Where are my MGLs?!?!

[Lord Jimbob]

The most likely explanation is the laws of chance.

Just because you have a one in 16 chance of a MGL doesn't mean you *have* to get one. I have gone whole games without a single MGL. (Usually when not warmongering, though.) If you have a lot of battles with your elite units, you are *likely* to get on (or more), but it is NOT guaranteed!

There will also be runs of bad luck. There is a 1 in 16 chance ( for those who want the math, (15/16) ^ 43 ~= 1/16 ) that you'll have a run of 43 elite wins in a row before you get a single leader. -from his article on GLs

yeah, I had 7 GLs in GOTM31, including 3 in 2 turns (as a non-militaristic civ)! but I am playing a game as the japanese now and I had to win something like 25-30 elite victories before I got my 2nd GL (used the 1st to build HE!). SirP has a good article on GLs that discusses the probabilities of getting a leader/ having a drought of them, here.

 

[wilbill]

BassDude726, I think you can probably blame your lack of MGL in this game to the whims of the RNG. I don't play Always War, but I usually have at least a couple of decent wars in a game and usually create one or two MGL's. But I've gone through some streaks of two or three games without one. And lately I've been on a streak in which I seem to get an MGL everytime one of my Elites wins a battle. The game giveth, the game taketh...

 

[BassDude726]

Yeah, this must just be a case of the game being cheap. I was also wondering, I know that for each individual elite unit there is a certain algorithm that determines the appearance of an MGL based on the battles they win. But in the case of multiple Elites, how do the chances of getting an MGL change?

For example: One elite cavalry without HE has a 1/16 chance. If four elite cavalry win in a turn, even though the 1/16 will stay the same, shouldn't the chances increase? I have a feeling the answer is pretty simple, but to be honest, I'm horrendous at math.

 

[Dell19]

Stupid post

 

[Arathorn]

So three elite cavalry should have a 3/16 chance of creating 1 Great leader.

False. They're independent events, but you can't just add them up.

There's a (1/16)^3 = 1/4096 chance of 3 GLs in those 3 victories.
There's a 3*(1/16)^2*15/16 = 45/4096 chance of 2 GLs in those 3 victories.
There's a 3*1/16*(15/16)^2 = 675/4096 chance of 1 GL in those 3 victories.
There's a (15/16)^3 = 3375/4096 chance of getting no GLs in those 3 victories.

Note that 3375/4096 > 13/16 as Dell19 claims.

Put another way, 3/16 is 18.75% of the time. But, in actuality, it's only about 18.33% of the runs will give at least one MGL. Seems subtle and unimportant, but it's not.

That's why SirPleb notes that a run of 43 elite victories with no MGL is about as likely as getting an MGL in one battle -- about 1/16th of the time. It's the same principle at work.

Arathorn

 

[Dell19]

That would be correct, I blame it on being ill... I should have even known the formula for calculating the probabilities since its a binomial distribution with n equal to 3, p equal to 1/16 and x being the number of leaders you want to find the probability of getting. The formula being:

n
C p^x(1-p)^n-x
p

 

[BassDude726]

HA, and I thought the answer would be simple! I'll just put it this way... I've only taken one math course in 3 years of college... and I got a C- in it! No wonder I couldn't figure it out. That's why I'm an English major.

Anyway, thanks to everyone who posted ideas, they've shed a lot of light on stuff I didn't really understand too well.

 

[Dell19]

I got an A in statistics and conveniently forgot about all those wonderful probability models. Ah we could even use hypothesis testing to see whether getting no leaders in x number of fights was acceptable in a 5% error range. Makes me think that I'm going to have to do some revision for the exam I have in stats on the 8th.

 

[dgfred]

@ Arathorn- Those stats are real eye-openers . I for one will not
complain about a MGL drought. Some of the SGers have had some doosies.
Most of the time I have had pretty good luck getting them- I play alot as
the Celts, Iroquois and Vikings- sometimes these civs get a bucketfull .

 

[Arathorn]

I think my longest string of elite victories without a MGL (was just a GL in those days...PTW, I think) was in the low-to-mid 50s. It happens sometimes. A string of 71 is just over a 1% chance, so if you played 100 games, you could expect one to start with a string of 71 elite victories with no MGL.

It's just the way the numbers work... It may FEEL like the game is out to get you (and I get that feeling, too...just like everybody else, but I don't let it change my gameplay), but it's really just odds.

Arathorn

 

[WackenOpenAir]

its just <Explitive deleted> luck. last game i also thought the game had to be bugged not to give me any leader. Just as i was about to give up paying attention to have my elite units fight the right battles, i got a leader again. sometimes you get them fast, sometimes it takes 100 elite wins.

Moderator Action: Watch your language, please. We don't appreciate vulgarities, even if they are *disguised*. --Padma
Please read the forum rules: http://forums.civfanatics.com/showthread.php?t=422889

 

[Dell19]

The best time to get a leader is on your last elite attack before your current enemy is wiped out Especially if it gives you the chance to form an army and then attack in a couple of turns to get the heroic epic for your next war...

 

[anarres]

I recently played a game with easily 100 elite victories and no MGL.

Kinda sucked as it was PTW and I based my strategy around a "perfect" RCP palace with a rushed FP far away.

It was the Industrial age before I *finally* got one - then to add insult to injury I got another the very next win.