2018 March Madness Prediction Results

Getting Started and Helper Functions

# Load necessary libraries
library(dplyr)
library(data.table)
library(knitr)

Input Model Predicted Spreads

# To input model predicted spreads
predict_ncaa_round <- function(round){
    # Subset data by round
    if(round == 0){rows <- c(1:4)}
    if(round == 1){rows <- c(5:36)}    
    if(round == 2){rows <- c(37:52)}
    if(round == 3){rows <- c(53:60)}
    if(round == 4){rows <- c(61:64)}    
    if(round == 5){rows <- c(65:66)}
    if(round == 6){rows <- 67}
    
    # Append model predicted spread for higher seeds
    bracket <- bracket[rows,] %>% 
        left_join(select(PredictionData, TeamName.x, TeamName.y, Prediction, Win_Prob), by = c('TeamName.x' = 'TeamName.x', 'TeamName.y'= 'TeamName.y'))  
    
    # Get the rows of lesser seeds
    na_rows <- which(is.na(bracket$Prediction))

    # Impute prediction data from lesser seeds
    for (i in na_rows) {
        bracket[i, 7] <- -PredictionData[which(bracket[i, 5] == PredictionData$TeamName.y & bracket[i,6] == PredictionData$TeamName.x), 9]
    
        bracket[i, 8] <- 1 - PredictionData[which(bracket[i, 5] == PredictionData$TeamName.y & bracket[i,6] == PredictionData$TeamName.x), 10]
    }
    
    bracket
}

Determining Model Choice

# To determine who model would bet on
ATSchoice <- function(bracket){
    bracket$ModelChoice <- NA
    for (i in c(1:length(bracket$Diff))){
        if(bracket$Diff[i] < -2) 
            {bracket$ModelChoice[i] <- bracket$TeamName.y[i]}
        if(bracket$Diff[i] > 2) 
            {bracket$ModelChoice[i] <- bracket$TeamName.x[i]}
        if(bracket$Diff[i] <= 2 & bracket$Diff[i] >= -2) 
            {bracket$ModelChoice[i] <- "No choice"}
    }
    bracket
}

Assessing Model Accuracy

ATSresults <- function(bracket){
    # Determine ATS winner
    bracket$ATSWinner <- NA
    for (i in c(1:length(bracket$Diff))){
        FinalDiff <- bracket$Team.x.score[i] - bracket$Team.y.score[i]
        if(!is.na(FinalDiff)) {
            if(FinalDiff < bracket$Spread[i]) 
                {bracket$ATSWinner[i] <- bracket$TeamName.y[i]}
            if(FinalDiff > bracket$Spread[i]) 
                {bracket$ATSWinner[i] <- bracket$TeamName.x[i]}
            if(FinalDiff == bracket$Spread[i]) 
                {bracket$ATSWinner[i] <- "Push"}
        }
    }
    # Determine if Model Prediction was Correct, Incorrect, or Not Made
    bracket$Accuracy <- NA
    for (i in c(1:length(bracket$Diff))){
        if(!is.na(bracket$ATSWinner[i])) {
            if(bracket$ModelChoice[i] == bracket$ATSWinner[i])      
                {bracket$Accuracy[i] <- "Correct"}
            if(bracket$ModelChoice[i] != bracket$ATSWinner[i]) 
                {bracket$Accuracy[i] <- "Incorrect"}
            if(bracket$ATSWinner[i] == "Push") 
                {bracket$Accuracy[i] <- "Push"}
        }
        if(is.na(bracket$ATSWinner[i])) 
            {bracket$Accuracy[i] <- NA}
        if(bracket$ModelChoice[i] == "No choice")
            {bracket$Accuracy[i] <- NA}
    }
    bracket
}

Putting winners into next round

# Function to advance winners from "x" final bracket into "bracket" object
advance_winners <- function(x){
    for (i in c(1:length(x$Slot))){
        slot <- x$Slot[i]
        if(x$Team.x.score[i] > x$Team.y.score[i]) {
            winner <- x$TeamName.x[i]
        } else{
            winner <- x$TeamName.y[i]
        }
    
        rownum <- which(bracket[,c(3:4)] == slot, arr.ind = TRUE)[1]
        colnum <- which(bracket[,c(3:4)] == slot, arr.ind = TRUE)[2]

        if(colnum == 1) 
            {bracket$TeamName.x[rownum] <- winner}
        else 
            {bracket$TeamName.y[rownum] <- winner}
    }   
    bracket
}

Creating Initial Bracket

The prediction model is a XGBoost model with 5-fold cross validation. The cleaning and preparation of the data and the training and testing of that model are in a script which can be found here [insert link here]. The prediction data for all possible pairs of opponents are in the file “saferesults.Rds”.

# Load prediction data from model
PredictionData <- readRDS("saferesults.Rds")
#kable(head(PredictionData))

# Load the 2018 bracket data and team names
tourneyslots <- fread("NCAATourneySlots.csv") %>% filter(Season == 2018)
tourneySeeds <- fread('NCAATourneySeeds.csv') %>% filter(Season == 2018)
teams <- fread('teams.csv')

# Attach team names to seed info
tourneySeeds <- tourneySeeds %>% 
    left_join(teams, by = 'TeamID') %>% 
    select(Season, Seed, TeamID, TeamName)

# Manually set the play-in winners
tourneyslots[5,4] <- "W16b"
tourneyslots[10,4] <- "W11a"
tourneyslots[18, 4] <- "X11b"
tourneyslots[29, 4] <- "Z16b"

# Complete Bracket
bracket <- tourneyslots %>% 
    left_join(select(tourneySeeds, -c(Season, TeamID)), by=c('StrongSeed'='Seed')) %>% 
    left_join(select(tourneySeeds, -c(Season, TeamID)), by=c('WeakSeed'='Seed'))

Play-in Round

Model Predictions

# Append model predicted spreads
playin <- predict_ncaa_round(round = 0)

# All Play-in Spreads are retroactive
# Spreads are positive if strong seed is favored, negative if underdogs
playinspreads <- data.frame(
    matrix(
        c(
"St Bonaventure", -3.5,
"Long Island", -4.5,
"Arizona St", 1,
"NC Central", -4.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(playinspreads) <- c("Team", "Spread")
playinspreads$Spread <- as.numeric(playinspreads$Spread)

# Append spreads on to bracket
playin <- playin %>% 
    left_join(playinspreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

A team needs to have a predicted spread more than 2 points beyond the sportsbook spread in order to have greater than a 55% chance of winning ATS (against the spread), which is the minimum winning percent needed to break even. In the ModelChoice variable, the model will return “No Choice” if the model prediction is within 2 points of the sportsbook spread or otherwise return which team the model predicts has a > 55% chance to cover the spread.

# Determine who model would bet on ATS
playin <- ATSchoice(playin)

Note: The Prediction variable outputs how many points the StrongSeed is expected to beat the WeakSeed by. So Prediction will be 3 if the StrongSeed team is favored by 3 points. This is the opposite of conventional wagering notation (-3). The Spread variable is not the conventional spread, but gives how many points the StrongSeed team is favored over the WeakSeed team by. The Win_Prob variable is the probability of the StrongSeed team winning the game straight up.

