was about to ask about anybody's interface changes' suggestions.
All I have now is:
- the game can be started with Starmap.exe alone
- the IRC, the new shield etc. improvements can be adjusted on all planets by ratios of 10,20,30% from the research breakthrough screen (was 25,50,75)
- the game is autosaved every turn in save7.gam
- a small rehaul of keyboard shortcuts, making SPACE the universal OK, DONE, GO-TO-MAIN button.

-------------
I must say I don't share your view of 'vital' status of that thing, jmas. Micro-managing the research is superfluous to the game itself, and it is only interesting and 'fun', because it is so hard to do. Once you put on the screen, how much, it will become tedious and boring to click every turn on the 'Research' screen.
Look at the example with 'scout II'. It was fantastic, when we found out first, but nowadays, I dislike having to design that very same ship at the start of every new game. It is boring, not cool, and I will correct the flaw, if I find it.
Besides, I am not sure I have the abilities to do that complicated thing you've asked. But thanks for the ideas, jmas.

-------------
Bombing POP / industry

maybe this is a place to experiment to see what ratio seems to work best in the modified game?

Yes. We don't want to make AI weaker. Any change to MOO is primarily doubtful. I will let the awkward computer impact (yes, there is one) in place as well.

[RefSteel]

Thanks for all the work you've put into this, kyrub! To add some suggestions and perhaps foment discussion, here are my interface requests for MoO - Iin each area, I'll try to list my ideas in order from what I assume is most doable to what I'd love to see but realize might be a pain to program

Fleet selection on Command Screen:
- "A" key would hit the "Accept" button for dispatching the fleet (also applicable for transports and RELOC). ESC would still cancel even if SPACE does so as well.
- TAB would move the cursor to the middle left ("reduce number of ships to be sent by 10%) button for the next ship type in the fleet (starting with the top ship type if the cursor is not on any of them yet).
- When the cursor is in place (i.e. via TAB above) the "<" and ">" keys would have the effect of hitting the "reduce by 10%" and "increase by 10%" buttons for that fleet, respectively.
- When multiple fleets that should be visible to the player are in the same location - e.g. when departing from the same planet - and one of these fleets is selected, the "U" key would cycle to the next fleet "up" (i.e. next-least-recently given orders) in the stack and/or the "D" key would cycle to the next fleet "down".

Yes, I love keyboard shortcuts.

Planets on Command Screen:
- When one of the player's colonies is selected, the "=" key would reduce ECO spending to the minimum clean (I suppose it would have to go into research by default, which is fine by me).
- ECO slider would display "CLEAN" when spending is greater than or equal to the amount needed to terraform the planet up to the maximum level available if that spending will not also add a population unit (just as it now does for Soil, etc.)
- IND slider would display the number of expected factories instead of "REFIT" unless robotic controls are made to work as advertised (built for base cost up to 2* pop, upgrade at stated cost, then add additional upgraded factories) - the latter would be better but, I imagine, extremely complicated to implement.
- ECO spending beyond maximum population would be transferred to the reserve (presumably at 1/2) as advertised.
- When a stargate completes, SHIP spending would be zeroed out, since that's what you spend to make the things, rather than DEF spending, as is done currently.
- Production quantities and ETAs (including "CLEAN") would correctly take outbound transports into account (instead of just reducing assumed available BC by one per transport as it does now). Ideally, it would also take expected pop growth into account, but that might be asking a bit much....

Planets Screen:
- When reserves are transferred to a world, the screen would not revert to its default position, but remain in the position reached via the up and down buttons at the bottom of the screen.
- "Page Up" and "Page Down" would have effects equivalent to hitting the up and down buttons at the bottom of the screen.
- The currently highlighted world (if any) would always be on the screen in its default position (e.g. when you hit "P" from the main command screen) and always at the top unless fewer than 11 planets are below it on the list.
- Planets screen would be sortable (by clicking on the top of the relevant columns) by planet number, name, population, factories, etc. Yes, I know I'm dreaming.

Fleets Screen:
- "Page Up" and "Page Down" would have effects equivalent to hitting the up and down buttons on the right side of the screen.
- "Home" key would bring you to the top screen of the fleet list; "End" key would take you to the bottom screen.
- Fleets would be sortable (by clicking on the top of the relevant columns) by location (i.e. as at present) or by ship type (largest number of that ship on top, ties broken by location or better yet, by most recent sort). Yes, I know I'm still dreaming!

Okay, that's a lot of stuff, especially considering that I really enjoy playing the game as is. Please don't feel obligated to respond to (never mind actually code!) all of these, but hopefully I'll inspire some ideas.

Anyone who reads this of course, please feel free to point out reasons that some or all of my interface ideas are foolish - or to add your own that might be better (and easier to implement) than some of mine!

