Best Speed to play at?

[cd0m]

What is the ideal speed to play this game at?

 

[Somebody613]

What is the ideal speed to play this game at?

This is very subjective.
But I'll just say that I myself only use two speeds in Civ: the fastest and the slowest, depending on what I'm aiming for in this particular run.
To be honest, I don't think I've ever used any other speeds after I became a good Civ player, lol.
Yet again - it's subjective, so you must find your own answer all on your own, lol.
A bit of info, though: "the slowest" means "tens of thousands of turns", and "the fastest" means "barely a few hundreds".
Yeah, that's a HUGE range of speeds, indeed.

 

[Toffer90]

What is the ideal speed to play this game at?

Long and Epic are good speeds to start out on. Avoid Ultrafast as it is mainly there for debugging and testing purposes, eternity is for certain enthusiasts.

 

[Somebody613]

Long and Epic are good speeds to start out on. Avoid Ultrafast as it is mainly there for debugging and testing purposes, eternity is for certain enthusiasts.

I totally don't understand the point of "debugging" at a different speed than what you suggest to actually be played by others.
And it's not exactly much faster than the preceding speed, so what's the point anyways?
Conversely, if I want to "test" an idea, I simply give myself "cheat bonuses" of the type I'm "debugging" at the moment - and then "my personal game speed" can be anything I want, lol.

 

[Toffer90]

I totally don't understand the point of "debugging" at a different speed than what you suggest to actually be played by others.
And it's not exactly much faster than the preceding speed, so what's the point anyways?
Conversely, if I want to "test" an idea, I simply give myself "cheat bonuses" of the type I'm "debugging" at the moment - and then "my personal game speed" can be anything I want, lol.

The point is that it is so extreme that it push everything down to far below 1 turn (but it will still be 1 turn as a minimum), the rounding errors will be way more significant and skew the game away from the intended experience (of course, multiple production and research mitigate this issue on some fronts when it is used actively and late game will be much more normal experience than prehistoric is). I.e. balance nuances will be somewhat washed away due to integer values being pushed down so far from the normal values making rounding significance bigger.

I'm not saying ultrafast should not be played by players, by all means play it to your hears content, I'm saying we most probably wouldn't have had the speed in the mod at all if it wasn't for it being quite useful for us impatient developers. Blitz is already considered a superfast gamespeed, so I wouldn't recomment ultrafast to new players at all because it is so far away from the intended experience, e.g that one can produce major unit stacks in 1 turn to defend against a unit stack the AI prepared for a weakly defended city, where the defending stack made in 1 turn means the AI would have needed 5x more units in its stack to even challenge the city.

 

[Somebody613]

The point is that it is so extreme that it push everything down to far below 1 turn (but it will still be 1 turn as a minimum), the rounding errors will be way more significant and skew the game away from the intended experience (of course, multiple production and research mitigate this issue on some fronts when it is used actively and late game will be much more normal experience than prehistoric is). I.e. balance nuances will be somewhat washed away due to integer values being pushed down so far from the normal values making rounding significance bigger.

I'm not saying ultrafast should not be played by players, by all means play it to your hears content, I'm saying we most probably wouldn't have had the speed in the mod at all if it wasn't for it being quite useful for us impatient developers. Blitz is already considered a superfast gamespeed, so I wouldn't recomment ultrafast to new players at all because it is so far away from the intended experience, e.g that one can produce major unit stacks in 1 turn to defend against a unit stack the AI prepared for a weakly defended city, where the defending stack made in 1 turn means the AI would have needed 5x more units in its stack to even challenge the city.

Sorry, but I totally didn't understand how this helps you DEBUG stuff any better than Blitz?

 

[Toffer90]

Sorry, but I totally didn't understand how this helps you DEBUG stuff any better than Blitz?

Basically, running 200 turns AI autoplay takes half the amount of time as 400 turns take (which is the blitz equivalent to 200 turns on Ultrafast). Eternity would be 8000 turns to see the same game development as 200 turns on Ultrafast (so no developer in their right mind would run test games on eternity unless they want to have a fun little game while testing (which is not exactly efficient work but more play mixed into it)). For some specific testing and debugging I chose blitz or normal speed, but I never go above that, 90% of the time I use ultrafast. Same is true for playing manually to see turn by turn how things develop (better it develop fast when debugging so there are less empty turns with nothing relevant to observe). There are many reasons why developers can debug and test stuff more efficiently with a tool like a ultrafast gamespeed. Main reasons would be to observe AI behaviour and performance.