#### Output of Model Predictions ATS
playinpredictionstable <- playin[1:4,] %>% select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Diff)
kable(playinpredictionstable)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Diff
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	6.1
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	5.4
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	0.3
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	13.8

Game Results and Model Performance

# Input Game Results Manually
# Eventually I need to scrape this data from web
playinresults <- data.frame(
    matrix(
        c(
"St Bonaventure", 65, 58, 
"Long Island", 61, 71,
"Arizona St", 56, 60,
"NC Central", 46, 64), ncol = 3, byrow = TRUE), stringsAsFactors = FALSE)

names(playinresults) <- c("Team", "Team.x.score", "Team.y.score")
playinresults$Team.x.score <- as.numeric(playinresults$Team.x.score)
playinresults$Team.y.score <- as.numeric(playinresults$Team.y.score)

# Append Game Results to Bracket
playin <- playin %>% 
    left_join(playinresults, by = c('TeamName.x' = 'Team'))

# Determine Model Accuracy
playin <- ATSresults(playin)

# Table of all outcomes
output_table_0 <- playin[1:4,] %>% 
    filter(!is.na(TeamName.x)) %>% 
    select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Team.x.score, Team.y.score, ATSWinner, Accuracy)
kable(output_table_0)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Team.x.score	Team.y.score	ATSWinner	Accuracy
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	65	58	St Bonaventure	Correct
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	61	71	Radford	Incorrect
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	56	60	Syracuse	NA
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	46	64	TX Southern	Incorrect

# Summary of results
table(output_table_0$Accuracy)

## 
##   Correct Incorrect 
##         1         2

The model’s record was 1-2 ATS for the play-in games.

Round 1

Model Predictions

# Append model predicted spreads
round1 <- predict_ncaa_round(1)

# All Round 1 Spreads as of 12:30pm Thursday
# Spreads are positive if strong seed is favored, negative if underdogs
round1spreads <- data.frame(
    matrix(
        c(
"Rhode Island", 1.5,
"Tennessee", 11.5,
"Gonzaga", 13.5,
"Kansas", 14.5,
"Duke", 20,
"Miami FL", 2,
"Ohio St", 8,
"Seton Hall", 3,
"Villanova", 23,
"Kentucky", 5,
"Houston", 4,
"Texas Tech", 11,
"Virginia Tech", 2.5,
"Arizona", 8.5,
"Michigan", 10,
"Florida", 5.5,
"Texas A&M", 2.5,
"Purdue", 20.5,
"Wichita St", 11.5,
"Cincinnati", 14,
"North Carolina", 19.5,
"Arkansas", -1.5,
"West Virginia", 10.5,
"Nevada", 0,
"Creighton", 1,
"Michigan St", 14.5,
"Xavier", 19.5,
"Auburn", 9,
"Virginia", 21,
"TCU", 4,
"Missouri", -1.5,
"Clemson", 4.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(round1spreads) <- c("Team", "Spread")
round1spreads$Spread <- as.numeric(round1spreads$Spread)

# Append spreads on to bracket
round1 <- round1 %>% 
    left_join(round1spreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

A team needs to have a predicted spread more than 2 points beyond the sportsbook spread in order to have greater than a 55% chance of winning ATS (against the spread), which is the minimum winning percent needed to break even. In the ModelChoice variable, the model will return “No Choice” if the model prediction is within 2 points of the sportsbook spread or otherwise return which team the model predicts has a > 55% chance to cover the spread.

# Determine who model would bet on ATS
round1 <- ATSchoice(round1)

There is something wrong with my predictions for the top seeds, especially 1s and 2s, so I stayed away from following my model on any of those. Other kagglers have had this same issue and have manually input values for them, some trouble I didn’t want to bother with until I had the rest of the model running well. I’ll focus on that at some point in the future. For now, I’m going to manually input “No Choice” for all 1 and 2 seeds of Villanova, Kansas, Duke, Purdue, Cincinnati, Xavier, UNC, and Virginia.

# Manually set 1 and 2 seeds to no choice
round1$ModelChoice[c(1,2,9,10,17,18,25,26)] <- "No choice"

# Output of Model Predictions ATS
round1predictionstable <- round1 %>% select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Diff)

kable(round1predictionstable)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Diff
W01	Villanova	W16b	Radford	23.0	9.5	No choice	-13.5
W02	Purdue	W15	CS Fullerton	20.5	13.4	No choice	-7.1
W03	Texas Tech	W14	SF Austin	11.0	11.4	No choice	0.4
W04	Wichita St	W13	Marshall	11.5	4.3	Marshall	-7.2
W05	West Virginia	W12	Murray St	10.5	1.1	Murray St	-9.4
W06	Florida	W11a	St Bonaventure	5.5	8.0	Florida	2.5
W07	Arkansas	W10	Butler	-1.5	-1.4	No choice	0.1
W08	Virginia Tech	W09	Alabama	2.5	9.2	Virginia Tech	6.7
X01	Kansas	X16	Penn	14.5	6.8	No choice	-7.7
X02	Duke	X15	Iona	20.0	15.0	No choice	-5.0
X03	Michigan St	X14	Bucknell	14.5	7.8	Bucknell	-6.7
X04	Auburn	X13	Col Charleston	9.0	-0.1	Col Charleston	-9.1
X05	Clemson	X12	New Mexico St	4.5	-0.3	New Mexico St	-4.8
X06	TCU	X11b	Syracuse	4.0	1.0	Syracuse	-3.0
X07	Rhode Island	X10	Oklahoma	1.5	4.7	Rhode Island	3.2
X08	Seton Hall	X09	NC State	3.0	1.1	No choice	-1.9
Y01	Virginia	Y16	UMBC	21.0	3.5	No choice	-17.5
Y02	Cincinnati	Y15	Georgia St	14.0	7.1	No choice	-6.9
Y03	Tennessee	Y14	Wright St	11.5	5.4	Wright St	-6.1
Y04	Arizona	Y13	Buffalo	8.5	-7.1	Buffalo	-15.6
Y05	Kentucky	Y12	Davidson	5.0	1.7	Davidson	-3.3
Y06	Miami FL	Y11	Loyola-Chicago	2.0	6.7	Miami FL	4.7
Y07	Nevada	Y10	Texas	0.0	6.2	Nevada	6.2
Y08	Creighton	Y09	Kansas St	1.0	-1.1	Kansas St	-2.1
Z01	Xavier	Z16b	TX Southern	19.5	15.6	No choice	-3.9
Z02	North Carolina	Z15	Lipscomb	19.5	13.1	No choice	-6.4
Z03	Michigan	Z14	Montana	10.0	5.2	Montana	-4.8
Z04	Gonzaga	Z13	UNC Greensboro	13.5	7.8	UNC Greensboro	-5.7
Z05	Ohio St	Z12	S Dakota St	8.0	8.9	No choice	0.9
Z06	Houston	Z11	San Diego St	4.0	5.1	No choice	1.1
Z07	Texas A&M	Z10	Providence	2.5	2.3	No choice	-0.2
Z08	Missouri	Z09	Florida St	-1.5	-11.0	Florida St	-9.5

