20250227 - clean up

Author: isaactpetersen

basic-statistics.qmd

@@ -456,7 +456,7 @@ For instance, when evaluating whether Quarterbacks have <u>longer</u> careers th
 #| layout-ncol: 2
 #| fig-cap: "Interpretation of *p*-Values When Examining The Differences Between Groups. The vertical black lines reflect the group means."
 #| fig-alt: "Interpretation of *p*-Values When Examining The Differences Between Groups. The vertical black lines reflect the group means."
-#| fig-subcap: 
+#| fig-subcap:
 #|   - "What is the probability my data would look like this..."
 #|   - "...if in the population, the groups were really this?"
 #| code-fold: true
@@ -555,7 +555,7 @@ For instance, when evaluating whether number of carries is <u>positively</u> ass
 #| layout-ncol: 2
 #| fig-cap: "Interpretation of *p*-Values When Examining The Association Between Variables."
 #| fig-alt: "Interpretation of *p*-Values When Examining The Association Between Variables."
-#| fig-subcap: 
+#| fig-subcap:
 #|   - "What is the probability my data would look like this..."
 #|   - "...if in the population, the association was really this?"
 #| code-fold: true
@@ -711,7 +711,7 @@ z_crit <- qnorm(1-(0.05/2), m1, sd1)
 min1 <- m1-sd1*4
 max1 <- m1+sd1*4
 min2 <- m2-sd2*4
-max2 <- m2+sd2*4          
+max2 <- m2+sd2*4
 # create x sequence
 x <- seq(min(min1,min2), max(max1, max2), .01)
 # generate normal dist #1
@@ -726,52 +726,53 @@ df2 <- data.frame("x" = x, "y" = y2)
 # Alpha polygon
 y.poly <- pmin(y1,y2)
 poly1 <- data.frame(x=x, y=y.poly)
-poly1 <- poly1[poly1$x >= z_crit, ] 
+poly1 <- poly1[poly1$x >= z_crit, ]
 poly1<-rbind(poly1, c(z_crit, 0))  # add lower-left corner
  
 # Beta polygon
 poly2 <- df2
-poly2 <- poly2[poly2$x <= z_crit,] 
+poly2 <- poly2[poly2$x <= z_crit,]
 poly2<-rbind(poly2, c(z_crit, 0))  # add lower-left corner
  
 # power polygon; 1-beta
 poly3 <- df2
-poly3 <- poly3[poly3$x >= z_crit,] 
+poly3 <- poly3[poly3$x >= z_crit,]
 poly3 <-rbind(poly3, c(z_crit, 0))  # add lower-left corner
  
-# combine polygons. 
+# combine polygons
 poly1$id <- 3 # alpha, give it the highest number to make it the top layer
 poly2$id <- 2 # beta
 poly3$id <- 1 # power; 1 - beta
 poly <- rbind(poly1, poly2, poly3)
-poly$id <- factor(poly$id,  labels=c("power","beta","alpha"))
+poly$id <- factor(poly$id,  labels = c("power", "beta", "alpha"))
 
