Jarrett Karnibad

Sports Articles and Data Analysis

Recent Articles

  • The 2026 MLB Karn Prediction Model

    The 2026 MLB Karn Prediction Model

    Right around Memorial Day of this year, I became focused on trying to create a game prediction model for MLB games. I had attempted this in the past with limited success, but I felt that now was a good time to try again. There is an abundance of publicly available data to use, and building prediction models is something I love to do (see March Madness). Within a week, I had created a model that I decided to track for the rest of the season. And, that is what I did. I input the teams, starting pitchers, and odds for all 1,557 games from May 31st to the last regular season games on September 27th. Each day, my data was refreshed so that all games were run through the model accurately. Before I get into the results, I want to explain what the model is, and what it is trying to accomplish.

    The Model

    So, let’s begin by examining the inputs of the model. For a given game, I enter the two teams (away and home), and each team’s starting pitchers. The model pulls the starting pitcher’s average length of outing (some teams use relievers to start games, and thus it is better to use innings per appearance rather than innings per start). This also gives the bullpen its expected innings for this game (9 innings – Starter Average Innings / Game). I use a special blend of pitching statistics to give the starter and the bullpen their expected number of runs allowed per 9 innings. This blend is then adjusted for the amount of innings expected for the starter and the bullpen. This is the basis for the “Pitcher Runs” for that game. For “Batter Runs,” I use teams’ wRC+ and scale it to league average runs. But to make it more in-depth, I use a team’s wRC+ vs Righties (if the opposing starting pitcher is right-handed), and wRC+ vs Lefties (if the opposing starting pitcher is left-handed). For the bullpen portion, I simply use the teams’ total wRC+. This gives an expected number of “Batter Runs” per team for that specific game. Also involved in “Batter Runs,” is Fielding Run Value, to measure the defensive effects of the opposing team. To find the expected final score of the game, I use a weighted average formula with Pitching Runs and Batter Runs. And finally, I use Park Factor to scale the final result to the park that it is being played in (Playing at Coors Field leads to a lot more runs than at Globe Life Field).

    Without giving away too much, that is the model. It gives a predicted final score for any game between any two pitchers. As a game prediction model, I wanted to use odds to help track its success. I input each team’s moneyline odds, as well as the game’s total. This gave me more data to track the success of my model. i was focused on tracking the following statistics/categories: moneyline win percentage, 10% expected value win percentage, predicted underdog win percentage, and total (o/u) win percentage. These are the main four categories, but the winning percentage does not tell the full story. I wanted to track profit/loss based on a standard unit. This is much more telling in determining the success of a model. Now let’s get into the results:

    Results
    Predicting the Winner of Each Game

    I am going to start with the most fundamental thing this model was built to predict: the winning team. Yes, over/unders are inherent to the model as well, but the only important thing in real baseball is winning the game. Did my model do a good job of accurately predicting the winning team? I would say “yes.” Here are its results compared to two less-complex methods of predicting the winning team:

    As shown above, my model was better than the oddsmakers at predicting the winning team in a given game. The Karn Model won 57.48% of its games, whereas picking the favored team only won 56.82% of the team. And although that might seem small, look the the next column to see how big of a difference that really is. The Karn Model was over 63 units better than simply betting one unit on the team favored to win. This is not totally unexpected though. Because you are betting a standard one unit on every game, the team that is favored to win will always return less than 1 unit back. This means that the breakeven win percentage is much higher because there are no underdogs to return over 1 unit per win. The Home Team Only method performed the worst of the three models by far. Not only did it have the lowest win percentage, it also lost the most units.

    Some may say that my model was a failure because it lost units over the course of the season (May 31-End of Regular Season), but I would disagree for two main reasons. First, putting one unit on all 1,557 games is not realistic. Secondly, it discounts the fact that the model can favor some games more than others. This is the basis for the 10% Expected Value category.

    10% Expected Value

    As I mentioned above, I needed a way to sort which games that model favored the most. To accomplish this, I used an adjusted Pythagorean Win Expectancy formula. Because this formula is meant to project a winning percentage across a whole season’s worth a runs and runs allowed, and my model only predicts the score of a single game, I had to increase the exponents to 3.5. This felt like a good balance of creating separation in win probability between the two teams. I did some testing in May and June, and exponents of 3.5 performed the best. This gave me my win probability per game. When compared to the implied win probability from the sportsbook odds, I could then use the expected value formula to see which games were more favorable than others. From there, I isolated only the expected value percentage of the team my model projected to win. There is no point in positive expected value in a team that was not projected to win the game. I then tracked all games in which the projected winning teams’ expected value was above 10%.

