March Madness: Civ4 AI Edition

[superjm]

So I guess there's a lesson to be learned here.

Traits, AI aggressiveness and peace flavors are all important, but most of all, Never Bet Against Praets

[Azza]

I'm shocked that Lincoln wasn't utterly incompetent for once. Thought Augustus would be tough to dislodge due to Praets, but didn't expect him to single-handedly tear Ragnar to shreds.

[Lord Parkin]

#1 - Ragnar x
#2 - Lincoln x
#3 - Kublai x
#4 - Washington x
#5 - Augustus x
#6 - Hammurabi ✓

1/6 compared to 3/7 last time... I'm getting worse.

[Sian]

Other than betting on the opposite outcome of the Ragnar-Augustus war, i was consistently close but failing the mark

[Brian Shanahan]

The lesson to be learned is "not that the AI is more idiotic than we believe, but that it is more idiotic than we can believe". How in the name of Sid could Ragnar have thrown away that double gold start? Seriously, any member of this forum, given the equal handicaps of the game, would have run away with that start and roflstomped the competition.

I think from now own I'm going to just let random.org make my picks for me.

Oh, and one turn off on the victory score

[regoarrarr]

Winner: Washington
I actually like both of the American leaders' positions - going with Washington as he at least has average traits instead of Lincoln's horrible ones.

Grrr I meant Lincoln. Lincoln! Yeah, that's the ticket!
[/quote]

[Dantski]

Was I the only person to predict a Lincoln win?

[Molach]

Huh, what's that? Praetorians are actually useful?

I never.

Live and learn, I guess.

----

"Ragnaaaaar Looooodebroooooook!"
"Goddammit Ragnar! He has Praets now"
"We got 'im, we got 'im"
"Why do you do this shit Ragner? Take him before he hooks up iron!"
"It's not my fault"
"Or at least use catapults effectively. You are stupid as hell"
"At least I got chicken"

Awaiting next one eagerly.

[Sullla]

OK, we ended up with 42 total predictions this time. That's slightly down from the 44 we had for the last game, still an excellent number. First of all, stating the obvious: Ragnar failed nearly all of us. That includes me, I absolutely would have picked him as the winner as well. I'm still not sure how the AI manages to screw up a double gold start with Financial trait (!!!) Seriously, that dude owes us all money.

Anyway, let's look at some of the more interesting written predictions.

I see Ragnar and Kublai Khan teaming up to squeeze out Hammurabi in the midgame, piled on by some of the other leaders. Ragnar will be able to keep up with the peaceniks by way of being FIN while Khan falls behind the main lead. I actually like Rome's position if he didn't have that huge belt of tundra to his east - I think it's going to be hard for him to get enough land to stay competitive.

I actually like both of the American leaders' positions - going with Washington as he at least has average traits instead of Lincoln's horrible ones. Hopefully they'll figure out to send their 2nd settlers inland instead of into the back lands (as it does look like he is if I look closely. Actually it looks like Kublai Khan has sent out units to the northern arctic...

This was a pretty good prediction, only having Washington come out on top instead of Lincoln (and of course Ragnar to come in second). Washington was the smarter pick to emerge from the two Americans in my opinion: better traits, more land, more aggressive AI personality. The fact that Lincoln went to war while Washington sat around doing nothing still baffles me.

Winner - Lincoln
Runner Up - Washington
First to Die - Hammurabi
Number of Civs Surviving - 4
Victory Type and Victory Date - Space T327

I think the 2 Americans will get ahead of the others who will be warring for significant periods. Lincoln (who's capital is amusingly called Washington, not confusing at all!) is more isolated and I think his land is good enough to put him ahead.

Also not following the trend and picking Ragnar.

The brave soul who went against the trend! Also one of only two people who picked Lincoln to win.

Winner: Lincoln
Runner Up: Hammurabi
First to Die: Kublai Khan
Number of Civs Surviving: 4
Victory Type and Victory Date: Culture, T300

Ragnar and Kublai think they're the wolves in the sheepfold, but these rams have horns. The peaceful four grow tired of the warmongers' antics and gang up on them. Once put under pressure, they fold like the paper tigers they are. Yes, we did metaphors in English class this week.

Anyway, with the troublesome two out of the way, the remaining civs just tech along in peace. Washington has the most land, but due to his weak midgame, Hammurabi looks like the best candidate for space. Lincoln pips them both by switching to culture too soon for anyone to launch.

