Welcome to my 2022 Projections for Pitchers 6x6

Projections using Hypertuned model through XGboost

All data is from FanGraphs. I have no affiliation with FanGraphs, but please consider contributing to their website if you found this project informative.

1 Project Scope

1.1 Objective

This project is designed to showcase how Using a Percentile Based Worth System values Fantasy Baseball Players through a Inning Pitched (IP) weighted projection

The Categories used for prediction valuation are year-end rankings for the following metrics:

• Wins
• Saves
• Strike Outs
• ERA ( 9 * Earned Runs per Inning Pitched)
• WHIP (Walks and Hits per Inning Pitched)
• Holds

---

---

2 Processing the Data

2.1 Getting Data Into R

2.1.1 Load Libraries

First we need to load the packages that R needs to run the analysis

library(sqldf) #SQL in R
library(skimr) #Summaries and useful for removing low % data
library(ggplot2) #Plotting Functions
library(plyr) #slightly deprecated data cleaning
library(dplyr) #slightly updated data cleaning
library(tidyverse) #tidyverse data cleaning universe
library(caret) #wrapper for creating, tuning and validating models
library(xgboost) #package for creating regression tree model
library(vtreat) # useful package for treating data before modeling 
library(Matrix) #creating matricies for xgboost
library(mgcv)
library(moments) #for measuring skewness
library(data.table) #alternative to dplyr we use to create lags
library(pdp) #partial dependence graphs
library(vip) #variable importance 
library(grid) #put multiple plots on one grid
library(gridExtra) #additional grid functionality
library(janitor) #one function used to clean transposed data set
library(ggpubr) #for qq plot 
library(owmr) #Removing Prefixes
library(kableExtra) # formatting HTML Tables
library(formattable) # formatting HTML Tables

The # comments generally explain what additional functionality each library adds to R

2.1.2 Load in Data

All data is downloaded from Fan Graphs from this location. The data is also available on my Github here. There are player level and team data sets

#data read-in
pitcher_data <- read_csv("FanGraphs Leaderboard_Pitching20IP.csv")

#Team datasets
FDG_Team = read_csv("FanGraphs Leaderboard_Team.csv")

#Create a prefix for all team stats that starts with T_
FDG_Team2 <- FDG_Team %>% 
  rename_with( ~ paste0("T_", .x))

2.1.3 Checking Team Data

str give information about an object, while skim provides a customizable summary

#Output not shown for space
#str(FDG_Team2)

skim(FDG_Team2) %>%  
  tibble::as_tibble() #Remove this option for a normal HTML table

---

2.2 Understanding the Dataset

2.2.1 Exploring the dataset

skim let’s us see how the data was imported into R. Documentation can be found here

#Full Dataset dimensions

skimr::skim(pitcher_data) %>% 
  tibble::as_tibble() %>%  #Remove this option for a normal HTML table
  select(skim_type,skim_variable,complete_rate) %>% 
  filter(complete_rate >0.30) #250 Variables

#skim_type - character or numeric
#skim_variable - name of variable
#complete_rate - % of data that is not missing
#filter - only keep variables that have 30% of data populated

---

Additionally let’s look at how variables vary by year to see if there are any discrepancies there

#It looks like one year, there were fewer games played, and there is a clear drop off in home runs
pitcher_data_dist =
pitcher_data %>% 
 group_by(Season) %>% 
  summarize (Max_Games = max(G),
             Avg_W= mean(W)
             )

pitcher_data_dist

#Plot Win Data by Year
ggplot(pitcher_data_dist, aes(Season, Avg_W)) +
  geom_col()+
  ggtitle("Average Wins by Year")+
  theme(plot.title = element_text(hjust = 0.5,size = 22,color ="steel blue"))

---

2.3 Cleaning and Creating Initial Dataset for Model

What are some issues with the data?

1. Many of Variables, such as K%, are being read in as characters

   • Only Team and Player Name should be characters

2. There is spotty data coverage in some of the variables (~Variables have less than 30% Coverage)

3. 2020 Data only includes 60 games worth of data

   • This was a season shortened due to Covid-19

4. Team Data needs to be appended to pitcher Data by Team Name

---

2.3.1 Cleanly Changing all Variables that are characters to numeric.

There are several ways to do this, we will identify the variables we want to change that are mis-identified. parse_number can be used to pull numbers from these variables. Additional ways to tackle this can be found here.

#Select Column names that are characters but not Team or Name, These should be percentages
pitcher_data_chars_to_convert <- pitcher_data %>% 
  select_if(is.character)%>% select(-Team,-Name) %>% 
  mutate_all (function(x) as.numeric(readr::parse_number(x))/100)
#Note : There are additional ways to do this, this is just one solution


#We can exclude the variables we converted and reintroduce them
pitcher_data_num <- pitcher_data %>% select(-colnames(pitcher_data_chars_to_convert))

pitcher_data2 = cbind(pitcher_data_num,pitcher_data_chars_to_convert) %>% 
  select (colnames(pitcher_data)) %>%  #preserve original order 
  dplyr::rename(flyball_perc = `FB%...50`,fastball_perc = `FB%...74`) #rename two ambiguous columns
  
skim(pitcher_data2) %>% 
  as_tibble() %>% 
  group_by(skim_type) %>% 
  count()

#Logical variables are R's best guess, in our case they are all NA's and will be removed at a later step

The same can be done for the Team Data that is loaded

#Select Column names that are characters but not Team or Name, These should be percentages
FDG_Team2_chars_to_convert <- FDG_Team2 %>% 
  select_if(is.character)%>% select(-T_Team) %>% 
  mutate_all (function(x) as.numeric(readr::parse_number(x))/100)
#Keep in mind, parse number may make actual characters into numerical variables so carefully check your data before using

#We can exclude the variables we converted and reintroduce them
FDG_Team2_num <- FDG_Team2 %>% select(-colnames(FDG_Team2_chars_to_convert))

FDG_Team3 = cbind(FDG_Team2_num,FDG_Team2_chars_to_convert) %>% 
  select (colnames(FDG_Team2)) %>%  #preserve original order
dplyr::rename(T_flyball_perc = `T_FB%...45`,T_fastball_perc = `T_FB%...72`)  #rename two ambiguous columns

skim(FDG_Team3) %>% 
  as_tibble() %>% 
  group_by(skim_type) %>% 
  count()

---

2.3.2 Filtering Data with Low Coverage

I choose 30% coverage of data necessary but this can be adjusted up or down. This will also get rid of columns that are all NA.

# Keep variables with enough values (Need 30% data coverage rate here)
Player_cols_to_keep =
skim(pitcher_data2) %>% 
  dplyr::select(skim_type, skim_variable, complete_rate) %>% 
  filter (complete_rate > 0.30)

#Transpose Rows to get column names as skim melts the data
Player_cols_to_keep_transpose = t(Player_cols_to_keep) 

#extract the colnames we would like to keep
Player_cols_to_keep = colnames(janitor::row_to_names(Player_cols_to_keep_transpose,row_number = 2))

#Only keep the columns designated to have over 30% of their data populated or greater
pitcher_data3 = pitcher_data2 %>% 
  select(one_of(Player_cols_to_keep)) 

Repeat the process for Team Variables

Team_cols_to_keep =
skim(FDG_Team3) %>% 
  dplyr::select(skim_type, skim_variable, complete_rate) %>% 
  filter (complete_rate > 0.30)


#Transpose Rows to get column names as skim melts the data
Team_cols_to_keep_transpose = t(Team_cols_to_keep) 

#extract the colnames we would like to keep
Team_cols_to_keep = colnames(janitor::row_to_names(Team_cols_to_keep_transpose,row_number = 2))

#Only keep the columns designated to have over 30% of their data populated or greater
FDG_Team4 = FDG_Team3 %>% 
  select(one_of(Team_cols_to_keep)) 

---

2.3.3 Creating Variables Normalized by Year

Some Variables will need to be normalized by Innings_Pitched (IP) if they aren’t a percentage already. Remaining Variables are percentages or indices so will not need to be transformed. The full data dictionary for these variables can be found on FanGraph’s website here. for pitching variables and here. for hitting variables.

pitcher_data4 = pitcher_data3 %>% 
  mutate( #create new variables based on existing variables
    W_IP = W/IP,
    L_IP =  L/IP, 
    ShO_IP = ShO/IP,
    SV_IP = SV/IP,
    BS_IP = BS/IP,
    TBF_IP = TBF/IP,
    H_IP = H/IP,
    R_IP = R/IP,
    ER_IP = ER/IP,
    HR_IP=HR/IP,
    BB_IP=BB/IP,
    IBB_IP=IBB/IP,
    HBP_IP=HBP/IP,
    WP_IP= WP/IP,
    BK_IP=BK/IP,
    SO_IP=SO/IP,
    GB_IP = GB/IP,   #Groundballs
    FB_IP =  FB/IP,  #FlyBalls
    LD_IP = LD/IP,   #LineDrives
    IFFB_IP = IFFB/IP,  #Infield Fly balls
    Balls_IP= Balls/IP,
    Strikes_IP= Strikes/IP,
    Pitches_IP= Pitches/IP,
    RS_IP= RS/IP,
    IFH_IP= IFH/IP,
    BU_IP= BU/IP,
    BUH_IP= BUH/IP,
    Pulls_IP= Pulls/IP,
    HLD_IP= HLD/IP,   
    SD_IP= SD/IP,    
    MD_IP= MD/IP,    
    Barrels_IP= Barrels/IP,
    HardHits_IP= HardHit/IP
  ) %>% select(-L,-G,-IP,-ShO,-BS,-(TBF:BK),-(GB:BUH),-Pulls,-(SD:MD),-Barrels,-HardHit,-Events)
               
#will be removed after data is lagged -FIP,-(RAR:WPA),,-(wFB:wCH),-(`ERA-`:`xFIP-`),-SIERA,-(`RA9-WAR`:`Age Rng`),-kwERA,-`wCH (pi)`:`wSL (pi)`,`K/9+`:`HR/FB%+`) 

#skim(pitcher_data4) %>% as_tibble()

Repeat the process for Team Variables

FDG_Team5 = FDG_Team4 %>% 
  mutate( #create new variables based on existing variables
    T_H_T_PA = T_H/T_PA,
    T_x1B_T_PA = T_1B/T_PA, #note: R can't have variables start with a number
    T_x2b_T_PA = T_2B/T_PA,
    T_x3b_T_PA = T_3B/T_PA,
    T_HR_T_PA = T_HR/T_PA,
    T_R_T_PA = T_R/T_PA,
    T_RBI_T_PA = T_RBI/T_PA,
    T_BB_T_PA = T_BB/T_PA,
    T_IBB_T_PA = T_IBB/T_PA,
    T_SO_T_PA=T_SO/T_PA,
    T_HBP_T_PA=T_HBP/T_PA,
    T_SF_T_PA=T_SF/T_PA,
    T_SH_T_PA=T_SH/T_PA,
    T_GDP_T_PA= T_GDP/T_PA,#ground into double play
    T_SB_T_PA=T_SB/T_PA,
    T_CS_T_PA=T_CS/T_PA,
    T_GB_T_PA = T_GB/T_PA,   #Groundballs
    T_FB_T_PA =  T_FB/T_PA,  #FlyBalls
    T_LD_T_PA = T_LD/T_PA,   #LineDrives
    T_IFFB_T_PA = T_IFFB/T_PA,  #Infield Fly balls
    T_Pitches_T_PA= T_Pitches/T_PA,
    T_Balls_T_PA= T_Balls/T_PA,
    T_Strikes_T_PA= T_Strikes/T_PA,
    T_IFH_T_PA= T_IFH/T_PA,
    T_BU_T_PA= T_BU/T_PA,
    T_BUH_T_PA= T_BUH/T_PA,
    T_PH_T_PA= T_PH/T_PA,
    T_Barrels_T_PA= T_Barrels/T_PA,
    T_HardHits_T_PA= T_HardHit/T_PA
  ) %>% select(-(T_H:T_CS),-(T_GB:T_BUH),-T_PH,-T_Barrels,-T_HardHit,-T_Events) #Drop the old variables


#skim(FDG_Team5) %>% as_tibble()

---

2.3.4 Creating Lagged Variables

There are several ways to lag a dataset BY GROUP.
* Dplyr way is here..
* The data.table (the method used below) is here.

#Note we will only be lagging the player level data, as the previous year's team performance shouldn't impact current performance


#Order the dataset by lag columns
pitcher_data5 =  arrange(pitcher_data4, playerid,Season) #playerid is the Fangraph id assigned to each player

# Convert dataframe to data.table format
DT_pitcher = data.table(pitcher_data5)

#designate columns to lag - which is all of them
cols1 = colnames(pitcher_data5)
anscols = paste("lag", cols1, sep="_")
DT_pitcher[, (anscols) := data.table::shift(.SD, 1, NA, "lag"),by ='playerid', .SDcols=cols1] #Create 1 period lags by year

pitcher_data6 = as.data.frame(DT_pitcher) %>% select(-lag_playerid, -lag_Team, -lag_Season, -lag_Age,-lag_Name)

ncol(pitcher_data5) #251 - no lags

[1] 251

ncol(pitcher_data6) #497 - lagged data ~ (251 * 2)-5

[1] 497

---

2.3.5 Merging Team and Player Data

We can use either the merge function or the SQL functionality provided by the sqldf package to join the lagged player level data to the Team level data

df_pitching_init = sqldf(
  "
  select a.*, b.*
  from pitcher_data6 a
  left join FDG_Team5 b
  on a.Team = b.T_Team and a.Season = b.T_Season
  
  "
)  %>% select(-T_Team,-T_Season,-T_Age,-T_G,-T_AB)# Unncessary Team Variables


nrow(df_pitching_init) - nrow(pitcher_data6) #check if any rows are duplicated

[1] 0

---

3 Creating Rankings for Players Based On Percentiles

We can use Percentile based ranking to get rankings for players from the 2021 season.

