BOTM Gauntlet #11

[Lexad]

As Grey Cardinal has solved the issues with his 1520AD submission, I'm adding him and Ronnie1 to the final roster.

 

[Skurn]

Well, I'm still last! Do I get a wooden shield or something?

 

[Lexad]

You seem to have one already

 

[Skurn]

That's right, I need to earn the sword and the helm now!

 

[jesusin]

(0-16-18) - it must be a typo... it must be a typo...

Jesusin, if you include 18th GA, which saves you 1 turn, the average effect for three is (85-71)/3 = 14/3, or 2 turns per GA on Quick.

????????

I don't understand.

In my example the GP 17th saves 13 turns. The 18th GP saves 1 turn. A 19th GP would have saved another 8 turns.

It's 14/2, not 14/3

 

[Lexad]

????????

I don't understand.

In my example the GP 17th saves 13 turns. The 18th GP saves 1 turn. A 19th GP would have saved another 8 turns.

It's 14/2, not 14/3

Well, I modeled your example in my Excel (attached), so 16th saves 3 turns, 17th - 10, 18th - 1. That's 14 for three, if we want a representative averaging (if not - just take 10 from 17th). I have yet to see a person get 19th GA, though

 

[jesusin]

Well, I modeled your example in my Excel (attached), so 16th saves 3 turns, 17th - 10, 18th - 1. That's 14 for three, if we want a representative averaging (if not - just take 10 from 17th). I have yet to see a person get 19th GA, though

If you don't take 15-16-17 but 16-17-18 you'll get a different value. These are only exmaples, they don't make the rule.



Look at bostich's thread in HOF forum, and somme of my games there too. 22-23 GA is not so uncommon. It is not optimum, though.

 

[Lexad]

Look at bostich's thread in HOF forum, and somme of my games there too. 22-23 GA is not so uncommon. It is not optimum, though.

Thanx, I've seen now. And yes, it is the whole point, it hurts regular cpt so much it is not optimal - cause you do not see these games in HOF top10. The question is where is the optimum.

If you don't take 15-16-17 but 16-17-18 you'll get a different value

Yes, still not 14/2.
Let's take a broader average for GA speeding up 3 cities (from say 12 to 19) to decrease the outliers' effect. Let's use your example so I'm not rigging the result by picking self-constructed case.

12th GA saves 2 turns
13 - 4
14 - 4
15 - 6
16 - 3
17 - 10
18 - 1
19 - 8
20 - 1
The average is 4,33 ~ 1,9 turns on Quick. Adding 21 (5) will not change it perceptibly. The point is there would be no "positive" outliers like 17 and 19, if not for the "negative" outliers 18 and 20.

 

[jesusin]

Splitting hairs, please, ignore if not worth it: why is 17th in your table 10 and not 13?

 

[Lexad]

Cause it is not 13 in the model I constructed on your data, in none of my previous posts. If you compare our 16-GA distribution you'll see that optimal is not 3-3-10 in your example, but 2-4-10 (as in mine).

 

[jesusin]

Hmm, what a bad example I've chosen!

 

[jesusin]

Sure. I'll report back whether I improve the result or not. It just won't happen this week, sorry, too busy

I still it would be better if it was himself who tried to improve, with the general idea of "sooner pacif, more GA" in mind.

Sorry to let you down, guys.

I am sorry to report that I won't be trying to replay your game.

You were right. Your civ is geared in such a different way from what I usually do, that it would take too much time and effort to give shape to my try.

I thought I'd do it today (1.5 hour free time, at last!) but I have just realised it can't be done in so little time.

In particular, the detail that has decided me not to try it is the NE in Bibracte. How can I move it to Fish?

Please forgive me.

Replaying from 4000BC would be too time consuming and it wouldn't be the kind of experiment we wanted to perform anyway.

 

[PutCashIn]

Not to over simplify matters, but if you had 3 GA's on Standard/Normal that would be 12,000 CP's for your lowest CPT city, ja?

So at 800 time averaged CPT your saving 15 turns.

Thus your getting 800 CPT's (at least) on the last turn in your lowest earning CPT city?

Crikey, no wonder you guys win so early!

 

[Lexad]

You oversimplified. Usually you'll use those 3 GA on at least 2 cities.