bad great generals

[snipafist]

This is why you should attach great generals to units that can withdraw and/or give his unit the tactics upgrade. That way when you throw his unit at an enemy, there's always a decent chance you can get out of there if things start to go sour. Not to mention with 50+% retreat odds, you can throw him at units with ~80% chance to win and not feel too worried, and get tons of XP with Leadership as well.

 

[TrailblazingScot]

The main function of a Warlord as Great General is to put all the exp. into one unit so it levels up to allow West Point and the Heroic Epic to be built.

Otherwise he is being used as instructors...far more valuable.

That said, the chap suggesting using him as an auxillary player in terms of medic promotions is speaking sense.

Also, if I use a Great General on a sub I should be able to manage 100% withdrawal with the correct choice of promotions - got to do that in game.

Hopefully Firaxis won't nerf that.

 

[Gyathaar]

Also, if I use a Great General on a sub I should be able to manage 100% withdrawal with the correct choice of promotions - got to do that in game.

Hopefully Firaxis won't nerf that.

They already did.. you cant promote units to have more than 90% retreat chance.

 

[jkp1187]

Somewhat on-topic: I've noticed that, since I installed Warlords, I have seemed to lose a good number of battles in which I had a 90+% percent, including a couple where I lost a Warlord unit. (yeah, I know, boo hoo.)

Still, the battles odds seem kind of... out-of-whack. Has anyone else noticed this or have a possible explanation?

Honestly, I've found that I've WON a few battles where my odds of winning were as low as 24%. I've also lost a few battles where odds of winning were in the 80's.

I also lost a great general, but he was with a cavalry unit that was down to 75% strength and was attacked by a French cavalry unit that was at full strength, so them's are the breaks, I suppose. Besides, a few turns later I sacked Paris and the French surrendered, so I can't complain too much.


Random chance, maybe?

 

[Older than Dirt]

I have not lost any battles with my 42 strength gunship with GG attached but have suffered rather high damage when attacking longbowmen in cities when I have 100% chance of winning. (This unit had 202 experience points before I inadvertently won with domination. I was using it almost exclusively to attack for about a hundred years.) I don't think you can lose with a 100% but I bet you would lose about a third of the time with only 99.9%. )

 

[Cookie Crumbs]

Yeah, and with Blitz I believe Gunships can attack 4 times per turn. I normally try to get Blitz on my Gunships but they do tend to get damaged too easily. March Gunships, now that's something.
In my first and only game so far I've been quite lucky in that I don't tend to lose many battles when odds are above 70%, touch wood.

However, I think half the reason why people think get cheated by the RNG is because whenever there's a battle you're expecting to win, whenever you do you don't take note of it, but when you lose, you do, and so you remember losing as much as winning. With lower odds, you expect to lose more, so you don't take note of losses that much. I don't dispute that the numbers are at the very least off, though.

 

[Mango]

Nope. It's 120*0.00000093 = 0.00011, or about 1 in 9000. Still too small to be likely.

Damn math majors.

Also, that is just for 99%. Assuming it's >99%, that number jumps significantly. I would venture to say that it would double to 1 in 18000 if it were just 99.5%, correct?

 

[porkdiesel]

I was churchill (Protective) and I had a level 4 or 5 longbowman.
A great general popped out, and i attached him to that longbowman for 20+ XP. So this longbowman had maxed out City Defense, maxed out Drill, and 1 or 2 +strength upgrades. The zulus stacked up a crapload of units outside a city i conquered in French territory, then of course declared war on me.

That general unit must have killed about 15 to 20 units while defending the city before finally dying, basically saving me from total annihilation! Too bad i couldn't use that XP for more promotions on the next turn

 

[DrewBledsoe]

I was churchill (Protective) and I had a level 4 or 5 longbowman.
A great general popped out, and i attached him to that longbowman for 20+ XP. So this longbowman had maxed out City Defense, maxed out Drill, and 1 or 2 +strength upgrades. The zulus stacked up a crapload of units outside a city i conquered in French territory, then of course declared war on me.

That general unit must have killed about 15 to 20 units while defending the city before finally dying, basically saving me from total annihilation! Too bad i couldn't use that XP for more promotions on the next turn

At least all the xp he earned must have got you close to another GG, sounds like he deserves his very own Heroic Epic

 

[gettingfat]

I don't have enough experience with Great Generals to comment yet, but after probably 100 games I've got a strong feeling I've lost more sea battles to the AIs than the odds indicator suggest.

