Welcome to the first ever super bowl game and player prop projections. I have leveraged my Machine Learning Algorithms to project the outcome of the game, player props, and also go over a few other super bowl props I find interesting.
Keep checking back throughout the week as I add info and props to this write-up. If you have something you want me to take a look into, shoot me a message using the form on the bottom of the page. I plan on deep diving into the props starting Wed to see if there is anything I want to put money on, so expect the “Game Props” section to increase through the later part of the week.
The models predictions on this varied quite a bit for some stats, while others were all pretty consistent. I have never done this before so I didn’t know what model would work the best. To try to account for these factors if the stats were fairly close, I just used the average of all the models’ projections and, if there was inconsistencies, I threw out the outliers and averaged the other 3-5 models that had fairly similar projections. This throw-out only happened for a few of the same models for the same stats (for example Random Forest consistently projected 0 – 0.5 catches for Hill, Kelce, and others, while the other models all consistently projected between 7-10, so I just didn’t use Random Forest for any of the projected catches for players). I will make notes on these throw-outs for the various stats in the writeup.
This took a ton of work, hope you all find something useful. Enjoy!
I wanted to compile a list of observations and little side notes I took while doing the research and compiling all the data to publish this. So here are my observations (running list):
- Tampa Bay Buccaneers statistics have the largest variance in the NFL (least consistent team in the NFL). This made it incredibly hard to project Kansas City receivers targets and rushing attempts (more details in the sections below).
- The Bucs have attempted 31 field goals this season while the Chiefs have attempted 27 field goals.
- Bucs don’t run trick plays. Hardly any.
- Chiefs are 3 for 3 on 2pt attempts this season. Bucs are 0 for 2.
- Mahomes has thrown 2 pick-6’s. Brady has thrown 1.
- Each team has recorded 1 Safety.
- *There is a high chance for rain at the Super Bowl Sat night and into Sunday.
1. Mahomes Over 330.5 Pass Yards
My Logic: first and foremost the projections are on the over. Last time these teams squared off Mahomes threw for 462 Yards, but he should have had even more than that. Late in the game he threw a deep pass that hit Sammy Watkins in stride, although it hit him in a bad spot, his hands. All joking aside, the drop from Sammy Watkins cost Mahomes another TD and another 75 yards on his passing total. Factor in Tampa Bay’s stout run defense, I don’t think KC is going to run the ball at all and will have to rely heavily on the pass if they want to win. Rain or Shine, I like this going over.
2. Sammy Watkins Over 37.5 Yards
My logic: I have the projection working against me here, but going over all the data, not only did he go over that total last game, he should have gone way over with the deep ball drop. Tampa Bay is going to do everything they can to try to slow down Tyreek and Kelce (good luck), leaving the other receivers in favorable matchup’s without any help from other defenders. Watkins should have no problem getting open for at least 3 receptions, which should be all he needs to go over the total. Plus is Mahomes will go over, all those yards have to go to somebody and I doubt Hill is going to have another 269 yard game. Hopefully if it rains Watkins doesn’t let another slip through his hands. My Worry: Watkins re-aggravates his calf injury and is limited in his number of snaps.
3. Fournette Over 11.5 Rushing Attempts and Over 50.5 Rush Yards
My Logic: First and foremost, the model is on my side so that’s always a good thing. Second, the Buccaneers have to find a way to slow down the KC offense, what better way to do that than pounding the rock and chewing clock. Fournette has been the go to back late in the season and in the playoffs, so I expect him to continue his role as lead back and get a majority of the carries. Kansas City is ranked 19th in rush defense allowing 121 yards per game. My worry: KC goes up big early and the Bucs are forced to abandon the run Or Jones II busts off a few good runs and Arians does what he usually does and sticks with the hot hand.
