[GS] Ok so what do Rock Bands actually do?

[ggmoyang]

In my experience, the band level is very important because they affect the concert results. So lower chance of unit dying and higher chance of unit promoting.

My level 4 rock band with Album Cover Art promotion has 4% to lose on wonders (level 6) and gets 6 stars quite often.


What I expect is the game adds a random number to rock band level, and compare that to BaseProbability values to determine concert results.
(the XML says <!--BaseProbability needs to be within 3 and 18. Must be specified from highest BaseProbability to lowest-->)

3~18 sounds like sum of 3 dices. You know what? The chance to get 10 or lower number with 3d6 is 50%.
And chance to get 9 or lower is 37.5%.
8... 25.93%
7... 16.2%
6... 9.26%
5... 4.63%

I think these numbers look familiar.

 

[Victoria]

I think these numbers look familiar

I know why Korea gets +4 adjacency now... thanks!
The random number is 196 BTW, you can see the rolls in the RandCalls.csv in the log directory.

 

[Disgustipated]

My best band was Deep Purple

Good choice.

Seems strange performing at a shipyard is more effective than an amphitheatre.

 

[tiamats4esgares]

I just played a Kupe Rock Band game. I only built wonders fopr tourism and when I got to rock bands I had 18,000 faith which was just enough to win
Here is the before and after pictures.... notice Montezuma was where I got most of my tourists due to rock bands. Alsop notice how much his domestic tourists dropped.
So many rock band deaths... My best band was Deep Purple and they died on 650 albums.... unless you have the +2 promotion for wonders you will always be around the 25% chance of fail.
View attachment 518768View attachment 518767

Also stats show my culture output was not affected by my rock bands
View attachment 518771

Wow. Proof that now faith is more important than tourism lol

 

[ggmoyang]

Wow. Proof that now faith is more important than tourism lol

Or you can just focus on culture and get Cultural Hegemony civic for Hallyu policy, then your elite rock bands will do the job. (And get more gold than you spend)

It's crazy that you can completely ignore other tourism and win by culture.

 

[tiamats4esgares]

@Victoria Would you say that Arabia or Russia are the best civs for culture victories now?

 

[Victoria]

@Victoria Would you say that Arabia or Russia are the best civs for culture victories now?

If you can correctly define best I can try and give you an answer. To try and short circuit that lets talk about a few “bests”

I love playing Gorgo, she is by no means the best but her cultural kick start can be great, however you have to take than into the context of too much culture early means more expensive districts and also a lot of culture that could have been bypassed on inspirations, Rome has the same issue. Gorgo’s real strength comes in an early wildcard slot for writers and a cheap theatre with better adjacency and an extra envoy for everyone you build.
Her ‘good friend’ Pericles however goes one better, with those extra envoys he gets more suze and therefore more culture. Both Greek civs are fast cultural but Pericles is much better.

They can all be pipped by a Kandy game like the current GOTM where a CV is being completed by Kupe in under 100 turns. I think this is the fastest CV for a Kupe but not the most fun. Kupe suits kandy 100%.

If we now look at Ones with UI like Sphinx(been buffed) or all those others... you can spam them for huge amounts of tourism at Flight, it is not just the maori’s. The maoris are sorta not the best at fast cultural but fun cultural, oh yeah they are up there.

Faith makers like Russia are great and Russias Great writers early are great. You have to understand that getting to Cold War early is not easy and that is where Russia and rock bands have trouble. They are still a great civ for CV, Arabia less so.

If you can get to broadcast centres fast then America is OP amd needs no rock.

What is nice about Rock bands is they do not require science, pure culture push, no more relying on computers.

The Chinese are expert fast CV players and use the salted fish strategy... roughly translates to standard theatre or boring simple way.

Is best the fastest or the most fun? When we talk about fun, everyone has a different take, and that is great... play the civs you like.

I’ll try a Russian game after my current Eleanor and see.

 

[Disgustipated]

Do your military units kill rock bands when at war? I just bopped one with my destroyer. It disappeared and I don't see it, so I'm assuming I killed it?

edit: Nope, it seems they do get kicked back to the nearest city. Too bad.

 

[Victoria]

Nope, it seems they do get kicked back to the nearest city. Too bad.

Y, like great people.

 

[tiamats4esgares]

Why would you say Arabia isn't as good as Russia?

 

[Victoria]

Why would you say Arabia isn't as good as Russia?

It is all quite contraversial and people will disagree with civ choices which is why I was not that keen on replying... The Mameluk is really cool and strong and can really help.
However to answer your question Russia gtets lots of great writers through its Lavra. One of the kets to a fast CV is not only getting great writers early but stopping other civs getting them.
With regards to the mechanics of a CV Russia far outstrips Arabia

 

[ggmoyang]

)

I know why Korea gets +4 adjacency now... thanks!
The random number is 196 BTW, you can see the rolls in the RandCalls.csv in the log directory.

196 is such a weird number... 196 is indivisible by 16 or 18, nor is 197.

BTW I did some tests about those BaseProbability number.

I changed BaseProbability of RESULT_OPENING_ACT from 15~5 then checked level 1 Rock Band's displayed final gig chance. Here are the results.