[kyrub]

Refsteel,
your contribution is magnanimous. Thanks for ideas.
Several corrections will be quite easy. I repaired the stargates - shp% clumsiness in two minutes.
Some are very, very hard to do. In fact, everything that requires inflating the game code (inserting a lot of stuff), is beyond my capacties. I am just an amateur, no IT.
I almost fell from my chair while reading your fleet overview dream

- IND slider would display the number of expected factories instead of "REFIT" unless robotic controls are made to work as advertised (built for base cost up to 2* pop, upgrade at stated cost, then add additional upgraded factories) - the latter would be better but, I imagine, extremely complicated to implement.

1) I did not know. This is most interesting. It actually looks more like a bug than interface, no?
2) It seems repairable, because the code is definitely there, just not working.
3) What exactly is the 'advertised cost' in OSG, please?

[RefSteel]

Cool - as I hope I conveyed, I didn't really believe that all my suggestions could be implemented without an insane amount of work, nor that insane amounts of work should go into it; I included some of the crazy ones just in case I was wrong about how crazy they were. For a game I already enjoy, anything that is possible among my suggestions is just icing on the cake. Also, amateur though you may be, you're way ahead of me; I wouldn't even know where to start with the code for this game.

Yeah, the effects of RC on factory costs are definitely bugged. My interface suggestion was simply that if it isn't reasonably possible to fix the bug, the IND slider display would hopefully be changed to reflect the bugged reality. Fixing the bug itself would obviously be better, so: I don't have the strategy guide, but here's my best current guess on the basis of in-game graphics, things I've read here, and Sargon's save file mapping:

Current (bugged) system for non-Meklar:
(Meklar already work correctly, as RC refit costs aren't supposed to affect them in the first place, and in fact do not)
- Each population unit can work one factory for each level of the race's RC regardless of the planet's (or individual facotries'?) RC level (at game start, race RC = 2)
- Cost to build any factory is always eqaul to the race's IIT level times half the race's RC level, even if it's the first factory of a new colony for a race with RC7.
- There are no refit costs, ever, in spite of the IND display.

Advertised system for non-Meklar as best I can understand it:
- Each population unit can work a number of factories equal to the RC level of the planet. (Resets to 1* on conquest?)
- Cost to build each factory up to (current population x current planet RC level) factories is equal to the race's IIT level times half the planet's RC level. If (race RC > planet RC) then spending beyond this number of factories contributes to "refitting" the planet's existing factories instead of building more.
- Cost to refit a planet's factories is equal to the race's IIT level times half the number of factories on the planet. Completing this project increases the planet's RC level by 1 (up to a maximum of the race's RC level obviously).
- If I'm interpretting Sargon's savefile mapping correctly, I would guess that "factory partial (BC)" might also be intended to serve as "refit partial (BC)" such that a partially-completed refit project might turn into completed lower-RC factories for a planet with growing population that has not yet completed its latest refit.

*- Down to 1, not 2, on conquest, according to Sargon's save file mapping. This might mean newly-conquered planets are supposed to need refitting for the conquerer before they even reach the default RC2(?) which actually would be kind of cool/appropriate if true; it would mean you can't turn alien factories around at full strength without paying to fix them up for your own race.

I hope this answers your question - and is at least slightly comprehensible!

[kyrub]

Thanks. 2-in-1 buggie.

1) They forgot to multiply the refit cost by the number of factories. This makes the 'refit' a one turn show-up with a negligible effect.

2) They put the factories' cost straight at the race level.

The comical thing is that they cancel each out, in a way. At the end of the day, you pay the same.
Although you have the factories later than you should.

Both can be corrected, but 2) will have a big impact on the game.
(I could also make the factories' cost basic only for the basic level of RC=2 thus protecting the undeveloped planets - the way you suggested in your previous post, with a meaning: The base factories control is not dependent upon robotics, but on manpower. There was something strategically interesting in current status: you had to choose whether it is the right time ro research RC and risk stopping the growth of your empire. This effect will be much smaller now.)

[jmas]

@RefSteel and kyrub: Good work, guys. I think I would prefer having a lesser significance (which kyrub mentioned) to the question of whether to acquire robotic controls at a particular time. It may not be so bad when one is considering whether to research them or not, but it seems pretty unfair that one could be set back by discovering them at an artifacts planet, and given how rare it's been in my games to get a meaningful tech stealing opportunity through espionage, I would like it to be clearly a good thing if I managed to steal robotic controls.

@kyrub:

Regarding adding information relevant to interest to the tech screen:

"Vital" might be a bit strong of a word, however I believe the ability to earn 15% interest--compounded every turn--is very significant for tech, so "superfluous" seems a bit strong of a word on the other side. You could say the "truth" is somewhere in the middle. I think the mechanics of interest could be a LOT easier for less experienced players to understand (plus, having it mentioned on the tech screen alerts players to the possibility of earning interest, and evidence that it is being earned, without having to read the manual, not to mention all the threads here about how things really work ), and it could be easier even for more experienced players to use on a regular basis without so much calculating. I see your point about the Scout 2 ships, however I disagree that it's the difficulty of calculating interest that makes it "interesting" (and I do realize you may have meant "interesting" with tongue-in-cheek). I think I tech pretty well, but I've always been impressed by sargon's teching, and I think his success with it is at least partly due to him managing the interest better than I do. Which leads to the following point: players of course would not be not obligated to do calculations or even concern themselves with interest at all if they don't care to, just as it is now.

