A way to score starting positions?

Lazteuq

Chieftain
Joined
Jun 26, 2014
Messages
75
Location
Fairbanks, Alaska
The starting position obviously makes a huge difference in how easy a game is. A win on Immortal with a start with a plains cow and a floodplain, plus a bunch of plains and a few hills is far more difficult than a 2 gold, wet corn, riverside, lots of forests start.
I'm thinking of a scoring system for estimating how difficult a start is, maybe something like this...

Score starts at zero if all just grasslands
-1 per plains after the first 5
-1 per desert after the first 5, except oasis
-1 per ocean tile
+2 per seafood
+2 per floodplain
+1 per forest
+3 if river running through BFC
+3 for sheep, cows, horse, pigs
+3 per corn,rice,wheat +1 more if wet
+5 or more? per gold, +3 per silver
+2 per camp resource (Elephants, Deer, Furs)
+1 calendar resources and wine

EXAMPLE: The current NC game, NC181 (Image shows a few tiles around start, not a big spoiler)

Spoiler :



IF you Settle In Place,

-1 Plains
-0 Deserts
+- 0 Oceans
+0 Seafood
+6 Floodplains
+13 forests
+3 from river
+3 from cow(maybe not though because its on plains?)
+0 from farm resources
+10 gold
+0 from camp resources
+1 wine

TOTAL SCORE: 35. Meanwhile, a HOF level, insane, 1 in 1000 start might have a score of 50+, and a really atrocious start might be well under 20



QUESTIONS:
How to account for hills because an ideal amount of them is somewhere in the middle.

I think this method overvalues forests, but it does keep the math simple.s

How to account for having sea tiles? Are they even a good thing?

Are you civfanatics members actually interested in this? It would be useful when comparing game scores and finish dates. I'm imagining games like the Noble's Club having a score for the start. Maybe we could turn the start score in to a multiplier for the game winning score, or the finish date.

The numbers are mostly just temporary placeholders.
EDIT: How do you make the image bigger?
 
I think the river thing should be +N (1 or 2 probably) for each riverside tile, not just a flat 3 for a river. A river that touches 2 tiles at the edge of the BFC is worth way less than one that zig-zags through the middle.
 
Corn > Wheat > Rice
Coastal Fish > Coastal Clam/Shellfish > Ocean Fish
Plains > Desert/Mountain > Jungle
Pigs > Green Cows > Plains Cows > Sheep
Copper >>> Horse > Iron, Ivory > no pre-modern military resource
Riverside crop > riverside Flood Plain (capped at (five + number of crops) - fifteen FP with no health is a slow start) > riverside Grassland > riverside Hill / Plains (I'm thinking maybe 1 : 3/4 : 1/2 : 1/4 ?)
Gold > Gems > Silver
Increase Jungle penalty to 1.5 for Ivory and 50% of Resource value for other resources
 
Modern military resources +1 if on a hill or water tile, otherwise 0. I hate bulldozing a riverside Town to get some Coal I could easily get in a colony or expo.
River connections to resources need an additional bonus, +0.5 for food & Iron, +1 for luxes, +2 for Copper & Horses.
Plains hill starting tile +2. Non-starting tile that retain all starting resources in BFC need accounting for: +1 Plains Hill, +2 Plains River Wine hill, +1 Plains Ivory, Banana or dry Rice, +2 Plains Hill Stone, +3 Plains Hill Marble. And additional +0.5 for the 2H and +1 for the 3H tiles for each seafood.
+0.5 x number of Seafood for having a 3H tile you won't settle on.
Food resources have diminishing returns, a five seafood start is going to effectively be a four seafood start once the whip starts cracking.
…unless you're CHA, which opens up the whole 'FIN river' and 'PHI Stone' can of worms.
 
Forests are amazing

Wine is poor on a hill but moderately useful on a river

A few sea tiles are great. They get you superior trade routes, enable your Bureaucap to contribute to the navy, make intercontinental espionage a bit easier, and open up GLH. They're worth more if you have Marble and lots of forests & food, putting GLH-Artemis combo within reach. A dozen empty sea tiles are a frustrating waste, though. I guess something like +4 for being coastal and -0.5 per coastal tile and -0.75 per ocean tile. -0.25 for coastal tiles if you're FIN, no harm in running a few of them to get you up to happy cap.
 
Thanks for your input, Lexicus and Lindsay. I like your ideas, especially this most recent thing^ about adding for coastal and then subtracting per coastal tile.
I have a concern that this will become too complicated to be actually used, though. If we made a system that really accurately scores starts, it will take 20 minutes to do, so there has to be some kind of balance point, and I'm not sure where that is.
 
Lazteuq said:
I have a concern that this will become too complicated to be actually used, though. If we made a system that really accurately scores starts, it will take 20 minutes to do, so there has to be some kind of balance point, and I'm not sure where that is.

I think you have the rub right here. Part of the reason Civ 4 is such a good game is, it takes 20 minutes to do the analysis properly!
 