    This is really where the model shined. Out of the 1,557 games, there were 350 in which the expected value (from here on out, “of the winning team” is implied) was above 10%. That’s roughly 22.5% of all games. Of the 350 games, the model correctly predicted the winner in 203 of them, for a winning percentage of 58.00%. While that isn’t significantly above the winning percentage of predicting every game, the margins were far better in this section. If a bettor placed a 1 unit bet on all 350 games, they would have finished the season with a profit of 32.15 units. Not only is that a very strong return, but it is also much more realistic for actual bettors.

    You might be asking, “how is it possible that only around half of a percentage point can make such a big difference in the profit/loss of each model?” The answer: “Return.” Let’s look at presumptive AL Cy Young Award Winner: Cam Schlittler and the New York Yankees. Schlittler pitched in 20 games while I was tracking games. The Yankees won 11 of those games. But even though the Yankees won more than they lost in which he was the starting pitcher, they had a profit/loss of -2.51 units. This is because the Yankees were favored in all 20 games, with average odds of -177. Meanwhile, Boston Red Sox won 13 of Sonny Gray’s 19 games on the season, and his games netted a profit of 3.69 units. The reason is because Gray was only favored by more than -177 (Schlittler’s average) three times all season. Another example is Cade Cavalli from the Washington Nationals. He was even less of a favorite than Sonny Gray: only favored -177 or more once in 20 games. The Nationals won 13 of Cavalli’s 20 starts, and netted bettors a profit 4.38 units with 1 unit bets on all 20 games. The Nationals were the underdog in half of his starts. This shows that there is profitability in being able to predict underdog teams accurately.

    Predicted Underdogs

    That brings us to the Predicted Underdogs section of my model. To measure this, I only counted games in which the team with “plus” odds (+100 or higher) was predicted to win the game by my model. This does not count the games in which the underdog team had odds lower than their opponent but still “minus” odds. An example of a game that would not be counted is if Team A was favored to team Team B with odds -112 for Team A and -108 for Team B. So in total, of the 1,557 games, only in 192 of them was a “plus” underdog predicted to win. That’s a percentage of 12.33%, so we can say roughly 1 out every 8 games, a plus underdog was predicted to win. In those 192 games, the predicted plus underdog won 93 games, a win percentage of 48.44%. Now on the surface, this does not look great: winning less than 1 out of every 2 games. However, because every team predicted to win in this section has plus odds, every win is worth more than 1 unit. This means that the breakeven point is much lower than with all the games included. In fact, if a bettor had placed 1 unit on all 192 predicted plus underdogs to win, he would have profited 6.3 units. I see this is an absolute success. Being able to predict underdog victories is difficult, and being able to profit off of them is even more difficult. This is definitely a plus for my model.

    Totals (Over/Under)

    Now here is where the model struggled. Predicting totals can be difficult because of sample sizes and averages. The model uses season-long averages to predict the results in one game. This can be a flawed concept because anything can happen in one game. However, as the season progresses that we should get closer and closer to the true averages. If a game was played 10,000 times, we should see the average of the results get close to my model’s prediction. When predicting the winner of the game, it does not matter how many runs score, or how much a team wins by, it is only about the team that wins the game. With totals, every run matters. There are a wider outcome of results, which makes it tougher to predict. Two other reasons that totals can be tough to predict in baseball: weather and park factor. Sometimes the wind will be blowing in, usually leading to fewer runs, or it will be colder, also usually leading to fewer runs. The weather has a big impact on the total. Additionally, the location of a game matters a great deal as well. The altitude at Coors Field leads to more runs than a park at sea level and on the water like Oracle Park in San Francisco. I used BallparkPal.com for the Park Factors of every game throughout this process. It combines weather and park effects to give each game a park factor. The park factor predicts how many more or fewer runs will be scored in that park than on average. This was perfect to help tailor my season-long averages to each individual game. Unfortunately, the results were not as good as I had hoped for. Of the 1,557 games I tracked, I correctly predicted the total in only 801 of them, a winning percentage of 51.45%. While still above 50%, it fell far below the breakeven line for a loss of 19.55 units. This definitely was not the result that I had hoped for, however my model was still able to predict the correct result more than half the time, so it was not completely wrong. There is certainly room for improvement.