 # plot with ggplot2
 ggplot(poly, aes(x,y, fill=id, group=id)) +
   geom_polygon(show.legend=F, alpha=I(8/10)) +
   # add line for treatment group
-  geom_line(data=df1, aes(x,y, color="H0", group=NULL, fill=NULL), linewidth=1.5, show_guide=F) + 
+  geom_line(data=df1, aes(x, y, color = "H0", group = NULL, fill = NULL), linewidth = 1.5, show_guide = FALSE) + 
   # add line for treatment group. These lines could be combined into one dataframe.
-  geom_line(data=df2, aes(color="HA", group=NULL, fill=NULL),linewidth=1.5, show_guide=F) +
+  geom_line(data=df2, aes(color = "HA", group = NULL, fill = NULL), linewidth = 1.5, show_guide = FALSE) +
   # add vlines for z_crit
   geom_vline(xintercept = z_crit, linewidth=1, linetype="dashed") +
-  # change colors 
-  scale_color_manual("Group", 
-                     values= c("HA" = "#981e0b","H0" = "black")) +
-  scale_fill_manual("test", values= c("alpha" = "#0d6374","beta" = "#be805e","power"="#7cecee")) +
+  # change colors
+  scale_color_manual(
+    "Group",
+    values = c("HA" = "#981e0b", "H0" = "black")) +
+  scale_fill_manual("test", values= c("alpha" = "#0d6374", "beta" = "#be805e", "power"="#7cecee")) +
   # beta arrow
-  annotate("segment", x=0.1, y=0.045, xend=1.3, yend=0.01, arrow = arrow(length = unit(0.3, "cm")), linewidth=1) +
+  annotate("segment", x = 0.1, y = 0.045, xend = 1.3, yend = 0.01, arrow = arrow(length = unit(0.3, "cm")), linewidth = 1) +
   annotate("text", label="beta", x=0, y=0.05, parse=T, size=8) +
   # alpha arrow
-  annotate("segment", x=4, y=0.043, xend=3.4, yend=0.01, arrow = arrow(length = unit(0.3, "cm")), linewidth=1) +
+  annotate("segment", x = 4, y = 0.043, xend = 3.4, yend = 0.01, arrow = arrow(length = unit(0.3, "cm")), linewidth = 1) +
   annotate("text", label="frac(alpha,2)", x=4.2, y=0.05, parse=T, size=8) +
   # power arrow
-  annotate("segment", x=6, y=0.2, xend=4.5, yend=0.15, arrow = arrow(length = unit(0.3, "cm")), linewidth=1) +
-  annotate("text", label=expression(paste(1-beta, "  (\"power\")")), x=6.1, y=0.21, parse=T, size=8) +
+  annotate("segment", x = 6, y = 0.2, xend = 4.5, yend = 0.15, arrow = arrow(length = unit(0.3, "cm")), linewidth = 1) +
+  annotate("text", label = expression(paste(1-beta, "  (\"power\")")), x = 6.1, y = 0.21, parse = TRUE, size = 8) +
   # H_0 title
-  annotate("text", label="H[0]", x=m1, y=0.28, parse=T, size=8) +
+  annotate("text", label = "H[0]", x = m1, y = 0.28, parse = TRUE, size = 8) +
   # H_a title
-  annotate("text", label="H[1]", x=m2, y=0.28, parse=T, size=8) +
+  annotate("text", label = "H[1]", x = m2, y = 0.28, parse = TRUE, size = 8) +
   ggtitle("Statistical Power") +
   # remove some elements
   theme(

causal-inference.qmd

@@ -168,7 +168,7 @@ Predicting within-person change provides stronger evidence consistent with causa
 However, predicting within-person change does not, by itself, control for time-varying [confounds](#sec-causalDiagramConfounding).
 So, it can also be useful to control for time-varying [confounds](#sec-causalDiagramConfounding), such as by use of [control variables](#sec-causalInferenceControlVariables).
 
-#### Control Variables {#sec-causalInferenceControlVariables} 
+#### Control Variables {#sec-causalInferenceControlVariables}
 
 One of the plausible alternatives to the inference that `X` causes `Y` is that there are third variable [confounds](#sec-causalDiagramConfounding) that influence both `X` and `Y`, thus explaining why `X` and `Y` are associated, as depicted in Figures [-@fig-correlationAndCausation3] and [-@fig-ZCausesXandY].
 Thus, another approach that can help increase [internal validity](#sec-internalValidity) is to include plausible [confounds](#sec-causalDiagramConfounding) as control variables.
@@ -372,7 +372,7 @@ plot(dagitty::graphLayout(mediationDag))
 dagitty::impliedConditionalIndependencies(mediationDag)
 