The model recommendations for round 1 are:

URI -2.5
Wright St +12
UNC Greensboro +12.5
Miami FL -2
Davidson +5
Va Tech -2.5
Buffalo +9
Florida -5.5
Montana +11
Marshall +11.5
Murray St +10.5
Bucknell +14.5 (I’m not putting my own money on this, but will consider it an outcome to evaluate the model on)
Col. Charleston +9
New Mexico St +4.5
Syracuse +4
Kansas St +1
Nevada EVEN
Florida St -1.5

Game Results and Model Performance

# Input Game Results Manually
# Eventually I need to scrape this data from web
round1results <- data.frame(
    matrix(
        c(
"St Bonaventure", 65, 58, 
"Long Island", 61, 71,
"Arizona St", 56, 60,
"NC Central", 46, 64,
"Rhode Island", 83, 78,
"Tennessee", 73, 47,
"Gonzaga", 68, 64,
"Kansas", 76, 60,
"Duke", 89, 67,
"Miami FL", 62, 64,
"Ohio St", 81, 73,
"Seton Hall", 94, 83,
"Villanova", 87, 61,
"Kentucky", 78, 73,
"Houston", 67, 65,
"Texas Tech", 70, 60,
"Virginia Tech", 83, 86,
"Arizona", 68, 89,
"Michigan", 61, 47,
"Florida", 77, 62,
"Texas A&M", 73, 69,
"Purdue", 74, 48,
"Wichita St", 75, 81,
"Cincinnati", 68, 53,
"North Carolina", 84, 66,
"Arkansas", 62, 79,
"West Virginia", 85, 68,
"Nevada", 87, 83,
"Creighton", 59, 69,
"Michigan St", 82, 78,
"Xavier", 91, 72,
"Auburn", 62, 58,
"Virginia", 54, 74,
"TCU", 52, 57,
"Missouri", 54, 67,
"Clemson", 79, 68), ncol = 3, byrow = TRUE), stringsAsFactors = FALSE)

names(round1results) <- c("Team", "Team.x.score", "Team.y.score")
round1results$Team.x.score <- as.numeric(round1results$Team.x.score)
round1results$Team.y.score <- as.numeric(round1results$Team.y.score)

# Append Game Results to Bracket
round1 <- round1 %>% 
    left_join(round1results, by = c('TeamName.x' = 'Team'))

# Determine Model's Accuracy ATS
round1 <- ATSresults(round1)

Note: The Prediction variable outputs how many points the StrongSeed is expected to beat the WeakSeed by. So Prediction will be 3 if the StrongSeed team is favored by 3 points. This is the opposite of conventional wagering notation (-3). The Spread variable is not the conventional spread, but gives how many points the StrongSeed team is favored over the WeakSeed team by. The Win_Prob variable is the probability of the StrongSeed team winning the game straight up.

# TABLE OF ALL RESULTS
output_table_1 <- round1 %>% 
    filter(!is.na(TeamName.x)) %>% 
    select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Team.x.score, Team.y.score, ATSWinner, Accuracy)

output_table_1 <- rbind(output_table_0, output_table_1)
kable(output_table_1)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Team.x.score	Team.y.score	ATSWinner	Accuracy
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	65	58	St Bonaventure	Correct
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	61	71	Radford	Incorrect
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	56	60	Syracuse	NA
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	46	64	TX Southern	Incorrect
W01	Villanova	W16b	Radford	23.0	9.5	No choice	87	61	Villanova	NA
W02	Purdue	W15	CS Fullerton	20.5	13.4	No choice	74	48	Purdue	NA
W03	Texas Tech	W14	SF Austin	11.0	11.4	No choice	70	60	SF Austin	NA
W04	Wichita St	W13	Marshall	11.5	4.3	Marshall	75	81	Marshall	Correct
W05	West Virginia	W12	Murray St	10.5	1.1	Murray St	85	68	West Virginia	Incorrect
W06	Florida	W11a	St Bonaventure	5.5	8.0	Florida	77	62	Florida	Correct
W07	Arkansas	W10	Butler	-1.5	-1.4	No choice	62	79	Butler	NA
W08	Virginia Tech	W09	Alabama	2.5	9.2	Virginia Tech	83	86	Alabama	Incorrect
X01	Kansas	X16	Penn	14.5	6.8	No choice	76	60	Kansas	NA
X02	Duke	X15	Iona	20.0	15.0	No choice	89	67	Duke	NA
X03	Michigan St	X14	Bucknell	14.5	7.8	Bucknell	82	78	Bucknell	Correct
X04	Auburn	X13	Col Charleston	9.0	-0.1	Col Charleston	62	58	Col Charleston	Correct
X05	Clemson	X12	New Mexico St	4.5	-0.3	New Mexico St	79	68	Clemson	Incorrect
X06	TCU	X11b	Syracuse	4.0	1.0	Syracuse	52	57	Syracuse	Correct
X07	Rhode Island	X10	Oklahoma	1.5	4.7	Rhode Island	83	78	Rhode Island	Correct
X08	Seton Hall	X09	NC State	3.0	1.1	No choice	94	83	Seton Hall	NA
Y01	Virginia	Y16	UMBC	21.0	3.5	No choice	54	74	UMBC	NA
Y02	Cincinnati	Y15	Georgia St	14.0	7.1	No choice	68	53	Cincinnati	NA
Y03	Tennessee	Y14	Wright St	11.5	5.4	Wright St	73	47	Tennessee	Incorrect
Y04	Arizona	Y13	Buffalo	8.5	-7.1	Buffalo	68	89	Buffalo	Correct
Y05	Kentucky	Y12	Davidson	5.0	1.7	Davidson	78	73	Push	Push
Y06	Miami FL	Y11	Loyola-Chicago	2.0	6.7	Miami FL	62	64	Loyola-Chicago	Incorrect
Y07	Nevada	Y10	Texas	0.0	6.2	Nevada	87	83	Nevada	Correct
Y08	Creighton	Y09	Kansas St	1.0	-1.1	Kansas St	59	69	Kansas St	Correct
Z01	Xavier	Z16b	TX Southern	19.5	15.6	No choice	91	72	TX Southern	NA
Z02	North Carolina	Z15	Lipscomb	19.5	13.1	No choice	84	66	Lipscomb	NA
Z03	Michigan	Z14	Montana	10.0	5.2	Montana	61	47	Michigan	Incorrect
Z04	Gonzaga	Z13	UNC Greensboro	13.5	7.8	UNC Greensboro	68	64	UNC Greensboro	Correct
Z05	Ohio St	Z12	S Dakota St	8.0	8.9	No choice	81	73	Push	NA
Z06	Houston	Z11	San Diego St	4.0	5.1	No choice	67	65	San Diego St	NA
Z07	Texas A&M	Z10	Providence	2.5	2.3	No choice	73	69	Texas A&M	NA
Z08	Missouri	Z09	Florida St	-1.5	-11.0	Florida St	54	67	Florida St	Correct

# Summary of results
table(output_table_1$Accuracy)

## 
##   Correct Incorrect      Push 
##        12         8         1

The model’s record was 11-6-1 ATS in round 1 action and is 12-8-1 ATS over the entire tournament. The model would have been 3-5 in the 1 and 2 seeded games.

