Re: [觀點] 一些有趣的數據

[iIvan (go O out)]

: 老實說這樣本數真的有點小，
: 其實所有刻意強調「關鍵時刻」的數據往往都有這個問題。XD
: 靠印象的話，本季的live文裡一再出現在關鍵時刻為何要放上防守黑洞的聲音，
: 所以或許「場上組合說」成立的機會比「運氣說」還來得小一點。
樣本數小沒錯阿，所以我支持運氣說XD

除了人工慢慢統計外，我找不到整理好的不同時段場上組合可以看，
所以場上組合說目前也只能靠印象。

如果有完整一點的數據（例如可以跨季比較）可以參考的話，
關鍵時刻發功說會比較有說服力。
: ElGee最近連續寫了好幾篇很有趣的文章在講這個問題。
: http://www.backpicks.com/2012/03/05/sloan-thoughts-revisiting-late-game-bias/
: Sloan Thoughts: Revisiting Late-Game Bias
: 接下來連續三篇分析「The Crunch Time Myth」
: http://www.backpicks.com/2012/03/17/the-crunch-time-myth-why-you-dont-need-a-closer-to-win/
: http://www.backpicks.com/2012/03/20/the-crunch-time-myth-why-closers-and-clutch-shooting-are-overrated-part-ii/
: http://www.backpicks.com/2012/03/23/the-crunch-time-myth-part-iii-overrating-closers-and-clutch-offense/
: 文章很長而且非常數據取向，我也還沒看完，
: 不過應該從標題就可以猜到他的結論了。
: 這邊貼一下他最後的總結：
: To recap the last three posts:
: On a league-wide level:
: ●Isolation basketball is the most inefficient offensive strategy
: ●Teams rack up fewer assists in crunch time
: ●Teams shoot poorer in the clutch (Probably, in some part at least, because of
: the first two bullet points.)
: Championship Teams:
: ●Can win without a closer
: ●Can win without shooting well in crunch time
: ●Can win without improving their shooting in crunch time
: On the relationship between clutch shooting and team success:
: ●There is no correlation (-0.04) between how large of a share of the clutch
: field goals one player makes and overall team eFG% in the clutch.
: ●There is only a small correlation between outperforming expected wins and
: clutch shooting (0.21)
: ●The correlation between non-clutch eFG% and wins (0.56) is much larger than
: clutch eFG% and wins (-0.32)
: ●There is almost no correlation between ast% and clutch shooting (0.15)
: ●Even the best teams only win about two extra games per year by outperforming
: their predicted records based on points
: ●The best clutch shooting teams are still upset in the playoffs – some miss
: the playoffs altogether
: ●Only one elite clutch shooting team has won a championship in the last 11
: years (2011 Dallas Mavericks)
: The conclusion: While it always helps to have great offensive players, and
: those who play well down the stretch, it is by no means necessary to win in
: basketball. Rarely, offensive players can increase their performance so much
: they spark a crunch time increase in team offense. (LeBron in Cleveland,
: Chris Paul, and arguably  Steve Nash are the only recent players to fit that
: bill.) But team offense still matters. Defense is a huge part of the game.
: The first 43 minutes decide most of the outcome.
這裡的結論有些的確也是很好想像與理解的，
所謂的強隊（冠軍隊），需要與對手拉鋸到靠關鍵時刻的爆發的比賽其實不多，
我實力比你強在進入關鍵時刻前就把分數拉開到安全距離了，
只要保持既有節奏打就會贏，何必搞什麼關鍵時刻的爆發？
: 另外他也寫了一篇討論到底進攻能不能奪冠的文章，
: 看起來他蠻支持老波的觀點XD。
: http://www.backpicks.com/2012/03/12/defense-wins-championships-except-offense-is-more-important/
: Defense Wins Championships…Except Offense Is More Important
這篇作為立論基礎的數據其實是有問題的，
他比較一支冠軍隊的優勢是進攻還是防守的依據是：
(該隊Off Rtg-聯盟平均Off Rtg)與(該隊Def Rtg-聯盟平均Def Rtg）分別取絕對值後比較
若是前者大就當作進攻為其優勢，反之亦然。

聯盟平均數這種東西很容易受到極端值的影響，
而且就以去年的冠軍小牛來說，兩個數據只差0.3就被歸類為進攻型冠軍隊，
但事實上去年的小牛在Off Rtg及Def Rtg的聯盟排名同樣都是第八，
很難說是靠進攻還是防守在贏球。

--
※ 發信站: 批踢踢實業坊(ptt.cc) 
◆ From: 118.166.229.21
推 Alfred:等有空的時候再把文章的摘要放上來好了。不過我的確也不是
→ Alfred:完全同意他使用進攻跟防守優勢的方式，有很多判定其實可以
→ Alfred:有討論空間，我覺得更好的結論應該是攻守均衡的球隊才能贏
推 Alfred:球。數據中另一個有趣的trend是近十年來防守的影響力增加了
→ Alfred:老波現在這樣逆風到底有沒有機會高灰呢？XD
→ Alfred:不過我倒是很同意他對於關鍵時刻的分析，不是說關鍵時刻真
→ Alfred:的不重要或無所謂，但現在太多人談論這個東西的方式都過度
→ Alfred:強調關鍵時刻的重要性，特別是炒作新聞，這會阻礙對球賽的
→ Alfred:的理解。
→ iIvan:要說近十年其實還是有希望的，2001季後賽的神鬼湖人隊例行賽
→ iIvan:的數據顯示是支進攻火力強大但防守卻淪落到聯盟後1/3的程度
→ iIvan:大概老波在這支馬刺隊身上看到了類似的潛能（誤）...XD
→ dogville:2001的湖人進攻太強大了