推 WillWaiting:Computer吧 06/14 12:41
推 greengreen42:cumpoter 06/14 13:00
推 Jasonsu:綠綠大..你不是認真的吧！Computer...... 06/14 15:33
推 tsming:cum put her ？XD 06/14 16:05
推 Kati:soga... 06/14 22:00
※ 引述《Kati (聚時歡樂，散後尤其冷清)》之銘言：
: 看到這邊有個疑問請教板上高手
: 日職到美職有Eqa的存在，
: 那國聯和美聯之間呢？
: 如果有的話又是怎樣算出來的
: 或是更誇張一點
: 兩聯盟間各分區的強度不同(純主觀)
: 那選手的成績有沒有Eqa的問題
: 並非要求個數字，
: 只是好奇這東西在地球上有沒有人討論
: (絕對是誠心發問，不要變成戰文XD)
sample size is problem.
small sample lets stats dirty and have many noises.
Eqa need long term stats to cover those problem.
MLB to NPB transfer has a lot of sample but PECOTA system
is still not 100% correct(even cant reach 80%).
so division transfer is lack of sample power,
and the most important is,there is not any confident evidence
can prove division difference is conspicuous enough XD.
But league difference like AL and NL exists and conspicuous.
--
my compoter has some problem,so i use my poor english to post my poor point.
--
※ 發信站: 批踢踢實業坊(ptt.cc)
◆ From: 218.170.71.81
※ 編輯: greengreen42 來自: 218.170.71.81 (06/14 12:03)
※ 編輯: greengreen42 來自: 218.170.71.81 (06/14 12:04)