One last interesting note is that a lot of 1 seeds seemed to not cover the first half spread. I heard this has some historical traction to it. Something to keep in mind next year.

Round 2

Model Predictions

# Recording Round 1 Winners into Round 2
bracket <- advance_winners(round1)

# Model Predictions for Round 2
round2 <- predict_ncaa_round(2)

# All Round 2 Spreads as of 1:00am Saturday
# Spreads are positive if strong seed is favored, negative if underdogs
round2spreads <- data.frame(
    matrix(
        c(
"Villanova", 11,
"Purdue", 3.5,
"Texas Tech", 0,
"Marshall", -11,
"Kansas", 4.5,
"Duke", 10,
"Michigan St", 9,
"Auburn", 1,
"UMBC", -12,
"Cincinnati", 7,
"Tennessee", 6,
"Buffalo", -7.5,
"Xavier", 4,
"North Carolina", 6,
"Michigan", 3,
"Gonzaga", 2.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(round2spreads) <- c("Team", "Spread")
round2spreads$Spread <- as.numeric(round2spreads$Spread)

# Append spreads on to bracket
round2 <- round2 %>% 
    left_join(round2spreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

Here are the model predictions for all round 2 games.

# Determine who model would bet on ATS
round2 <- ATSchoice(round2)

# Output of model recommendations
round2_output_table <- round2 %>% 
    select(StrongSeed, WeakSeed, TeamName.x, TeamName.y, Spread, Prediction, Diff, Win_Prob, ModelChoice)

kable(round2_output_table)

StrongSeed	WeakSeed	TeamName.x	TeamName.y	Spread	Prediction	Diff	Win_Prob	ModelChoice
R1W1	R1W8	Villanova	Alabama	11.0	8.3	-2.7	0.737	Alabama
R1W2	R1W7	Purdue	Butler	3.5	8.1	4.6	0.734	Purdue
R1W3	R1W6	Texas Tech	Florida	0.0	-0.7	-0.7	0.479	No choice
R1W4	R1W5	Marshall	West Virginia	-11.0	-4.6	6.4	0.368	Marshall
R1X1	R1X8	Kansas	Seton Hall	4.5	2.2	-2.3	0.563	Seton Hall
R1X2	R1X7	Duke	Rhode Island	10.0	3.8	-6.2	0.610	Rhode Island
R1X3	R1X6	Michigan St	Syracuse	9.0	5.2	-3.8	0.649	Syracuse
R1X4	R1X5	Auburn	Clemson	1.0	-0.6	-1.6	0.482	No choice
R1Y1	R1Y8	UMBC	Kansas St	-12.0	-3.5	8.5	0.399	UMBC
R1Y2	R1Y7	Cincinnati	Nevada	7.0	9.1	2.1	0.762	Cincinnati
R1Y3	R1Y6	Tennessee	Loyola-Chicago	6.0	5.8	-0.2	0.666	No choice
R1Y4	R1Y5	Buffalo	Kentucky	-7.5	2.2	9.7	0.565	Buffalo
R1Z1	R1Z8	Xavier	Florida St	4.0	-0.2	-4.2	0.492	Florida St
R1Z2	R1Z7	North Carolina	Texas A&M	6.0	3.3	-2.7	0.596	Texas A&M
R1Z3	R1Z6	Michigan	Houston	3.0	3.5	0.5	0.601	No choice
R1Z4	R1Z5	Gonzaga	Ohio St	2.5	10.6	8.1	0.803	Gonzaga

The model is still picking against all #1 seeds, but no longer against all #2 seeds. So I’m going to start evaluating model on all recommendations, while still hesitant to wager on #1 seed games. The recommendations are:

Alabama +11
Purdue -3.5
Marshall +11
Seton Hall +4.5
URI +10
Syracuse +9
UMBC +12
Cincinnati -7
Buffalo +7.5
Florida St +4
Texas A&M +6
Gonzaga -2.5

Game Results and Model Performance

# Manual input of round 2 results
# Need to automate this!
round2results <- data.frame(
    matrix(
        c(
"Villanova", 81, 58,
"Texas Tech", 69, 66,
"Duke", 87, 62,
"Tennessee", 62, 63,
"Buffalo", 75, 95,
"Michigan", 64, 63,
"Gonzaga", 90, 84,
"Kansas", 83, 79,
"Purdue", 76, 73,
"Michigan St", 53, 55,
"North Carolina", 65, 86,
"Cincinnati", 73, 75,
"Auburn", 53, 84,
"UMBC", 43, 50,
"Xavier", 70, 75,
"Marshall", 71, 94),  
    ncol = 3, byrow = TRUE), stringsAsFactors = FALSE
)
names(round2results) <- c("Team", "Team.x.score", "Team.y.score")
round2results$Team.x.score <- as.numeric(round2results$Team.x.score)
round2results$Team.y.score <- as.numeric(round2results$Team.y.score)

# Append Game Results to Bracket
round2 <- round2 %>% 
    left_join(round2results, by = c('TeamName.x' = 'Team'))

# Determine Model's Accuracy ATS
round2 <- ATSresults(round2)

# Table of results
output_table_2 <- round2 %>% 
    select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Team.x.score, Team.y.score, ATSWinner, Accuracy)

output_table_2 <- rbind(output_table_1, output_table_2)

