Playing around with the Slavs this week I started to wonder about how their farming bonus size changes during the game.
Slavs benefit more when farmers spend more time gathering and less time walking (in a hypothetical situation when vills don't have to walk at all Slavs would work ~28% faster rather than ~15%).
What happens to the walking time proportion when the game drags on? On the one hand, farms tend to be placed less and less efficiently. On the other hand, wheelbarrow and handcart push in the opposite direction.
I decided to gather the statistics from KOTD2 games and here are the results of my research.
Data
From each game, I extracted a lot of "farming trips" - periods of time when a farmer goes from the TC/mill to the farm, collects food and returns back to drop it off. For each trip I calculated:
Results
For the charts below:
Overall (generic civs):
Before wheelbarrow:
After wheelbarrow, before handcart:
After handcart:
Timestamp vs distance
Timestamp vs gathering time proportion
Timestamp vs efficiency
Thick lines - average efficiency; thin lines - median efficiency.
Blue: generic civ (all except Slavs, Aztecs, Mayans and Berbers); red: Slavs, green: Aztecs, brown: Vikings.
Conclusions
Slavs benefit more when farmers spend more time gathering and less time walking (in a hypothetical situation when vills don't have to walk at all Slavs would work ~28% faster rather than ~15%).
What happens to the walking time proportion when the game drags on? On the one hand, farms tend to be placed less and less efficiently. On the other hand, wheelbarrow and handcart push in the opposite direction.
I decided to gather the statistics from KOTD2 games and here are the results of my research.
Data
From each game, I extracted a lot of "farming trips" - periods of time when a farmer goes from the TC/mill to the farm, collects food and returns back to drop it off. For each trip I calculated:
- Timestamp - an average in-game time when the trip takes place
- Total time - duration of a trip
- Distance - an average distance from the dropoff point to the locations where a vill collects food
- Gathering time - time that a farmer spends gathering
- Capacity - an amount of food dropped off
- Efficiency - Capacity/Total time
Results
For the charts below:
- All the trips were separated into 10 equal size groups by X-value
- For each group calculated
- Average X-value
- Average Y-value
- Percentiles of Y value: 10, 20, … 50(median), …, 80, 90
- Thick lines represent the average Y value
- Thin lines represent percentiles
Overall (generic civs):
Before wheelbarrow:
After wheelbarrow, before handcart:
After handcart:
Timestamp vs distance
Timestamp vs gathering time proportion
Timestamp vs efficiency
Thick lines - average efficiency; thin lines - median efficiency.
Blue: generic civ (all except Slavs, Aztecs, Mayans and Berbers); red: Slavs, green: Aztecs, brown: Vikings.
Conclusions
- Average trip distance jumps from ~1.9 early on to ~2.3 on 25 mins and then slowly increases after that.
- Wheelbarrow increases the gathering rate by ~10%
- The effect of hand cart is slightly larger: ~13%
- Wheelbarrow and handcart influence the gathering time percentage more than inefficient farm placement. So, answering the initial question: yes, the Slav bonus is stronger later on. The size of the bonus is about 1-1/(0.54/1.28+(1-0.54)) = ~13% by minute 15; 1-1/(0.6/1.28+(1-0.6)) = ~15% by minute 30 and 1-1/(0.67/1.28+(1-0.67)) = ~17% by minute 45.
- Aztec extra carry capacity functions like a “demo version” of the Slav bonus: its effect is comparable (still slightly smaller) in feudal age but Aztec farmers become almost generic in the lategame.
- Vikings' free upgrades work similarly to the Aztec bonus. There is a difference though: Viking farming efficiency reaches its maximum upon hitting castle age.
- In general, average farming efficiency is smaller than the median one. This happens due to the tricky distribution of the farming distance.
Last edited: