Welcome to my first March Madness projections using similar Machine Learning algorithms used for the other major sports I have regularly projected over the past few years. All stats are based on team based stats from the season and all training data was taken from all games played this season by each respective team.

Somebody asked me about a Sharp report for the tournament. For big events like this and any professional sports playoff games, there is so much money being thrown around it can skew the betting data. For example, a millionaire/billionaire alumni loves basketball, especially the school they are alumni at, and heads to Vegas (or local sportsbook) and drops 50k for them to win that day and as a result the line moves. That’s not a Sharp or Sharp action but can look like Sharp action in the betting data. This happens a lot on big sporting events and playoff games (see Mattress Mack and the Astros). Since I don’t know what is or isn’t Sharp action, there will be no Sharp reports for the NCAA Tournament.

**Updated with Championship game Projections.

Best of Luck to everybody!

#### KenPom adjusted Power Rating Projections

I updated a modified model this morning (16 Mar) to use KenPom numbers to adjust for strength of schedule, ran some trials & tests, and then the first round projections. I will be posting the the KenPom adjusted projections and using those for the bracket building in the ESPN group rather than my custom power rating adjustments. Full Round 1: KenPom Adjusted Projections Here.

Round 2: Round 2 Projections

Round 3: Round 3 Projections

### Track the Model Brackets and Join In

If you want to check out each models bracket predictions and enter your own bracket to see how you do against the models, join the Bracket Challenge group on ESPN: B2SportsAnalytics Bracket Challenge.

ESPN Does not allow me to share the bracket openly until the games start so here are the Brackets. Selections for any of the First Four teams from Round 1 on are assuming the First Four game projections below are 100% accurate, if a team loses I will re-run and re-upload the bracket. [Models ordered below from best to worst performing in testing]

## Records

Records reflect results through the Final Four.

#### Consensus record

#### Overall Record

#### Model Rank

Here is how the models rank based on average win percentage across ATS, ML, and O/U Records.

## Championship Game

#### Summary of Projections

A “1” indicates that model selected that side (ATS, ML or O/U) with >55% probability. A Blank indicates that model didn’t project a side with >55% probability. And an “X” in the consensus column indicates every model selected that side with >55% probability.

#### Projections

All Models Projections for the Championship game, ordered by record through the tournament.

How to read the projections: The model type is at the very top. Matchups are denoted by the alternating white/green pairings, away teams are on top, home teams on bottom. The model analyzes the matchup and projects the home and away teams score (Proj Score). The difference between the home teams score and the away teams score give us what the model says the line should be (Proj Line). The “Line” column is what the Vegas line is at the time I run the model. The “Line Diff” is the difference in the projected line and the Vegas line. A positive line diff means the projected line is in the away teams favor compared to the Vegas line, negative means the projected line is in the home teams favor. The “Cover Prob” uses a normal distribution and the teams variance to project each teams probability to cover the listed Vegas line. Same thing for totals, “Proj Total” is the sum of the projected scores, “Line Total” is the listed Vegas total for the game, the difference between the two, and the probability to go over or stay under denoted by “O” and “U”.

#### Edge

If you don’t understand what you are looking at, I recommend reading my post about betting tips. The percentages show the betting Edge, which is the cover prob (from above) minus the implied probability (-110 odds implied prob is 52.4%). If a team has a 92.4% chance to win and a 52.4% implied probability (or -110 odds), the Edge is 40%. The Edge alone doesn’t mean you should blindly bet it. To quantify, >35% is great value, 20-35% is really good, 10-20% is decent, <10 is ok value, blanks are negative value. ML is moneyline.

## Final Four

#### Summary of Projections

A “1” indicates that model selected that side (ATS, ML or O/U) with >55% probability. A Blank indicates that model didn’t project a side with >55% probability. And an “X” in the consensus column indicates every model selected that side with >55% probability.

#### Projections

How to read the projections: The model type is at the very top. Matchups are denoted by the alternating white/green pairings, away teams are on top, home teams on bottom. The model analyzes the matchup and projects the home and away teams score (Proj Score). The difference between the home teams score and the away teams score give us what the model says the line should be (Proj Line). The “Line” column is what the Vegas line is at the time I run the model. The “Line Diff” is the difference in the projected line and the Vegas line. A positive line diff means the projected line is in the away teams favor compared to the Vegas line, negative means the projected line is in the home teams favor. The “Cover Prob” uses a normal distribution and the teams variance to project each teams probability to cover the listed Vegas line. Same thing for totals, “Proj Total” is the sum of the projected scores, “Line Total” is the listed Vegas total for the game, the difference between the two, and the probability to go over or stay under denoted by “O” and “U”.

#### Edge

If you don’t understand what you are looking at, I recommend reading my post about betting tips. The percentages show the betting Edge, which is the cover prob (from above) minus the implied probability (-110 odds implied prob is 52.4%). If a team has a 92.4% chance to win and a 52.4% implied probability (or -110 odds), the Edge is 40%. The Edge alone doesn’t mean you should blindly bet it. To quantify, >35% is great value, 20-35% is really good, 10-20% is decent, <10 is ok value, blanks are negative value. ML is moneyline.