kable(output_table_2)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Team.x.score	Team.y.score	ATSWinner	Accuracy
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	65	58	St Bonaventure	Correct
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	61	71	Radford	Incorrect
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	56	60	Syracuse	NA
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	46	64	TX Southern	Incorrect
W01	Villanova	W16b	Radford	23.0	9.5	No choice	87	61	Villanova	NA
W02	Purdue	W15	CS Fullerton	20.5	13.4	No choice	74	48	Purdue	NA
W03	Texas Tech	W14	SF Austin	11.0	11.4	No choice	70	60	SF Austin	NA
W04	Wichita St	W13	Marshall	11.5	4.3	Marshall	75	81	Marshall	Correct
W05	West Virginia	W12	Murray St	10.5	1.1	Murray St	85	68	West Virginia	Incorrect
W06	Florida	W11a	St Bonaventure	5.5	8.0	Florida	77	62	Florida	Correct
W07	Arkansas	W10	Butler	-1.5	-1.4	No choice	62	79	Butler	NA
W08	Virginia Tech	W09	Alabama	2.5	9.2	Virginia Tech	83	86	Alabama	Incorrect
X01	Kansas	X16	Penn	14.5	6.8	No choice	76	60	Kansas	NA
X02	Duke	X15	Iona	20.0	15.0	No choice	89	67	Duke	NA
X03	Michigan St	X14	Bucknell	14.5	7.8	Bucknell	82	78	Bucknell	Correct
X04	Auburn	X13	Col Charleston	9.0	-0.1	Col Charleston	62	58	Col Charleston	Correct
X05	Clemson	X12	New Mexico St	4.5	-0.3	New Mexico St	79	68	Clemson	Incorrect
X06	TCU	X11b	Syracuse	4.0	1.0	Syracuse	52	57	Syracuse	Correct
X07	Rhode Island	X10	Oklahoma	1.5	4.7	Rhode Island	83	78	Rhode Island	Correct
X08	Seton Hall	X09	NC State	3.0	1.1	No choice	94	83	Seton Hall	NA
Y01	Virginia	Y16	UMBC	21.0	3.5	No choice	54	74	UMBC	NA
Y02	Cincinnati	Y15	Georgia St	14.0	7.1	No choice	68	53	Cincinnati	NA
Y03	Tennessee	Y14	Wright St	11.5	5.4	Wright St	73	47	Tennessee	Incorrect
Y04	Arizona	Y13	Buffalo	8.5	-7.1	Buffalo	68	89	Buffalo	Correct
Y05	Kentucky	Y12	Davidson	5.0	1.7	Davidson	78	73	Push	Push
Y06	Miami FL	Y11	Loyola-Chicago	2.0	6.7	Miami FL	62	64	Loyola-Chicago	Incorrect
Y07	Nevada	Y10	Texas	0.0	6.2	Nevada	87	83	Nevada	Correct
Y08	Creighton	Y09	Kansas St	1.0	-1.1	Kansas St	59	69	Kansas St	Correct
Z01	Xavier	Z16b	TX Southern	19.5	15.6	No choice	91	72	TX Southern	NA
Z02	North Carolina	Z15	Lipscomb	19.5	13.1	No choice	84	66	Lipscomb	NA
Z03	Michigan	Z14	Montana	10.0	5.2	Montana	61	47	Michigan	Incorrect
Z04	Gonzaga	Z13	UNC Greensboro	13.5	7.8	UNC Greensboro	68	64	UNC Greensboro	Correct
Z05	Ohio St	Z12	S Dakota St	8.0	8.9	No choice	81	73	Push	NA
Z06	Houston	Z11	San Diego St	4.0	5.1	No choice	67	65	San Diego St	NA
Z07	Texas A&M	Z10	Providence	2.5	2.3	No choice	73	69	Texas A&M	NA
Z08	Missouri	Z09	Florida St	-1.5	-11.0	Florida St	54	67	Florida St	Correct
R1W1	Villanova	R1W8	Alabama	11.0	8.3	Alabama	81	58	Villanova	Incorrect
R1W2	Purdue	R1W7	Butler	3.5	8.1	Purdue	76	73	Butler	Incorrect
R1W3	Texas Tech	R1W6	Florida	0.0	-0.7	No choice	69	66	Texas Tech	NA
R1W4	Marshall	R1W5	West Virginia	-11.0	-4.6	Marshall	71	94	West Virginia	Incorrect
R1X1	Kansas	R1X8	Seton Hall	4.5	2.2	Seton Hall	83	79	Seton Hall	Correct
R1X2	Duke	R1X7	Rhode Island	10.0	3.8	Rhode Island	87	62	Duke	Incorrect
R1X3	Michigan St	R1X6	Syracuse	9.0	5.2	Syracuse	53	55	Syracuse	Correct
R1X4	Auburn	R1X5	Clemson	1.0	-0.6	No choice	53	84	Clemson	NA
R1Y1	UMBC	R1Y8	Kansas St	-12.0	-3.5	UMBC	43	50	UMBC	Correct
R1Y2	Cincinnati	R1Y7	Nevada	7.0	9.1	Cincinnati	73	75	Nevada	Incorrect
R1Y3	Tennessee	R1Y6	Loyola-Chicago	6.0	5.8	No choice	62	63	Loyola-Chicago	NA
R1Y4	Buffalo	R1Y5	Kentucky	-7.5	2.2	Buffalo	75	95	Kentucky	Incorrect
R1Z1	Xavier	R1Z8	Florida St	4.0	-0.2	Florida St	70	75	Florida St	Correct
R1Z2	North Carolina	R1Z7	Texas A&M	6.0	3.3	Texas A&M	65	86	Texas A&M	Correct
R1Z3	Michigan	R1Z6	Houston	3.0	3.5	No choice	64	63	Houston	NA
R1Z4	Gonzaga	R1Z5	Ohio St	2.5	10.6	Gonzaga	90	84	Gonzaga	Correct

# Summary of results
table(output_table_2$Accuracy)

## 
##   Correct Incorrect      Push 
##        18        14         1

In round 2, the model was 6-6 ATS. In total, the model is 18-14-1 through the first two rounds of the tournament for an overall success rate of 56.25%.

Sweet 16

Model Predictions

# Recording Round 2 Winners into Round 3
bracket <- advance_winners(round2)

# Model Predictions for Round 3
round3 <- predict_ncaa_round(3)

# All Round 3 Spreads as of 7:00pm Monday
# Spreads are positive if strong seed is favored, negative if underdogs
round3spreads <- data.frame(
    matrix(
        c(
"Villanova", 5.5,
"Purdue", 1.5,
"Kansas", 4.5,
"Duke", 11,
"Kansas St", -6,
"Nevada", 1.5,
"Florida St", -5.5,
"Texas A&M", -2.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(round3spreads) <- c("Team", "Spread")
round3spreads$Spread <- as.numeric(round3spreads$Spread)

# Append spreads on to bracket
round3 <- round3 %>% 
    left_join(round3spreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

Here are the model predictions for all round 3 games.

# Determine who model would bet on ATS
round3 <- ATSchoice(round3)

# Output of model recommendations
round3_predictions <- round3 %>% 
    select(StrongSeed, WeakSeed, TeamName.x, TeamName.y, Spread, Prediction, Diff, Win_Prob, ModelChoice)

kable(round3_predictions)

StrongSeed	WeakSeed	TeamName.x	TeamName.y	Spread	Prediction	Diff	Win_Prob	ModelChoice
R2W1	R2W4	Villanova	West Virginia	5.5	2.9	-2.6	0.583	West Virginia
R2W2	R2W3	Purdue	Texas Tech	1.5	0.6	-0.9	0.518	No choice
R2X1	R2X4	Kansas	Clemson	4.5	8.2	3.7	0.735	Kansas
R2X2	R2X3	Duke	Syracuse	11.0	4.3	-6.7	0.625	Syracuse
R2Y1	R2Y4	Kansas St	Kentucky	-6.0	-2.3	3.7	0.435	Kansas St
R2Y2	R2Y3	Nevada	Loyola-Chicago	1.5	-0.8	-2.3	0.477	Loyola-Chicago
R2Z1	R2Z4	Florida St	Gonzaga	-5.5	-5.9	-0.4	0.330	No choice
R2Z2	R2Z3	Texas A&M	Michigan	-2.5	-2.9	-0.4	0.416	No choice

The model recommendations for round 3 are not as strong as in previous rounds. The recommendations in order of strength are:

Syracuse +11
Kansas -4.5
Kansas St +6

Fringe picks:

West Virginia +5.5
Loyola-Chicago +1.5

Game Results and Model Performance

To be updated after games are played.