    Analysis

    Overall, the model performed quite well in its first partial season. I will be “running it back” next year with the same framework. However, there are some improvements that need to be made in the offseason. The first thing I will focus on is automating my data refreshing. Throughout the season, I would have to manually refresh my data each day to ensure that the model was handling the most up-to-date information to make predictions. If I can automate this, I can save myself around 5-10 minutes of copy and pasting. Secondly, regarding the model itself, I want to try to predict bullpen usage. This is going to be a tall task, but it’s not as accurate to use a team’s entire season long bullpen averages for a game in which only 3 or 4 of them may pitch. If I can find a way to predict usage, or even just use the averages of the “available” arms, I feel this will lead to an increase in accuracy. Finally, I want to find a way to introduce variance into the data for pitchers. The reason being that pitchers almost never throw exactly their season averages in a game. The same goes for hitters, but it should be much easier to track in pitchers. The more consistent the pitcher is around his season averages, the more confident the model is in predicting the result. This will require me to gather game-by-game data for the pitchers, but it could be well worth it if it increases prediction accuracy. I’m sure I will think of more upgrades to make over the offseason, but as of now, those are the three that are at the front of my mind.

    Conclusion

    Overall, I am very pleased with my model and its results. As I mentioned above, there are certainly things that can be improved, but I feel that it was a very good first attempt. I will bring the MLB Karn Model back next year, new and improved, and I will try to give updates along the way! Coming out over the course of the next couple of weeks will be my 2026 MLB Season Predictions Review, where I will go over all of my predictions that I made at the beginning of the season. Some are good, some are great, and some are down-right awful. Stay tuned for that article soon.

  • The New York Mets, or is it the New York Mercenaries?

    The New York Mets, or is it the New York Mercenaries?

    I love the New York Mets. I watch every game I can, and in the case of this season, that has meant watching a lot of uninspiring baseball. This post will essentially be an opportunity for me to vent about David Stearns and the moves of this offseason.

    The Collapse

    Let’s start all the way back at September 28th, 2025. Because that’s where this story truly begins. At 5:59pm, the Marlins defeated the Mets 4-0 and officially eliminated them from the 2025 Postseason. It was the culmination of a disastrous second-half of a season in which the Mets held the best record in the league at one point with a 45-21 record. The collapse mostly came at the expense of the starting pitching, which showed an astounding incapability of going deep into games, forcing the bullpen to cover large portions of games, which in turn wore them down as the season progressed. So, while the Mets’ supposed contemporaries were playing meaningful baseball in October, David Stearns was getting his sledgehammer ready.

    The Demolition

    David Stearns was thought of as something of a Boy Genius, when Steve Cohen hired him as President of Baseball Operations following the Mets’ disastrous 2023 season. Stearns’ first move was to fire Buck Showalter, and hire Carlos Mendoza as the Manager of the team. Stearns is known across the league for his ability to create baseball teams built on winning on the margins, and doing so without a massive payroll to help him. However, that skill set wouldn’t exactly be needed in New York as Cohen has “unlimited” pockets and would provide the financial backing for whatever Stearns needed.

    This offseason Stearns used the phrase “run prevention” as his justification for all the moves made. Let’s start with what I deem as the most egregious error: not even attempting to re-sign Pete Alonso. I’m going to give a set of statistics for two players, and you can be the judge on which is better.

    Seems pretty obvious which is the better player right? Well, Player A is the combined statistics of Mets players when playing 1st Base this season. Player B is Pete Alonso. Not only is Pete significantly outperforming all of the Mets’ replacements (in a down year by his standards), but he is having the best defensive season if his career. The irony of that last statement is just what makes this hurt even more. Stearns let Pete walk essentially because he wanted better defense at 1st Base. So you know who he signed to play 1st? Jorge Polanco, who prior to this season, had not even played a full inning of 1st Base in his career.

    That’s a good segue into the Jorge Polanco signing. As I mentioned above, Polanco is a career Shortstop and 2nd Baseman, without any real experience at 1st Base. Additionally, Polanco has a deep injury history, missing an average of over 50 games per year since establishing himself in the Major Leagues.1 Add in the fact that he is 33 years old, and coming off of one of the best seasons of his career, it just felt like buying high on a guy that is pretty late in his career to try to learn a new position. And what has the result been so far? He’s played in just 14 games due to a multitude of injuries, and in those games, he has been abysmal: .179 / .246 / .286 slash line and a 54 wRC+. Polanco should return from injury relatively soon, but it’s being reported that he will have to play through the pain for the rest of the season. That’s not exactly great news for a hitter who is already having an awful year.