3.1 Worth of each stat

3.1.1 Calculating past performance

Each player goes from a 0% to 100% on each percentile stat that is used for creating a scoring opportunity. Data is not normalized by IP as certain stats such as Wins will be worth more when we do.

#Categories I include are:
#Wins, Saves, WHIP, ERA, SOs, Holds

df_pitching_init2 =  df_pitching_init %>%
#  arrange(player_id,year) %>% 
  group_by(Season) %>% 
  mutate(
    Wins_share = order(order(rank(W_IP,ties.method = 'average'),decreasing = FALSE))/n(),
     SO_share = order(order(rank(SO_IP,ties.method = 'average'),decreasing = FALSE))/n(),
     SV_share = order(order(rank(SV_IP,ties.method = 'average'),decreasing = FALSE))/n(),
     WHIP_share = order(order(rank(WHIP,ties.method = 'average'),decreasing = FALSE))/n(),
     ERA_share = order(order(rank(ERA,ties.method = 'average'),decreasing = FALSE))/n(),
    HLD_share = order(order(rank(HLD_IP,ties.method = 'average'),decreasing = FALSE))/n(),
    Worth = Wins_share+SO_share+SV_share+WHIP_share+ERA_share+HLD_share
    ) %>% 
  ungroup() 

Chart of the Distribution of initial percentiles
As the chart below shows, the data is roughly normal.

skewness((df_pitching_init2$Worth))

[1] 0.1

ggplot2::qplot(df_pitching_init2$Worth, main="Total Pitching Worth Dataset") + geom_histogram(colour="black", fill="steelblue")

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

min(df_pitching_init2$Worth)

[1] 0.88

max(df_pitching_init2$Worth)

[1] 5.4

ggpubr::ggqqplot(df_pitching_init2$Worth)

shapiro.test(df_pitching_init2$Worth)

    Shapiro-Wilk normality test

data:  df_pitching_init2$Worth
W = 1, p-value = 0.0003

---

3.2 2021 Player Rankings - Per IP performance

3.2.1 2021 Player Rankings - Top Worth Players with Holds

Total Rankings for the players (Using 6x6 Scoring) can be found here. While it looks like many of the top players have low worth scores, it is because we haven’t applied a modifier for IP yet. Wins are harder to come by relative to any other stat and require more innings pitched.

df_pitching_init2_raw =  df_pitching_init %>%
#  arrange(player_id,year) %>% 
  group_by(Season) %>% 
  mutate(
    Wins_share_raw = order(order(rank(W,ties.method = 'average'),decreasing = FALSE))/n(),
     SO_share_raw = order(order(rank(SO,ties.method = 'average'),decreasing = FALSE))/n(),
     SV_share_raw = order(order(rank(SV,ties.method = 'average'),decreasing = FALSE))/n(),
     WHIP_share = order(order(rank(WHIP,ties.method = 'average'),decreasing = FALSE))/n(),
     ERA_share = order(order(rank(ERA,ties.method = 'average'),decreasing = FALSE))/n(),
    HLD_share_raw = 0,
    Worth = Wins_share_raw+SO_share_raw+SV_share_raw+WHIP_share+ERA_share+HLD_share_raw
    ) %>% 
  ungroup() %>% 
select(-W,-SO,-SV,-WHIP,-ERA,-HLD)



options(digits=2)

df_pitching_init2021_raw =
df_pitching_init2_raw %>% 
  group_by(Name) %>% 
  filter(Season == 2021) %>% 
  arrange(desc(Worth)) %>% 
  select(Name,Wins_share_raw,SO_share_raw,SV_share_raw,WHIP_share,ERA_share,Worth)


df_pitching_init2021_raw %>%
  filter (Worth>3.5) %>% 
  kbl() %>% 
 kable_material(c("striped", "hover","condensed","responsive"),full_width = F,fixed_thead = T)

Name	Wins_share_raw	SO_share_raw	SV_share_raw	WHIP_share	ERA_share	Worth
Daniel Bard	0.79	0.69	0.97	0.87	0.75	4.0
Garrett Richards	0.79	0.84	0.86	0.87	0.68	4.0
Jesus Luzardo	0.77	0.78	0.55	0.89	0.92	3.9
Brady Singer	0.72	0.89	0.66	0.82	0.69	3.8
Brad Keller	0.86	0.85	0.31	0.92	0.77	3.7
Jose Alvarado	0.82	0.60	0.90	0.87	0.51	3.7
Mitch Keller	0.69	0.76	0.40	0.96	0.87	3.7
Justus Sheffield	0.81	0.56	0.37	0.98	0.95	3.7
Nick Pivetta	0.89	0.95	0.72	0.51	0.60	3.7
Rafael Montero	0.65	0.34	0.91	0.81	0.91	3.6
Josh Fleming	0.93	0.58	0.78	0.59	0.73	3.6
Alec Mills	0.74	0.73	0.71	0.69	0.73	3.6
Joe Jimenez	0.75	0.49	0.73	0.80	0.85	3.6
Paul Fry	0.58	0.52	0.83	0.80	0.87	3.6
Wil Crowe	0.61	0.82	0.53	0.84	0.78	3.6
Erick Fedde	0.82	0.87	0.38	0.70	0.78	3.5
Adam Ottavino	0.77	0.62	0.93	0.70	0.51	3.5
Bryan Garcia	0.50	0.22	0.85	0.98	0.98	3.5
Alex Reyes	0.92	0.77	0.98	0.58	0.25	3.5

---

3.3 2021 Player Rankings - Actual Performance

3.3.1 2021 Player Rankings - Top Worth Players with Holds

While it looks like many of the top players have low worth scores, it is because we haven’t applied a modifier for IP yet.

options(digits=2)

df_pitching_init2021 =
df_pitching_init2 %>% 
  group_by(Name) %>% 
  filter(Season == 2021) %>% 
  arrange(desc(Worth)) %>% 
  select(Name,Wins_share,SO_share,SV_share,WHIP_share,ERA_share,HLD_share,Worth)


df_pitching_init2021 %>%
  filter (Worth>2.9) %>% 
  kbl() %>% 
 kable_material(c("striped", "hover","condensed","responsive"),full_width = F,fixed_thead = T)