Well of course, at the end of the day it is your decision, and figuring interest is something that is possible to do in the game currently; I just wanted to make a case for adding a piece of information.

[kyrub]

I think I would prefer having a lesser significance...

Okay, I'm on it. It will take some time.

ECO spending beyond maximum population would be transferred to the reserve (presumably at 1/2) as advertised.

The extra ECO is not waisted, actually. It goes automatically either to IND, or to TECH in case no more factories can be constructed - and the input is taken in account directly, so you don't lose anything.
Personnally, I am quite content with this design. What do you think?
Two things to change:
a) the pop-up text
b) there is a small bug in the code - if you leave only one click of waisteful ECO (4%), it is not automatically corrected. For anything more it pops up.
The fix is done & ready.

I believe the ability to earn 15% interest--compounded every turn--is very significant for tech... I just wanted to make a case for adding a piece of information.

Sure, jmas. I pushed my argument too far. And there could be some kind of info about the tech input, although maybe not the exact %.
I generally feel that Moo I should give us more info about what is running behind the scene. That could be another theme for the twillight patch, the research screen, the council screen, the detailed planet screen (!). Any idea for what could be where is welcome. Unfortunately, right now, the displaying functions remain mysterious to me... I could maybe swap some existing info with some new one. That's quite poor, isn't it?

I think I tech pretty well, but I've always been impressed by sargon's teching

Man, I privately hold sargon for a Psilon who infiltrated the human internet

[RefSteel]

On RC:
The two bugs actually don't quite cancel each other out. They do so for all future factories, but consider: If you start with a size 100 world and 200 factories, then get RC3, under the bugged system, you pay for 100 factories at (1.5 x IIT) each, for a total of 150 x IIT. Under the advertised system, you pay the same cost for new factories, but must first pay (100 x IIT) to refit your existing factories. So at IIT 10, you're paying 1000 BC extra, and doing so before you can even build your first new factory! Thus, the corrected system is much more expensive at developed worlds, but (though still equally or more expensive in the end) speeds early factory construction on less-developed colonies.

In any case, I agree that it's best to fix the bug and so reduce the significance of timing RC acquisition, especially because the bug applies equally to AI and human players, and the AI is not capable of intelligently making this currently-important decision.

On extra Eco going to IND or TECH:
I had no idea that it did this! I'll test it in-game at some point just so I can see it happening with my own eyes, but this sounds great; in some ways, it's a better solution than sending the excess to reserves. [EDIT: Only it doesn't seem to be working. See following post.]

On tech screen interest displays:
Wellllll ... I wouldn't be a huge advocate of actually doing this, especially if it's too much work, but I can think of a way to use existing design display functions to show whether you're taking full advantage of your investment in each field. Currently, when Tech Level in a field is at 99, the lightbulb display shows "MAX" instead of a bulb. "MAX" could be repurposed to also mean "This year's spending in this field is greater than or equal to the maximum amount that would be tripled." I wouldn't want that done though unless there were a way to turn the option on and off (for an individual bulb, e.g. by clicking on it, or for the whole screen) and I'll bet that would not be easy.

As for Sargon's teching abilities, it's not just a matter of knowing the interest in each field. Sargon macromanages his tech spending at the imperial level, deciding when to start and/or freeze research spending in each field on the basis of expected tech needs, the costs of each tech, expected imperial research capabilities, etc. In between techs coming in (as I understand it from his description in another thread) he doesn't have to touch the tech sliders - he just remembers how much he spent the previous turn empire-wide, and increases tech spending over that number by 22.5%. I have tried micromanaging tech for maximum interest turn-by-turn, using a spreadsheet to keep track of the numbers, and still fallen well short of what Sargon accomplishes; he's not just maxing his interest each turn, but planning his tech spending well in advance for all fields on the basis of that interest and the needs of his empire!

[RefSteel]

Hmmmmm. I started a game as the Sakkra, scrapped all ships, gave full reserves to Sssla, set slider to full ECO, set tech slider to full Planetology, and hit next turn a bunch of times. (I later also tried again with maxed factories.) In spite of doing some micromanagement to make sure I'd get massive overflow when population maxed, I never got even a single new factory, nor the 2 RP needed to open the Planetology field, nor anything in reserve, from ECO spending. The slider for the following turn of course switched over to 100% IND or TECH except for the minimum needed for CLEAN, but not the ECO spending itself. Perhaps the game intends to put ECO overflow into factories/tech, but somewhere along the way, the BCs fall into a black hole.