 dagitty::adjustmentSets(
-  mediationDag, 
+  mediationDag,
   exposure = "M1",
   outcome = "Y",
   effect = "total")
@@ -419,7 +419,7 @@ An example of confounding is depicted in @fig-counfounding:
 confounding <- ggdag::confounder_triangle(
   x = "Player Endurance",
   y = "Field Goals Made",
-  z = "Stadium Altitude") 
+  z = "Stadium Altitude")
 
 confounding %>%
   ggdag(
@@ -437,7 +437,7 @@ The output indicates that player endurance (`X`) and field goals made (`Y`) are
 
 ```{r}
 dagitty::adjustmentSets(
-  confounding, 
+  confounding,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -483,7 +483,7 @@ In other words, in this example, player preparation is the mechanism that fully
 
 ```{r}
 dagitty::adjustmentSets(
-  full_mediation, 
+  full_mediation,
   exposure = "x",
   outcome = "y",
   effect = "direct")
@@ -493,7 +493,7 @@ The output indicates that, to obtain an unbiased estimate of the *direct* causal
 
 ```{r}
 dagitty::adjustmentSets(
-  full_mediation, 
+  full_mediation,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -536,7 +536,7 @@ For instance, coaching quality could also influence player fantasy points throug
 
 ```{r}
 dagitty::adjustmentSets(
-  partial_mediation, 
+  partial_mediation,
   exposure = "x",
   outcome = "y",
   effect = "direct")
@@ -546,7 +546,7 @@ As with [full mediation](#sec-causalDiagramMediationPartial), the output indicat
 
 ```{r}
 dagitty::adjustmentSets(
-  partial_mediation, 
+  partial_mediation,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -588,13 +588,13 @@ In this example, `X` and `M` are conditionally independent with `Z` when account
 
 ```{r}
 dagitty::adjustmentSets(
-  descendentDag, 
+  descendentDag,
   exposure = "X",
   outcome = "Y",
   effect = "direct")
 
 dagitty::adjustmentSets(
-  descendentDag, 
+  descendentDag,
   exposure = "X",
   outcome = "Y",
   effect = "total")
@@ -633,7 +633,7 @@ In this example collision, diet (`X`) and coaching strategy (`Y`) are independen
 
 ```{r}
 dagitty::adjustmentSets(
-  colliderBias1, 
+  colliderBias1,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -672,7 +672,7 @@ In this example of collider bias, there are no conditional independencies.
 
 ```{r}
 dagitty::adjustmentSets(
-  colliderBias2, 
+  colliderBias2,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -718,7 +718,7 @@ As the output indicates, there are several conditional independencies.
 
 ```{r}
 dagitty::adjustmentSets(
-  mBias, 
+  mBias,
   exposure = "x",
   outcome = "y",
   effect = "total")
@@ -782,7 +782,7 @@ As the output indicates, there are several conditional independencies.
 
 ```{r}
 dagitty::adjustmentSets(
-  butterflyBias, 
+  butterflyBias,
   exposure = "x",
   outcome = "y",
   effect = "total")

cognitive-bias.qmd

@@ -224,7 +224,7 @@ Indeed, it is estimated that nearly half (~45%) of the variability in fantasy fo
 A manager who won their league in the prior season may believe they will perform better than they actually will (overestimation), will perform better than average (overplacement), and may hold excessive confidence regarding the accuracy of their predictions about which players will perform well or poorly (overprecision).
 These various types of overconfidence may lead them to draft high-risk players based on gut feeling, neglecting statistical analysis and expert consensus.
 
-People tend to focus on the role of skill and to neglect the role of luck when explaining the past and predicting the future, giving people an illusion of control [@Kahneman2011]. 
+People tend to focus on the role of skill and to neglect the role of luck when explaining the past and predicting the future, giving people an illusion of control [@Kahneman2011].
 Players' performance in fantasy football, and human behavior more generally, is complex and multiply determined (i.e., is influenced by many factors).
 Despite the bluster of so-called experts who pretend to know more than they can know, no one can consistently and accurately predict how all players will perform.
 Remain humble in your predictions; do not be more confident than is warranted.
@@ -344,7 +344,7 @@ Loss aversion can also influence trade negotiations.
 Risk aversion leads people to select safer options but may lead them to miss out on higher-gain opportunities.
 For instance, risk aversion may lead a fantasy manager to start players who are more steady (i.e., show greater game-to-game [consistency](#sec-evalHistoricalConsistency)) over players who are more volatile (i.e., show greater game-to-game variability) but have higher upside potential.
 