15: 91%
14: 83%
13: 74%
12: 62%
11: 50% (default)
10: 37%
09: 26%
08: 16%
07: 09%
06: 04%
05: 01%

Again, these numbers have the same pattern with 3d6 results. The chance to get 14 or lower from 3d6 is 90.74% and so on.
They might not actually roll 3d6 but it seems that they were at least inspired by it.

And yes, BaseProbability affects concert results, if you change BaseProbability of RESULT_LEGENDS_OF_ROCK to 8 (probably the lowest number allowed) you'll see those new bands doing 6-star concerts consistently.

With 3d6 theory, the chances of concert results look like this:

Sadly, this expectation doesn't fit perfectly with the actual test result in this thread(#47)

 

[Victoria]

Sadly, this expectation doesn't fit perfectly with the actual test result in this thread(#47)

However the quoted death rate does not fit the roll% either so I suspect they may be doing something odd with the result from RandCalls.csv. I love your theory and it is in mind, father in law coming over today so really busy... hopefully wandering dandelion can get some good quick stats out of LUA to give us a better idea,

 

[tiamats4esgares]

I think Arabia vs Russia comes down to how easy it is to keep your religion in your cities without expending too much faith. Probably not easy so Peter > Arabia.

 

[kaspergm]

Oh look, what happened to the mighty Kongo empire, did they get invaded by a foreign army?

No, that was just two Mali Indierockbands that stopped by.

 

[DanQuayle]

I personally don't care much about rock bands, but I did some additional research that may interest some of you...

I have been testing level 0 units only playing in campuses (different targets provide different tourism levels)
I have been logging the results for each call against the random number rolled (rand_calls.csv)
I'll see what the spread is like compared to the rand calls....
View attachment 518600

I checked out my rand_calls.csv file for "Rolling Concert Result"
It appears there are 5 different possible values for "Range": 196, 206, 212, 215, 216
However, it was from a game with rock bands of different levels.
After investigation, there is a pretty clearly relation between rock band level and the range(roll peak).
A level 1 rock band does a 196 roll.
A level 2 rock band does a 206 roll.
A level 3 rock band does a 212 roll.
A level 4 rock band does a 215 roll.
A level 5 or level 6 rock band does a 216 roll.
These numbers are not a coincidence: 216 (all possible rolls for 3d6), 215 (17 and under roll), 212 (16 and under roll), 206 (15 and under roll), 196 (14 and under roll)
It appears they are truncating the tail end of the distribution for lower level rock bands thus giving us very slightly better odds for survival and promotion (compared to a full 216 roll).
It mean the odds would be exactly the same as the one @ggmoyang posted for level 5 and level 6 rock bands since they are doing a "full" 216 roll in these cases.
It would also mean that the disband probability shown in the UI for disbanding are wrong and slightly higher than reality for lvl1-4 rock bands.

Caveat:
I only used 38 observations of my own and the 33 observations posted by @Victoria
The number of observations is particularly limited for higher level rock bands (for example, I have only one level 5 rock band observation).
My results are wrong, because of one result that I cannot explain: a lvl1 rock band 160 roll should theoretically be a 2-star performance and not a 1-star performance. It could be because the cutoffs point do not exactly follow the 3d6 theory, they were manually hardcoded, programmer error or something else entirely...
Maybe it is possible to roll a zero (which I have not seen), in that case the max range values are wrong and one above what they should theoretically be.
The only way to be sure would be to roll (for example in the case of a lvl1 rock band) all 196 possible results (and each result more than once to be sure they are consistent and that no additional randomness is at play, however I have not seen any roll with a different outcome in my observations).

For example, the most interesting rolls for a lvl1 rock band would be
4, 20, 56, 108, 160 (+ or - 1 or 2 to be sure)
These rolls should be the theoretical cutoffs between the different performances.

 

[Victoria]

My results are wrong, because of one result that I cannot explain: a lvl1 rock band 160 roll should theoretically be a 2-star performance and not a 1-star performance

Fantastic work @DanQuayle thank you very much.
It is very annoying how you need to get so many rolls and then you get odd results.
You start questioning if you recorded the results correctly.
I’ll try and get some more testing done at some stage.
They probably put in rock bands, found they were dying too fast and temporarily have buffed them.

 

[DanQuayle]

In this case, it definitely was not a misrecorded result since that lvl1 160 roll was done 4 times in the sample and every time it came out as a 1-star performance.

I attached my observations at the end of this post (import in a spreadsheet software or rename it with a .csv extension)
As I explained, only with more observations can we draw a better picture.
If 10 people play a rock band cultural game with around 30 concerts and report their results here, there is a good chance we can get most of the 196 possible observations for lvl 1.

Methodology
Write on a piece of paper or in a spreadsheet, every time you use a rock band: the turn number, the level the rock band will be playing at (including the promotion bonus if existent) and the number of stars you get for the performance.
After the game, open the "rand_calls.csv" file in the Civ VI My Document folder
Sort by Location and look for for "Rolling Concert Result"
Delete all the lines that do not have a "Rolling Concert Result" in Location
Add the data you have noted/wrote down in the next 2 columns "Level" and "Stars"
The entries are in chronological order (in the case you do multiple concerts on the same turn)

 

[Archon_Wing]

Wait, does different religion count towards all tourism for penalty, or just religious tourism.

 

[Disgustipated]

Is it better to send all your bands to the top domestic tourist leader or spread them around to anyone with more domestic tourists than you have tourists with?