Lazteuq said:
Score starts at zero if all just grasslands
-1 per plains after the first 5
-1 per desert after the first 5, except oasis
-1 per ocean tile
+2 per seafood
+2 per floodplain
+1 per forest
+3 if river running through BFC
+3 for sheep, cows, horse, pigs
+3 per corn,rice,wheat +1 more if wet
+5 or more? per gold, +3 per silver
+2 per camp resource (Elephants, Deer, Furs)
+1 calendar resources and wine
Gold is nice, but if your giving it +5 and wet corn only 4 then you are seriously undervaluing food. The screenshot you added also brings up a conundrum in that the value of tiles not being fixable, i.e. gold, is heavily dependant on the presence of good food tiles to effectively support them.
Similarly the plains cows value is severely diminished without an adequate food resource, especially as its the cause of that.
I would trade the cow, the wine, both golds and the floodplains for a pair of wet corn in that start :p
lindsay40k said:
Food resources have diminishing returns, a five seafood start is going to effectively be a four seafood start once the whip starts cracking.
Depends if the food is likely to be able to be given to an early city or not, at which point it can be very beneficial.
Lazteuq said:
River connections to resources need an additional bonus, +0.5 for food & Iron, +1 for luxes, +2 for Copper & Horses.
This is indeed really useful for luxes, copper and horses. Iron is too late for this to make half of the difference while foods may on rare occasion in FP heavy starts or when extremely early trading is possible
But to really ruin this for you, I think the OPs idea is as an initial assessment on starting a game, which means outside of advanced starts the only strategic resource you will be able to see is Ivory!

Another big potential benefit of river (and coast) is the ability to get trade routes set up between cities without the need for worker turns or even Sailing if the route lies within culture. It can make a significant difference at the start and you can be given a sniff of a longer river, or even see a likely settling site for a future city overlapping the capital on your first turn.
nate46 said:
you're missing a big advantage of ocean fish: can't be pillaged by barb galleys.
and valuing +1:commerce: over +1:food: :eek:
 
I think you have the rub right here. Part of the reason Civ 4 is such a good game is, it takes 20 minutes to do the analysis properly!
That's the difference between dirty players and everyone else--diety players will take that 20 mins, everyone else will do an approximate analysis that might be wrong.

But to really ruin this for you, I think the OPs idea is as an initial assessment on starting a game, which means outside of advanced starts the only strategic resource you will be able to see is ivory.
Unless you take a peek in WB. ;)
 
I think you are looking at it the wrong way. Why is cow worth more than deer when a forest deer has same yield as grassland cow? Someone suggested corn>wheat, when they are actually exactly the same, the only difference is that corn on the normal mapscripts appear on grassland while wheat appears on plains. On some mapscripts this isn't always the case.

Instead of scoring resources, you could score actual tile yield. Look at things like what is the food surplus at pop 2/3/4? What is the food surplus and hammer/commerce output when working all BFC resources? What is the food surplus and hammer/commerce output at pop 6/8/10 or what's the food surplus and hammer/commerce output when working all resources and all riverside tiles cottaged?
 
valuing +1:commerce: over +1:food: :eek:

Ah yes, to be fair I'm a Hannibal and Toku player who either likes an early +2:commerce: to get shooting towards Sailing & Masonry, or else gets naff-all :commerce: and appreciates every early source of it to counter AGG-PRO's economic weakness :3

I think the OPs idea is as an initial assessment on starting a game, which means outside of advanced starts the only strategic resource you will be able to see is Ivory!

Well, perhaps, but we all know how our early game assessment changes when we roll Toku and upon researching BW our hilltop mine on the river turns out to have Copper, whilst Huayna next door's surrounding Cuzco with Cottaged Flood Plains and half way to Stonehenge building the Mahabodhi…

Spoiler :
Time to TOCK!
 
Instead of scoring resources, you could score actual tile yield.
Yes I'd say that would be the simplest way of getting a baseline figure that could then be modified further with more in-depth stuff like the many suggestions so far. Use some standard conversion between :food:/:commerce:/:hammers:.

Other thoughts:
- A few excellent tiles is better than the same net positive yield spread over a larger number of goodish tiles.
- Any early happy is good regardless of the tile yield, and even if not in the BFC. Same I guess for other early usable stuff like stone and marble.
- Only seafood and not starting with fishing is awkward. Also a 3:hammers: tile is supposed to be ideal if you do start with fishing.
- Likewise forests not starting with mining, or grains and no ag, or livestock for any civ but especially ones without ag or hunting.
- Probably other stuff synergizing with leader traits like ORG or FIN getting extra benefit from coast, IMP from a hammer site like in the current Nobles club.
- The AI has an algorithm in it somewhere (maybe the settler unit AI, I can't remember) which scores city sites. That might suggest more ideas.

I definitely think this is an interesting problem anyway :goodjob:
 
I would trade the cow, the wine, both golds and the floodplains for a pair of wet corn in that start :p

I most likely would not, okay wine is unimportant ;)

But Deity (diff i would play on) plays different, while 2 wet corns are a very nice start too, gold and floodplains give so many possibilties.

We can take Izuul's game, HA rush with maths and powerful Bur. Cap.
Fast teching and production, with Imp. getting out settlers despite low food is not too difficult as well.

You can do much whipping with 2 wet corns, but you cannot get that kind of research and rush progress so fast. And in the end, it's usually more difficult without gold and food instead. To be fair, floodplains save that start from being much slower but you included them :)
 
Top Bottom