-In mixed gambles, in which it is possible for a person to experience either a gain or a loss, [loss aversion](#sec-cognitiveBiasesLossAversion) tends to lead to risk-averse choices [@Kahneman2011]. 
+In mixed gambles, in which it is possible for a person to experience either a gain or a loss, [loss aversion](#sec-cognitiveBiasesLossAversion) tends to lead to risk-averse choices [@Kahneman2011].
 By contrast, when all of a person's options are poor, people tend to engage in risk seeking, as has been observed in entrepreneurs and in generals [@Kahneman2011].
 In fantasy football, risk seeking may be more likely when a manager has a team full of underperforming players and a weak record.
 

data-visualization.qmd

@@ -114,7 +114,7 @@ confidenceLevel <- .95 # for 95% confidence interval
 player_stats_seasonal_offense_summary <- player_stats_seasonal %>%
   filter(position_group %in% c("QB","RB","WR","TE")) %>% 
   group_by(position_group) %>%
-  summarise( 
+  summarise(
     n = sum(!is.na(fantasyPoints)),
     mean = mean(fantasyPoints, na.rm = TRUE),
     sd = sd(fantasyPoints, na.rm = TRUE)
@@ -630,7 +630,7 @@ ggplot2::ggplot(
     ylim = c(0,NA),
     expand = FALSE) +
   scale_x_continuous(
-    breaks = seq(from = 20, to = 40, by = 5) 
+    breaks = seq(from = 20, to = 40, by = 5)
   ) +
   scale_y_continuous(
     breaks = seq(from = 0, to = 2500, by = 250)
@@ -664,7 +664,7 @@ ggplot2::ggplot(
     ylim = c(0,NA),
     expand = FALSE) +
   scale_x_continuous(
-    breaks = seq(from = 20, to = 40, by = 5) 
+    breaks = seq(from = 20, to = 40, by = 5)
   ) +
   scale_y_continuous(
     breaks = seq(from = 0, to = 2500, by = 250)
@@ -698,7 +698,7 @@ ggplot2::ggplot(
     ylim = c(0,NA),
     expand = FALSE) +
   scale_x_continuous(
-    breaks = seq(from = 20, to = 40, by = 5) 
+    breaks = seq(from = 20, to = 40, by = 5)
   ) +
   scale_y_continuous(
     breaks = seq(from = 0, to = 2500, by = 250)
@@ -738,7 +738,7 @@ plot_ypcByPlayerAge <- ggplot2::ggplot(
     ylim = c(0,NA),
     expand = FALSE) +
   scale_x_continuous(
-    breaks = seq(from = 20, to = 40, by = 5) 
+    breaks = seq(from = 20, to = 40, by = 5)
   ) +
   scale_y_continuous(
     breaks = seq(from = 0, to = 2500, by = 250)
@@ -778,7 +778,7 @@ plot_ypcByPlayerAge <- ggplot2::ggplot(
     ylim = c(0,NA),
     expand = FALSE) +
   scale_x_continuous(
-    breaks = seq(from = 20, to = 40, by = 5) 
+    breaks = seq(from = 20, to = 40, by = 5)
   ) +
   scale_y_continuous(
     breaks = seq(from = 0, to = 2500, by = 250)
@@ -1231,7 +1231,7 @@ ggplot2::ggplot(
     y = def_epa)) +
   nflplotR::geom_mean_lines(
     aes(
-      x0 = off_epa ,
+      x0 = off_epa,
       y0 = def_epa)) +
   nflplotR::geom_nfl_logos(
     aes(

