turn time since patch

[Tatran]

I got 18.82 seconds before the patch and 24.06 afterwards (both numbers are average of three test runs, using i5-3570K@4.2GHz). Not as drastic increase as yours, but it's there.

Could you be more specific? What stage of the game (early, mid or late) and what map size (standard, large or huge)?

 

[Janskey]

Could you be more specific? What stage of the game (early, mid or late) and what map size (standard, large or huge)?

I, as well as the OP, used the AI benchmark included in the game. You can find it in the main menu.

 

[Hajee]

I got 18.82 seconds before the patch and 24.06 afterwards (both numbers are average of three test runs, using i5-3570K@4.2GHz). Not as drastic increase as yours, but it's there.

Thats good to know that others are seeing the same thing. Mine is more drastic, as I dont have a K processor, I sit at 3.4, and you have a stable 4.2 OC.

I would start to assume then that the winter patch has some bug built into the engine. One thing I have noticed is there is a Hang time in between turns. The decrease in performance would mean they were working on something in the engine but caused an error some where.

 

[Gub]

used the AI benchmark included in the game. You can find it in the main menu.

That is an interesting feature, that I wasn't aware off, it suggest that they have robust performance profiling tools. They certainly would have better tools and use dedicated test machines before any release, and thus aware of this data and consider it runs "well" on its stated minimum system.

The problem is that we (or at least I) have no idea what this test entails ( is it some average or absolute worst case scenario e.g. playing huge maps, with specific layout, max civs, hundreds of units ...) nor what it represents ( what 19->24 means e.g. will a player on normal map scenario notice anything?). To try and make sense of that, it is best that we provide as much info as possible (e.g. to avoid any outliners) so its good that you provided extra info on your system. Also i seen on reddit some very nice\informative test with educated guess that might be of interest to you.

Overall developers always need to balance playability with usability. I don't know what the increase in this 'AI benchmark' represent, whether they added or broke something, only that stating its bug as mater of fact is foolish.

 

[Hajee]

That is an interesting feature, that I wasn't aware off, it suggest that they have robust performance profiling tools. They certainly should have better tools that are run on dedicated test machines before any release, and thus aware of this data and consider it runs "well" on its stated minimum system.

The problem is that we (or at least I) have no idea what this test entails ( is it some average or absolute worst case scenario e.g. playing huge maps, with specific layout, max civs, hundreds of units ...) nor what it represents ( what 19->24 means e.g. will a player on normal map scenario notice anything?). To try and make sense of that, it is best that we provide as much info as possible (e.g. to avoid any outliners) so its good that you provided extra info on your system. Also i seen on reddit some very nice\informative test with educated guess that might be of interest to you.

Overall developers always need to balance playability with usability. I don't know what the increase in this 'AI benchmark' represent, whether they added or broke something, only that stating its bug as mater of fact is foolish.

The benchmark it self only has 4 civs and few city states.
The turns them self does not take along time to happen, maybe 15 secs to complete. It is the time between the turns that is concerning this takes anywhere from 15 to 20 secs to happen....ie start new turn

 

[bladex]

45.93 mine came in! what's going on firaxis?!

 

[c4c6]

Startup finished in 17.142s (firmware) + 5.181s (loader) + 20.637s (kernel) + 3.307s (userspace) = 46.269s
What the hell is going on playtex???

 

[Mustakrakish]

Well I see... was wondering what the hell happened to performance. Seems like I'm not playing the game till next patch.

 

[orasis]

Same for me, I just do not run DX12. Too new, still buggy.

 

[bladex]

What the hell is going on playtex

who's playtex?

 

[tekjunkie28]

For me the turn time only increased by 1-2 seconds. I was getting 15.7 before and now im getting about 16.5-17.2 seconds. Directx version has to difference and the DX12 version works flawlessly for me.

 

[dstar]

As stated above, the only variable considered in a given calculation is the random seed, the rest is scripted. IE, the seed generates a random number which is applied to a table (with static modifiers) to see where a unit moves, whether a civ offers peace etc. All you can do is script the actions better and change the % chance of a given decision. Nothing else is possible under this system, which can't/won't be changed now. So expect refinements but not miracles. Given the additional movement restrictions present now I doubt they could even reach the level of Civ 5 without a *lot* more effort though.

Wait, are you actually saying that they really _did_ make a multiplayer-only game and just slapped a something that looks enough like an AI on it that they could get away with claiming it also had single-player?

Because if what you're saying above is correct, that's _exactly_ what they did.

 

[need my speed]

Wait, are you actually saying that they really _did_ make a multiplayer-only game and just slapped a something that looks enough like an AI on it that they could get away with claiming it also had single-player?

Because if what you're saying above is correct, that's _exactly_ what they did.

That isn't necessarily a bad thing; Civilization IV was designed as a multiplayer game, because it made it easier to test and because it is far easier to convert a multiplayer game to a single player game. Of course, the intention, then, was to make a proper single player game, and you're implying that this isn't the case for Civilization VI.

 

[TheMeInTeam]

Why?
Of course, if the AI shows just exactly the same "performance" (functionality, not speed), it is worse to waste now 1/3 of the computing power. But who knows, what the AI can, what is achieved (or not) exactly?

Theoretical question to all: Would you be satisfied / satisfiable with waiting 8 MINUTES between turns provided, the AI plays 'ok' in your standards?

You get diminishing marginal utility per AI time increase.

The answer to 8 minute turns is no. The answer to 30 second turns is no for most of the game.

