讓台灣棒球的數據分析與美國齊頭並進

Improving Taiwanese Baseball Data Analysis Goes Hand in Hand with U.S.A

棒球打擊者能力的評估方法眾多，最常使用的傳統指標莫過於打擊率，一般而言，打擊率高的球員被認定打擊能力優秀，對職業棒球選手而言，更具有談薪的優勢，然而這一切的理所當然，建立於打擊率的可靠性，此文章想以數據分析提出更全面的評估方法，試想，兩個打擊率相同的球員，但其中一位更常出現雙殺打，是否應該有不同的評價？本文章所提出的新指標概念為：改善 D’Esopo 和 Lefkowitz （1977）的跑者推進模型，逐打席檢查此新的推進模型與正式結果的吻合情形，接著，選取固定的 2017 年明星賽前中華職棒富邦悍將打席次數最多的 8 人名單，在所登錄的一軍 26 位打者名單當中，逐位選取 1 人加入以形成 9 人完整打線，執行 26 次新的跑者推進模型，並排序此 26 個值即可評估特定打者貢獻得分的能力。並提出特定的出局數、壘上狀況、對戰中信兄弟投手的不同情況，呈現富邦打者的能力排序，所得的結論為：在所有的情況下，楊承駿為能力排序第一的球員。得點圈有人的情況下是胡金龍。在主力投打評估方面，面對投手鄭凱文時，胡金龍表現佳。投手為鄭錡鴻時，林哲瑄表現佳。投手為羅曼時，林益全表現佳。 There are many ways to evaluate the ability of baseball batters, the most commonly used traditional indexes is AVG. In general, high AVG batters were identified as excellent players. The professional baseball players also have the advantages of salaries. However, the reliability of AVG is questionable. This article does not negate the effect of traditional indexes, but propose a more comprehensive assessment index. The concept of the new index proposed in this article is to improve the runner advancement model by D'Esopo and Lefkowitz (1977). We also check the coincidence of the new model and the formal results. In the list of 26 roster players, select one of them to a fixed 8 batters one by one. Finally, we obtain 26 values from the new model and sort them to evaluate the ability of batters. Moreover, we rank players by dividing game conditions into the location of runners on base, outs, and pitchers. The conclusions are: Yang is the best batter all the times. Hu is the best batter in scoring position. When batters face pitcher Kevin Cheng, Hu is the best batter. When face pitcher Cheng, Che-Hsuan Lin is the best batter. When face pitcher Roman, YiQuan Lin is the best batter.

There are many ways to evaluate the ability of baseball batters, the most commonly used traditional indexes is AVG. In general, high AVG batters were identified as excellent players. The professional baseball players also have the advantages of salaries. However, the reliability of AVG is questionable. This article does not negate the effect of traditional indexes, but propose a more comprehensive assessment index. The concept of the new index proposed in this article is to improve the runner advancement model by D'Esopo and Lefkowitz (1977). We also check the coincidence of the new model and the formal results. In the list of 26 roster players, select one of them to a fixed 8 batters one by one. Finally, we obtain 26 values from the new model and sort them to evaluate the ability of batters. Moreover, we rank players by dividing game conditions into the location of runners on base, outs, and pitchers. The conclusions are: Yang is the best batter all the times. Hu is the best batter in scoring position. When batters face pitcher Kevin Cheng, Hu is the best batter. When face pitcher Cheng, Che-Hsuan Lin is the best batter. When face pitcher Roman, YiQuan Lin is the best batter.