Our other prediction to get the winner correct, and mostly solid elsewhere. The lategame was indeed a field of peaceful sheep.

There's basically no hope for Hammarubi this game, he's surrounded by both war leaders, and Kublai can only expand into him. Ragnar will either eat Hammarubi or Augustus first, then continue into the Americans.

Halfway correct!

Nobody giving Augustus much of a chance but I like him for runner up: Ragnar swallows up Hammurabi and then continues past him, Augustus' traits and personality lead him to a more wonder + expansion game and he possibly buddy-buddies up to him via religion: Praetorians, even without aggressive, will help him survive if he lives that long. Then Ragnar calls diplo to win after failing to hit Domination for a long time and before he launches a ship.

Also halfway correct, although with the runner up instead of the winner. Better than most of us did.

For kicks, I predict Augustus will win, followed by Lincoln. Ragnar will attack Augustus early and be the first to die (to Praets). Kublai Khan will attack Hammurabi, while Washington will squander his settlers on the tundra, allowing Lincoln to claim the fertile land to his southwest. Hammurabi will hold off the squeezed Kublai Khan for a long time, and the game will otherwise be largely peaceful. Augustus will build the spaceship and win on turn 357, with four players remaining.

Another near miss, had the two leaders that advance to the playoffs, only in the wrong order. Half credit in the points system.

I can't believe how far off you guys are on this! There is only one possible outcome: Ammurica.






Never bet against Ammurica.

I'm shedding a red, white, and blue tear right now.

Hey, I'm flattered that my nonsensical ramblings are being taken seriously enough that someone thinks there's something to copy.

I think this perfectly captures the spirit of this game. And the fact that RB members can turn everything into a competition.

This post is getting really long, let's put the results in a separate one.

[Sullla]

Here were our results in table form:



51 total participants thus far, and 42 for this particular games. On to the picks.

- Winner and runner up were pretty much a disaster for the field. Ragnar screwed us all over, he was picked to win by 29 people. Washington was the next most popular choice there. Lincoln was picked by only two people, Dantski and Becko. Not a single person picked Hammurabi to win. Washington and Kublai Khan were the most popular picks for the runner up spot. I think we were all shocked that Lincoln and Augustus ended up taking the playoff positions.

- In direct contrast, EVERYONE picked Hammurabi as the first to die, correctly in this case. 31 out of 42, 74%. If Rangar had lived three more turns, we'd have had a total wipeout here. Jabah was the only one who picked Ragnar as the first to die.

- Once again, we had 3 surviving civs. At some point the streak has to end, right? Right? 62% of you nailed this correctly.

- Spaceship was picked by about a quarter of the submissions. Domination was the most popular choice for victory type, with 20 entries.

- The field zeroed in on victory date with a high degree of accuracy, and this was a major reason why the scores for this game were still pretty good. No one hit the exact date, and that was a surprise since it was Turn 315, a round number. We had picks for Turn 309, 310, 311, 312, 313, 314, 316, 317, and 318 - but not 315. Very, very weird. Nevertheless, 24 entries were within 10 turns of the victory date, and another 13 within 25 turns. All told, 88% of our picks were within 25 turns. Looks like the field is starting to narrow in on the rough timing of when the AI claims victory under these settings.

With regard to our individual scores:

* High scores for the round were Dantski, tiredkitten, Lord Parkin, and regoarrarr all at 16 points. We didn't have any truly huge scores because of the massive flameout for the first two spots.

* TheHumanHydra is our special case with picking the two successful leaders, only in the wrong order. It's worth half credit, or 5 points.

* Bobchillingworth and Hashoosh had 100% identical entries - apparently by accident! - until Hashoosh edited his victory date. It didn't matter which one of the two dates I picked, since they both scored the same point value.

- In the race for first place, we have a log jam between 47 and 45 points. Katon is our leader despite only picking for three games, powered by that exact victory date from the last game. Ceiliazul and Commodore are a single point behind, while dazedroyalty / Hesmyrr / von Adlercreutz are marching in lockstep at 45 points. All of them had identical 13 point games this time. Quite a race here for first - no pressure!

That's all I've got for now. Tomorrow we'll do the leader draw for Game Six and post the overall bracket. Thanks again for your efforts, it's been a lot of fun so far.