 

[Pantastic]

If that really happened, I don't think it was purely random. The odds are about 1 in 10,000. More likely that there's a bug than that you just happened to get the 1 in 10,000 outcome.

How many people have Civ4? How many chains of 10 fights do each of them play on average? If we assume 50,000 active players who each average 1000 fights apiece, then we'd expect 5000 occurences of this 1 in 10,000 outcome. Doesn't sound like a bug to me that he saw something you'd expect to see thousands of times in civ.

 

[Pantastic]

I don't have enough experience with Great Generals to comment yet, but after probably 100 games I've got a strong feeling I've lost more sea battles to the AIs than the odds indicator suggest.

Yes, people always do. They could bias the RNG to give secretly cut your chances of losing in half, and people would still get strong feelings that they're getting screwed by the RNG because the human mind tends to focus on and remember the losses.

 

[DaviddesJ]

How many people have Civ4?

Many more than the number that post here, which is the relevant number. There are probably only a few hundred who might post in this thread.

 

[Soneji]

bugger that maths. but yes my odds have deff decreased, my offensives are much slower now with warlord, vanilla I would normally have continent on huge by 1800ish now I struggle. And I know its down to the battle odds its got to be my strategy hasn't changed that much in many years

 

[ohcrapitsnico]

I can't wait to see what Firaxis will come up with in their first patch with all these problems people are having.

On topic, I can say too that warlords are slowing me down. It's strange to think that the warlord which was intended to dominate the battlefield and change an outcome of a war but is really the opposite. My warlord hides behind a stack of units and only attacks other units when they are heavily damaged and much weaker since it dies pretty quickly even after I stack flanking, tactics, and so forth to help it withdraw. None of my warlords have ever withdrawn from a battle and I have had experience with about 30 of them. All my warlords died in combat too a much weaker unit usually after 4-6 fights, many times even less.

 

[Pantastic]

Many more than the number that post here, which is the relevant number. There are probably only a few hundred who might post in this thread.

No, it's not the relevant number. There are tens or hundreds of thousands of people playing Civ4, and having hundreds or thousands of strings of ten battles at a time. If you know statistics, you know that you should expect events that have "only" a one in ten thousand chance of occuring to happen to a lot of people. The fact that something happened that you'd statistically expect to happen to thousands or tens of thousands of players of the game does not indicate a bug, and is not even suprising if you understand the math.

 

[DaviddesJ]

If you know statistics, you know that you should expect events that have "only" a one in ten thousand chance of occuring to happen to a lot of people.

I have a PhD in information theory, and I certainly understand the mathematics. The fact is, it doesn't matter how many thousands of people it happens to, if those aren't among the people who post here. If there are only a few hundred people who might post here, and it happens to one of those small number of people, that's significant, if the odds are against it.

I don't know whether there's a bug, or not. One anecdote doesn't prove or even strongly suggest it---we don't even know if that anecdote is accurate. But it wouldn't be terribly surprising for there to be such bugs: implementing a correct combat calculator is not all that easy, and it wouldn't be a big surprise if there are glitches that cause it to sometimes (or even often) report incorrect odds. In the early versions of Civ4, the combat calculator would sometimes report a probability over 100%.

 

[scottcstoness]

No, it's not the relevant number. There are tens or hundreds of thousands of people playing Civ4, and having hundreds or thousands of strings of ten battles at a time. If you know statistics, you know that you should expect events that have "only" a one in ten thousand chance of occuring to happen to a lot of people. The fact that something happened that you'd statistically expect to happen to thousands or tens of thousands of players of the game does not indicate a bug, and is not even suprising if you understand the math.

I think the point was that only the people who post here are likely to post here about losing 3 out of 10 battles....

Of course, noticing that statistic might prompt one's first post, so perhaps half the readership might be a better estimate of the sample size (e.g. half the people reading have either posted or would post if they noticed this happening).

 

[DaviddesJ]

Of course, noticing that statistic might prompt one's first post

FWIW, the poster who made that particular claim has 28 posts over a period of 3.5 years.

 

[Jastrow]

Losing 3 encounters where you have a 99% chance of winning is odd.

Unless you have fought over 300 battles.

Of course the odds of losing 3 out of 10 battles like that is quite unfathomable. It could be a bug, but it's probably more of a psychological thing.

Edit... posted before reading further pages... I see this has been covered already... ignore....

Not all that unfanthomable...

In fact, the odds of that happening are better than 1 in 10000. Compare that to many situations where probabilities come up (lotteries, etc), and 1 in 10000 is pretty good odds.