4. Tom Brady Over 0.5 Rush yards
My Logic: Tom Brady has a lot of attempted rushes for not a lot of yards, in fact a good majority of those rushes were for negative yards. But 3rd and 1, 4 and 1, everybody knows its coming and nobody can ever seem to stop it, the Tom Brady sneak up the gut for 2 yards. At plus money this is an easy bet for me to take a flyer on (plus my model is on my side). My worry: Brady is pressured, tucks the ball and “runs” for a 3 yard loss, which counts as rush attempt for negative yards and ruins the prop.
5. Either team to miss a FG and longest FG Under 47.5
My Logic: I didn’t run any models for kickers because there was no easy way for me to get an accurate projection. But at + money and a good possibility for rain in the forecast, there is good enough likelihood one of these normally great kickers hooks one wide from 40+ yards out. They should also get plenty of chances if they are putting up a bunch of touchdowns to shank and extra point too. The longest FG Under 47.5 I will wait until closer to game time, if the field looks wet and a little sloppy I’m taking the Under. If the Field looks good I’m staying away.
6. Player to Have More Passing Yards: Mahomes (-150)
My Logic: My logic here is also explained in #1. I think KC is going to have to throw the ball and Tampa doesn’t have to throw the ball. At -150 I like the price. My worry: KC goes up big early forcing Tampa to pass for 3 quarters giving Brady time to rack up a ton of yards.
7. Player to have most rushing yards: Fournette (+175)
My Logic: For all the reasons I have said in #1 and #3. I don’t see KC being able to run the ball and think Tampa will have some success and Fournette has been the primary back for the Bucs late in the season and in the playoffs. My worry: Jones II busts off a few good runs and Arians does what he usually does and sticks with the hot hand.
8. Cameron Brate Over 30.5 Receiving Yards and Anytime TD scorer
My Logic: He has been a popular target for Brady recently and my projections support him going over. KC has allowed the 5th most receiving yards to tight ends and the 7th most touchdowns to tight ends & I trust Brate a lot more than I trust Gronk at this stage in Gronks career. I saw a couple of ESPN experts today had picked Brate under, so there is another reason to go for Brate.
9. Scott Miller Under 18.5 Rec Yards
My Logic: First and foremost, he is projected to go under his yards total by my model, but not by much. More importantly, when Antonio Brown signed with the Bucs, no other receiver was impacted as much as Miller. Miller averages 2 Targets, 1.11 Receptions, and 16 Yards per game when Antonio Brown plays. With the latest news that Brown was a full participant at practice it appears that he will go, Miller’s snaps will be limited and with weapons across the board for Tampa, there are just too many better options to go to than Miller. My worry: Brown re-aggravates his injury early and Miller gets more snaps and subsequently more targets.
10. Darrell Williams Under 7.5 attempts & Under 28.5 Rushing yards
My Logic: In games when Edwards-Helaire is healthy, Williams averages 3.3 carries for 12.1 yards per game. Couple that with the fact that Tampa has an outstanding rushing defense, I don’t see Williams getting many chances or many yards.
These NFL game projections are based on season stats (including postseason) and are heavily weighted on recent games. They do account for their weekly opponents, meaning scoring more points against a top ranked defense scores better for the offense than scoring a bunch on the worst defense in the league. Offense, defense, and special teams are weighted based on NFL averages to account for each team’s strengths and weaknesses for themselves and their opponents each week. For those of you familiar with Football Outsiders, this is my approach to replicating their DVOA metrics to project NFL scores.
On the season, the NFL model went 76-44 (63.3%) ATS and 18-18 (50%) Over/Under. Each was rebuilt multiple times throughout the season to optimize outcome, hence the difference in records.
Game outcome projection
The model projection shows the Tampa Bay Buccaneers should be -1.10 favorites, but instead find themselves at +3. The Buccaneers are wildly inconsistent statistically on both sides of the ball. When their variance is factored in to the model, the Chiefs range anywhere from -5 favorites to +10 point underdogs. So what Bucs team shows up on Sunday will determine the final outcome. When the Chiefs played the Bucs on 29 Nov, also in Tampa Bay, the Chiefs won 27-24, so projections show this being a little bit higher score than that game (I don’t have the projections from that game saved).
