[外絮] atthehive 18場比賽數據分析

[akainorei (赤いの靈)]

http://tinyurl.com/nuygksf

18 games into the season, what trends we can see

Eighteen games in, sample size still small, but worth to take a look already.
18場比賽, 數據還不是很足夠, 但可以拿來檢視一下了

Again, all numbers were coming from NBA.com. I pulled the game log numbers, and
plot individual player numbers vs. game date, and trying to look for any
meaningful trend or split. I did found something worth to talk about. BTW, I
only look into rotation players, and sorry, I did not include PJ.
數據全來自NBA, 我只看輪值球員, PJ不在裡面抱歉

1. Play time.
上場時間

In this graph, player's play time was graphed against game date, red dot
indicate winning games and blue ones were games we lost. I connected dots with
lines to have an indication of game sequence. I could see the following:
紅點就是贏球, 藍點是輸球, 把這些連起來就有一些順序

Big AI and Spencer Hawes get the most consistent PT with minimum variation.
Big Al 跟 Hawes 的上場時間很穩定沒有啥改變
Other than the game AI got hurt thus played single digit min, he played steady
min. And Hawes almost fall out of rotation, but after AI got injured, he went
back to his 17 min routine.
Big Al 有一場受傷所以上場時間不多外都很穩定, Hawes 原本要調出輪值外
Big Al 受傷就重新站回每場17分鐘
Look at Kaminsky's PT one can easily tell when he cracked into the rotation,
and once he is in, he stays there.
看司機的時間你就可以觀察到最近他上場越來越多, 而且時間很穩定
After Lamb got his contract extension, his PT never dropped under 20 min.
羊一拿到合約後他的時間沒有低於20min
Kemba/Marvin/Batum were our highest min player, and Kemba's minutes were
trending upwards.
砍八/Marvin/巴圖是上場時間最久的球員, 砍八有上去的趨勢
Cody and Lin were the players had the biggest PT swing without clear trend or
correlation to something that I'm aware of. Literally they had back to back
games with 15 min play time swing up or down. Looks like those two players do
not have steady role on this team yet, thus got adjusted back and forth
frequently, unfortunately.

Cody 跟林的時間是很劇烈晃動, 沒有一個明顯的趨勢, 基本上就是15分鐘然後一下飆高
一下又往下. 看來這兩個球員目前在隊伍上還沒有一個穩定的定位, 所以才會
一直調整上場時間, 很遺憾
No player's PT change has clear correlation to team win or loose, maybe Lamb a
little bit, when Lamb play more than 25 min/game, we were 6-1. However, he was
not playing much in the first three loosing games could screw the split a
little bit.
沒有一個球員的上場時間跟勝負非常有關, 可能羊有吧
只要他打超過25分鐘, 戰績是6-1. 但因為賽季最初三場他沒有打
這樣有可能會影響統計
2. Field Goal Attempt:

I think FGA is a rough indication of scorer priority.
我認為FGA可以大概推估出球員出手的優先值

AI/Kemba/Batum were the top three priorities on this team to score. Followed by
the Jeremys.
Al/Kemba/Batum 三大優先, 再來就是兩個Jeremy
For games Kemba had more than 14 FGA, we were 2- 7, for games Kemba had less
than 14 FGA, we were 8-1. I do think this split is significant. Interesting
thing is, every game we loose except for the first 3, Kemba will dial back his
FGA to under 14
只要Kemba 出手超過14次, 戰績2-7, 沒超過戰績是8-1.
這是很明顯的數據, 有趣的是(除了最初三場) 只要上一場輸了
下一場Kema 就不會出手太多
Not sure whether it was coincidence or causation, we win all the games AI shoot
more than 18 or Lin shoot 11 times or above. We loose all the games Hawes shoot
more than 8.
不知道這是不是巧合還是還是真的有關係, 只要AL 出手超過18次或林出手超過11次
比賽就會贏, 只要Hawes 出手超過8次就會輸
This one is fun, we were 6-0 when Lamb shoot exactly 10 times. Give it a little
wiggle room, we were 8-1 when Lamb shoot 10 +/- 1 FGA.
Lin's FGA had a sharp drop in recent games, correlates to his play time too.
Not sure that's a trend or variation.
還有這個有趣的數據, 只要羊出手10次(不多不少) 戰績6-0
如果出手10 +/- 一次, 戰績8-1

林最近的出手數急速下降, 跟他上場時間吻合
不知道這是一個趨勢還是只是暫時的

3. Field Goal Percentage:

FG% is looked at to see how stable an player is playing at.
命中率可以看出球員的穩定度

In the graph, the solid line indicate the player's average FG% and the dotted
line were the FG% average +/- SD. The range indicated the game to game
variation.
我懶得翻這句XD
For Zeller/Kaminsky/Hawes, since they don't shoot much, thus their FG% could
have bigger swing, that's normal.
Cody/司機/Hawes, 因為出手數不多, 命中率變化很大, 正常
AI had bigger variation than I anticipated, since he shoots closer to the rim,
thus I anticipate he had higher and more steady FG%, He had quite some games
shooting at lower than 50%.
AL 的命中率變化比我想像中的大, 因為他靠近籃框我認為他有一個穩定又高的命中率
但他有好多比賽命中率低於50%
Lamb has apparent bigger variation but that's due to couple of his crazy
efficient games with more than 70% FG%, so that's all good. When Lamb is
shooting at 50% FG% or better, we had 8-2 record. The fact that he had 10 out
of 17 games he played with higher than 50% FG% is pretty impressive.
羊的命中率變化很大, 因為他有幾場超級瘋狂的比賽(命中率超過七成)
當他的命中率超過50%, 戰績8-2.
這17場比賽中有10場命中率超過五成很引人注目
Lin wasn't shooting as good as his previous seasons, but somehow he was
shooting with the smallest variation on the team.
林的命中率沒有熱身賽時來得好, 但他是最穩定的