    Now I want to get into the Brandon Nimmo-Marcus Semien trade. When this was first announced, I was very disappointed. Not because I thought Marcus Semien would have a 68 wRC+ through 54 games (which he does), but because Brandon Nimmo was a career Met who loved playing in New York and was beloved by the fanbase. Yes, the Brandon Nimmo contract from the 2022 Offseason was definitely an overpay that was going to be tough to stomach in a few years, but that is part of what comes with retaining homegrown players. Going into this season, Nimmo had 5 years and $102.5 million left on his deal, whereas Semien had 3 years and $72 million left on his. This trade was completed to accomplish two things for the Mets: long-term flexibility in the outfield, and improve defense. Semien has been a much better defender than Nimmo throughout their careers, and he is coming off of a Gold Glove winning season at 2nd Base. Additionally, after trading Jett Williams and Luisangel Acuña, the positional strength of the Mets Minor League system is their Outfield (as we have seen with Carson Benge, A.J. Ewing, and Nick Morabito all making their MLB Debuts before Memorial Day this year. From those two perspectives, one could argue this trade makes a lot of sense. Unfortunately, hitting is massive aspect of a positional player’s worth, and Marcus Semien cannot hit right now. Let’s look at the comparison for this season between Nimmo and Semien:

    Nimmo is outperforming Semien in almost every category, and in most cases, doing so in a major way. The most jarring thing about this comparison is that Nimmo is better in all three fielding statistics. It is absolutely unfathomable that Brandon Nimmo could be playing better defense than a two-time Gold Glove winner. But hey, that’s how it goes when it comes to the Mets.

    Next up are two trades I want to address together. The first trade was Luisangel Acuña (and Truman Pauley) for Luis Robert Jr. This trade was to address the Mets’ glaring hole in Centerfield. The second trade was Jeff McNeil (and cash) for Yordan Rodriguez. While I was not the biggest Luisangel Acuña fan, he was a highly ranked prospect in the Mets organization, and they flipped him for Luis Robert Jr, a guy who has dealt with injuries his entire career, and has never put together a full great season outside of 2023. He’s never even played a full season outside of 2023: missing an average of almost 50 games per year, excluding his rookie season in the shortened 2020 season. Now for the Jeff McNeil trade. This was essentially a throw-away of another longtime New York Met. The return player in this trade, Yordan Rdoriguez, was not even a Top-30 prospect in the A’s organization. So the two trades merge to become essentially Luis Robert Jr for Jeff McNeil at the big league level. Guess who is having a more productive season?

    While it is close, I’m going to take the player who is better in almost every hitting statistic, and has played in 27 more games. Everyone loves to talk about Luis Robert Jr.’s ceiling and ability, but one of my favorite sayings is, “the best ability is availability.” Chalk this up to another poor decision for Stearns and the Front Office.

    And finally, the icing on the cake; the supposed crown jewel of this New York Mets’ offseason: Bo Bichette. I do not even know where to begin. But let’s start with his position. Bo Bichette has been a shortstop for his entire career. Outside of the 2025 World Series run with the Blue Jays, in which he played 2nd Base due to coming off of an injury, Bo Bichette has never played anywhere besides Shortstop. But the New York Mets already have a superstar Shortstop in Francisco Lindor. Can Bo play 2nd Base for the Mets? If that was the plan, then why trade for a Gold Glove winning 2nd Baseman? Ok, let’s have Bo Bichette play 3rd Base, a spot he has never played. Not even in the Minor Leagues did Bichette play 3rd Base. Furthermore, Brett Baty was coming off of an amazing second half of the 2025 season, and it seemed like he was primed to breakout in 2026. But instead, the Mets sign Bichette, kick Baty out of his natural home at 3rd Base, and force him to learn 2 new positions: Right Field and 1st Base. In case you were keeping count, that is now four key players that are playing new positions going into this season: Bichette, Baty, Polanco, and Vientos.2 Ok, sure. Maybe Bichette will struggle a bit at 3rd Base to start the season, but he’s a lifetime 0.294 hitter who averages 23 home runs per 162 games (going into this season). That bat isn’t just going to go away, right? Right??

    This season, Bo Bichette is leading the league in at-bats, but is hitting a measly .227 / .274 / .323. That’s an OPS of 0.597. His OPS+ is 72, meaning his OPS is 28% worse than a league-average hitter. His wRC+ is 72, as well. It is an unfathomable drop-off for a player with such amazing bat-to-ball talent. Throw in the fact that, defensively, he has 5 errors on the season, and it’s not hard to believe that Bichette has been one of the worst players in the league this season. And what makes it worse, is he’s making $42 million this season. And, because of the way his contract is structured with two player options over the next two years, he can simply opt-in this offseason, and make $42 million next year as well. What a disaster of a signing he has been.