Name	Wins_share	SO_share	SV_share	WHIP_share	ERA_share	HLD_share	Worth
Paul Fry	0.83	0.87	0.82	0.80	0.87	0.87	5.1
Jose Alvarado	0.96	0.85	0.89	0.87	0.51	0.93	5.0
Daniel Bard	0.93	0.85	0.97	0.87	0.75	0.61	5.0
Joe Jimenez	0.98	0.86	0.76	0.80	0.85	0.72	5.0
Tanner Rainey	0.22	0.93	0.89	0.93	0.97	0.95	4.9
Adam Ottavino	0.95	0.76	0.94	0.70	0.51	0.98	4.8
Rafael Dolis	0.63	0.84	0.89	0.95	0.81	0.68	4.8
Tanner Scott	0.87	0.90	0.40	0.84	0.74	0.93	4.7
Ben Bowden	0.83	0.82	0.57	0.97	0.92	0.53	4.6
Ryan Hendrix	1.00	0.73	0.51	0.82	0.85	0.68	4.6
James Karinchak	0.97	0.96	0.94	0.35	0.47	0.88	4.6
Sean Newcomb	0.62	0.93	0.79	0.93	0.65	0.62	4.5
Pete Fairbanks	0.71	0.92	0.90	0.69	0.35	0.96	4.5
Bryan Abreu	0.83	0.52	0.78	0.73	0.82	0.83	4.5
Jeurys Familia	0.99	0.84	0.72	0.66	0.45	0.81	4.5
Brad Brach	0.24	0.69	0.81	0.87	0.89	0.96	4.5
Rafael Montero	0.91	0.26	0.91	0.81	0.91	0.65	4.5
Rex Brothers	0.55	0.96	0.73	0.68	0.75	0.77	4.4
Lucas Sims	0.92	0.99	0.92	0.20	0.56	0.83	4.4
Bryan Garcia	0.76	0.21	0.85	0.98	0.98	0.64	4.4
Trevor May	0.94	0.92	0.86	0.43	0.35	0.90	4.4
Enyel De Los Santos	0.55	0.94	0.44	0.95	0.90	0.60	4.4
Ryan Helsley	0.97	0.52	0.75	0.67	0.61	0.85	4.4
Aaron Bummer	0.86	0.92	0.81	0.44	0.33	0.98	4.3
Sam Howard	0.66	0.92	0.33	0.73	0.80	0.88	4.3
Amir Garrett	0.04	0.89	0.92	0.83	0.86	0.77	4.3
Tyler Chatwood	0.22	0.80	0.80	0.72	0.81	0.95	4.3
Sean Poppen	0.39	0.79	0.83	0.96	0.74	0.58	4.3
Adam Morgan	0.80	0.72	0.88	0.56	0.53	0.78	4.3
Gregory Soto	0.89	0.83	0.96	0.59	0.28	0.71	4.3
Alex Reyes	0.98	0.90	0.98	0.58	0.25	0.57	4.3
David Hess	0.91	0.32	0.32	1.00	1.00	0.69	4.2
Devin Williams	0.99	0.99	0.86	0.31	0.09	0.99	4.2
Aroldis Chapman	0.93	1.00	0.99	0.52	0.27	0.51	4.2
Trevor Megill	0.35	0.90	0.40	0.98	0.99	0.57	4.2
Phil Maton	0.87	0.88	0.42	0.72	0.65	0.63	4.2
Kyle Finnegan	0.75	0.56	0.93	0.74	0.34	0.84	4.2
Jake Diekman	0.45	0.95	0.90	0.56	0.43	0.87	4.2
Seth Lugo	0.84	0.82	0.75	0.49	0.32	0.92	4.1
Jake Brentz	0.78	0.80	0.80	0.47	0.38	0.88	4.1
J.B. Wendelken	0.87	0.33	0.83	0.78	0.54	0.75	4.1
Jeffrey Springs	0.95	0.96	0.83	0.20	0.30	0.87	4.1
Paul Sewald	0.99	0.99	0.93	0.10	0.19	0.90	4.1
Hansel Robles	0.36	0.70	0.95	0.62	0.58	0.89	4.1
Paul Campbell	0.76	0.49	0.63	0.86	0.91	0.43	4.1
Jacob Barnes	0.26	0.78	0.87	0.73	0.88	0.55	4.1
Josh Sborz	0.67	0.78	0.72	0.67	0.45	0.77	4.1
Carlos Estevez	0.44	0.48	0.94	0.76	0.56	0.89	4.1
Mychal Givens	0.79	0.62	0.92	0.61	0.27	0.85	4.1
Hirokazu Sawamura	0.89	0.77	0.68	0.71	0.19	0.82	4.0
Victor Gonzalez	0.84	0.39	0.78	0.70	0.35	1.00	4.0
Jesus Luzardo	0.63	0.57	0.55	0.89	0.92	0.48	4.0
Mike Mayers	0.66	0.82	0.77	0.49	0.43	0.87	4.0
Sean Doolittle	0.60	0.66	0.74	0.72	0.60	0.70	4.0
Joely Rodriguez	0.36	0.54	0.75	0.81	0.63	0.92	4.0
Anthony Misiewicz	0.87	0.47	0.43	0.64	0.63	0.97	4.0
Brett de Geus	0.59	0.21	0.65	0.95	0.98	0.60	4.0
Chris Stratton	0.86	0.67	0.90	0.50	0.37	0.67	4.0
Matt Foster	0.49	0.55	0.77	0.70	0.85	0.59	4.0
Heath Hembree	0.25	0.96	0.92	0.30	0.79	0.73	4.0
Camilo Doval	1.00	0.94	0.90	0.13	0.17	0.81	4.0
Kyle Zimmer	0.74	0.26	0.81	0.65	0.68	0.81	4.0
Shane Greene	0.01	0.57	0.83	0.91	0.96	0.66	3.9
Darwinzon Hernandez	0.47	0.93	0.53	0.77	0.28	0.94	3.9
Archie Bradley	0.98	0.17	0.82	0.68	0.39	0.90	3.9
Brad Boxberger	0.78	0.90	0.86	0.15	0.26	0.98	3.9
Sean Reid-Foley	0.90	0.89	0.36	0.77	0.75	0.24	3.9
Matt Andriese	0.33	0.59	0.74	0.88	0.75	0.62	3.9
Reid Detmers	0.44	0.39	0.67	0.97	0.97	0.46	3.9
Kevin Ginkel	0.09	0.70	0.54	0.82	0.89	0.85	3.9
Stefan Crichton	0.05	0.11	0.93	0.99	0.96	0.86	3.9
Luis Oviedo	0.24	0.62	0.63	0.99	0.99	0.42	3.9
Jose Cisnero	0.64	0.53	0.87	0.54	0.38	0.93	3.9
Codi Heuer	0.92	0.22	0.79	0.53	0.54	0.90	3.9
Glenn Otto	0.10	0.83	0.61	0.93	0.99	0.41	3.9
Aaron Ashby	0.90	0.86	0.80	0.28	0.61	0.44	3.9
Daniel Norris	0.26	0.54	0.72	0.76	0.87	0.73	3.9
Greg Holland	0.52	0.43	0.91	0.57	0.68	0.76	3.9
Nick Mears	0.36	0.50	0.66	0.90	0.71	0.74	3.9
Brooks Raley	0.32	0.92	0.82	0.32	0.67	0.81	3.8
Joakim Soria	0.17	0.66	0.92	0.61	0.72	0.74	3.8
Sam Coonrod	0.42	0.75	0.84	0.53	0.47	0.82	3.8
Jacob Webb	0.98	0.45	0.79	0.79	0.50	0.31	3.8
Richard Lovelady	0.91	0.74	0.85	0.15	0.32	0.84	3.8
Austin Voth	0.70	0.57	0.29	0.74	0.76	0.73	3.8
Corey Knebel	0.99	0.81	0.90	0.06	0.08	0.91	3.8
Trevor Stephan	0.42	0.81	0.71	0.66	0.57	0.59	3.8
Griffin Canning	0.80	0.51	0.54	0.75	0.80	0.35	3.8
Hector Neris	0.51	0.91	0.93	0.26	0.37	0.77	3.8
Demarcus Evans	0.10	0.87	0.60	0.81	0.74	0.64	3.8
Bryan Shaw	0.77	0.36	0.77	0.62	0.32	0.90	3.7
Anthony Castro	0.32	0.91	0.82	0.44	0.66	0.57	3.7
Brad Hand	0.88	0.41	0.97	0.45	0.44	0.58	3.7
Taylor Rogers	0.46	0.97	0.95	0.23	0.27	0.84	3.7
Tim Hill	0.91	0.41	0.72	0.40	0.37	0.93	3.7
Chad Green	0.96	0.81	0.87	0.02	0.20	0.86	3.7
Kyle Keller	0.21	0.67	0.47	0.83	0.92	0.61	3.7
Mason Thompson	0.33	0.41	0.63	0.98	0.46	0.89	3.7
Alex Lange	0.19	0.71	0.78	0.77	0.47	0.79	3.7
Diego Castillo	0.84	0.89	0.96	0.08	0.13	0.80	3.7
Anthony Kay	0.20	0.79	0.59	0.92	0.81	0.39	3.7
Josh Taylor	0.15	0.87	0.75	0.68	0.29	0.95	3.7
Conner Greene	0.30	0.58	0.35	0.95	0.95	0.56	3.7
Scott Barlow	0.67	0.85	0.95	0.32	0.07	0.82	3.7
Aaron Slegers	0.64	0.20	0.30	0.98	0.95	0.62	3.7
Kyle Funkhouser	0.92	0.37	0.71	0.65	0.30	0.74	3.7
Tommy Nance	0.26	0.63	0.47	0.55	0.96	0.80	3.7
Spencer Howard	0.11	0.62	0.65	0.88	0.97	0.44	3.7
Edwin Uceta	0.10	0.86	0.59	0.80	0.93	0.39	3.7
Michael Lorenzen	0.25	0.10	0.91	0.63	0.80	0.98	3.7
Brusdar Graterol	0.87	0.20	0.58	0.66	0.62	0.73	3.7
Jordan Romano	0.94	0.93	0.97	0.13	0.04	0.64	3.7
Michael Fulmer	0.72	0.61	0.94	0.47	0.17	0.74	3.7
Alex Colome	0.61	0.31	0.96	0.64	0.49	0.64	3.7
Yimi Garcia	0.69	0.60	0.96	0.25	0.51	0.63	3.7
Spencer Patton	0.42	0.75	0.84	0.32	0.42	0.90	3.6
Alex Claudio	0.22	0.38	0.79	0.86	0.79	0.61	3.6
Matt Barnes	0.94	0.98	0.98	0.21	0.41	0.11	3.6
Matt Wisler	0.62	0.89	0.74	0.15	0.39	0.85	3.6
Rowan Wick	0.04	0.86	0.95	0.58	0.54	0.66	3.6
Wander Suero	0.42	0.60	0.31	0.66	0.89	0.75	3.6
Justin Garza	0.71	0.56	0.53	0.84	0.65	0.34	3.6
Cionel Perez	0.34	0.59	0.51	0.93	0.90	0.34	3.6
Blake Treinen	0.82	0.80	0.89	0.07	0.04	0.99	3.6
Giovanny Gallegos	0.74	0.80	0.93	0.02	0.18	0.94	3.6
Wandy Peralta	0.90	0.24	0.89	0.62	0.27	0.69	3.6
Max Kranick	0.50	0.23	0.63	0.94	0.89	0.43	3.6
Michael Rucker	0.08	0.63	0.81	0.80	0.95	0.33	3.6
J.D. Hammer	0.47	0.70	0.51	0.88	0.70	0.34	3.6
Tim Mayza	0.89	0.65	0.73	0.08	0.29	0.97	3.6
Zac Lowther	0.24	0.56	0.57	0.91	0.93	0.38	3.6
Ryne Stanek	0.37	0.84	0.79	0.34	0.30	0.95	3.6
Dylan Floro	0.88	0.45	0.96	0.36	0.14	0.79	3.6
Jackson Kowar	0.10	0.44	0.62	1.00	1.00	0.42	3.6
Cesar Valdez	0.36	0.47	0.93	0.91	0.84	0.06	3.6
Justus Sheffield	0.85	0.17	0.37	0.98	0.95	0.25	3.6
Lucas Gilbreath	0.72	0.60	0.76	0.52	0.28	0.68	3.6
Albert Abreu	0.52	0.44	0.78	0.42	0.74	0.65	3.5
Lou Trivino	0.89	0.34	0.97	0.41	0.22	0.71	3.5
Michael Feliz	0.02	0.70	0.85	0.91	0.96	0.10	3.5
Phillips Valdez	0.47	0.29	0.76	0.58	0.84	0.59	3.5
Garrett Richards	0.49	0.24	0.75	0.87	0.68	0.50	3.5
Chris Rodriguez	0.68	0.50	0.64	0.71	0.38	0.63	3.5
Junior Guerra	0.76	0.38	0.05	0.94	0.86	0.54	3.5
Trevor Richards	0.93	0.84	0.71	0.05	0.32	0.68	3.5
Keynan Middleton	0.23	0.16	0.91	0.85	0.69	0.69	3.5
Daniel Lynch	0.58	0.20	0.62	0.90	0.81	0.41	3.5
Edward Cabrera	0.10	0.64	0.62	0.90	0.83	0.42	3.5
Tyler Zuber	0.09	0.36	0.58	0.84	0.88	0.76	3.5
Adam Plutko	0.13	0.17	0.73	0.90	0.94	0.67	3.5
Brent Suter	1.00	0.40	0.70	0.52	0.20	0.71	3.5
Jose Quijada	0.07	0.98	0.48	0.60	0.61	0.78	3.5
Shawn Armstrong	0.18	0.84	0.17	0.77	0.94	0.60	3.5
Garrett Crochet	0.53	0.83	0.67	0.46	0.15	0.86	3.5
Garrett Whitlock	0.94	0.72	0.77	0.20	0.03	0.83	3.5
Cole Sulser	0.80	0.77	0.91	0.21	0.12	0.68	3.5
Jose Quintana	0.02	0.93	0.13	0.94	0.91	0.54	3.5
Josh Fleming	0.89	0.04	0.69	0.59	0.73	0.54	3.5
A.J. Minter	0.56	0.68	0.46	0.37	0.41	0.99	3.5
Brady Singer	0.29	0.54	0.66	0.82	0.69	0.46	3.5
Alex Young	0.29	0.22	0.44	0.96	0.92	0.64	3.5
Genesis Cabrera	0.56	0.70	0.39	0.44	0.40	0.99	3.5
Brandon Bielak	0.59	0.36	0.74	0.63	0.59	0.56	3.5
Alex Vesia	0.75	0.94	0.77	0.08	0.06	0.86	3.4
Patrick Murphy	0.08	0.58	0.52	0.72	0.74	0.81	3.4
Andrew Wantz	0.28	0.95	0.66	0.40	0.70	0.45	3.4
Emilio Pagan	0.63	0.68	0.27	0.26	0.68	0.91	3.4
Craig Kimbrel	0.67	1.00	0.98	0.02	0.06	0.70	3.4
Bailey Falter	0.59	0.55	0.56	0.32	0.81	0.61	3.4
Wade Davis	0.01	0.32	0.84	0.74	0.94	0.58	3.4
Austin Davis	0.29	0.64	0.37	0.62	0.73	0.78	3.4
Derek Holland	0.60	0.59	0.03	0.83	0.72	0.65	3.4
Will Vest	0.19	0.16	0.53	0.88	0.87	0.79	3.4
Kendall Graveman	0.86	0.68	0.94	0.08	0.02	0.84	3.4
Jay Jackson	0.88	0.91	0.07	0.40	0.40	0.75	3.4
Dinelson Lamet	0.34	0.83	0.37	0.76	0.57	0.52	3.4
J.P. Feyereisen	0.72	0.41	0.86	0.38	0.12	0.91	3.4
Yency Almonte	0.14	0.50	0.29	0.86	0.97	0.62	3.4
Brandon Kintzler	0.68	0.13	0.10	0.96	0.90	0.63	3.4
Josh Staumont	0.61	0.71	0.88	0.16	0.16	0.89	3.4
Austin Adams	0.56	0.97	0.24	0.32	0.48	0.83	3.4
Phil Bickford	0.79	0.77	0.73	0.16	0.14	0.80	3.4
Keegan Thompson	0.54	0.58	0.73	0.74	0.28	0.51	3.4
Bruce Zimmermann	0.62	0.29	0.58	0.79	0.72	0.39	3.4
Miguel Diaz	0.72	0.68	0.76	0.31	0.37	0.52	3.4
Anthony Banda	0.59	0.44	0.26	0.81	0.53	0.72	3.4
Humberto Mejia	0.10	0.33	0.60	0.98	0.96	0.40	3.4
Edwin Diaz	0.80	0.96	0.99	0.13	0.31	0.18	3.4
Buck Farmer	0.04	0.61	0.27	0.94	0.90	0.60	3.4
Jonathan Loaisiga	0.97	0.49	0.87	0.10	0.05	0.88	3.4
Shane McClanahan	0.81	0.76	0.62	0.46	0.31	0.41	3.4
Josiah Gray	0.19	0.67	0.66	0.60	0.78	0.45	3.4
Jeff Hoffman	0.32	0.66	0.38	0.85	0.61	0.53	3.4
Caleb Baragar	0.84	0.08	0.88	0.59	0.02	0.94	3.4
Miguel Sanchez	0.77	0.31	0.56	0.85	0.49	0.37	3.4
Hoby Milner	0.03	0.96	0.20	0.79	0.77	0.58	3.3
Braxton Garrett	0.20	0.40	0.63	0.97	0.71	0.42	3.3
Luke Farrell	0.32	0.57	0.29	0.92	0.66	0.57	3.3
Dylan Cease	0.79	0.94	0.45	0.42	0.44	0.28	3.3
A.J. Alexy	0.97	0.11	0.64	0.51	0.65	0.43	3.3
Nick Wittgren	0.23	0.48	0.71	0.41	0.72	0.76	3.3
Sergio Romo	0.12	0.48	0.84	0.40	0.64	0.83	3.3
Andrew Kittredge	0.96	0.66	0.90	0.07	0.03	0.69	3.3
JC Mejia	0.13	0.33	0.59	0.88	0.99	0.40	3.3
Cody Poteet	0.66	0.62	0.47	0.57	0.71	0.29	3.3
Daniel Hudson	0.90	0.97	0.07	0.17	0.26	0.95	3.3
Tyler Duffey	0.43	0.48	0.84	0.36	0.22	0.97	3.3
Mitch Keller	0.46	0.35	0.40	0.96	0.87	0.26	3.3
Ryan Borucki	0.97	0.34	0.34	0.38	0.70	0.57	3.3
Clay Holmes	0.95	0.72	0.22	0.26	0.36	0.78	3.3
Tejay Antone	0.59	0.87	0.88	0.02	0.04	0.88	3.3
Nick Pivetta	0.57	0.74	0.68	0.51	0.60	0.20	3.3
Blake Taylor	0.89	0.46	0.34	0.66	0.22	0.72	3.3
Michael Kopech	0.57	0.98	0.37	0.22	0.32	0.82	3.3
Tarik Skubal	0.51	0.69	0.64	0.45	0.55	0.43	3.3
Justin Steele	0.70	0.58	0.38	0.58	0.53	0.50	3.3
Joe Smith	0.92	0.28	0.02	0.64	0.71	0.70	3.3
Sam Hentges	0.12	0.51	0.45	0.96	0.93	0.28	3.3
Brandon Workman	0.27	0.32	0.14	0.99	0.78	0.76	3.2
Raisel Iglesias	0.91	0.98	0.99	0.04	0.10	0.25	3.2
Andrew Heaney	0.62	0.77	0.30	0.53	0.84	0.19	3.2
Sammy Long	0.45	0.40	0.66	0.48	0.79	0.46	3.2
Ryan Pressly	0.78	0.87	0.98	0.06	0.06	0.50	3.2
Josh Tomlin	0.81	0.13	0.09	0.77	0.92	0.52	3.2
Kohei Arihara	0.45	0.03	0.67	0.69	0.93	0.46	3.2
Eduardo Rodriguez	0.82	0.79	0.19	0.64	0.66	0.13	3.2
Chad Kuhl	0.62	0.38	0.28	0.68	0.68	0.59	3.2
Robert Stephenson	0.36	0.74	0.75	0.50	0.20	0.66	3.2
Travis Lakins Sr.	0.27	0.27	0.46	0.69	0.83	0.71	3.2
Dan Winkler	0.17	0.54	0.16	0.82	0.75	0.77	3.2
David Peterson	0.21	0.60	0.57	0.65	0.79	0.38	3.2
Carson Fulmer	0.06	0.42	0.43	0.80	0.93	0.56	3.2
Tylor Megill	0.39	0.72	0.61	0.48	0.60	0.41	3.2
Connor Brogdon	0.85	0.29	0.72	0.23	0.30	0.80	3.2
Caleb Thielbar	0.93	0.83	0.10	0.26	0.25	0.82	3.2
Taylor Clarke	0.16	0.33	0.40	0.80	0.71	0.79	3.2
Tony Gonsolin	0.73	0.79	0.50	0.59	0.25	0.32	3.2
Yusmeiro Petit	0.92	0.01	0.77	0.12	0.44	0.92	3.2
Anthony Bender	0.44	0.78	0.84	0.15	0.13	0.84	3.2
David Price	0.68	0.18	0.69	0.67	0.46	0.49	3.2
Austin Warren	0.99	0.50	0.85	0.11	0.02	0.69	3.2
Wil Crowe	0.25	0.42	0.53	0.84	0.78	0.35	3.2
Alexander Wells	0.42	0.03	0.53	0.89	0.94	0.35	3.2
Luis Patino	0.64	0.43	0.65	0.46	0.54	0.44	3.2
David Bednar	0.46	0.88	0.85	0.07	0.05	0.86	3.2
Chris Martin	0.41	0.15	0.76	0.45	0.45	0.94	3.2
John King	0.99	0.28	0.63	0.25	0.33	0.66	3.1
Sam Clay	0.05	0.14	0.34	0.93	0.80	0.89	3.1
Daniel Ponce de Leon	0.21	0.10	0.86	0.89	0.87	0.21	3.1
Cristian Javier	0.30	0.88	0.74	0.29	0.34	0.59	3.1
Randy Dobnak	0.14	0.01	0.74	0.82	0.98	0.45	3.1
Sean Guenther	0.09	0.12	0.56	0.99	1.00	0.37	3.1
Brad Keller	0.59	0.32	0.31	0.92	0.77	0.20	3.1
Jordan Holloway	0.54	0.53	0.60	0.60	0.46	0.40	3.1
Shaun Anderson	0.07	0.21	0.51	1.00	0.99	0.33	3.1
Logan Gilbert	0.47	0.65	0.64	0.27	0.65	0.43	3.1
Wes Benjamin	0.07	0.26	0.48	0.99	0.99	0.30	3.1
Jake Cousins	0.24	0.97	0.62	0.27	0.12	0.87	3.1
Collin McHugh	0.88	0.77	0.71	0.04	0.02	0.68	3.1
Chris Paddack	0.64	0.35	0.56	0.45	0.73	0.37	3.1
Jorge Alcala	0.48	0.56	0.72	0.07	0.45	0.81	3.1
Packy Naughton	0.11	0.01	0.65	0.97	0.90	0.44	3.1
Junior Fernandez	0.44	0.12	0.44	0.99	0.81	0.28	3.1
Luis Garcia	0.71	0.65	0.66	0.30	0.32	0.45	3.1
Yusei Kikuchi	0.38	0.59	0.60	0.54	0.57	0.40	3.1
Dominic Leone	0.75	0.39	0.81	0.19	0.01	0.92	3.1
Matt Peacock	0.57	0.02	0.58	0.83	0.69	0.38	3.1
Dean Kremer	0.07	0.31	0.50	0.90	0.98	0.32	3.1
Vladimir Gutierrez	0.79	0.16	0.49	0.66	0.66	0.31	3.1
Pierce Johnson	0.49	0.91	0.21	0.44	0.24	0.78	3.1
Shane Bieber	0.73	0.95	0.51	0.34	0.22	0.33	3.1
Alek Manoah	0.81	0.75	0.67	0.13	0.24	0.46	3.1
Joe Kelly	0.39	0.74	0.83	0.07	0.15	0.87	3.1
Griffin Jax	0.43	0.18	0.57	0.59	0.90	0.38	3.1
Matt Manning	0.41	0.06	0.58	0.79	0.83	0.38	3.1
Hyeon-jong Yang	0.11	0.09	0.67	0.91	0.80	0.47	3.1
Zack Littell	0.65	0.56	0.80	0.23	0.16	0.65	3.0
Erick Fedde	0.50	0.44	0.38	0.70	0.78	0.25	3.0
Kenta Maeda	0.54	0.63	0.45	0.51	0.64	0.28	3.0
Cody Ponce	0.06	0.40	0.41	0.95	0.95	0.26	3.0
Jharel Cotton	0.66	0.50	0.22	0.65	0.33	0.69	3.0
Liam Hendriks	0.95	0.99	0.99	0.01	0.09	0.02	3.0
Jorge Lopez	0.16	0.36	0.25	0.89	0.86	0.50	3.0
Caleb Smith	0.26	0.69	0.28	0.62	0.68	0.51	3.0
Taylor Widener	0.19	0.59	0.48	0.71	0.56	0.49	3.0
Ryan Thompson	0.86	0.68	0.35	0.11	0.07	0.96	3.0
Art Warren	0.98	0.99	0.43	0.02	0.01	0.59	3.0
Ryan Weathers	0.34	0.15	0.69	0.63	0.76	0.45	3.0
Tyler Rogers	0.84	0.07	0.92	0.16	0.05	0.98	3.0
Andrew Miller	0.01	0.72	0.06	0.85	0.67	0.72	3.0
Keegan Akin	0.14	0.28	0.50	0.85	0.93	0.32	3.0
Jarlin Garcia	0.85	0.51	0.70	0.05	0.11	0.79	3.0
Joe Ryan	0.76	0.75	0.61	0.01	0.47	0.41	3.0
Eli Morgan	0.54	0.34	0.57	0.43	0.76	0.37	3.0
Tyler Kinley	0.35	0.46	0.43	0.34	0.66	0.76	3.0
Yohan Ramirez	0.27	0.89	0.88	0.18	0.44	0.33	3.0
Taylor Hearn	0.56	0.31	0.43	0.53	0.63	0.48	3.0
Yacksel Rios	0.94	0.25	0.28	0.39	0.53	0.55	2.9
Tyler Clippard	0.30	0.23	0.96	0.50	0.23	0.72	2.9
Carlos Hernandez	0.71	0.28	0.65	0.48	0.38	0.44	2.9
Kris Bubic	0.40	0.30	0.61	0.63	0.58	0.41	2.9
Johan Oviedo	0.11	0.21	0.65	0.84	0.69	0.44	2.9
Dane Dunning	0.35	0.46	0.50	0.70	0.59	0.32	2.9
Nick Sandlin	0.20	0.97	0.59	0.23	0.17	0.77	2.9
Kyle Muller	0.53	0.54	0.56	0.42	0.50	0.37	2.9
Luis Garcia	0.21	0.56	0.86	0.08	0.25	0.96	2.9
Luis Gil	0.25	0.90	0.60	0.55	0.20	0.40	2.9
Robert Dugger	0.07	0.13	0.47	0.97	0.96	0.30	2.9