 

[Somebody613]

Basically, running 200 turns AI autoplay takes half the amount of time as 400 turns take (which is the blitz equivalent to 200 turns on Ultrafast). Eternity would be 8000 turns to see the same game development as 200 turns on Ultrafast (so no developer in their right mind would run test games on eternity unless they want to have a fun little game while testing (which is not exactly efficient work but more play mixed into it)). For some specific testing and debugging I chose blitz or normal speed, but I never go above that, 90% of the time I use ultrafast. Same is true for playing manually to see turn by turn how things develop (better it develop fast when debugging so there are less empty turns with nothing relevant to observe). There are many reasons why developers can debug and test stuff more efficiently with a tool like a ultrafast gamespeed. Main reasons would be to observe AI behaviour and performance.

What do you call "debugging", though?
I mean, I get the part about fishing out actual outright programming ERRORS, yeah, but if you are talking about pure gameplay... that ain't gonna work like this.
Simply said, hunting and leveling units on Eternity works VERY differently from Ultrafast - and the difference is technically unpredictable to begin with.
Each additional turn (raw number) literally lets you "roll" for another "subdued animal" and/or "free promotion" - and these DON'T scale with speed.
Meaning, if your game follows the same "events", you'd have EXACTLY the same number of promotions AND animals, REGARDLESS of the game speed in use.
But "100 turns in Eternity" is EONS away from "100 turns in Ultrafast" - while leaving you with the SAME "achievements" in promotions and animals (in the copy-situation case).
And even in Blitz, DOUBLE the amount of turns technically means the potential for DOUBLE the amount of promotions and animals.
Which still makes quite a difference, even so.
Hence, really, I don't see how one speed SCALES to another speed for "debugging" the actual gameplay - it just doesn't.

 

[Toffer90]

Hence, really, I don't see how one speed SCALES to another speed for "debugging" the actual gameplay - it just doesn't.

Well I don't have to explain to you the benefit of having a very fast gamespeed to test out changes and see if it works as intended or not, e.g. see how a specific code change to AI affect the AI in the short and long run.

But here's an example, let's say I modified the the AI with the intent of making them better at getting folklores from hunting; well sometimes it is hard to gauge the effect of the code changes without having an initial turn 0 save start and run e.g. 100 turns without the code change 10 times from the same save; and then 100 turns with the code change 10 times from the same save. If I were to do the same test on Normal game speed I would have to run 400 turns 20 times which is 8000 turns, but the fastest gamespeed this test is only 2000 turns of testing.

Logging can often be implemented to observe the actual thing one is interested in so one can look over the log to see if AI were more effective at city conquest with some code changes than they are without the code change after having run a couple different games multiple times with and without the code change and look at the average effect of the change.
Some times I just want a game that has reached a specific tech point in a natural game so I run AI autoplay in 100 turn intervals until the tech has been reached and then I can manually play from there and observe AI turn from turn and see if some tech unlocked mechanic works as it should for both the player and the AI.

That's my last attempt to explain to you why we added the ultrafast gamespeed for development reasons.

 

[Somebody613]

Well I don't have to explain to you the benefit of having a very fast gamespeed to test out changes and see if it works as intended or not, e.g. see how a specific code change to AI affect the AI in the short and long run.

Dude, we aren't fighting, eh?
I'm just saying that CERTAIN (obviously not ALL) features utterly fail to scale properly via "turn number increase" - while some OTHERS do it absolutely alright.
That's it, nothing more.
Never mind, okay?

 

[Toffer90]

Dude, we aren't fighting, eh?
I'm just saying that CERTAIN (obviously not ALL) features utterly fail to scale properly via "turn number increase" - while some OTHERS do it absolutely alright.
That's it, nothing more.
Never mind, okay?

^^ We are not fighting, but you were challenging my statement about ultrafast being added in for testing purposes, it's usefulness for mod development, I couldn't not react to it.
I guess the conversation started around whether ultrafast was a good speed for newcomers to start out with (which I for several reasons don't think it is).

 

[ranagrande]

The most useful thing about faster game speeds for this is if you're testing something that comes into play in later eras. That way you can get to the part you want to work on sooner.

 

[tramwajg]

when I first played with mod, I used epic, I think it's good balance I wish it was possible to have varying speed during game