# Manual input of round 3 results
# Need to automate this!
round3results <- data.frame(
    matrix(
        c(
        "Villanova", 90, 78,
        "Purdue", 65, 78,
        "Kansas", 80, 76,
        "Duke", 69, 65,
        "Kansas St", 61, 58,
        "Nevada", 68, 69,
        "Florida St", 75, 60,
        "Texas A&M", 72, 99),  
    ncol = 3, byrow = TRUE), stringsAsFactors = FALSE
)
names(round3results) <- c("Team", "Team.x.score", "Team.y.score")
round3results$Team.x.score <- as.numeric(round3results$Team.x.score)
round3results$Team.y.score <- as.numeric(round3results$Team.y.score)

# Append Game Results to Bracket
round3 <- round3 %>% 
    left_join(round3results, by = c('TeamName.x' = 'Team'))

# Determine Model's Accuracy ATS
round3 <- ATSresults(round3)

# Table of results
round3_results_table <- round3 %>% 
    select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Team.x.score, Team.y.score, ATSWinner, Accuracy)

round3_results_table <- rbind(output_table_2, round3_results_table)

kable(round3_results_table)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Team.x.score	Team.y.score	ATSWinner	Accuracy
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	65	58	St Bonaventure	Correct
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	61	71	Radford	Incorrect
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	56	60	Syracuse	NA
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	46	64	TX Southern	Incorrect
W01	Villanova	W16b	Radford	23.0	9.5	No choice	87	61	Villanova	NA
W02	Purdue	W15	CS Fullerton	20.5	13.4	No choice	74	48	Purdue	NA
W03	Texas Tech	W14	SF Austin	11.0	11.4	No choice	70	60	SF Austin	NA
W04	Wichita St	W13	Marshall	11.5	4.3	Marshall	75	81	Marshall	Correct
W05	West Virginia	W12	Murray St	10.5	1.1	Murray St	85	68	West Virginia	Incorrect
W06	Florida	W11a	St Bonaventure	5.5	8.0	Florida	77	62	Florida	Correct
W07	Arkansas	W10	Butler	-1.5	-1.4	No choice	62	79	Butler	NA
W08	Virginia Tech	W09	Alabama	2.5	9.2	Virginia Tech	83	86	Alabama	Incorrect
X01	Kansas	X16	Penn	14.5	6.8	No choice	76	60	Kansas	NA
X02	Duke	X15	Iona	20.0	15.0	No choice	89	67	Duke	NA
X03	Michigan St	X14	Bucknell	14.5	7.8	Bucknell	82	78	Bucknell	Correct
X04	Auburn	X13	Col Charleston	9.0	-0.1	Col Charleston	62	58	Col Charleston	Correct
X05	Clemson	X12	New Mexico St	4.5	-0.3	New Mexico St	79	68	Clemson	Incorrect
X06	TCU	X11b	Syracuse	4.0	1.0	Syracuse	52	57	Syracuse	Correct
X07	Rhode Island	X10	Oklahoma	1.5	4.7	Rhode Island	83	78	Rhode Island	Correct
X08	Seton Hall	X09	NC State	3.0	1.1	No choice	94	83	Seton Hall	NA
Y01	Virginia	Y16	UMBC	21.0	3.5	No choice	54	74	UMBC	NA
Y02	Cincinnati	Y15	Georgia St	14.0	7.1	No choice	68	53	Cincinnati	NA
Y03	Tennessee	Y14	Wright St	11.5	5.4	Wright St	73	47	Tennessee	Incorrect
Y04	Arizona	Y13	Buffalo	8.5	-7.1	Buffalo	68	89	Buffalo	Correct
Y05	Kentucky	Y12	Davidson	5.0	1.7	Davidson	78	73	Push	Push
Y06	Miami FL	Y11	Loyola-Chicago	2.0	6.7	Miami FL	62	64	Loyola-Chicago	Incorrect
Y07	Nevada	Y10	Texas	0.0	6.2	Nevada	87	83	Nevada	Correct
Y08	Creighton	Y09	Kansas St	1.0	-1.1	Kansas St	59	69	Kansas St	Correct
Z01	Xavier	Z16b	TX Southern	19.5	15.6	No choice	91	72	TX Southern	NA
Z02	North Carolina	Z15	Lipscomb	19.5	13.1	No choice	84	66	Lipscomb	NA
Z03	Michigan	Z14	Montana	10.0	5.2	Montana	61	47	Michigan	Incorrect
Z04	Gonzaga	Z13	UNC Greensboro	13.5	7.8	UNC Greensboro	68	64	UNC Greensboro	Correct
Z05	Ohio St	Z12	S Dakota St	8.0	8.9	No choice	81	73	Push	NA
Z06	Houston	Z11	San Diego St	4.0	5.1	No choice	67	65	San Diego St	NA
Z07	Texas A&M	Z10	Providence	2.5	2.3	No choice	73	69	Texas A&M	NA
Z08	Missouri	Z09	Florida St	-1.5	-11.0	Florida St	54	67	Florida St	Correct
R1W1	Villanova	R1W8	Alabama	11.0	8.3	Alabama	81	58	Villanova	Incorrect
R1W2	Purdue	R1W7	Butler	3.5	8.1	Purdue	76	73	Butler	Incorrect
R1W3	Texas Tech	R1W6	Florida	0.0	-0.7	No choice	69	66	Texas Tech	NA
R1W4	Marshall	R1W5	West Virginia	-11.0	-4.6	Marshall	71	94	West Virginia	Incorrect
R1X1	Kansas	R1X8	Seton Hall	4.5	2.2	Seton Hall	83	79	Seton Hall	Correct
R1X2	Duke	R1X7	Rhode Island	10.0	3.8	Rhode Island	87	62	Duke	Incorrect
R1X3	Michigan St	R1X6	Syracuse	9.0	5.2	Syracuse	53	55	Syracuse	Correct
R1X4	Auburn	R1X5	Clemson	1.0	-0.6	No choice	53	84	Clemson	NA
R1Y1	UMBC	R1Y8	Kansas St	-12.0	-3.5	UMBC	43	50	UMBC	Correct
R1Y2	Cincinnati	R1Y7	Nevada	7.0	9.1	Cincinnati	73	75	Nevada	Incorrect
R1Y3	Tennessee	R1Y6	Loyola-Chicago	6.0	5.8	No choice	62	63	Loyola-Chicago	NA
R1Y4	Buffalo	R1Y5	Kentucky	-7.5	2.2	Buffalo	75	95	Kentucky	Incorrect
R1Z1	Xavier	R1Z8	Florida St	4.0	-0.2	Florida St	70	75	Florida St	Correct
R1Z2	North Carolina	R1Z7	Texas A&M	6.0	3.3	Texas A&M	65	86	Texas A&M	Correct
R1Z3	Michigan	R1Z6	Houston	3.0	3.5	No choice	64	63	Houston	NA
R1Z4	Gonzaga	R1Z5	Ohio St	2.5	10.6	Gonzaga	90	84	Gonzaga	Correct
R2W1	Villanova	R2W4	West Virginia	5.5	2.9	West Virginia	90	78	Villanova	Incorrect
R2W2	Purdue	R2W3	Texas Tech	1.5	0.6	No choice	65	78	Texas Tech	NA
R2X1	Kansas	R2X4	Clemson	4.5	8.2	Kansas	80	76	Clemson	Incorrect
R2X2	Duke	R2X3	Syracuse	11.0	4.3	Syracuse	69	65	Syracuse	Correct
R2Y1	Kansas St	R2Y4	Kentucky	-6.0	-2.3	Kansas St	61	58	Kansas St	Correct
R2Y2	Nevada	R2Y3	Loyola-Chicago	1.5	-0.8	Loyola-Chicago	68	69	Loyola-Chicago	Correct
R2Z1	Florida St	R2Z4	Gonzaga	-5.5	-5.9	No choice	75	60	Florida St	NA
R2Z2	Texas A&M	R2Z3	Michigan	-2.5	-2.9	No choice	72	99	Michigan	NA

# Summary of results
table(round3_results_table$Accuracy)

## 
##   Correct Incorrect      Push 
##        21        16         1

The model went 3-2 ATS in the Sweet 16 and is now 21-16-1 throughout the tourney.