---

4 Creating Model File

4.1 Additional Data Prep

4.1.1 Remove Variables which are based off current hitting numbers

Not all variables can be used for predictive modeling. Variables that go into the percentile ranking or are non-normalized metrics created after the fact (such as WAR - Wins above Replacement or RS - Raw Run Support) should be removed. However, metrics that are normalized by a per pitch basis (such as wFB/C) can remain as we expect pitchers to have similar performance in these metrics one year out.

#Be careful about RS - Run Support and RS/9

#Creating a new dataset to keep original intact
df_pitching_init3 = df_pitching_init2 %>% 
  select (-Name)

Lagged Percentile (_share) Variables can be used for predictive modeling. However since these variables were created for the Worth metric they must also be removed for modeling purposes.

#Order the dataset by lag columns
df_pitching_init4 =  arrange(df_pitching_init3, playerid,Season) #playerid is the Fangraph id assigned to each player

# Convert dataframe to data.table format
DT_pitcher2 = data.table(df_pitching_init4)

#designate columns to lag - just the new shares
cols1 = (c('Wins_share','SO_share','SV_share', 'ERA_share','WHIP_share','HLD_share','Worth'))
anscols = paste("lag", cols1, sep="_") 
DT_pitcher2[, (anscols) := data.table::shift(.SD, 1, NA, "lag"),by ='playerid', .SDcols=cols1] #Create 1 period lags by year

df_pitching_final = as.data.frame(DT_pitcher2) %>% 
  select(-c(Wins_share,SO_share,SV_share, ERA_share,WHIP_share,HLD_share))%>%
select(-FIP,-(RAR:WPA),-(wFB:wCH),-(`ERA-`:`xFIP-`),
       -SIERA,-(`RA9-WAR`:`Age Rng`),-kwERA,-(`wCH (pi)`:`wSL (pi)`),-(`K/9+`:`HR/FB%+`)) %>% select(-W,-SO,-SV,-HLD,-W_IP,-SO_IP,-SV_IP,-WHIP,-ERA,-HLD_IP)

4.1.2 Creating Training/Test Split

We split the data into Training Data (which is used to create the model) and test data (which is used to validate the model)

set.seed(15674)  # For reproducibility
# Create index for testing and training data
inTrain <- createDataPartition(y = df_pitching_final$Worth, p = 0.80, list = FALSE)
# subset pitching data for training
tr_2021 <- df_pitching_final[inTrain,]
# subset the rest to test and validate trained model
te_2021 <- df_pitching_final[-inTrain,]

nrow(tr_2021)/nrow(df_pitching_final) #check if split is 0.8

[1] 0.8

4.1.3 Treat Missing Data by Imputing Mean Value

Vtreat Package in R is excellent for treating data before using for modeling. Additional documentation can be found here.

treat_plan_2021 <- vtreat::designTreatmentsZ(
  dframe = tr_2021, # training data
  varlist = colnames(tr_2021) %>% .[. != "hitting_score1"], # input variables = all training data columns, except random
  codeRestriction = c("clean", "isBAD", "lev"), # derived variables types (drop cat_P)
  verbose = FALSE) # suppress messages

#clean stands for cleaned numerical variable, isBAD indicates that a value replacement has occurred (which indicates a missing value in this case), and lev is a binary indicator whether a particular value of that categorical variable was present.  

#### Checking Scoreframe

score_frame <- treat_plan_2021$scoreFrame %>% 
  select(varName, origName, code)

head(score_frame)

tr_treated_2021 <- vtreat::prepare(treat_plan_2021, tr_2021)
te_treated_2021 <- vtreat::prepare(treat_plan_2021, te_2021)


treat_plan_2021 <- vtreat::designTreatmentsZ(
  dframe = DT_pitcher2, # training data
  varlist = colnames(DT_pitcher2) %>% .[. != "hitting_score1"], # input variables = all training data columns, except random
  codeRestriction = c("clean", "isBAD", "lev"), # derived variables types (drop cat_P)
  verbose = FALSE) # suppress messages


total_treated_2021_pitching <- vtreat::prepare(treat_plan_2021, DT_pitcher2)

#tr_treated = tr
#te_treated = te

dim(tr_treated_2021) #note there are dummies for each player and team

[1] 3197 1416

---

4.1.4 Check Distribution of Training Population

The population used for Training should be indicative of Total Population

ggplot2::qplot(tr_treated_2021$Worth, main="Training Set") + geom_histogram(colour="black", fill="steelblue") + theme_bw()

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

#The skewness is actually a bit better than the overall data set
skewness(tr_treated_2021$Worth) 

[1] 0.1

---

5 Running XGboost Model

To keep things simple with modeling, we’ll turn the training data into simple input variables for caret::train, dropping the response variable and converting the data frame to a matrix. Documentation for this approach to XGboost can be found here.

5.1 Tuning the Model

5.1.1 Initial Non-Tuned Model

Break the data set into x and y inputs with x being a matrix. "_isBAD" is a category created by the Vtreat package in case you want to identify rows

input_x <- as.matrix(((tr_treated_2021))%>%
   select(-Worth) %>%                      
   select(!ends_with ("_isBAD")))

input_y <- tr_treated_2021$Worth

XGBoost with Default Hyperparameters:
The Variable Importance (caret::varImp(xgb_base_2021, scale = F) from the caret package shows the contribution of each variable to the initial model. Since this is untuned, we can expect the percentage imporantance to change as the models iterate through potential hyperparameters.
XGBoost documentation can be found for more general models here.

#Defaults for xgboost model
grid_default <- expand.grid(
  nrounds = 100,
  max_depth = 6,
  eta = 0.3,
  gamma = 0,
  colsample_bytree = 1,
  min_child_weight = 1,
  subsample = 1
)

#This is a blank train_control set, this will be updated after
train_control <- caret::trainControl(
  method = "none",
  verboseIter = FALSE, # no training log
  allowParallel = TRUE # FALSE for reproducible results 
)

xgb_base_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = train_control,
  tuneGrid = grid_default,
  method = "xgbTree",
  verbose = TRUE
)

caret::varImp(xgb_base_2021, scale = F  )

xgbTree variable importance

  only 20 most important variables shown (out of 774)

---

5.2 Further Variable Selection

5.2.1 Remove redundant and highly correlated variables

Selection Removal Step 1: Check for high correlations
Normally, this step is done early, but those steps were reserved for preparing the data

dep_cor1 <- t(as.data.frame(cor(tr_treated_2021[ , colnames(tr_treated_2021) != "Worth"],
                tr_treated_2021$Worth)))
dep_cor1 <-
as.data.frame(t(as.data.frame(dep_cor1)%>% 
  select(!starts_with("lag")) %>% #remove lag variables
  select(!contains("_isBAD")))) 

dep_cor1 <- tibble::rownames_to_column(dep_cor1,"VARIABLES")%>% #remove indicators for missing data
  filter(V1 > 0.40|V1 < -0.3)

dep_cor1

dep_cor2 <- colnames(row_to_names(t(dep_cor1),row_number = 1))

Let’s Remove variables with high correlation to worth metric, and metrics that are calculated after a player’s performance (such as WPA/RE24) or redundant (RS_IP)

input_x <- as.matrix(((tr_treated_2021))%>%
   select(-Worth) %>% #Remove some variables variables
     select (-RS_IP,-ER_IP,-R_IP,-REW,-RE24,-Clutch,-WPA_slash_LI,-Season #Remove redundant variables or non/weighted variables
) %>%      
select(!ends_with ("_isBAD"))) #indicator variable for missing data

input_y <- tr_treated_2021$Worth

Run the model on the new dataset to make sure the variable importances look fine

#Note Training parameters were set in initial model set up
xgb_base_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = train_control,
  tuneGrid = grid_default,
  method = "xgbTree",
  verbose = TRUE
)

caret::varImp(xgb_base_2021, scale = F  )

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.3 Model with new data

5.3.1 Tuning All Hyperparameters

A tune grid allows us to test a large amount of hyper-parameters and find the model with the lowest RMSE for predictions.
However, The more values you want to test and the greater the amount of Cross-Fold Validations (method = "cv"), the greater the computational time it will take. More information on the specific parameters can be found here.