Although this is projected to be a high scoring affair, typically NFL playoff games trend heavily to the under. But if the Buccaneers have another day like they did when they played KC in Nov, this is looking like a high scoring affair.
The Betting edge is showing good value on the Buccaneers ATS and Moneyline. If you are confident in the Bucs, take a look at the Bucs moneyline, this will be priced at plus money and a better pick than taking the ATS price (in my opinion).
1. National Anthem
Jasmine Sullivan and Eric Church are singing the National Anthem. The average length of time for all Super Bowl National Anthems is about 1 Min 55 Sec. Sullivan sung the National Anthem in 2016 at the NHL Stadium Series game (another large televised game) that lasted only 1 min 38 Sec. Eric Church is extremely talented but not known to belt out long drawn out songs like an Alicia Keys type singer would. 9 of the last 15 have all been under the listed 1 min 55 Sec, so all indications are pointing to the Under 1 Min 59 Sec.
2. Coin Toss
As I said in my betting tip write-up, the coin toss has a 50% chance of landing on either heads or tails. Once you factor in Implied probability you will see that the implied odds to land on either side is between 50.7% (-103) and 51.7% (-107), making this a losing bet in the long run… but its still fun to bet on for the Super Bowl.
The Tails is winning the toss at 29 – 25 (53.7%). If you expect it to approach the 50/50 mark, Heads is your bet. Otherwise, Tails has been the outcome in 6 of the last 7 Super Bowls, so ride the hot hand and take Tails.
Or, you can fade the Chiefs who are 0-3 in Super Bowl coin tosses who are up against the undefeated Super Bowl coin toss champ Bucs (1-0). Also, the NFC is 37-17 all time for being on the winning side of the coin toss. If you like Parlays, the team who has won the coin toss is 24-30 in the actual game.
3. Players to Attempt a Pass (2.5)
Only Tom Brady and Gabbart have attempted a pass for the Bucs in a game (1 attempt for Gabbart if you exclude the garbage time appearance for Brady in the blowout win over the Lions). Bruce Arians is also not known for being a creative offensive mind.
If this prop is going to hit, it will almost certainly have to be from the Chiefs. Andy Ried won’t hesitate to pull out all of the tricks against an aggressive Bucs defense. I wouldn’t be surprised to see a player like Kelce or Watkins attempt a pass.
4. Opening kickoff to go for a Touchback
Butker (KC) has had 78% of kickoffs go for a touchback, while Succop (TB) has kicked 85% of kickoffs for a Touchback. BUT, 17 of the last 19 opening kickoffs have been returned. The ball used for opening kickoffs at the super bowl is a ceremonial ball and not a typical kicking ball. Kickers have said the ceremonial ball feels “Harder” and typically does not travel as far as the typical kicking ball.
KC Chief’s Props
If you don’t remember the last time the Chiefs played the Bucs, let me give you a synopsis. Tyreek Hill had 200+ and 2 TDs…IN THE FIRST QUARTER! he finished with 13 Rec, 269 yds, and 3 TD’s.
The hardest part about predicting the the Chiefs players props was the Bucs. Not the Chiefs players themselves, but the Bucs’ inconsistency on defense. The Bucs have been (by far) the most inconsistent defensive team in all aspects of the passing game, doing things like allowing Tyreek Hill to have the game he did but containing Michael Thomas and the Saints and slowing Devontae Adams (Packers). The Bucs have been diced up in the pass defense by the likes of Taylor Heinicke, Jared Goff, Matt Ryan (twice), and Derek Carr. But on the flip side they have also held up fairly well against other decent players like Drew Brees 2 weeks ago. All of this to say, the Bucs inconsistency skewed the Chiefs player prop projections, especially receiving.
None of the models had any issues projecting Passing stats and most were all extremely close.