Elite 8

Model Predictions

# Recording Round 3 Winners into Round 4
bracket <- advance_winners(round3)

# Model Predictions for Round 4
round4 <- predict_ncaa_round(4)

# All Round 4 Spreads as of 7:00pm Monday
# Spreads are positive if strong seed is favored, negative if underdogs
round4spreads <- data.frame(
    matrix(
        c(
"Villanova", 7,
"Kansas", -3.5,
"Kansas St", 1.5,
"Florida St", -4.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(round4spreads) <- c("Team", "Spread")
round4spreads$Spread <- as.numeric(round4spreads$Spread)

# Append spreads on to bracket
round4 <- round4 %>% 
    left_join(round4spreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

Here are the model predictions for all round 4 games.

# Determine who model would bet on ATS
round4 <- ATSchoice(round4)

# Output of model recommendations
round4_predictions <- round4 %>% 
    select(StrongSeed, WeakSeed, TeamName.x, TeamName.y, Spread, Prediction, Diff, Win_Prob, ModelChoice)

kable(round4_predictions)

StrongSeed	WeakSeed	TeamName.x	TeamName.y	Spread	Prediction	Diff	Win_Prob	ModelChoice
R3W1	R3W2	Villanova	Texas Tech	7.0	-5.1	-12.1	0.353	Texas Tech
R3X1	R3X2	Kansas	Duke	-3.5	-3.4	0.1	0.402	No choice
R3Y1	R3Y2	Kansas St	Loyola-Chicago	1.5	-1.9	-3.4	0.445	Loyola-Chicago
R3Z1	R3Z2	Florida St	Michigan	-4.5	-1.7	2.8	0.453	Florida St

The model recommendations for round 4 are …

Game Results and Model Performance

To be updated after games are played.

# Manual input of round 4 results
# Need to automate this!
round4results <- data.frame(
    matrix(
        c(
        "Villanova", 71, 59,
        "Kansas", 85, 81,
        "Kansas St", 62, 78,
        "Florida St", 54, 58),  
    ncol = 3, byrow = TRUE), stringsAsFactors = FALSE
)
names(round4results) <- c("Team", "Team.x.score", "Team.y.score")
round4results$Team.x.score <- as.numeric(round4results$Team.x.score)
round4results$Team.y.score <- as.numeric(round4results$Team.y.score)

# Append Game Results to Bracket
round4 <- round4 %>% 
    left_join(round4results, by = c('TeamName.x' = 'Team'))

# Determine Model's Accuracy ATS
round4 <- ATSresults(round4)

# Table of results
round4_results_table <- round4 %>% 
    select(StrongSeed, TeamName.x, WeakSeed, TeamName.y, Spread, Prediction, ModelChoice, Team.x.score, Team.y.score, ATSWinner, Accuracy)

round4_results_table <- rbind(round3_results_table, round4_results_table)

kable(round4_results_table)