Kemba is the opposite to Lin. He is shooting the best percentage in his career,
but it could go as hot as 70% or as cold as 10%, or any place in between.
However, with his current FG% and 3FG%, I will take it.
Kemba 完全相反, 目前是生涯命中率最好, 但有時七成有時一成 (或者在這兩個
極端值中間). 但是, 看他目前的命中率跟三分命中率, 我可以接受
4. Plus minus:
+/- 值
Some people love to quote +/- for a single game, and some people just totally
ignore it, but I do think when you put them together collectively, it could
still mean something.
有些人喜歡看一場就做結論, 有些人完全不管
我認為看連續好幾場是可以看出一點端倪

The clearest trend is Jeremy Lin. For all the games he got more than +5 in +/-,
we win. For all the games he got less than +5 in +/-, we loose!
最明顯的是林. 只要他+/-高於5, 我們就會贏
只要他低於5, 我們就輸!
Cody Zeller had equally clear split, just the split is at -6. When he had
better than -6 in +/-, we win, worst than that, we loose.
Cody 也是, 只是分界點是-6
只要他高於-6我們就會贏
In all the winning games but 1, Lamb had positive +/-, in all the loosing games
but 1, Lamb had negative +/-
在所有贏的比賽裡(除了一場) 羊的數值是正的
在所有輸的比賽裡(除了一場) 數值是負的
In all the winning games but 2, Hawes had positive +/-, and in all the loosing
games but 2, Hawes had negative +/-
在所有贏的比賽裡(除了2場) Hawes的數值是正的
在所有輸的比賽裡(除了2場) 數值是負的
So if you don't trust individual +/- number, look it collectively, don't you
appreciate more about our BENCH FORCE ONE?
如果你不相信正負值, 就把所有比賽捆再一起看
覺得我們的板凳很爛嗎? 再看清楚一點
I think this is the cleanest split I have seen in this 18 games data.

--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 27.246.131.142
※ 文章網址: https://www.ptt.cc/bbs/Jeremy_Lin/M.1449549365.A.7A4.html
推 gold97972000: 感謝翻譯分享
推 casman: 原文竟然一直寫 "Big AI" XD
→ casman: 感謝翻譯~~~
→ akainorei: 我的結論是，因為砍八今年狀況好，加上一堆新人，教練
→ akainorei: 還在摸索如何運用林。羊的定位清楚，就是投。Cody and
→ akainorei:  Lin教練還在嘗試
→ littlehouse: 推，球迷還是可以寫出很棒的分析文的:)
推 gidapops: 根本不用摸 林是USB 甚麼都能插 怎樣都好用
推 djviva: 推, 也推AI(x)是Al啦哈哈
推 aaron123: 我是覺得羊在決勝時刻的防守實在太差了 不然他真的很準
推 casman: 林的使用方法一直很明確, 給他球權給他時間, 就會給你成效
推 gold97972000: 所以球給Lin控 Lin多出手 對球隊有幫助
→ gold97972000: 話說今天升上第5了
→ akainorei: 其實不難啦，砍八出手少一點，時間分個五分鐘給林，林
→ akainorei: 出手最少11次就好。
→ aaron123: 另外林真的是進攻組織的第一選擇 沒問題的
→ aaron123: 穩又有效率 可惜是命中率不好
→ akainorei: 寇弟現在是吃big al時間，等big al回來，教練如何配置
→ akainorei: 寇弟就挺傷腦筋的
推 WEDBLACK: Big data 的概念 酷
推 gidapops: 跟今天類似 batum多打 walker 30+- lin 26+-
→ roger7: "previous seasons"應該是之前的球季
→ gidapops: lamb 24~26 別多 關鍵時刻拼防守
→ casman: 勝率其實可以排到東區第四了, 跟第一名騎士只有一場勝差!
推 gratitude: 代表今年找Lin來有助Walker的表現
→ akainorei: 哦對，感謝更正，我趁中午休息快速翻一下而已
推 inschool: 就只怕lin自己習慣了USB用法 獲得穩定空間反而表現下滑
→ inschool: 希望只是杞人憂天而已
推 gidapops: 2上4上都給林控 別怕 重點是要掛一個zeller lamb
→ akainorei: 出手順序是前五也就是板凳前二選擇，沒啥好擔心，重點
→ akainorei: 是命中率，最近真的很差
推 turbomons: 感謝翻譯 蠻有意思的
→ peggie: 作者22NN4是林迷喔，按照酸酸的標準，本文可信度先打3折XD
→ bluesunflowe: 只有三折嗎??未看先猜 數據正確 解讀錯誤
推 minishrike: Lamb之前不太準，現在準鳩鏡是因為簽約還是純進步呢?
推 b551122: 感謝翻譯!!
推 jt: 感謝翻譯
推 avril0625: 謝謝翻譯
→ OCEANBOY7329: 東區競爭蠻像過往西區.. 第一跟第八勝差都沒很大，
→ OCEANBOY7329: 林感覺真的對自己信心還是沒找回尼克時期那樣.. 希
→ OCEANBOY7329: 望時間跟教練能幫助他找回來吧
推 yachuyachu33: 謝謝翻譯
推 xajx: Al其實已經不太適合現在的黃蜂了.
推 xajx: Lin這季上場時間是22.9分, 場均出手8.4. 以這上場時間來說,
→ xajx: 出手次數不會特別少. 比較需要提升的是命中率.
→ xajx: 至於上場時間就不是Lin可以決定的了.
推 miha80425: 黃蜂現在不缺進攻手 Al的低位進攻就沒那麼必要
→ miha80425:             *得分手