For the Chiefs, Patrick Mahomes has done most of the passing this year, and here is how his statistics for Super Bowl Sunday are project out.
What this is saying is, based on the defensive stats for the Bucs and the offensive stats of Mahomes and Co., Mahomes is expected to attempt 45 passes, completing between 28 and 29 for 333 yard, between 2-3 touchdowns, small chance of an INT, and his longest completion (Lng) to be about 45 Yards.
As I was just complaining about, the Bucs pass defense is inconsistent so projections for these same stats ranged as follows; Attempts (41-49), Completions (24-37), Yds (254-351), TD (2.3-3.5), and Lng (33-60).
Neural Network (NN) had issues projecting attempts and TD rushing stats.
The backfield for the Chiefs has seen a healthy dose of Edwards-Helaire, although he came back from an injury in the championship game. I don’t anticipate he will be limited in any way based on recent reports.
The unfortunate truth for the Chiefs rushing over bettors is that the Bucs rush defense is stout and they’re consistent. Also, the model projections support that and don’t project a big ground day for the Chiefs. In the regular season meeting between these 2 teams, the Bucs held the Chiefs in check in the running game. Edwards-Helaire was the leading rusher with 11 carries for 37 yards. The models this time all say he will have roughly 13 carries (Att) for 59 yards and is unlikely to see the end zone (0.36 means if the game was played 100 times he would score a total of about 36 TD’s).
The projections also show Bell and Mahomes contributing to the run game, although Mahomes projected rushing stats will likely be due to scrambles. Tyreek Hill has attempted quite a few rushing attempts this season (multiple games with 3 att) and may be a good value target for prop bets. Mecole Hardman also had 6 games with a rushing attempt and in 1 of those games he had multiple attempts, another potential prop to get value on.
Darrell Williams saw his usage increase while Edwards-Helaire was out and even out-touched Bell over the last 3 games, but he is listed lower than Bell on the Depth Chart so his numbers are lower.
The longest projected rush is 14 yards by Edwards-Helaire.
Projection variance notes: Edwards-Helaire rushing attempts ranged from 10-19, with yards ranging from 31-64.
Random Forest (R.F.) had issues on projecting targets and touchdown projections for all receiving stats.
The Chiefs have a wide variety of weapons, but last time they played the Bucs, Tyreek Hill gashed them for 13 Rec, 269 Yds, and 3 TD’s. Here are the projections for the Chiefs receiving core.
Tyreek Hill is projected to be the leading receiver although not by much over Travis Kelce. It is no surprise to see those two on top of the stat sheet. Hill has gone over 100 yards 5 times and Kelce going over 100 yards 8 times. They have both gone over 100 yards 4 times (including playoffs). Travis Kelce has scored a TD in 13 of 18 games (0 in the Bucs game), but only has 2 games with multiple TD’s. Tyreek Hill has scored in 12 of 18 games and has 3 games with 2+ TD’s.
Sammy Watkins is a decent option for prop bets, his projections aren’t stellar, but it is likely the Bucs will focus on not letting Hill beat them like he did last game. Watkins has only scored twice all season. Mecole Hardman has scored 5 times this season and may be worth a flyer on the TD prop if the price is good. Robinson has scored 3 times this season.
Edwards-Helaire has the only receiving touchdown for the running backs.
Projection variance notes: Kelce’s yards variance was high ranging from 65-145 yards. Sammy Watkins also had a large yards variance ranging from 11-41 yards.
- Non-QB’s with a pass attempt: Kelce (2), Watkins (1), Tommy Townsend (1).
- 1 Kick return for a TD all season
- 2 Defensive Touchdowns all season
TB Buccaneer’s Props
Last time the Bucs faced the the Chiefs, it was lopsided earlier in the game, but the Bucs made it close with a 27-24 Final. The Bucs scored 3 times in the second half and held the Chiefs to 1 TD. The Bucs did score 2 fourth quarter TD’s to try to complete a comeback but ultimately came up short. Unlike the Bucs defense, the Chiefs defense was consistent, making the Bucs offensive projections have less variance.