# maximum number of trees
nrounds <- 1000

# note to start nrounds from 200, as smaller learning rates result in errors so
# big with lower starting points that they'll mess the scales
tune_grid <- expand.grid(
  nrounds = seq(from = 100, to = nrounds, by = 50),
  eta = c(0.01, 0.025, 0.05, 0.075, 0.1),
  max_depth = c(2, 4, 6, 8, 10),
  gamma = 0,
  colsample_bytree = 1,
  min_child_weight = 1,
  subsample = 1
)

tune_control <- caret::trainControl(
  method = "cv", # cross-validation
  number = 5, # with n folds 
  ## Note this was # out in the original code
  #index = createFolds(tr_treated$Id_clean), # fix the folds
  verboseIter = FALSE, # no training log
  allowParallel = FALSE # FALSE for reproducible results 
)

Running the initial tuning model

#Note I will be timing these runs to give an estimate on how long this model takes to run
start_time <- Sys.time()

xgb_tune_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = tune_control,
  tuneGrid = tune_grid,
  method = "xgbTree",
  verbose = FALSE
  ,verbosity = 0
)

end_time <- Sys.time()

end_time - start_time

Time difference of 21 mins

Tuning Plot and Variable Importance

varImp(xgb_tune_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

# helper function for the plots
tuneplot <- function(x, probs = .90) {
  ggplot(x) +
    coord_cartesian(ylim = c(quantile(x$results$RMSE, probs = probs), min(x$results$RMSE))) +
    theme_bw()
}

tuneplot(xgb_tune_2021)

---

5.3.2 Fine Tuning Model

5.3.2.1 Second Tuning: Maximum Depth and Minimum Child Weight

After fixing the learning rate to the best tune from the previous iteration and we’ll also set maximum depth to 3 +-1 (or +2 if max_depth == 2) to experiment a bit around the suggested best tune in previous step. Then, well fix maximum depth and minimum child weight.

tune_grid2 <- expand.grid(
  nrounds = seq(from = 50, to = nrounds, by = 50),
  eta = xgb_tune_2021$bestTune$eta,
  max_depth = ifelse(xgb_tune_2021$bestTune$max_depth == 2,
    c(xgb_tune_2021$bestTune$max_depth:4),
    xgb_tune_2021$bestTune$max_depth - 1:xgb_tune_2021$bestTune$max_depth + 1),
  gamma = 0,
  colsample_bytree = 1,
  min_child_weight = c(1, 2, 3),
  subsample = 1
)

xgb_tune2_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = tune_control,
  tuneGrid = tune_grid2,
  method = "xgbTree",
  verbose = TRUE
)

tuneplot(xgb_tune2_2021)

xgb_tune2_2021$bestTune

varImp(xgb_tune2_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.3.2.2 Third Tuning: Column and Row Sampling

tune_grid3 <- expand.grid(
  nrounds = seq(from = 50, to = nrounds, by = 50),
  eta = xgb_tune_2021$bestTune$eta,
  max_depth = xgb_tune2_2021$bestTune$max_depth,
  gamma = 0,
  colsample_bytree = c(0.4, 0.6, 0.8, 1.0),
  min_child_weight = xgb_tune2_2021$bestTune$min_child_weight,
  subsample = c(0.5, 0.75, 1.0)
)

xgb_tune3_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = tune_control,
  tuneGrid = tune_grid3,
  method = "xgbTree",
  verbose = TRUE
)

tuneplot(xgb_tune3_2021, probs = .95)

xgb_tune3_2021$bestTune

varImp(xgb_tune3_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.3.2.3 Fourth Tuning: Gamma

Next, we again pick the best values from previous step, and now will see whether changing the gamma has any effect on the model fit:

tune_grid4 <- expand.grid(
  nrounds = seq(from = 50, to = nrounds, by = 50),
  eta = xgb_tune_2021$bestTune$eta,
  max_depth = xgb_tune2_2021$bestTune$max_depth,
  gamma = c(0, 0.05,0.1, 0.2,0.4, 0.5, 0.7, 0.9, 1.0),
  colsample_bytree = xgb_tune3_2021$bestTune$colsample_bytree,
  min_child_weight = xgb_tune2_2021$bestTune$min_child_weight,
  subsample = xgb_tune3_2021$bestTune$subsample
)

xgb_tune4_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = tune_control,
  tuneGrid = tune_grid4,
  method = "xgbTree",
  verbose = TRUE
)

tuneplot(xgb_tune4_2021)

Warning: The shape palette can deal with a maximum of 6 discrete values because more than 6
becomes difficult to discriminate; you have 9. Consider specifying shapes manually if
you must have them.
Warning: Removed 60 rows containing missing values (geom_point).

xgb_tune4_2021$bestTune

varImp(xgb_tune4_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.3.2.4 Fifth Tuning: Reducing the Learning Rate

Now, we have tuned the hyperparameters and can start reducing the learning rate to get to the final model:

start_time <- Sys.time()

tune_grid5 <- expand.grid(
  nrounds = seq(from = 100, to = 10000, by = 75),
   eta = c(0.01, 0.015, 0.025,0.035, 0.05,0.75, 0.1),
  max_depth = xgb_tune2_2021$bestTune$max_depth,
  gamma = xgb_tune4_2021$bestTune$gamma,
  colsample_bytree = xgb_tune3_2021$bestTune$colsample_bytree,
  min_child_weight = xgb_tune2_2021$bestTune$min_child_weight,
  subsample = xgb_tune3_2021$bestTune$subsample
)



xgb_tune5_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = tune_control,
  tuneGrid = tune_grid5,
  method = "xgbTree",
  verbose = TRUE
)

#tuneplot(xgb_tune5_2021)

end_time <- Sys.time()

end_time - start_time

Time difference of 17 mins

xgb_tune5_2021$bestTune

varImp(xgb_tune5_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.3.2.5 Fitting Final Model

(final_grid_2021 <- expand.grid(
  nrounds = xgb_tune5_2021$bestTune$nrounds,
  eta = xgb_tune5_2021$bestTune$eta,
  max_depth = xgb_tune5_2021$bestTune$max_depth,
  gamma = xgb_tune5_2021$bestTune$gamma,
  colsample_bytree = xgb_tune5_2021$bestTune$colsample_bytree,
  min_child_weight = xgb_tune5_2021$bestTune$min_child_weight,
  subsample = xgb_tune5_2021$bestTune$subsample
))

(xgb_model_2021 <- caret::train(
  x = input_x,
  y = input_y,
  trControl = train_control,
  tuneGrid = final_grid_2021,
  method = "xgbTree",
  verbose = TRUE
))

eXtreme Gradient Boosting 

3197 samples
 766 predictor

No pre-processing
Resampling: None 

varImp(xgb_model_2021, scale = F  ) 

xgbTree variable importance

  only 20 most important variables shown (out of 766)

---

5.4 Model Performance

5.4.1 Checking Model on Test Split Data

We don’t need to look too closely at are training data as Xgboost will heavily overfit the model based on that data. The more important part is how the model performs on in predicting our Test Sample that was not included.

y_pred_test <- predict(xgb_model_2021, data.matrix(te_treated_2021))

test_stats= cbind((te_treated_2021$Worth),y_pred_test)

test_statsR2 = cor(test_stats[,1],test_stats[,2])^2

print(test_statsR2)

[1] 0.79

y_pred_train <- predict(xgb_model_2021, data.matrix(tr_treated_2021))

train_stats = cbind((tr_treated_2021$Worth),y_pred_train)

train_statsR2 = cor(train_stats[,1],train_stats[,2])^2

print(train_statsR2)

[1] 0.94

#test dataset
x <- select(te_treated_2021, -Worth)
y <- (te_treated_2021$Worth)

(xgb_model_rmse <- ModelMetrics::rmse(y, predict(xgb_model_2021, newdata = x)))

[1] 0.33

holdout_x <- select(tr_treated_2021, -Worth)
holdout_y <- tr_treated_2021$Worth

(xgb_model_rmse <- ModelMetrics::rmse(holdout_y, predict(xgb_model_2021, newdata = holdout_x)))

[1] 0.18

5.4.1.1 Graphical Representation of Model

ggplot2::ggplot() +
  aes(x = test_stats[,1], y = test_stats[,2]) +
  geom_jitter() +
  xlab("Predicted Values") +
  ylab("Actual Values") +
  ggtitle("Results of Pitching Model on Test Data")+
  theme(plot.title = element_text(hjust = 0.5,size = 22,color ="steel blue"))+
  geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

---

6 Creating 2022 Projections from Model

6.1 Re-fit model for Important Variables

Now that we have an acceptable model, we can use it to create projections for how well we think players should do in 2022 based on their hitting statistics in 2021. First let’s reduce

Step 1: Only keep variables with high enough importance in model

vip(xgb_model_2021, num_features = 30)  # 10 is the default, 30 gives a visual on the top 30 most important features of the model

unscalevi = vi(xgb_model_2021, method="model") #shows the numbers behind the plot

unscalevi$Importance_perc = with(unscalevi,Importance/sum(Importance)) #adds percentages 

unscalevi # importance by variables

variables_to_keep_2021 = subset(unscalevi, Importance_perc > 0.0010) %>% select(Variable) #Keep Variables that explain at least a small amount [0.1%] of the model. This is a low threshold for inclusion ,but you can adjust this

variables_to_keep_2021b = t(variables_to_keep_2021)

variables_to_keep_2022 = colnames(row_to_names(variables_to_keep_2021b,row_number = 1))

tr_treated_2022 = tr_treated_2021 %>%  select(Worth,one_of(variables_to_keep_2022),starts_with("Team_lev_x_")) #keep modeled important variables along with team indicator variables

te_treated_2022 = te_treated_2021 %>%  select(Worth,one_of(variables_to_keep_2022),starts_with("Team_lev_x_"))

input_x_2022 = as.matrix(select(tr_treated_2022, -Worth))

input_y_2022 = tr_treated_2022$Worth

---

Step 2: Re-fit model with reduced variable scope
Note from the best tune below the nrounds - is the max I set above and eta is at the lowest possible value. This could cause potential overfitting issues, but from our Actual vs. Predicted Graph we know this not to be the case.

(final_grid_2021 <- expand.grid(
  nrounds = xgb_tune5_2021$bestTune$nrounds,
  eta = xgb_tune5_2021$bestTune$eta,
  max_depth = xgb_tune5_2021$bestTune$max_depth,
  gamma = xgb_tune5_2021$bestTune$gamma,
  colsample_bytree = xgb_tune5_2021$bestTune$colsample_bytree,
  min_child_weight = xgb_tune5_2021$bestTune$min_child_weight,
  subsample = xgb_tune5_2021$bestTune$subsample
))

(xgb_model_2022 <- caret::train(
  x = input_x_2022,
  y = input_y_2022,
  trControl = train_control,
  tuneGrid = final_grid_2021,
  method = "xgbTree",
  verbose = TRUE
))

eXtreme Gradient Boosting 

3197 samples
  93 predictor

No pre-processing
Resampling: None 

vip(xgb_model_2022, num_features = 30)

unscalevi24 = vi(xgb_model_2022, method="model")

unscalevi24$Importance_perc = with(unscalevi24,Importance/sum(Importance)) 

unscalevi24

# Save work for later prediction

save(xgb_model_2022,file = '2022_Pitching6x6_Model.Rdata')

pitching6x6 = xgb_model_2022

pitchinginput6x6 = input_x_2022

---

6.2 Get 2022 list of players

6.2.1 Arrange the Data so the Columns are in the exact order as the model

First let’s prepare a file for predicting based on our model object

variableslag6x= row_to_names(as.data.frame(t(variables_to_keep_2022)),row_number = 1)  %>% select (starts_with("lag"))

variables_nolag6x = (owmr::remove_prefix(variableslag6x,"lag" , sep = "_"))

Data_Predict_2022a6x = total_treated_2021_pitching %>% select (one_of(colnames(variables_nolag6x)),Season,playerid)

colnames(Data_Predict_2022a6x) <- paste0("lag_", colnames(Data_Predict_2022a6x))

Data_Predict_2022b6x = total_treated_2021_pitching %>% select (one_of(colnames(variables_nolag6x)))
colnames(Data_Predict_2022b6x) = colnames(variableslag6x)

variables_to_keep_2022_nolag6x = total_treated_2021_pitching %>% select(one_of(variables_to_keep_2022),Season,playerid,starts_with("Team_lev_x_"))%>% select(-one_of(colnames(Data_Predict_2022b6x)))


Data_predict_20226x = sqldf(
  "
  select a.*,b.* from
  Data_Predict_2022a6x a,
  variables_to_keep_2022_nolag6x b
  on b.playerid = a.lag_playerid
  and b.Season = a.lag_Season
  "
) %>% select(-lag_playerid,lag_Season) %>%
  filter(Season == 2021) %>% 
  select(one_of(variables_to_keep_2022),starts_with("Team_lev_x_"))

---

6.3 Create Predictions for Model

6.3.1 Run Projections on Players who Played in 2021

This is the raw prediction score per IP for each pitcher

pitching_predictions6x = as.data.frame(predict(xgb_model_2022,Data_predict_20226x))

names(pitching_predictions6x) = c("Predict_Score")

Data_predict_2022_w_Pitching_Predictions6x = cbind(Data_predict_2022,pitching_predictions6x) %>% select(playerid,Predict_Score)

head(Data_predict_2022_w_Pitching_Predictions6x)

NA

---

6.3.2 Load in Latest 2022 Projections for Innings Pitched

Downloaded from FanGraphs here.