## Elite Eight

#### Summary of Projections

A “1” indicates that model selected that side (ATS, ML or O/U) with >55% probability. A Blank indicates that model didn’t project a side with >55% probability. And an “X” in the consensus column indicates every model selected that side with >55% probability.

#### Projections

How to read the projections: The model type is at the very top. Matchups are denoted by the alternating white/green pairings, away teams are on top, home teams on bottom. The model analyzes the matchup and projects the home and away teams score (Proj Score). The difference between the home teams score and the away teams score give us what the model says the line should be (Proj Line). The “Line” column is what the Vegas line is at the time I run the model. The “Line Diff” is the difference in the projected line and the Vegas line. A positive line diff means the projected line is in the away teams favor compared to the Vegas line, negative means the projected line is in the home teams favor. The “Cover Prob” uses a normal distribution and the teams variance to project each teams probability to cover the listed Vegas line. Same thing for totals, “Proj Total” is the sum of the projected scores, “Line Total” is the listed Vegas total for the game, the difference between the two, and the probability to go over or stay under denoted by “O” and “U”.

#### Elite Eight Edge

If you don’t understand what you are looking at, I recommend reading my post about betting tips. The percentages show the betting Edge, which is the cover prob (from above) minus the implied probability (-110 odds implied prob is 52.4%). If a team has a 92.4% chance to win and a 52.4% implied probability (or -110 odds), the Edge is 40%. The Edge alone doesn’t mean you should blindly bet it. To quantify, >35% is great value, 20-35% is really good, 10-20% is decent, <10 is ok value, blanks are negative value. ML is moneyline.

## First Four – Round 3 Data

## West & East

Round 3: Round 3 Projections

#### Summary of Projections:

#### West & East Projections:

#### West & East Edge

## South & Midwest

Round 3: Round 3 Projections

#### Summary of Projections:

#### South & Midwest Projections

#### South & Midwest Edge

## First Four – Round 2 Data

## Model Notes

#### Model Results

Below is the number of models that selected each team to make each round.

- Champion:
- 6 – Gonzaga
- 1 – Iowa

- Championship Game:
- 3 – Gonzaga vs Illinois
- 2 – Gonzaga vs Arkansas
- 1- Gonzaga vs Houston
- 1 – Iowa vs Ohio State

- Final Four:
- 7 – Gonzaga, Alabama
- 6 – Illinois
- 3 – Arkansas
- 3 – Ohio State
- 1 – Baylor
- 1 – Houston

- Elite Eight:
- 7 – Gonzaga, Iowa, Alabama, Illinois
- 6 – Colorado, Houston
- 4 – Ohio State, Baylor
- 3 – Arkansas
- 2 – Purdue
- 1 – Michigan, Morehead State, Wisconsin

- First Round Upsets (4 or more models, excluding 8-9 matchup):
- All 7 Models Like:
- (12) UCSB over (5) Creighton

- (11) Drake over (6) USC
- (10) Maryland over (7) UConn
- (10) Rutgers over (7) Clemson
- (13) Liberty over (4) Oklahoma

- 6 Models Like:
- (12) Winthrop over (5) Villanova
- (13) Ohio over (4) Virginia
- (11) Utah State over (6) Texas Tech

- 5 – (14) Morehead State over (3) West Virginia

- All 7 Models Like:
- Sweet 16 Cinderella’s:
- 7 – (13) Liberty
- 4 – (14) Morehead State, (13) Ohio
- 3 – (12) Winthrop
- 2 – (12) UCSB, (11) Drake

## South

Round 2 Projections: Round 2 Projections

Round 1 Projections: KenPom Adjusted Projections Here (Old Round 1 Projections)

#### Summary of Projections:

#### South Projections

#### Betting Edge

## Midwest

Round 2 Projections: Round 2 Projections

Round 1 Projections: KenPom Adjusted Projections Here (Old Round 1 Projections)

#### Summary of Projections:

#### Midwest Projections

#### Betting Edge

## West

Round 2 Projections: Round 2 Projections

Round 1 Projections: KenPom Adjusted Projections Here (Old Round 1 Projections)

#### Summary of Projections:

#### West Projections

#### Betting Edge

## East

Round 2 Projections: Round 2 Projections

Round 1 Projections: KenPom Adjusted Projections Here (Old Round 1 Projections)

#### Summary of Projections:

#### East Projections

#### Betting Edge

## First Four Games

*Updated the projections using KenPom power ratings for the First Four games and loaded it on the google sheet below.

First Four games start on Friday! Thursday evening I will update the latest odds and update the projections here and on Google sheets at: First Four Projections.

Check spread for Syracuse. Projections show line at -6, but should be plus 6

Yeah, somebody notified me of that yesterday. I am out of town and will correct my typo and repost the correct projection when i can (tomorrow evening at the latest).