I want to play the game, not wait on the AI. 150x30s turns is an hour and 15 minutes of doing nothing but waiting on the AI to move, should this last 150 turns. Civ 6 is awful enough with during turn tedium of required actions w/o required decisions. Making players wait even longer to make choices that matter makes the game worse still.

The AI is not 50% better than pre-patch or whatever either, or any proportion better befitting significantly increasing wait time. This was a problem in civ 5 too, just like UI was a problem back then (and before). Waiting to play a game is not the same thing as playing the game. It seems Firaxis forgot this after Civ 4, and it makes a lot of the other mechanical considerations moot/unfair comparisons.

 

[Tatran]

150x30s turns is an hour and 15 minutes of doing nothing but waiting on the AI to move,.

This is another serious (growing) problem in the civ series. Time goes by really fast without noticing it and what has the human player done during that time?
Anything beyond 30 seconds is unacceptable, even playing large or huge maps. Turn times should be 5-15 seconds.
Players losing hours doing nothing during one game. And the total number of wasted hours becomes really scary when multiplied by the amount of played games.
People should check their turn times with a stopwatch like I did. And 1 minute per turn around turn 600 at epic speed (standand or large map?) was a reason, besides the bad and boring game itself, to delete whole civ5.

 

[c4c6]

You get diminishing marginal utility per AI time increase.

I know. I know the concrete numbers of exponential growing combinations. I know the general life experience, that 80% of anything I do is done in 20% of the time and the remaining 20% of anything need 80% of the time, because they contain the problems & exception handling etc. etc. etc.

The AI is not 50% better than pre-patch or whatever either, or any proportion better befitting significantly increasing wait time.

All I understand is, that a process needs significant more cycle time now. I have no idea what was achieved before & after the change or how buggy both implementations are _and_ how that compares to what has to be achieved in order to provide an 'acceptable' gaming experience.

The answer to 8 minute turns is no. The answer to 30 second turns is no for most of the game.
I want to play the game, not wait on the AI. 150x30s turns is an hour and 15 minutes of doing nothing but waiting on the AI to move, should this last 150 turns.

Ok, interesting. You perceive the waiting time as "doing nothing", as a complete waste of time!
Of course I want to play the game too without pauses, but given a significant stronger playing AI, I probably could use the 'in between'-time being freed up for doing something else, especially if I could prolong the pause along my needs and return to the game whenever I want. (Probably thats a leftover from old PBEM days.)

Another aspect is the "even playing field". When playing I like the illusion, that the computer players are playing the same game. But I know, they simulate and simply try to entertain me.
So I have NO PROBLEM at all with "cheating" (extra resources, magic view etc.), alone the word is wrong, "they" just have other rules. One of them is that they have much lesser time.

I play slowly and need many minutes per turn (enough hectic on the job). If done right, I could do in the pauses "neutral" city-management ... Also could the animations of the first civ be shown while the moves of the second are calculated and so on.

I do not like waiting. I don't want to "defend" long waiting times (I'm in no way responsible or something).
And I know, that not all can happen as quickly as I wish.

 

[bladex]

i can play the game fairly quickly with 6 civs on a large map but as soon as i move past that onto either a huge map or more civs the turn time is horrendous..

 

[Hajee]

The game is just poorly optimized. Even out high clock speed it does not reduced the turn time
START at 8:15

There you can see even with the new kaby lake cpus (which is only a high clock speed then skylake) you can see that it really does not make much of a difference.

I still think the whole issue is the post/pre turn lag. when you bench mark the AI there is this lag in between turns. it would nice to know what the game is think in that time

 

[Janskey]

I ran the AI benchmark again with the new 1.0.0.110 patch, three times like before and got an average of 20.22 seconds (was 18.82 with patch 1 and 24.06 with patch 2).

 

[Mustakrakish]

Why?
Of course, if the AI shows just exactly the same "performance" (functionality, not speed), it is worse to waste now 1/3 of the computing power. But who knows, what the AI can, what is achieved (or not) exactly?

Theoretical question to all: Would you be satisfied / satisfiable with waiting 8 MINUTES between turns provided, the AI plays 'ok' in your standards?

In civ IV turn times are almost instantaneous, while AI performs very much OK by my standards and very good with K-Mod. All this while playing on bigger maps with more civs, about twice as many cities and 10 times the units. What makes VI so slow? - Question to all. Civ VI is a very pretty game and I want to believe it has potential to be a good game after final expansion and maybe a few good mods. But, and this is a big BUT for me, even if the AI will become more or less tolerable after all of this (it is not now) while turn times keep the same as now (which is very slow for me), that potential is completely lost. For me it's like owning an expensive cool car without really being able to afford gas for it (which it eats like mad) or the taxes. Useless.

For me it was always number one problem and reason for stopping playing eventually (as with V as well as with VI), how game runs. Second being UI, but that's unrelated to the topic. I appreciate the pretty graphics, but I care not for them, If I had a noticeable decrease in turn times by playing on the lowest settings, I would, never looking back. Since I have a good graphics card, it makes no difference whether I play on max or low (it did before I upgraded the card).

So the answer is no I wouldn't, I'm not even now. I'm not not playing it now because it's boring and AI is useless (which it is, but these 2 will inevitably change), I'm not playing because I find turn times simply unacceptable to bare when playing on standard size or higher and / or 9 or more civs. Especially when compared to what I get out of it. Nor should I. Poorly optimized game.