Latest_2022_pitchingdata_FP = read_csv("FanGraph_Fantasy_Baseball_Pitching.csv")

Rows: 817 Columns: 27
-- Column specification -------------------------------------------------------------------------
Delimiter: ","
chr  (3): Name, Team, playerid
dbl (24): GS, G, IP, W, L, QS, SV, HLD, H, ER, HR, SO, BB, WHIP, K/9, BB/9, ERA, FIP, WAR, RA...

i Use `spec()` to retrieve the full column specification for this data.
i Specify the column types or set `show_col_types = FALSE` to quiet this message.

Latest_2022_pitchingdata_FP

NA

---

As you can see from the chart below there aren’t many elite pitchers in the 87+ Predict score range.

Pitching_Data_NonAdj_Projections6x = sqldf(
  "
  select a.*,b.Predict_Score
  from Latest_2022_pitchingdata_FP a 
  left join 
  Data_predict_2022_w_Pitching_Predictions6x b
  on a.playerid = b.playerid
  "
) %>% filter(ADP<370 | is.na(Predict_Score)==F)


Pitching_Data_Adj_Projections6x =
Pitching_Data_NonAdj_Projections6x %>% 
  mutate(
    Avg_IP = 60,
    AdjPredict_Score_raw = ifelse(is.na(Predict_Score),NA,Predict_Score*(IP/Avg_IP)),
    max_predscore= max(AdjPredict_Score_raw,na.rm = T),
    AdjPredict_Score = ifelse (is.na(AdjPredict_Score_raw),NA,AdjPredict_Score_raw *100/max_predscore),
    WAR_rank = order(order(rank(WAR,ties.method = 'average'),decreasing = TRUE)),
    AdjPredict_Score_Rank = order(order(rank(AdjPredict_Score,ties.method = 'average'),decreasing = TRUE))-sum(is.na(AdjPredict_Score)),
        Ranks_Above_ADP = ADP - AdjPredict_Score_Rank
  ) %>% select (Name,ADP,WAR, WAR_rank,AdjPredict_Score ,AdjPredict_Score_Rank,Ranks_Above_ADP)


  

ggplot2::qplot(Pitching_Data_Adj_Projections6x$AdjPredict_Score, main="Predictions") + geom_histogram(colour="black", fill="grey") + theme_bw()

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

---

7 2022 Projections Full

7.1 Table of Pitching Projections (Players who Didn’t Play in 2021 - Recieve an NA)

AdjPredict_Score are normalized to 100

tableexport =
Pitching_Data_Adj_Projections6x %>%
  arrange (ADP,WAR) %>% 
  kbl() %>% 
 kable_material(c("striped", "hover","condensed","responsive"),full_width = F,fixed_thead = T)

save_kable(tableexport,file = "Pitching6x6.html")

#tableexport

This is a better formatted Table

ft_dt <- Pitching_Data_Adj_Projections6x[1:nrow(Pitching_Data_Adj_Projections6x), 1:ncol(Pitching_Data_Adj_Projections6x)] %>% 
  filter(AdjPredict_Score_Rank>0)%>%  arrange((AdjPredict_Score_Rank))

ft_dt$ADP <- color_tile("white", "red")(ft_dt$ADP)

ft_dt$WAR <- color_bar("lightblue")(ft_dt$WAR)

ft_dt$AdjPredict_Score<- color_bar("lightblue")(ft_dt$AdjPredict_Score)

ft_dt$WAR_Rank <- color_tile("green","orange")(ft_dt$WAR_rank)

ft_dt$Predict_Rank <- color_tile("green","orange")(ft_dt$AdjPredict_Score_Rank) 


ft_dt$Ranks_Above_ADP <- 
  ifelse(
  ft_dt$Ranks_Above_ADP < 0,
  cell_spec(round(ft_dt$Ranks_Above_ADP,2), color = "red", italic = T),
  cell_spec(round(ft_dt$Ranks_Above_ADP,2), color = "green", italic = T)
)


ft_dt2 <- ft_dt[c("Name", "ADP", "WAR", "AdjPredict_Score", "WAR_Rank","Predict_Rank","Ranks_Above_ADP")]



table_export = 
kbl(ft_dt2, escape = F) %>% 
 kable_material(c("striped", "hover","condensed","responsive"),full_width = F,fixed_thead = T) %>%   column_spec(6, width = "3cm") %>%
  add_header_above(c(" ", "Scores" = 3, "Ranks" = 2," "))

save_kable(table_export,file = "Pitching6x6_updated.html")
  