StrongSeed	TeamName.x	WeakSeed	TeamName.y	Spread	Prediction	ModelChoice	Team.x.score	Team.y.score	ATSWinner	Accuracy
W11a	St Bonaventure	W11b	UCLA	-3.5	2.6	St Bonaventure	65	58	St Bonaventure	Correct
W16a	Long Island	W16b	Radford	-4.5	0.9	Long Island	61	71	Radford	Incorrect
X11a	Arizona St	X11b	Syracuse	1.0	1.3	No choice	56	60	Syracuse	NA
Z16a	NC Central	Z16b	TX Southern	-4.5	9.3	NC Central	46	64	TX Southern	Incorrect
W01	Villanova	W16b	Radford	23.0	9.5	No choice	87	61	Villanova	NA
W02	Purdue	W15	CS Fullerton	20.5	13.4	No choice	74	48	Purdue	NA
W03	Texas Tech	W14	SF Austin	11.0	11.4	No choice	70	60	SF Austin	NA
W04	Wichita St	W13	Marshall	11.5	4.3	Marshall	75	81	Marshall	Correct
W05	West Virginia	W12	Murray St	10.5	1.1	Murray St	85	68	West Virginia	Incorrect
W06	Florida	W11a	St Bonaventure	5.5	8.0	Florida	77	62	Florida	Correct
W07	Arkansas	W10	Butler	-1.5	-1.4	No choice	62	79	Butler	NA
W08	Virginia Tech	W09	Alabama	2.5	9.2	Virginia Tech	83	86	Alabama	Incorrect
X01	Kansas	X16	Penn	14.5	6.8	No choice	76	60	Kansas	NA
X02	Duke	X15	Iona	20.0	15.0	No choice	89	67	Duke	NA
X03	Michigan St	X14	Bucknell	14.5	7.8	Bucknell	82	78	Bucknell	Correct
X04	Auburn	X13	Col Charleston	9.0	-0.1	Col Charleston	62	58	Col Charleston	Correct
X05	Clemson	X12	New Mexico St	4.5	-0.3	New Mexico St	79	68	Clemson	Incorrect
X06	TCU	X11b	Syracuse	4.0	1.0	Syracuse	52	57	Syracuse	Correct
X07	Rhode Island	X10	Oklahoma	1.5	4.7	Rhode Island	83	78	Rhode Island	Correct
X08	Seton Hall	X09	NC State	3.0	1.1	No choice	94	83	Seton Hall	NA
Y01	Virginia	Y16	UMBC	21.0	3.5	No choice	54	74	UMBC	NA
Y02	Cincinnati	Y15	Georgia St	14.0	7.1	No choice	68	53	Cincinnati	NA
Y03	Tennessee	Y14	Wright St	11.5	5.4	Wright St	73	47	Tennessee	Incorrect
Y04	Arizona	Y13	Buffalo	8.5	-7.1	Buffalo	68	89	Buffalo	Correct
Y05	Kentucky	Y12	Davidson	5.0	1.7	Davidson	78	73	Push	Push
Y06	Miami FL	Y11	Loyola-Chicago	2.0	6.7	Miami FL	62	64	Loyola-Chicago	Incorrect
Y07	Nevada	Y10	Texas	0.0	6.2	Nevada	87	83	Nevada	Correct
Y08	Creighton	Y09	Kansas St	1.0	-1.1	Kansas St	59	69	Kansas St	Correct
Z01	Xavier	Z16b	TX Southern	19.5	15.6	No choice	91	72	TX Southern	NA
Z02	North Carolina	Z15	Lipscomb	19.5	13.1	No choice	84	66	Lipscomb	NA
Z03	Michigan	Z14	Montana	10.0	5.2	Montana	61	47	Michigan	Incorrect
Z04	Gonzaga	Z13	UNC Greensboro	13.5	7.8	UNC Greensboro	68	64	UNC Greensboro	Correct
Z05	Ohio St	Z12	S Dakota St	8.0	8.9	No choice	81	73	Push	NA
Z06	Houston	Z11	San Diego St	4.0	5.1	No choice	67	65	San Diego St	NA
Z07	Texas A&M	Z10	Providence	2.5	2.3	No choice	73	69	Texas A&M	NA
Z08	Missouri	Z09	Florida St	-1.5	-11.0	Florida St	54	67	Florida St	Correct
R1W1	Villanova	R1W8	Alabama	11.0	8.3	Alabama	81	58	Villanova	Incorrect
R1W2	Purdue	R1W7	Butler	3.5	8.1	Purdue	76	73	Butler	Incorrect
R1W3	Texas Tech	R1W6	Florida	0.0	-0.7	No choice	69	66	Texas Tech	NA
R1W4	Marshall	R1W5	West Virginia	-11.0	-4.6	Marshall	71	94	West Virginia	Incorrect
R1X1	Kansas	R1X8	Seton Hall	4.5	2.2	Seton Hall	83	79	Seton Hall	Correct
R1X2	Duke	R1X7	Rhode Island	10.0	3.8	Rhode Island	87	62	Duke	Incorrect
R1X3	Michigan St	R1X6	Syracuse	9.0	5.2	Syracuse	53	55	Syracuse	Correct
R1X4	Auburn	R1X5	Clemson	1.0	-0.6	No choice	53	84	Clemson	NA
R1Y1	UMBC	R1Y8	Kansas St	-12.0	-3.5	UMBC	43	50	UMBC	Correct
R1Y2	Cincinnati	R1Y7	Nevada	7.0	9.1	Cincinnati	73	75	Nevada	Incorrect
R1Y3	Tennessee	R1Y6	Loyola-Chicago	6.0	5.8	No choice	62	63	Loyola-Chicago	NA
R1Y4	Buffalo	R1Y5	Kentucky	-7.5	2.2	Buffalo	75	95	Kentucky	Incorrect
R1Z1	Xavier	R1Z8	Florida St	4.0	-0.2	Florida St	70	75	Florida St	Correct
R1Z2	North Carolina	R1Z7	Texas A&M	6.0	3.3	Texas A&M	65	86	Texas A&M	Correct
R1Z3	Michigan	R1Z6	Houston	3.0	3.5	No choice	64	63	Houston	NA
R1Z4	Gonzaga	R1Z5	Ohio St	2.5	10.6	Gonzaga	90	84	Gonzaga	Correct
R2W1	Villanova	R2W4	West Virginia	5.5	2.9	West Virginia	90	78	Villanova	Incorrect
R2W2	Purdue	R2W3	Texas Tech	1.5	0.6	No choice	65	78	Texas Tech	NA
R2X1	Kansas	R2X4	Clemson	4.5	8.2	Kansas	80	76	Clemson	Incorrect
R2X2	Duke	R2X3	Syracuse	11.0	4.3	Syracuse	69	65	Syracuse	Correct
R2Y1	Kansas St	R2Y4	Kentucky	-6.0	-2.3	Kansas St	61	58	Kansas St	Correct
R2Y2	Nevada	R2Y3	Loyola-Chicago	1.5	-0.8	Loyola-Chicago	68	69	Loyola-Chicago	Correct
R2Z1	Florida St	R2Z4	Gonzaga	-5.5	-5.9	No choice	75	60	Florida St	NA
R2Z2	Texas A&M	R2Z3	Michigan	-2.5	-2.9	No choice	72	99	Michigan	NA
R3W1	Villanova	R3W2	Texas Tech	7.0	-5.1	Texas Tech	71	59	Villanova	Incorrect
R3X1	Kansas	R3X2	Duke	-3.5	-3.4	No choice	85	81	Kansas	NA
R3Y1	Kansas St	R3Y2	Loyola-Chicago	1.5	-1.9	Loyola-Chicago	62	78	Loyola-Chicago	Correct
R3Z1	Florida St	R3Z2	Michigan	-4.5	-1.7	Florida St	54	58	Florida St	Correct

# Summary of results
table(round4_results_table$Accuracy)

## 
##   Correct Incorrect      Push 
##        23        17         1

The model went 2-1 ATS in the Elite 8 and is now 23-17-1 through the entire tourney for a 57.5% success rate. Considering the sample size of 40 games, a 95% confidence interval for the true model’s accuracy is between 42.2% and 72.8%.

Final 4

Model Predictions

# Recording Round 5 Winners into Round 5
bracket <- advance_winners(round4)

# Model Predictions for Round 5
round5 <- predict_ncaa_round(5)

# All Round 5 Spreads as of 8:30pm Thursday
# Spreads are positive if strong seed is favored, negative if underdogs
round5spreads <- data.frame(
    matrix(
        c(
"Villanova", 5,
"Loyola-Chicago", -5.5), 
    ncol = 2, byrow = TRUE), stringsAsFactors = FALSE
)

names(round5spreads) <- c("Team", "Spread")
round5spreads$Spread <- as.numeric(round5spreads$Spread)

# Append spreads on to bracket
round5 <- round5 %>% 
    left_join(round5spreads, by = c('TeamName.x' = 'Team')) %>% 
    mutate(Diff = Prediction - Spread)

Here are the model predictions for all round 4 games.

# Determine who model would bet on ATS
round5 <- ATSchoice(round5)

# Output of model recommendations
round5_predictions <- round5 %>% 
    select(StrongSeed, WeakSeed, TeamName.x, TeamName.y, Spread, Prediction, Diff, Win_Prob, ModelChoice)

kable(round5_predictions)

StrongSeed	WeakSeed	TeamName.x	TeamName.y	Spread	Prediction	Diff	Win_Prob	ModelChoice
R4W1	R4X1	Villanova	Kansas	5.0	1.9	-3.1	0.554	Kansas
R4Y1	R4Z1	Loyola-Chicago	Michigan	-5.5	-6.8	-1.3	0.305	No choice

The only model recommendations for the Final Four is Kansas.

Game Results and Model Performance

To be updated after games are played.

2018 March Madness Prediction Results

Going into Final Four

Billy Jackson

3/29/2018

Getting Started and Helper Functions

Input Model Predicted Spreads

Determining Model Choice

Assessing Model Accuracy

Putting winners into next round

Creating Initial Bracket

Play-in Round

Model Predictions

Game Results and Model Performance

Round 1

Model Predictions

Game Results and Model Performance

Round 2

Model Predictions

Game Results and Model Performance

Sweet 16

Model Predictions

Game Results and Model Performance

Elite 8

Model Predictions

Game Results and Model Performance

Final 4

Model Predictions

Game Results and Model Performance