How Speed Affects Isolated Power by Jeff Zimmerman March 14, 2014 Billy Butler and Ben Zobrist had a very different mix of hits (pics via Keith Allison). People define a hitter’s power in many different forms, with Isolated Power (ISO) being a fairly common method. Isolated Power looks at the bases beyond first base that a player gets per at bat. While how far and hard a player hits a ball factors into his ISO, so does speed. If a player is able to turn a hit into the gap into a double or a hit to the wall into a triple, that depends a lot on his legs. I was able to find out how much speed factors into a player’s ISO value. Let’s start by looking at two players. Butler vs. Zobrist Comparison Player AB PA H 1B 2B 3B HR ISO Billy Butler 582 668 168 126 27 0 15 .124 Ben Zobrist 612 698 168 117 36 3 12 .127 Watching each play last season, Butler seemed like he hit the ball harder and further. The ISO values seems to say otherwise. Each player had 168 hits and while Butler was able to hit three more home runs, Zobrist had nine more doubles and three more triples. Here is their ISO split apart into the double, triple and home run components. With inside the park home runs so rare, just the effect of extra doubles and triples can be changed by a player’s speed. Butler vs. Zobrist Comparison Player 2B & 3B ISO HR ISO ISO Billy Butler .046 .077 .124 Ben Zobrist .069 .059 .127 Even though the players had almost identical ISO values, Butler’s is home run driven while Zobrist’s comes largely from doubles and triples. To show the difference in speed, here are Butler and Zobrist hitting balls into the outfield gap. While the hits aren’t exactly the same, the difference in player speed is plainly obvious. Butler took about 6.3 seconds to get around first base. Zobrist was standing up at second in 8.4 seconds. I wanted to find a good and easy rule of thumb for showing how much Zobrist’s speed contributed to his power. I cut and diced the data a hundred different ways. I need to find how I tried to use the ratio of singles to doubles and doubles to triples to measure speed. I had in my mind this was the best method. In the end, the main issue ended up being not enough triples. I tried adding stolen bases, nothing. I was trying hard to stay away from Ultimate Base Running or Speed Score because of the complexity in calculating the values. But I eventually had to give and found a workable solution. Some really not so gory math: From the past 12 seasons, I took all the players with a minimum 400 plate appearances and grouped then by Speed Score. The Speed Scores were then grouped into different buckets, 0 to 1, 1 to 2, 2 to 3 and so on. Initially, I didn’t want to use Speed Score because one of its five inputs is triples and I worried about too much overlap. Eventually, it was needed to find a solution. I averaged the bucketed players’ doubles and triples2B&3B ISOs to get the following results: Speed Score Effects on ISO Speed Score Range Avg. Speed Score 2B & 3B ISO < 1 0.8 .064 >= 1 and < 2 1.6 .073 >= 2 and < 3 2.5 .079 >= 3 and < 4 3.5 .083 >= 4 and < 5 4.5 .087 >= 5 and < 6 5.4 .089 >= 6 and < 7 6.4 .094 >= 7 and < 8 7.5 .100 >= 8 8.3 .107 So for every increase or decrease in a player’s Speed Score by one, a hitter’s doubles and triples ISO changes 0.005 on average in the same direction. While the player’s overall ISO should change also, his home run total may mask the speed based change. Take Jose Reyes for example. In 2012 his ISO was .146 and it dropped to .131 in 2013 even through he increased his home run rate by more than 50 percent. The problem was a early season twisted ankle which limited his mobility. His Speed Score went from 7.1 to 4.4. His doubles and triples ISO dropped from .095 (.091 career value) to .052. So adjusting his ISO up because of the drop in speed, Reyes would see his 2013 ISO stand at .144, almost the same as his previous season’s value. Looking just at Reyes’ ISO, it may seem like he lost some power. After taking his drop in speed into account, his power stayed the same between the two seasons. Some other players saw their doubles and triples ISO drop as they slowed down. Micheal Bourn’s Speed score dropped from 7.5 to 6.5 and his doubles-triples ISO dropped 0.011. The speed-based power drop was a little more than expected. The same happened to Coco Crisp, but his home run power spike masked the decline. Crisp’s ISO jumped from .158 to .183 while his Speed Score dropped 8.5 to 6.5. Dividing his ISO out, the doubles-triples component dropped from .085 to .054. A player can see his ISO get pulled in two directions as his strength increases but his speed declines. To understand how each component ages, I created an aging curve with doubles and triples and home run ISO components. Additionally, I added plain old ISO. The doubles-triples component begins to age immediately. This drop is not a surprise since speed peaks in the early 20s. The doubles-triples curve then slowly begins a consistent decline. The home run component plateaus from age 22 to age 30. After age 30, home runs begins a steep descent. For example, here are the ages 29-31 home run leaders from 2008 and here is what they did in 2013: Ages 29-31 Home Run Leaders, 2008 Name 2008 HR PA per HR, 2008 PA per HR, 2013 Ryan Ludwick 37 6.0% 1.4% Jack Cust 33 5.5% Gone Chase Utley 33 4.6% 3.4% Aubrey Huff 32 4.8% Gone Jason Bay 31 4.6% 4.7% Carlos Pena 31 5.1% 2.4% Kevin Youkilis 29 4.7% 1.7% Only Bay was able to maintain a constant homer rate, but the rest of his game was substandard (.298 OBP). With players aging at a different rates, a person can use the speed component of ISO to determine if an ISO decline is power or speed related. I am going to help save the baseball world by not creating another stat to remember. There is not enough reason to adjust ISO for every player depending on his speed. Rather, it is important to remember that speed is a factor of ISO and if a player’s speed drops quite a bit, his power is likely to go down. Going back to our original two players, Butler and Zobrist: In 2013, Zobist had a Speed Score of 4.4 while Butler was at 0.8. Butler’s Speed Score was 3.6 points lower than Zobrist’s which equals an .018 lower doubles and triples ISO. Adding .018 to Butler’s .046 doubles-triples ISO, it jumps to .064, or about the same as Zobrist’s .069 doubles-triples ISO. After adding in the home run ISO values, here is how the two compare Butler vs. Zobrist Comparison Name Adjusted ISO Billy Butler .142 Ben Zobrist .127 Butler’s sloth-like speed makes it look like he and Zobrist hit at the same power level. If they both could run at the same speed, Butler’s ISO would be higher. Isolated Power (ISO) is a decent measure of a player’s power. The problem is that there is some speed component to ISO that needs to be taken into account with extra base hits. By looking at a player’s Speed Score and the doubles-triples component of ISO, a Speed Score increase or drop by one equals a .005 increase or drop in the same direction for ISO. If a player’s power is on the decline, a person can use the information to determine if the player is not hitting the ball as hard and far or is just getting slower. References and Resources All data was taken from FanGraphs.com.