table_export  

	Scores			Ranks
Name	ADP	WAR	AdjPredict_Score	WAR_Rank	Predict_Rank	Ranks_Above_ADP
Eduardo Rodriguez	150.1	3.5	100.0	15	1	149.1
Brady Singer	492.4	2.2	97.1	53	2	490.4
Dylan Cease	81.2	3.3	95.4	20	3	78.2
Josiah Gray	290.7	1.6	93.1	89	4	286.7
José Quintana	575.5	1.0	92.8	143	5	570.5
Shane Bieber	31.2	4.4	91.9	4	6	25.2
Tarik Skubal	197.6	2.0	90.9	61	7	190.6
Aaron Nola	41.0	4.2	88.3	7	8	33
Nick Pivetta	350.8	1.7	87.5	86	9	341.8
Alek Manoah	93.0	2.8	87.5	33	10	83
Shane McClanahan	109.6	2.6	87.3	40	11	98.6
Luis Castillo	103.2	3.3	85.4	21	12	91.2
Zac Gallen	154.3	2.3	85.2	50	13	141.3
Logan Webb	69.9	3.7	84.7	11	14	55.9
Tyler Mahle	132.2	2.6	84.7	39	15	117.2
Logan Gilbert	161.8	2.2	84.6	52	16	145.8
Corbin Burnes	10.8	5.3	84.5	1	17	-6.2
Jon Gray	236.9	2.3	84.4	51	18	218.9
Brad Keller	582.1	1.6	83.8	91	19	563.1
Germán Márquez	261.7	2.9	83.5	29	20	241.7
Walker Buehler	18.6	3.8	80.5	8	21	-2.4
Yusei Kikuchi	301.3	1.7	80.2	83	22	279.3
Jesús Luzardo	289.5	0.8	79.8	158	23	266.5
Gerrit Cole	7.1	5.0	79.0	3	24	-16.9
Luis Garcia	144.7	2.0	78.9	64	25	119.7
Sean Manaea	141.4	2.8	78.9	31	26	115.4
Freddy Peralta	54.0	3.6	78.8	12	27	27
Trevor Rogers	92.9	3.1	78.7	25	28	64.9
Lucas Giolito	44.2	3.7	77.9	10	29	15.2
Michael Kopech	161.1	2.5	77.9	43	30	131.1
Ian Anderson	151.1	2.1	77.0	56	31	120.1
Tanner Houck	206.4	2.4	76.5	47	32	174.4
José Berríos	77.5	3.2	76.4	23	33	44.5
Andrew Heaney	299.4	1.4	76.3	102	34	265.4
Framber Valdez	149.6	3.1	76.3	26	35	114.6
Luis Patiño	290.7	1.5	75.5	99	36	254.7
Patrick Sandoval	196.0	2.4	75.4	46	37	159
Frankie Montas	92.8	3.3	75.2	19	38	54.8
Cole Irvin	532.2	1.5	74.8	97	39	493.2
Kris Bubic	536.0	1.1	73.7	125	40	496
Nathan Eovaldi	135.1	3.6	73.5	13	41	94.1
Julio Urías	34.7	3.5	72.8	14	42	-7.3
Dane Dunning	484.6	1.9	72.4	72	43	441.6
Jordan Montgomery	208.7	2.7	72.4	35	44	164.7
Patrick Corbin	456.0	1.6	72.2	95	45	411
Steven Matz	249.2	1.9	72.2	67	46	203.2
Blake Snell	113.4	2.4	71.9	49	47	66.4
Zac Lowther	999.0	0.8	71.9	159	48	951
Aaron Civale	269.5	1.5	71.8	96	49	220.5
Zach Plesac	346.6	1.5	71.6	100	50	296.6
Bruce Zimmermann	600.8	1.3	71.6	114	51	549.8
Carlos Hernández	412.5	1.2	71.5	121	52	360.5
Mitch Keller	531.5	1.2	71.5	123	53	478.5
Kyle Hendricks	279.8	1.7	70.6	79	54	225.8
Cristian Javier	284.0	1.2	70.6	118	55	229
Sandy Alcantara	43.4	3.3	70.4	18	56	-12.6
Bailey Ober	261.3	2.2	70.2	55	57	204.3
Shohei Ohtani	9.2	2.8	70.1	32	58	-48.8
Kevin Gausman	71.4	3.2	69.9	24	59	12.4
Kyle Gibson	393.9	1.9	69.4	71	60	333.9
Robbie Ray	49.9	3.0	69.4	27	61	-11.1
Michael Lorenzen	543.3	0.6	69.3	203	62	481.3
Yu Darvish	100.3	2.7	69.3	34	63	37.3
Matt Manning	543.4	1.1	69.2	124	64	479.4
Sonny Gray	168.2	2.7	68.8	38	65	103.2
Joe Ryan	221.5	2.2	68.5	54	66	155.5
Huascar Ynoa	245.8	1.8	68.0	76	67	178.8
Mike Minor	532.2	2.0	67.8	63	68	464.2
Brandon Woodruff	21.2	4.2	67.2	6	69	-47.8
Triston McKenzie	229.0	1.6	67.2	93	70	159
JT Brubaker	537.7	1.4	66.6	105	71	466.7
Hyun Jin Ryu	207.3	2.4	66.3	45	72	135.3
Dylan Bundy	498.5	1.3	66.2	109	73	425.5
Kyle Freeland	574.4	1.5	66.0	98	74	500.4
Max Fried	69.8	3.5	65.9	16	75	-5.2
Keegan Akin	600.0	1.0	65.9	139	76	524
Chris Flexen	413.6	1.9	65.3	73	77	336.6
Casey Mize	279.4	1.7	65.0	84	78	201.4
Ranger Suárez	179.3	2.4	64.9	44	79	100.3
Joe Musgrove	70.6	3.2	64.7	22	80	-9.4
Spencer Howard	575.8	0.9	64.4	150	81	494.8
Reid Detmers	427.4	1.1	64.2	134	82	345.4
Drew Rasmussen	282.4	1.6	63.5	88	83	199.4
Antonio Senzatela	585.0	2.0	63.1	62	84	501
Taylor Hearn	591.2	0.7	62.7	183	85	506.2
Dallas Keuchel	567.3	1.2	62.7	119	86	481.3
Austin Gomber	551.9	1.2	62.4	116	87	464.9
Pablo López	140.6	2.7	62.3	36	88	52.6
Max Scherzer	19.7	3.8	62.1	9	89	-69.3
Jordan Lyles	578.9	0.6	61.5	196	90	488.9
Ryan Yarbrough	541.5	1.4	61.5	106	91	450.5
Cal Quantrill	267.3	1.7	61.0	82	92	175.3
Merrill Kelly	526.0	1.7	60.8	81	93	433
José Suarez	535.1	1.3	60.6	113	94	441.1
Carlos Carrasco	284.6	1.3	60.3	110	95	189.6
Zack Wheeler	31.0	4.4	60.3	5	96	-65
Alex Wood	219.9	1.7	59.8	78	97	122.9
Vladimir Gutierrez	585.4	0.2	59.7	322	98	487.4
Carlos Rodón	110.2	2.9	59.5	28	99	11.2
Eric Lauer	306.3	1.4	59.5	103	100	206.3
Chris Bassitt	128.5	2.4	59.1	48	101	27.5
Aaron Ashby	264.1	1.5	58.9	101	102	162.1
Lance Lynn	70.0	3.4	58.9	17	103	-33
Alex Cobb	242.9	1.9	58.8	70	104	138.9
Tony Gonsolin	300.6	0.9	58.7	148	105	195.6
Elieser Hernandez	339.6	1.0	58.4	145	106	233.6
José Urquidy	212.1	2.0	58.2	65	107	105.1
James Kaprielian	448.3	1.1	58.2	126	108	340.3
Wil Crowe	600.8	0.1	57.0	379	109	491.8
Wade Miley	502.0	1.2	56.9	120	110	392
Chris Sale	85.9	1.9	56.8	66	111	-25.1
Charlie Morton	94.3	2.9	56.3	30	112	-17.7
Bryse Wilson	600.5	0.5	56.3	223	113	487.5
Zack Greinke	312.8	1.6	56.1	90	114	198.8
Marco Gonzales	319.6	1.7	56.0	80	115	204.6
Chris Paddack	436.8	1.2	55.9	117	116	320.8
Zach Davies	595.9	0.2	55.7	343	117	478.9
J.A. Happ	597.0	0.6	55.2	199	118	479
Erick Fedde	589.7	0.6	54.9	188	119	470.7
Daniel Bard	579.3	0.4	54.8	266	120	459.3
Marcus Stroman	187.5	2.6	54.7	41	121	66.5
Zach Eflin	485.0	1.9	54.3	69	122	363
Tylor Megill	334.2	1.1	54.0	129	123	211.2
Martín Pérez	598.8	0.8	53.8	168	124	474.8
Jack Flaherty	115.4	1.6	53.7	87	125	-9.6
Michael Wacha	549.7	0.9	53.6	149	126	423.7
Adam Ottavino	598.1	0.2	53.3	328	127	471.1
Alec Mills	594.1	0.6	53.3	211	128	466.1
Daniel Lynch	559.5	0.7	52.7	174	129	430.5
Anthony DeSclafani	212.6	1.8	52.7	77	130	82.6
Jameson Taillon	299.6	2.1	52.7	59	131	168.6
Corey Kluber	324.4	1.1	52.4	130	132	192.4
Luke Weaver	514.4	1.1	52.1	128	133	381.4
Garrett Whitlock	239.2	1.2	52.0	122	134	105.2
José Alvarado	587.6	0.5	51.8	237	135	452.6
John Means	217.1	2.7	50.8	37	136	81.1
Tanner Rainey	371.9	0.2	50.6	325	137	234.9
Gregory Soto	194.3	0.5	50.5	228	138	56.3
Drew Smyly	577.4	0.6	50.4	208	139	438.4
Héctor Neris	493.9	0.4	50.0	242	140	353.9
Glenn Otto	582.9	0.9	50.0	152	141	441.9
Taijuan Walker	414.5	0.6	49.6	185	142	272.5
James Karinchak	468.6	0.7	49.5	171	143	325.6
Randy Dobnak	600.4	0.8	49.4	162	144	456.4
Madison Bumgarner	507.1	1.1	49.4	131	145	362.1
Nestor Cortes	333.1	1.4	49.3	107	146	187.1
Jacob deGrom	19.7	5.1	49.0	2	147	-127.3
Adrian Houser	465.0	1.4	49.0	104	148	317
Kyle Finnegan	351.5	0.2	48.5	336	149	202.5
Edward Cabrera	523.6	0.6	47.8	202	150	373.6
Paul Sewald	289.0	0.5	47.7	234	151	138
Chad Green	403.2	1.0	47.7	147	152	251.2
Devin Williams	287.1	1.1	47.6	133	153	134.1
Scott Barlow	159.3	0.9	47.5	155	154	5.3
Clayton Kershaw	149.4	2.5	47.4	42	155	-5.6
Pete Fairbanks	517.3	0.6	47.4	194	156	361.3
Jake Diekman	585.9	0.5	47.2	232	157	428.9
Amir Garrett	586.6	0.2	47.1	333	158	428.6
Carlos Estévez	528.6	0.2	47.0	337	159	369.6
Phil Maton	600.4	0.4	46.6	248	160	440.4
Stephen Strasburg	294.0	1.3	46.5	111	161	133
Tanner Scott	599.0	0.7	46.4	182	162	437
Lucas Sims	243.6	0.7	46.1	180	163	80.6
Tyler Alexander	586.8	0.8	45.9	167	164	422.8
Trevor May	565.3	0.4	45.9	239	165	400.3
Ryne Stanek	598.8	0.2	45.8	330	166	432.8
Alex Colomé	378.0	0.1	45.8	387	167	211
Adam Wainwright	194.7	2.1	45.8	58	168	26.7
Kyle Funkhouser	999.0	0.1	45.8	386	169	830
A.J. Alexy	596.2	0.6	45.7	186	170	426.2
Aroldis Chapman	83.3	0.9	45.6	151	171	-87.7
Justin Steele	598.7	0.4	45.6	245	172	426.7
Aaron Bummer	595.4	1.0	45.4	137	173	422.4
Josh Fleming	600.8	0.5	45.4	222	174	426.8
Brad Boxberger	588.3	0.2	45.4	331	175	413.3
Mychal Givens	549.7	0.1	45.2	419	176	373.7
Zach Thompson	546.8	1.0	45.2	146	177	369.8
Bryan Shaw	597.8	-0.1	45.1	507	178	419.8
Hansel Robles	596.0	-0.1	45.1	505	179	417
Blake Treinen	147.6	0.8	44.9	166	180	-32.4
Mike Mayers	594.7	0.5	44.6	221	181	413.7
Chris Stratton	519.4	0.3	44.3	284	182	337.4
Giovanny Gallegos	113.5	1.0	44.3	140	183	-69.5
Miles Mikolas	512.8	1.6	43.8	94	184	328.8
Brent Suter	587.6	0.4	43.8	268	185	402.6
Garrett Crochet	549.8	0.9	43.6	154	186	363.8
Rich Hill	477.4	0.7	43.6	173	187	290.4
Dillon Peters	600.7	0.0	43.4	421	188	412.7
Joe Jiménez	999.0	0.2	43.3	329	189	810
Lou Trivino	237.2	0.2	43.2	316	190	47.2
Domingo Germán	494.6	1.1	43.1	132	191	303.6
Daniel Norris	999.0	0.1	43.1	398	192	807
Génesis Cabrera	596.6	0.6	43.1	205	193	403.6
Tyler Wells	454.6	0.9	43.1	153	194	260.6
David Peterson	595.2	0.7	43.0	175	195	400.2
Heath Hembree	600.8	0.2	42.9	338	196	404.8
Joely Rodríguez	600.9	0.5	42.8	214	197	403.9
Tyler Anderson	564.5	0.7	42.7	172	198	366.5
Matt Barnes	245.2	0.8	42.7	164	199	46.2
Jonathan Loáisiga	393.7	1.0	42.6	142	200	193.7
Johnny Cueto	585.8	0.3	42.5	283	201	384.8
Seth Lugo	598.1	0.5	42.5	236	202	396.1
Yimi García	581.1	0.2	42.4	352	203	378.1
José Cisnero	999.0	0.3	42.2	281	204	795
Austin Voth	600.4	0.0	42.1	420	205	395.4
Michael Fulmer	347.0	0.6	42.1	192	206	141
Rowan Wick	302.0	0.4	42.0	249	207	95
Camilo Doval	159.8	0.5	41.9	229	208	-48.2
Cole Sulser	427.2	0.7	41.6	177	209	218.2
Justus Sheffield	599.5	0.4	41.4	269	210	389.5
J.B. Wendelken	600.1	0.2	41.3	341	211	389.1
Jordan Romano	88.3	0.8	41.2	160	212	-123.7
David Price	581.5	0.5	41.2	216	213	368.5
Matt Wisler	599.6	0.4	41.1	270	214	385.6
Diego Castillo	409.2	0.6	41.0	191	215	194.2
Justin Dunn	577.6	0.6	40.6	198	216	361.6
Hirokazu Sawamura	589.6	0.0	40.6	443	217	372.6
Brad Hand	520.0	0.2	40.2	351	218	302
Emilio Pagán	551.7	0.1	40.1	364	219	332.7
Jaime Barría	999.0	0.5	40.1	238	220	779
Anthony Banda	999.0	-0.1	40.1	514	221	778
Anthony Bender	412.8	0.7	40.1	184	222	190.8
Edwin Díaz	63.5	1.1	39.9	135	223	-159.5
Michael Pineda	465.2	1.4	39.9	108	224	241.2
Tyler Rogers	534.3	0.5	39.8	230	225	309.3
Sam Hentges	999.0	0.3	39.8	306	226	773
Jeurys Familia	597.5	0.3	39.8	305	227	370.5
Griffin Canning	594.2	0.5	39.8	213	228	366.2
Lance McCullers Jr.	260.0	1.7	39.8	85	229	31
Daniel Hudson	550.2	0.5	39.7	225	230	320.2
Clay Holmes	600.3	0.6	39.7	201	231	369.3
Andrew Wantz	999.0	0.4	39.7	262	232	767
Taylor Rogers	172.5	1.1	39.6	127	233	-60.5
Caleb Smith	592.3	0.0	39.6	441	234	358.3
Jorge Alcala	459.0	0.6	39.6	209	235	224
Luis Gil	525.9	0.8	39.5	165	236	289.9
Dylan Floro	221.3	0.4	39.5	251	237	-15.7
Josh Staumont	505.8	0.4	39.2	244	238	267.8
Spencer Patton	599.0	0.1	39.1	410	239	360
Tyler Duffey	577.8	0.4	39.0	260	240	337.8
Jake Brentz	600.8	0.3	39.0	301	241	359.8
Josh Sborz	600.9	0.3	38.9	290	242	358.9
Raisel Iglesias	47.6	1.0	38.8	138	243	-195.4
Chad Kuhl	600.7	0.2	38.8	349	244	356.7
Ryan Pressly	64.3	1.0	38.8	136	245	-180.7
J.P. Feyereisen	597.6	0.0	38.7	487	246	351.6
Keegan Thompson	999.0	0.0	38.3	429	247	752
J.C. Mejía	999.0	0.1	38.2	389	248	751
Liam Hendriks	32.0	1.6	38.1	92	249	-217
Sam Howard	999.0	0.3	38.1	310	250	749
Robert Stephenson	590.5	0.0	38.0	466	251	339.5
Tim Mayza	599.3	0.6	37.9	197	252	347.3
José Ureña	999.0	0.0	37.8	467	253	746
Craig Kimbrel	164.4	0.7	37.6	176	254	-89.6
Jeff Hoffman	999.0	0.0	37.5	426	255	744
Garrett Richards	598.7	0.4	37.4	272	256	342.7
Anthony Misiewicz	999.0	0.3	37.3	295	257	742
Tony Santillan	600.3	0.4	37.3	263	258	342.3
Brusdar Graterol	567.4	0.7	37.2	178	259	308.4
Andrew Kittredge	257.6	0.8	37.2	169	260	-2.4
A.J. Minter	596.5	0.5	37.1	215	261	335.5
Trevor Stephan	999.0	0.1	37.1	368	262	737
Ryan Helsley	600.2	0.1	37.0	380	263	337.2