download-football-data.qmd

@@ -2065,8 +2065,8 @@ nfl_advancedStatsPFR_seasonal <- nfl_advancedStatsPFR_seasonal %>%
     by = c("pfr_id","season")
   ) %>% 
   select(
-    pfr_id, season, pfr_player_name, pos, age, team, g, gs, 
-    contains(".pass"), contains(".rush"), contains(".rec"), contains(".def"), 
+    pfr_id, season, pfr_player_name, pos, age, team, g, gs,
+    contains(".pass"), contains(".rush"), contains(".rec"), contains(".def"),
     everything())
 ```
 
@@ -3601,7 +3601,7 @@ Note: the following code takes a while to run.
 #nfl_actualFantasyPoints_weekly_raw <- ffanalytics:::actual_points_scoring(
 #  season = 2023,
 #  summary_level = c("week"),
-#  stat_type = c("player", "dst", "team"), 
+#  stat_type = c("player", "dst", "team"),
 #  season_type = c("REG", "POST", "REG+POST"),
 #  scoring_rules = scoring_obj,
 #  vor_baseline = NULL,
@@ -3619,7 +3619,7 @@ for(i in 1:length(seasons)){
     ffanalytics:::actual_points_scoring(
       season = seasons[i],
       summary_level = c("week"),
-      stat_type = c("player"), 
+      stat_type = c("player"),
       #season_type = c("REG"),
       scoring_rules = scoring_obj,
       vor_baseline = NULL,
@@ -3629,7 +3629,7 @@ for(i in 1:length(seasons)){
     ffanalytics:::actual_points_scoring(
       season = seasons[i],
       summary_level = c("week"),
-      stat_type = c("dst"), 
+      stat_type = c("dst"),
       #season_type = c("REG"),
       scoring_rules = scoring_obj,
       vor_baseline = NULL,
@@ -3767,7 +3767,7 @@ Note: the following code takes a while to run.
 #nfl_actualFantasyPoints_seasonal_raw <- ffanalytics:::actual_points_scoring(
 #  season = 2023,
 #  summary_level = c("season"),
-#  stat_type = c("player", "dst", "team"), 
+#  stat_type = c("player", "dst", "team"),
 #  season_type = c("REG"),
 #  scoring_rules = scoring_obj,
 #  vor_baseline = NULL,
@@ -3785,7 +3785,7 @@ for(i in 1:length(seasons)){
     ffanalytics:::actual_points_scoring(
       season = seasons[i],
       summary_level = c("season"),
-      stat_type = c("player"), 
+      stat_type = c("player"),
       season_type = c("REG"),
       scoring_rules = scoring_obj,
       vor_baseline = NULL,
@@ -3795,7 +3795,7 @@ for(i in 1:length(seasons)){
     ffanalytics:::actual_points_scoring(
       season = seasons[i],
       summary_level = c("season"),
-      stat_type = c("dst"), 
+      stat_type = c("dst"),
       season_type = c("REG"),
       scoring_rules = scoring_obj,
       vor_baseline = NULL,

draft.qmd

@@ -81,7 +81,7 @@ In general, Kickers and Defenses tend to have the lowest dropoff (i.e., the lowe
 Defenses, in particular, appear to be among the least predictable of the positions [@Lee2022].
 
 Another important concept is a player's [value over a typical replacement player](#sec-fantasyValueVORP) at that position (shortened to "value over replacement player"; VORP), which is described in @sec-fantasyValueVORP.
-A player's [value over a typical replacement player](#sec-fantasyValueVORP) provides a way to more fairly compare (and thus rank) players across different positions. 
+A player's [value over a typical replacement player](#sec-fantasyValueVORP) provides a way to more fairly compare (and thus rank) players across different positions.
 
 Another important concept is a player's [uncertainty](#sec-fantasyValueUncertainty), which is described in @sec-fantasyValueUncertainty.