None of the models had any issues projecting Bucs passing stats and most were extremely close.
Tom Brady is the primary passer for the Bucs, with Blaine Gabbart being the only other player to attempt a pass in a game. Tom Brady is projected to complete about 27 of 42 for 322 yards and almost 3 TD’s. He is projected to to throw an interception and did throw 2 last time they played the Chiefs.
Projection variance notes: Tom Brady’s projected touchdowns ranged from 1.1 to 3.2.
Neural Network (NN) had issues projecting attempts and TD rushing stats.
The Bucs have been pretty evenly split in the backfield between Leonard Fournette and Ronald Jones II. Lately, Fournette has seen a majority of the carries, although Jones still see’s a decent number of carries.
The big thing I notice between Fournette and Jones here is that the difference in carries vs the difference in yards. Jones seems like the better big play (or home run) option where as Fournette is the more physical back. I wouldn’t be surprised if the Bucs end up in a goal to go situation and Fournette pounds one in. Jones has 7 Touchdowns, but only 3 since the mid point of the season. Fournette has 9 touchdowns, with 7 coming since the mid point and riding a 3 game TD streak (all 3 playoff games). Fournette certainly appears to be the better target for prop bettors.
The Bucs have run the ball with the 3 listed receivers, but with Antonio Brown’s status uncertain it appears that Scott Miller is the best option for WR to rush the ball. No WR has rushed for a TD all season.
I was very surprised to see how many times Tome Brady rushed the ball each game. Sure some of them may have been scrambles to avoid a sack, but they are credited as a rush attempt nonetheless. Most of his rush attempts were for -2 to 2 yards, so he wont be ripping off anything large. However, if there is a good price on Brady attempting a rush, it might be worth looking into. He has rushed the ball in 15 of 19 games and averages 2.6 attempts per game, but only averages 0.12 yards per attempt. He has only gone over 3 yards rushing once and has 3 rushing TD’s on the season.
The longest projected rush is 25 yards by Ronald Jones II.
Projection variance notes: Fournette and Jones attempt numbers varied a lot. I settled on using the last 8 games as the base which gives the stats above. Using the entire season shows Jones out carrying Fournette 17 to 11.
Random Forest (R.F.) had issues on projecting targets and touchdown projections for all receiving stats.
Antonio Brown is currently listed as doubtful, although many people expect his status to change by the time the Super Bowl rolls around. If he cant go, Chris Godwin and Scott Millers numbers are projected to go up (I will repost those here later in the week when Brown’s status is better known). The Bucs have very talented receivers and a lot of depth.
Mike Evans is listed as the main target for Tom Brady with Chris Godwin not far behind him. Mike Evans scored twice last time these 2 teams met, both in the 4th quarter. Both Evans and Godwin will certainly be the favorite targets for Brady as they have been all season, even more so if Brown doesn’t play. Godwin has scored 8 time this season with 6 of them coming since the middle point in the season. Evans has scored 15 times, including three 2 TD games, with 8 of his touchdowns coming since the mid point.
Last time these teams met, Gronk had a breakout game, going for 106 yards. BUT, this was the only time all season Gronk has eclipsed the 100 yard mark and he has only caught for more than 62 yards twice all season. He has caught a fair number of touchdowns (7), but only has 2 in the last 6 games, both coming against the hapless Detroit Lions.
As for the running backs, Fournette is certainly the bigger receiving threat averaging 4.5 targets in the last 6 games, Jones only has 1 target in that same timeframe. Each running back has 1 TD catch all season with Fournette’s coming against the Saints in the playoffs.
Cameron Brate might be a guy to target in the Prop market averaging 4 targets, 3 receptions, and 40 yards over the last 6, catching a TD in the game against Green Bay.
Projection variance notes: Nothing significant for the Bucs receiving projections
- 1 Defensive touchdown all season.