Ryan Tepera	581.3	0.4	36.8	246	264	317.3
David Bednar	190.8	0.9	36.8	156	265	-74.2
Jackson Kowar	594.6	0.5	36.8	233	266	328.6
Dean Kremer	999.0	0.6	36.7	210	267	732
Jake Odorizzi	533.5	0.8	36.6	163	268	265.5
Patrick Murphy	999.0	0.4	36.6	254	269	730
Yusmeiro Petit	601.0	-0.3	36.5	531	270	331
Sergio Romo	600.5	0.0	36.4	450	271	329.5
Tyler Kinley	999.0	0.0	36.3	433	272	727
Kendall Graveman	531.0	0.4	36.3	240	273	258
Griffin Jax	600.8	0.1	36.3	363	274	326.8
Trevor Richards	600.5	0.4	36.3	261	275	325.5
Rafael Dolis	999.0	0.0	36.2	434	276	723
Sam Coonrod	600.9	0.4	36.1	271	277	323.9
Emmanuel Clase	57.5	1.3	36.0	112	278	-220.5
Caleb Thielbar	999.0	0.6	36.0	195	279	720
Dinelson Lamet	400.5	0.9	36.0	157	280	120.5
Tyler Matzek	573.2	0.7	36.0	181	281	292.2
Corey Knebel	151.3	0.6	35.9	204	282	-130.7
Josh Taylor	999.0	0.5	35.4	217	283	716
Paolo Espino	599.2	0.2	35.2	327	284	315.2
Nick Wittgren	999.0	0.1	35.2	374	285	714
Tim Hill	999.0	0.3	35.1	293	286	713
Wandy Peralta	999.0	0.2	35.1	340	287	712
Will Smith	120.8	0.5	35.1	227	288	-167.2
Anthony Bass	600.9	0.1	35.1	412	289	311.9
Alex Vesia	600.0	0.3	35.0	289	290	310
Josh Hader	30.5	1.3	35.0	115	291	-260.5
Archie Bradley	600.7	0.2	35.0	348	292	308.7
Eli Morgan	600.5	0.2	34.6	346	293	307.5
JT Chargois	600.3	0.3	34.5	279	294	306.3
Pierce Johnson	433.3	0.6	34.4	200	295	138.3
Duane Underwood Jr.	999.0	0.1	34.4	417	296	703
Ross Stripling	583.7	0.4	34.3	256	297	286.7
Brooks Raley	600.9	0.3	34.0	296	298	302.9
Miguel Castro	600.9	0.1	33.8	372	299	301.9
Lucas Gilbreath	596.4	0.0	33.7	471	300	296.4
José Quijada	999.0	0.5	33.7	212	301	698
Connor Brogdon	600.4	0.4	33.6	252	302	298.4
Nick Sandlin	599.8	0.6	33.6	207	303	296.8
Jhoulys Chacín	999.0	-0.4	33.5	533	304	695
Jorge López	599.9	0.4	33.4	265	305	294.9
Dillon Tate	600.9	0.4	33.3	264	306	294.9
Jake Cousins	600.1	0.5	33.3	226	307	293.1
Luke Jackson	598.1	0.4	33.1	267	308	290.1
Adbert Alzolay	444.0	0.7	33.0	179	309	135
Kenley Jansen	90.0	0.6	33.0	187	310	-220
Jarlín García	600.8	0.1	32.9	411	311	289.8
Darwinzon Hernandez	999.0	0.2	32.7	324	312	687
Trevor Williams	601.0	0.2	32.5	326	313	288
Rafael Montero	999.0	0.1	32.5	394	314	685
Drew Steckenrider	401.5	0.3	32.5	294	315	86.5
Austin Warren	999.0	0.5	32.4	231	316	683
Matthew Boyd	591.0	1.0	32.3	144	317	274
Collin McHugh	545.5	0.6	32.3	189	318	227.5
John King	999.0	0.5	32.1	219	319	680
Vince Velasquez	599.4	0.3	32.1	308	320	279.4
Alex Lange	999.0	0.1	32.0	366	321	678
Justin Wilson	999.0	0.1	32.0	377	322	677
Mark Melancon	131.8	0.4	31.9	247	323	-191.2
Tyler Gilbert	599.9	0.4	31.9	255	324	275.9
Zach Pop	999.0	0.2	31.9	345	325	674
Sean Doolittle	598.5	0.1	31.8	402	326	272.5
Bailey Falter	599.8	0.5	31.8	218	327	272.8
Zack Littell	599.3	0.1	31.6	367	328	271.3
Kyle Zimmer	999.0	-0.2	31.3	528	329	670
Steve Cishek	595.3	0.0	31.3	439	330	265.3
Ian Kennedy	409.8	-0.1	31.2	512	331	78.8
Jake Woodford	599.5	0.0	31.1	460	332	267.5
Kwang Hyun Kim	593.0	0.3	31.1	303	333	260
Joe Smith	999.0	0.1	31.0	361	334	665
Ryan Hendrix	999.0	0.0	31.0	468	335	664
Cionel Pérez	999.0	0.5	31.0	235	336	663
Jeffrey Springs	999.0	0.3	31.0	287	337	662
Brett Martin	999.0	0.4	30.9	275	338	661
Max Kranick	999.0	0.3	30.8	297	339	660
Jakob Junis	600.1	0.2	30.7	332	340	260.1
Adam Morgan	999.0	0.1	30.3	391	341	658
Ryan Thompson	999.0	0.3	30.3	278	342	657
Taylor Widener	600.5	0.0	30.2	424	343	257.5
Jake McGee	251.1	0.4	30.1	241	344	-92.9
Andrew Chafin	587.9	0.6	29.9	206	345	242.9
Blake Taylor	999.0	0.2	29.8	353	346	653
Dominic Leone	600.8	0.1	29.7	371	347	253.8
Phil Bickford	600.9	0.4	29.6	258	348	252.9
Johan Oviedo	999.0	0.2	29.6	356	349	650
Chris Martin	600.3	0.4	29.6	277	350	250.3
Dennis Santana	999.0	0.0	29.4	442	351	648
Craig Stammen	599.4	0.4	29.4	259	352	247.4
Tony Watson	999.0	0.0	28.7	479	353	646
Taylor Clarke	999.0	0.0	28.6	437	354	645
Luis Cessa	588.8	0.3	28.5	299	355	233.8
Austin Adams	600.0	0.4	28.3	273	356	244
Joe Barlow	216.1	0.2	28.3	344	357	-140.9
Brett Anderson	600.9	0.6	28.2	190	358	242.9
Joe Kelly	600.4	0.4	28.2	243	359	241.4
Reynaldo López	557.4	0.4	28.2	257	360	197.4
Sean Newcomb	600.1	0.1	28.1	400	361	239.1
Austin Davis	999.0	0.2	28.1	320	362	637
Mike Foltynewicz	600.4	-0.1	28.0	511	363	237.4
Sammy Long	596.1	0.3	27.9	307	364	232.1
Deolis Guerra	599.9	0.2	27.9	342	365	234.9
Albert Abreu	999.0	-0.1	27.9	496	366	633
Josh Tomlin	999.0	-0.3	27.8	532	367	632
Aaron Loup	594.4	0.4	27.8	274	368	226.4
Bryan Garcia	999.0	-0.2	27.7	530	369	630
Joe Mantiply	999.0	0.2	27.6	334	370	629
Greg Holland	600.2	0.0	27.5	473	371	229.2
Tucker Davidson	584.9	0.6	27.4	193	372	212.9
Joe Ross	599.6	0.7	27.3	170	373	226.6
Alex Reyes	395.7	0.2	27.2	335	374	21.7
Alexander Wells	999.0	0.4	26.8	276	375	624
Wander Suero	999.0	0.2	26.7	317	376	623
Demarcus Evans	600.6	0.0	26.7	472	377	223.6
Michael King	600.5	0.5	26.6	220	378	222.5
Richard Rodríguez	556.5	0.1	26.5	414	379	177.5
Richard Bleier	600.7	0.5	26.4	224	380	220.7
Ryan Weathers	600.8	0.3	26.4	300	381	219.8
Yency Almonte	999.0	-0.2	26.3	524	382	617
Matt Harvey	999.0	0.3	26.1	286	383	616
Logan Allen	600.8	0.1	25.9	418	384	216.8
Art Warren	482.3	0.8	25.9	161	385	97.3
Ralph Garza Jr.	999.0	-0.1	25.9	501	386	613
Andrew Miller	999.0	-0.1	25.8	510	387	612
Brad Brach	999.0	0.0	25.4	436	388	611
Nick Mears	999.0	0.0	25.4	480	389	610
Sean Reid-Foley	999.0	0.0	25.3	438	390	609
Domingo Tapia	999.0	0.0	25.0	492	391	608
Adam Cimber	600.6	0.3	24.9	282	392	208.6
Sean Poppen	999.0	0.2	24.8	319	393	606
Kolby Allard	601.0	0.2	24.6	312	394	207
Hunter Strickland	599.9	0.0	24.6	423	395	204.9
Brandon Kintzler	999.0	0.0	24.5	464	396	603
Matt Shoemaker	999.0	0.0	24.5	457	397	602
Ross Detwiler	999.0	0.0	24.4	474	398	601
Steven Brault	600.9	0.0	24.2	432	399	201.9
Steven Okert	999.0	0.3	24.2	304	400	599
Josh Rogers	600.8	-0.1	24.1	506	401	199.8
Noé Ramirez	600.9	0.1	24.1	390	402	198.9
Jesse Chavez	999.0	0.1	24.0	408	403	596
Wily Peralta	600.8	0.1	23.9	388	404	196.8
Tyler Clippard	600.8	-0.2	23.8	529	405	195.8
Jacob Webb	999.0	0.0	23.7	447	406	593
Drew Smith	999.0	0.1	23.4	360	407	592
Chi Chi González	999.0	0.1	23.3	396	408	591
Erik Swanson	600.1	0.3	23.0	291	409	191.1
Packy Naughton	999.0	0.3	23.0	292	410	589
Andres Machado	999.0	-0.1	22.8	513	411	588
Derek Holland	999.0	-0.1	22.6	498	412	587
Jay Jackson	999.0	0.2	22.6	347	413	586
T.J. McFarland	999.0	0.1	22.4	378	414	585
Mitch White	562.7	0.3	22.3	288	415	147.7
Buck Farmer	999.0	-0.2	22.2	519	416	583
Blake Parker	999.0	0.1	22.1	393	417	582
Junior Guerra	999.0	0.0	22.0	490	418	581
Touki Toussaint	598.6	0.1	22.0	369	419	179.6
Ryan Borucki	999.0	0.2	21.9	355	420	579
Trevor Bauer	208.4	1.0	21.6	141	421	-212.6
Jharel Cotton	600.8	0.1	21.6	376	422	178.8
Kodi Whitley	999.0	0.2	21.6	350	423	576
Bryan Abreu	999.0	0.1	21.5	392	424	575
Charlie Barnes	999.0	0.2	21.5	339	425	574
Michael Rucker	999.0	0.0	21.5	427	426	573
Ben Bowden	999.0	0.1	21.4	395	427	572
Kyle Muller	590.8	0.3	21.3	302	428	162.8
Nabil Crismatt	999.0	0.1	21.1	403	429	570
Brandon Bielak	999.0	0.1	21.0	362	430	569
Cody Ponce	999.0	0.2	20.8	354	431	568
Mason Thompson	999.0	0.0	20.8	448	432	567
Trevor Cahill	999.0	0.2	20.0	313	433	566
Alex Claudio	999.0	0.0	19.9	463	434	565
Carlos Martínez	596.0	0.0	19.8	477	435	161
Ryan Burr	999.0	0.0	19.7	440	436	563
Paul Campbell	999.0	-0.1	19.6	495	437	562
Brandon Workman	999.0	-0.2	19.6	521	438	561
Jake Arrieta	999.0	0.0	19.5	481	439	560
Joel Payamps	999.0	0.3	19.0	285	440	559
Trent Thornton	999.0	0.1	18.9	401	441	558
Jordan Holloway	999.0	-0.1	18.8	502	442	557
Juan Minaya	999.0	0.1	18.6	397	443	556
Casey Sadler	600.5	0.3	18.6	280	444	156.5
Kyle McGowin	999.0	0.2	18.4	321	445	554
Matt Foster	999.0	0.1	18.3	373	446	553
Humberto Castellanos	999.0	0.1	18.2	370	447	552
Paul Blackburn	999.0	0.2	18.0	315	448	551
José Álvarez	600.2	0.2	17.7	359	449	151.2
Tyler Zuber	999.0	0.0	17.6	462	450	549
Yohan Ramirez	599.8	0.1	17.4	413	451	148.8
Jordan Sheffield	999.0	-0.2	17.4	520	452	547
Miguel Sánchez	999.0	0.0	17.3	484	453	546
Jacob Barnes	999.0	0.1	17.1	381	454	545
Justin Garza	999.0	-0.2	16.7	523	455	544
Drew Pomeranz	582.3	0.1	16.7	383	456	126.3
Phillips Valdez	999.0	-0.1	16.4	493	457	542
Ryne Harper	999.0	0.1	16.3	416	458	541
Danny Coulombe	999.0	0.1	16.2	382	459	540
José Ruiz	999.0	0.2	16.1	357	460	539
JD Hammer	999.0	0.0	16.0	483	461	538
Michael Feliz	999.0	0.0	15.9	425	462	537
Ervin Santana	999.0	-0.5	15.8	534	463	536
Tommy Nance	999.0	0.1	15.6	385	464	535
Rex Brothers	999.0	0.0	15.6	461	465	534
Junior Fernández	999.0	0.1	15.5	415	466	533
Anthony Castro	999.0	0.1	15.2	375	467	532
Dan Winkler	999.0	-0.1	15.1	515	468	531
Chasen Shreve	999.0	-0.1	15.0	509	469	530
Héctor Santiago	999.0	-0.2	14.7	525	470	529
Brett de Geus	999.0	0.0	14.6	475	471	528
Hoby Milner	999.0	0.1	14.3	405	472	527
Wade LeBlanc	999.0	0.2	14.1	358	473	526
Matt Peacock	999.0	0.0	14.1	488	474	525
Shane Greene	999.0	-0.1	13.7	499	475	524
Nick Neidert	999.0	-0.2	13.7	517	476	523
Aaron Sanchez	600.6	0.0	13.6	451	477	123.6
Matt Moore	999.0	-0.1	13.5	503	478	521
Anthony Kay	999.0	0.0	13.4	422	479	520
Sean Guenther	999.0	0.2	13.4	323	480	519
Victor González	999.0	0.1	13.3	365	481	518
Keynan Middleton	999.0	-0.1	13.1	500	482	517
Humberto Mejía	999.0	0.0	13.1	453	483	516
Sam Selman	999.0	0.0	12.9	444	484	515
Danny Duffy	587.9	0.2	12.3	318	485	102.9
Cody Poteet	999.0	0.1	12.1	407	486	513
Chris Mazza	999.0	0.3	11.9	309	487	512
Kohei Arihara	999.0	0.0	11.9	449	488	511
Yennsy Díaz	999.0	-0.1	11.9	516	489	510
Braxton Garrett	999.0	0.0	11.4	428	490	509
Louis Head	600.5	0.3	11.4	311	491	109.5
Ashton Goudeau	999.0	-0.2	11.4	527	492	507
Chase Anderson	999.0	0.0	11.2	491	493	506
Cam Bedrosian	999.0	0.0	10.7	458	494	505
César Valdez	600.3	0.0	10.6	431	495	105.3
Sam Clay	999.0	0.0	10.3	430	496	503
Dustin May	559.5	0.3	10.1	298	497	62.5
Daniel Ponce de Leon	600.7	0.0	9.6	456	498	102.7
Dillon Maples	999.0	0.1	9.3	409	499	500
Erasmo Ramírez	999.0	-0.2	9.2	526	500	499
Caleb Baragar	999.0	-0.2	9.2	522	501	498
Robert Gsellman	999.0	-0.2	8.6	518	502	497
Marcos Diplán	999.0	0.0	8.6	459	503	496
Miguel Diaz	999.0	-0.1	8.5	494	504	495
Edwin Uceta	999.0	0.0	8.1	455	505	494
Enyel De Los Santos	999.0	0.1	7.6	404	506	493
Wes Benjamin	999.0	-0.1	7.4	504	507	492
Thomas Eshelman	999.0	0.0	7.3	476	508	491
Reiss Knehr	999.0	0.0	7.3	446	509	490
Jake Faria	999.0	0.0	6.9	469	510	489
Luke Farrell	999.0	0.0	6.8	489	511	488
Tayler Saucedo	999.0	0.1	6.7	399	512	487
Robert Dugger	999.0	0.0	6.2	478	513	486
Shaun Anderson	999.0	0.0	6.2	445	514	485
Adrian Sampson	999.0	-0.1	6.1	508	515	484
Drew Hutchison	999.0	-0.1	5.7	497	516	483
Kyle Crick	999.0	0.0	5.5	465	517	482
Shawn Armstrong	999.0	0.0	5.4	454	518	481
Edgar Santana	999.0	0.0	5.3	452	519	480
Kevin Ginkel	999.0	0.0	5.1	470	520	479
Sean Nolin	999.0	0.0	5.0	485	521	478
Carson Fulmer	999.0	0.0	4.8	482	522	477
Tyler Glasnow	598.8	0.2	4.6	314	523	75.8
Daniel Castano	999.0	0.0	4.0	486	524	475
John Gant	600.8	0.0	4.0	435	525	75.8
Kenta Maeda	600.7	0.1	3.3	384	526	74.7
Chris Ellis	600.8	0.1	3.0	406	527	73.8

NA
NA
NA
NA
NA
NA

---