Welcome to the moe-data.github.io wiki!

---

What is moe-data?

This is a wiki for the moe-data project. It is a data analyze tool of Kantai Collection.

This project collects data from the game and analyze it.

The main function of this website is multi-equipment development querying and unpacking data viewing.

Why moe-data?

The fascinating point of Kantai Collection is construction warships by costing resources.

Different recipe of resources can build different warships.

However, The game publisher never reveals the relationship between the recipes and construction results officially.

Moe-data is a project to analyze the relationship between recipes and construction results from big data generated by players.

How it works?

The project is based on the data collected from the players.

The data is collected by a game plugin. It collects EVERY recipe and the construction results used by players from the game.

The server calculates each probability of the construction results with multi-ship construction/ multi-equipment development and responds to every query from the website.

How to use moe-data?

visit moe-data.github.io and click the button Search(查询) .

Then you can see the information of warship recipes.

What is Kantai Collection?

Kantai Collection is a Japanese free-to-play browser game developed by Kadokawa Games. It is a naval warfare themed game, where the player commands a fleet of warships from the Imperial Japanese Navy during World War II.

the player can collect the warships as cards. The warships are called "Kantai" in Japanese. The game is also known as KanColle.

---

FAQ

和 POI-DB 有什么区别？

1、多装备概率叠加统计。
2、相同公式下，对秘书舰进行分类，不同旗舰出货率显示。
3、时间段筛选（对新装备很有用）。
4、可查看失败率（在下方设置里选择，部分提督希望开发公式失败率高以节省资材）

秘书舰如何分类？

请见 秘书舰分类表

数据来源？

poi 为本站提供数据支持。
欢迎使用poi浏览器来为本站的数据添砖加瓦
如果您使用的是poi浏览器
您的数据将在次月1~5日匿名同步到本网站

查询结果一直在转圈？

在设置里减少选择的时间段或换个浏览器

---

Technical Details

Raw data is too large for putting on github pages

Here shows how to convert data from poi to the parsed json in moe-data

Data source comes from https://github.com/poooi/poi-server/wiki

1. download the .gz files

Construction

URL: http://[SERVER_NAME]/api/report/v2/create_ship

Field	Type	Explanation
items	[Number]	Items for construction, from api_item1 to api_item5 in KanColle game api
kdockId	Number	Dock ID, examples: [0, 3], note the index is 0-based, if you get it from KanColle API's api_kdock_id field, you need to minus 1 from it.
secretary	Number	Secretaryship's ship_id
shipId	Number	Result ship id
highspeed	Number	use highspeed. [0, 1, 10]
teitokuLv	Number	Game player's level
largeFlag	Boolean	Set to true if it was LSC(大型舰建造)
origin	String	(Optional) It's the name of your report plugin, requests' User-Agent will be used if not present

Development

URL: http://[SERVER_NAME]/api/report/v2/create_item

Field	Type	Explanation
items	[Number]	(Same as construction)
secretary	Number	(Same as construction)
itemId	Number	api_slot_item.api_slotitem_id on success. If development failed, api_fdata will be a comma separated list, get the second element from it (e.g. if api_fdata = "2,33", you should fill this field with 33)
teitokuLv	Number	(Same as construction)
successful	Boolean	(Successful or unsuccessful.)
origin	String	(Same as construction)

2. convert gz to SQL.(tool: mongodb)

I have renamed the files
the json file is named like "d1.json" or "c1.json"
where d for develop, c for construction.
and the number indicates date:

Number	filename(in poi)	date	days	size
1	createitemrecords-5cf1c000.gz	2019/6/1		188417481
2	createitemrecords-5d194d00.gz	2019/7/6	35	13895626
2	createitemrecords-5d422b80.gz	2019/8/1	26	11084170
2	createitemrecords-5d929700.gz	2019/10/15	75	9712275
2	createitemrecords-5dbb7580.gz	2019/11/1	17	2986889
2	createitemrecords-5de30280.gz	2019/12/1	30	4747394
2	createitemrecords-5e0be100.gz	2020/1/1	31	3395150
2	createitemrecords-5e34bf80.gz	2020/2/1	31	6903967
2	createitemrecords-5e5afb00.gz	2020/3/1	29	8088439
2	createitemrecords-5e83d980.gz	2020/4/1	31	7904881
3	createitemrecords-5eab6680.gz	2020/5/1	30	7095492
3	createitemrecords-5ed44500.gz	2020/6/1	31	6933353
3	createitemrecords-5efbd200.gz	2020/7/1	30	6568448
3	createitemrecords-5f24b080.gz	2020/8/1	31	7322584
3	createitemrecords-5f4d8f00.gz	2020/9/1	31	7542877
8	createitemrecords-5f9dfa80.gz	2020/11/3	63
9	createitemrecords-5fc58780.gz	2020/12/3	30
10	5fee6600.gz	2021/1/1	29
11		2021/2/1	31
12		2021/3/1	28
13		2021/4/1	31

I seperated them according to the release date of items
(I didn't seperate it for ship construcitons since don't know the release date, here I need help)

          "2020/11/13":[78,147],
          "2020/10/16":[194,242,249,250],
          "2020/03/27":[120],
          "2020/02/07":[51,90,207,226],
          "2017/03/17":[168],
          "2016/06/30":[181],
          "2016/03/11":[49],
          "2016/02/29":[163],
          "2015/06/12":[65,66],
          "2014/07/04":[73],
          "2014/06/06":[72],
          "2014/05/23":[61],
          "2014/03/14":[59],
          "2014/02/26":[60,75],
          "2013/08/26":[47,57],
          "2013/05/15":[54,55],
          "2013/05/08":[44,45,46,52],

3. run SQL query:(tool: SQL)

UPDATE d2 set itemId=-1 WHERE successful="false";
UPDATE d2 set itemId=-1 WHERE itemId=0;
DELETE FROM d2 WHERE secretary=0;

4. run SQL query:

the table cstype is from https://github.com/moe-data/moe-data.github.io/blob/master/parsed/cstype.json

for develop:

SELECT * FROM(
select ifnull(items,-2) i, ifnull(b.dtype,ifnull(secretary,-2)) s, ifnull(itemid,-2) o, ifnull(count(*),-2) n, ifnull(min(teitokuLv),-2) l
from d2 LEFT JOIN cstype b on secretary=b.id
group by items,itemid,b.dtype
ORDER BY items,b.dtype,itemid)t
WHERE s!=-2 and i!='[null,null,null,null]' and i!='' and i!='[]'

for construction:

SELECT * FROM(
select ifnull(items,-2) i, ifnull(secretary,-2) s, ifnull(shipid,-2) o, ifnull(count(*),-2) n, ifnull(min(teitokuLv),-2) l
from c10
group by items,secretary,shipid
ORDER BY items,secretary,shipid) t
WHERE s!=-2 and i!='[null,null,null,null]' and i!='' and i!='[]'

5. export the query result in 4 to json format

---