AU Mod: Min / Max Science Turns

[joncnunn]

I think that the current limits on Min/ Max turns to reserach science should be set to 1 min turn and 1000 max turns.

This is because:
1. The AI knows nothing about how to use the 1 beaker @ 50 turns science exploit to get expensive techs at very low cost. (Most notably, Writing when either Commerical or Seafaring, but also several other techs on huge maps and/or at high difficulty levels)

2. The AI also knows nothing about how extra science past 4 turns does no good, and mindlessly sets it as high as possible without running a deficit under Despotism, even if currently in a GA on a high difficulty level when without the four turn min limit it could research faster, but with it, it is throwing away tons of money.

My own testing in the ancient era shows that this does indeed totally elimatinate the one beaker exploit. (330 turns for Writing on turn 1 as a Seafaring civ on a huge map on Monarch level)

[Modo44]

The maximum number of turns need not be so huge. I did a similar thing, and it was enough to change it to 80~100, to gain this effect. Such a chage would leave a strategic decision open, while a 1000-turn limit would outright kill it. Plus, of course, it is closer to stock.

As to the lower turn limit, this is highly level-dependant. It would help the AI on higher difficulty levels, where a human player seldom reaches the 4-turn mark. However, it would favor the human on lower levels, because there he is the one making quicker research. A double-edged sword, I think.

[nbarclay]

The 50-turn cap plays a valuable role in limiting how far behind players fall in the research race if they get stuck in bad starting positions. The only real problem I see with it is that its calibration isn't geared well for the relatively slow early-game research pace on huge maps, which means Seafaring and Commercial civs can often exploit it in researching Writing from relatively normal starting positions (as I've done myself on quite a number of occasions). But since we can't change the limit for huge maps without the change affecting all map sizes, I'd rather leave it alone.

Regarding the 4-turn lower limit, do you have evidence that AIs are hurt by it in more than a very tiny number of very special cases, if at all? Unless there is a significant problem, I don't like the idea of making a change.

Nathan

[joncnunn]

In a bad starting position, there is still the war mongoling option with lots of whatever units you do have as early as possible to improve your starting position. Or at least ruin your neighbors starting position.

Not to mention in SP games the restart with a new map option.

Well, it depends upon how the AI sets it's slider on during cash-rush govts [particularly Republic / Democracy combined with lots of Libraries and a GA. Under non-cash govts its harder to determine how the AI sets science, but there is a clear danger in a GA in the middle ages of wasting beakers by any well developed empire across map sizes.

Originally posted by nbarclay
The 50-turn cap plays a valuable role in limiting how far behind players fall in the research race if they get stuck in bad starting positions. The only real problem I see with it is that its calibration isn't geared well for the relatively slow early-game research pace on huge maps, which means Seafaring and Commercial civs can often exploit it in researching Writing from relatively normal starting positions (as I've done myself on quite a number of occasions). But since we can't change the limit for huge maps without the change affecting all map sizes, I'd rather leave it alone.

Regarding the 4-turn lower limit, do you have evidence that AIs are hurt by it in more than a very tiny number of very special cases, if at all? Unless there is a significant problem, I don't like the idea of making a change.

Nathan

[joncnunn]

Well, that range does on a normal size map largely eliminate the cheap Writing tech. It's still a huge discount on huge maps.

Yes, min turn limits are a double edge sword affecting both humans and AI, I have found myself on a no-optional tech mod wishing to research a bypassed middle age tech faster than 4 turns when I clearly had the economy to do so in 2 turns becuase my slider was at 40% and still needed tweaked down to 30% on the last turn.

It should also be noted that neither Civ II nor SMAC had min/max reserach turn features. You could research Transcedant Thoughts 5, 6, and 7 in a single turn on the Transdendant difficulty level if your science output was high enough. And in the other direction, you could set science so low it would have taken 100s of turns to finish a tech.

Originally posted by Modo44
The maximum number of turns need not be so huge. I did a similar thing, and it was enough to change it to 80~100, to gain this effect. Such a chage would leave a strategic decision open, while a 1000-turn limit would outright kill it. Plus, of course, it is closer to stock.

As to the lower turn limit, this is highly level-dependant. It would help the AI on higher difficulty levels, where a human player seldom reaches the 4-turn mark. However, it would favor the human on lower levels, because there he is the one making quicker research. A double-edged sword, I think.

[Pfeffersack]

I like the idea of getting rid of both the upper and lower limt.It just feels like am unrealitsic barrier to me (lower level) and makes the single-scientist exploit (IMO, it is) possible(upper limit).I don't think upping it would kill stratgic decisions, you can still get the tech in less turns if you create more scientists.At the moment you have no benefit of creating say three scientists instead of one.
And one-turn research would make future tech a valuable option, especially if combined with lower future tech cost (would suggest to reduce it for the AU mod, since it is generally rated as completly useless, because you can get always a better score with maxing out population and happiness)

[lockstep]

The single-scientist strategy comes at least close to an exploit, and I like the idea of setting the maximum turn no. to 80 or 100.

I'm against a minimum turn no. of 1 or 2 - making it harder to get a very large tech lead IMO improved Civ3's game balance over SMAC's, even if I thought the constraint to be somewhat unrealistic in the beginning. However, if a 4-turn limit causes the AI to waste large number of beakers during a GA, it might be advisable to set the minimum turn no. to 3.

[joncnunn]

A 3 turn min would be sufficent to almost entirely eliminating the danger of AI wasting a large number of beakers during a GA.
(Possible exception of tiny map, but that's not very popular)

My own mod needs a lower min in the 1 / 2 range because I've also made every single tech required and it appears that even the AI is smart enough to bypass the Iron Clad tech.