Quoted from Trium3:

I'm inclined to go with my original figure of 8 ticks unloading. ... I think I need to re-instate my original formula.

You're correct, and I was wrong. I was looking at wharf cart pushers delivering to a warehouse, where it takes 12 ticks to unload. I didn't consider the possibility that unloading times would be different at a granary and at a warehouse. (I had determined a while back that a warehouse cart pusher could deliver food to a granary 8 times per month if the delivery distance was 2 or 3 tiles, but didn't think about that when posting reply #21.)

Absolute minimum wharfs to feed 90,000 would be 141, assuming that all can be positioned where they travel 0 or 1 tile to the grounds and contention for the centre tile does not cause them to travel further. You'd also have to have all of them within 8 tiles (to the centre) of a granary. I don't know if you could contrive this on a purpose built map - I seem to remember the editor limits the number of fishing points to eight (?)

It would probably be easier to use substantially more wharfs at a greater distance (reducing contention) but as joshofet says the sprite limit is probably the limiting factor. I'd say 73,500 is pretty good going.

The whole crux of the problem is designing a layout that allows for the maximum output with the finite number of wharves the game allows you to have. The last few wharves in each group are quite a long way from the fishing point and thus nowhere near peak efficiency. In doing this, you are lucky to feed 250 per wharf (again a seat of the pants figure). Philon's strategy appeared to be only to include fishing wharves that could outproduce a wheat farm. Thus his city was in Central climate and needed prefects as well, which in a fish only city you can do without.

You can build a very efficient fishing city of 50 or 60 thousand, but returns start to diminish fairly rapidly after that.

[This message has been edited by goonsquad (edited 09-21-2008 @ 03:04 PM).]

Trium,

Your research on this subject looks very good. I'm using your numbers for the new city I'm building.

I had already an excel file with fishing block designs and production calculations. I only had to alter the numbers to the following:

distance/production
00: 38.400
01: 38.400
02: 32.000
03: 27.428
04: 27.428
05: 24.000
06: 24.000
07: 21.333
08: 19.200
09: 19.200
10: 17.454
11: 17.454
12: 16.000
13: 14.769
14: 14.769
15: 13.714
16: 13.714

Edit:
Thinking again, I'm not sure if the game actually uses 3 digits after the decimal. Maybe it uses only two like you did.

I believe this is exactly the kind of subject where excel is the perfect tool. You do the design and you automate the calculations at the same time. So when you change the number of wharves at a certain distance you can instantly see how it effects the total production.

For up to 7 wharves it is fine to use a single granary because the granary is always within the required distance and you don't use any production.

So first, here are blocks with a single granary up to 7 wharves:

The pink points are road access tiles of wharves. On the left side it is down, on the right side it is up. North is top right as usual.





In 5 and 7 wharf blocks with maximum production, the granary and the single wharf is at the bottom. If you move them up you can't get maximum production anymore. The blocks on the left show what you can get. You lose about 6 cartloads.

Trium, I think you are spot on about the importance of using the granary centre as the target. On Caesar Alan's site it says 5 tiles to granary but it would be better to say 6 tiles to the granary centre.

Edit: Although we now know that it is not 6 but either 7 or 8 tiles to granary centre depending on the distance as calculated in my next message.

I just did some testing and indeed 8 tiles to granary works fine when the boat is 3 tiles from fishing point. I think the reason why it was expected to be 6 tiles before is because the cartpusher gets back after the boat if it is 7 or 8 tiles. But I checked and the production isn't effected.

Said that, this doesn't make any difference. With 6 tiles to granary, and using a single granary, the maximum wharves you can have without losing production is 7. With 8 wharves, the granary is too far and you lose production. With the new distances it is exactly the same: max. 7 wharves with single granary.

Therefore with 8 or more wharves it is much better to use two granaries

The block on the right shows where the fishing point should be for maximum production. The granary is at the correct place as well. But there is no way to connect the roads to the granary within required tiles. You lose by 1 tile.

The block on the left shows what you can have with 1 granary.



With two granaries the distances are never a problem. You don't need to bother calculating distances anymore.


Here are two more common size blocks:

[This message has been edited by Philon (edited 07-21-2012 @ 05:50 PM).]

Time (in ticks) per cartload = (((tiles to grounds) * 30) + 210) rounded to next 50

Trium in your formula "rounded to next 50" means you will need to use the next 50 even if the number doesn't need rounding. For example 8 tiles gives us: 8*30+210=450 which doesn't need rounding but we still need to use 500 to get to the correct result. Other than that small detail it looks very good.

I think you did a great job discovering the 50 tick intervals for spawning wharf cartpushers. The game appears to use a city-wide timeline. I have seen many wheat farms produce at the same tick and many market ladies spawn at the same tick.

This helps with Brugle's gatehoused forced walkers. That was a very cool idea by Brugle. Here are a few of my designs:


The distances between these gatehoused forced market ladies will be maintained at all time even if one market has to travel one more tile. Even though they disappear into the market with 15 ticks difference, they appear at the same time. This is similar to cartpushers waiting for the next 50.th.

Once I had a problem with market ladies and I couldn't figure out why they suddenly decided not to maintain the distance between them. Luckily a few days earlier I had read your message above about the 50.th tick and I was able to solve the problem by making the loop one tile longer.

Anyway, returning to the subject,

Boat to fishing point is 0 tiles:
time per cartload = 0*30+210=210 > rounded up > 250 ticks
wharf to granary = (250 - 8)/30 > 8.06 > rounded down > 8 tiles

0: 0*30+210= 210> 250// (250 - 8)/30 > 08.06 > 08// 08-0= 8
1: 1*30+210= 240> 250// (250 - 8)/30 > 08.06 > 08// 08-1= 7
2: 2*30+210= 270> 300// (300 - 8)/30 > 09.73 > 09// 09-2= 7
3: 3*30+210= 300> 350// (350 - 8)/30 > 11.40 > 11// 11-3= 8
4: 4*30+210= 330> 350// (350 - 8)/30 > 11.40 > 11// 11-4= 7
5: 5*30+210= 360> 400// (400 - 8)/30 > 13.06 > 13// 13-5= 8
6: 6*30+210= 390> 400// (400 - 8)/30 > 13.06 > 13// 13-6= 7
7: 7*30+210= 420> 450// (450 - 8)/30 > 14.73 > 14// 14-7= 7
8: 8*30+210= 450> 500// (500 - 8)/30 > 16.40 > 16// 16-8= 8
9: 9*30+210= 480> 500// (500 - 8)/30 > 16.40 > 16// 16-9= 7

It looks like the production can be more for single granary blocks because the distance from wharf to granary can be 7 tiles in all cases, not 6 like I used above. It doesn't make a difference for two granary blocks though.

[This message has been edited by Philon (edited 07-09-2012 @ 07:40 PM).]