Randomized Leader choices

[Grimus]

Question: When the computer chooses your "random" leaders (Note: I'm not talking about random personalities), does it choose your opponents by civilization or by the actual leader?

Each have their downfalls...

If the game chooses by civilization:

...then that means that civs with multiple leaders will have much less of a chance than other leaders of appearing in your games. Example: Montezuma (single leader of the Aztecs) will appear much more often than say Stalin (1 of many Russian leaders).

If the game chooses by leader:

...then that means some civs will have a much greater chance of appearing in your games over other civs with only 1 leader. Example: America has a much greater chance of appearing in games than does Japan.


This is why I think most other civilizations need more representation by other leaders to balance this particular aspect out.

 

[Gaius Octavius]

Which is the lesser of the two evils in your opinion?

 

[azzaman333]

Question: When the computer chooses your "random" leaders (Note: I'm not talking about random personalities), does it choose your opponents by civilization or by the actual leader?

Each have their downfalls...

If the game chooses by civilization:

...then that means that civs with multiple leaders will have much less of a chance than other leaders of appearing in your games. Example: Montezuma (single leader of the Aztecs) will appear much more often than say Stalin (1 of many Russian leaders).

If the game chooses by leader:

...then that means some civs will have a much greater chance of appearing in your games over other civs with only 1 leader. Example: America has a much greater chance of appearing in games than does Japan.


This is why I think most other civilizations need more representation by other leaders to balance this particular aspect out.

Bolded what actually occurs, underlined what I would prefer.

 

[Grimus]

Ah, I see... thanks azzaman. That's what I thought it did.

@Gaius: I would have to say the lesser of the two evils would be if the game chooses by "civilization", that way, you'd have the same % chance to play all the great civs included in the game... even if you don't get to see some of the leaders that often. Luckily, that's apparently how it works!

 

[apenpaap]

I would prefer it to be chosen by leader. Lots of annoying leaders (Monty, Mansa, Izzy, Shaka, Suryavarayamara... whatever) are on the only leaders of their civ.

 

[LegionSteve]

If the game chooses by civilization:

...then that means that civs with multiple leaders will have much less of a chance than other leaders of appearing in your games. Example: Montezuma (single leader of the Aztecs) will appear much more often than say Stalin (1 of many Russian leaders).

If the game chooses by leader:

...then that means some civs will have a much greater chance of appearing in your games over other civs with only 1 leader. Example: America has a much greater chance of appearing in games than does Japan.

Not just a question to Grimus but to everybody:

If it was made to work the 2nd way, and you started a game where all the leaders are selected randomly (as I often do) would you expect it to allow more than one occurence of each Civ? E.g. first leader slot rolls Cathy, 2nd leader slot rolls Stalin, you have 2 Russias in the game? You can do this if you set it up that way yourself, so I know the game can handle it, but I wouldn't like it - I would find it confusing

Alternatively, if it was made to do random leaders that way, it could be done so that if the first slot rolls Cathy then Stalin and Peter are excluded from the remaining rolls. But then wouldn't you have the same problem - Monty is still more likely to appear in your games than Stalin?

 

[PieceOfMind]

Not just a question to Grimus but to everybody:

If it was made to work the 2nd way, and you started a game where all the leaders are selected randomly (as I often do) would you expect it to allow more than one occurence of each Civ? E.g. first leader slot rolls Cathy, 2nd leader slot rolls Stalin, you have 2 Russias in the game? You can do this if you set it up that way yourself, so I know the game can handle it, but I wouldn't like it - I would find it confusing

Alternatively, if it was made to do random leaders that way, it could be done so that if the first slot rolls Cathy then Stalin and Peter are excluded from the remaining rolls. But then wouldn't you have the same problem - Monty is still more likely to appear in your games than Stalin?

There is a solution to this dilemma. But it's fairly complicated. I might work it out and show you tomorrow.

 

[MqsTout]

How does it behave with unlocked leaders? Even with that on, I still see the same leaders all the time. (I can't remember the last time I saw Wang Kong, or a game without Mansa Musa and Tokugawa).

 

[LegionSteve]

There is a solution to this dilemma. But it's fairly complicated. I might work it out and show you tomorrow.

Having thought about it some more, I think I can answer my own question. You would still pick the Civs first, but you would add up the number of leaders, and then say russia would have a 3 out of X number of leaders chance of being picked, and aztec would have 1 out of X chance. Then if russia gets picked, randomly pick one of the three leaders. That (I think) gives all the leaders a fair chance of being picked. Not sure I have explained that very well but to show it properly would, as you say, be complicated.

 

[Onagan]

I prefer the current way.
Otherwise you get to much America (or Russia, England & France)

I prefer 3x Montezuma over 1x Washington (I will except that).

 

[PieceOfMind]

Having thought about it some more, I think I can answer my own question. You would still pick the Civs first, but you would add up the number of leaders, and then say russia would have a 3 out of X number of leaders chance of being picked, and aztec would have 1 out of X chance. Then if russia gets picked, randomly pick one of the three leaders. That (I think) gives all the leaders a fair chance of being picked. Not sure I have explained that very well but to show it properly would, as you say, be complicated.

What I claim is that you can give every leader an exactly equal weight (probability) of appearing in every game AND assuming no two leaders from the same civ appear. Technically this would make the civs with more leaders appear more often but I don't think that bothers people. What bothers people is when they see some particular leaders too often and some not often enough.

 

[MqsTout]

I believe I saw a tool (spreadsheet or javascript web page) that was built for people to 'randomize' their selections using it in different ways, that you could then select in the game creation screen.

 

[PieceOfMind]

Having thought about it some more, I think I can answer my own question. You would still pick the Civs first, but you would add up the number of leaders, and then say russia would have a 3 out of X number of leaders chance of being picked, and aztec would have 1 out of X chance. Then if russia gets picked, randomly pick one of the three leaders. That (I think) gives all the leaders a fair chance of being picked. Not sure I have explained that very well but to show it properly would, as you say, be complicated.

I thought about it a bit more, and while I was hesitant at first, I'm fairly confident now that you're right and this is all you need to do. As you pick a random civ from the list of ones that haven't already been placed, weight each civ according to its number of leaders. 1 for 1 leader, 2 for 2 leaders etc.

So for example, Hathseput would have 2/(number of remaining leaders) chance of being picked. And yes, to show it works is way too complicated so I'm not going to try.