    The Mercenaries

    I think I speak for the vast majority of Mets fans when I say that we would rather just watch the Mets lose with homegrown guys, rather than all of these hired guns who we are not attached to. David Stearns’ promise was to break everything down, in order to build this team back up better and stronger. Many Mets fans, myself included, came to terms with losing Nimmo, McNeil, and Alonso because the expectations for this 2026 team were so high. I bought in to the “run prevention” motto, and I did what Mets fans always do, I believed. And, what did I get instead? I got an uninspiring team of mercenaries that don’t care about the orange and blue, and have no connection to the city or the fan base. Ask any Met fan on the street, if they would rather have Brandon Nimmo or Marcus Semien. Ask them if they would rather have Jeff McNeil or Luis Robert Jr. Ask any rational person on the planet, if they would rather have Pete Alonso or Jorge Polanco. If the Mets were going to be 22-32 through 54 games, I would have rather done it with my guys.

    The Silver Lining

    There is only one silver lining for Mets fans right now. While it may not feel like it, this 2026 team is in a similar spot to 2024. The OMG Mets of 2024 were 22-33, and nothing was working. An auspicious team-meeting following another embarrassing loss to the Dodgers turned the whole season around. The emergence of Mark Vientos, a resurgent starting pitching staff, the blaring trumpets of Edwin Diaz, the exuberance of Jose Iglesias, and of course, the super-stardom of clubhouse leader Francisco Lindor turned a lost season into a magical one. Will this 2026 team be the same? The odds are against them, but all Mets fans know that if you want to be a true fan, Ya Gotta Believe.

    1. This does not include his first two seasons in the Majors in which he played 5 and 4 games respectively. ↩︎
    2. Vientos had played 1st Base before, but not much in the Major Leagues, and certainly not very well. ↩︎

  • The NCAA Expands March Madness to 76 Teams

    The news broke yesterday that it was official: the NCAA Men’s and Women’s Basketball Tournament will be expanding from 68 teams to 76 teams. It is effective immediately, meaning the next March Madness we see, will feature 76 teams. Throughout this article, I will go over the pros, cons, and the consequences (not all negative) of this move to expand the tournament. But first, I will explain how the expansion will look, and who will play in the extra games.

    To accommodate the extra 8 teams that will be making the tournament, there will now be twelve “play-in” games, up from the “First Four.” These games will be played over the course of Tuesday and Wednesday, six each day, in Dayton, Ohio, the previous host of the “First Four,” and another site to be determined. It has already been announced that the second site will likely be in the Western half of the United States to help with logistics and travel arrangements. The First Round of the tournament will begin, as it always has in recent history, on Thursday. The teams that will play in these games will be the twelve lowest-seeded automatic bids and the last twelve at-large bids. The play-ins featuring the last automatic bids will feature eight teams seeded 16, and four teams seeded 15. The play-ins featuring the the last at-large bids will feature eight teams seeded 12, and four teams seeded 11. That is the official format of the tournament moving forward. Before getting into everything that is wrong with this new format, I want to go over the benefits.

    The obvious beneficiaries of this new format are the Power Conferences. They will undoubtedly be getting some, if not most, of the extra eight at-large spots into the tournament. Last year’s First Four Out were Auburn, Oklahoma, San Diego State, and Indiana. Three of the four teams were from Power Conferences, and San Diego State played in the National Championship game only four years ago in 2023. The true mid-major schools might be left out. With that said, there is one more positive to this change, and it’s probably not one that many will expect.

    In my opinion, the low-major conferences will be the ones that actually benefit the most. The reason for this is the NCAA Tournament Unit System. This is the system for how conferences, not teams, are financially rewarded for their performance in the tournament. Here is how it works:

    • Each conference is awarded 1 unit per team that makes the tournament, therefore each conference is guaranteed at least 1 unit per season.
    • Teams earn additional units by playing in more tournament games. Each win in the tournament guarantees a team another unit. New for this past season, the National Champion receives an additional unit.
    • A unit equates to roughly $350,000 per year paid out over six years
    • Units are paid to the conferences, not the teams, and the conferences are allowed to distribute the money however they want
      • Most conferences split the money evenly among all teams, including those that did not make the tournament.

    The reason the expanded tournament will help low-major teams is because there are more attainable units available. For a low-major conferences such as the Northeast Conference, America East, and the Mid-Eastern Athletic Conference, if they do not play in the First Four, where they would play another low-major team, they would play against a 1-seed and likely lose, only earning their conference 1 unit. With the expansion to 68 teams, and the addition of six play-in games for the bottom 12 automatic bids, these small, and traditionally weak conferences will have the opportunity to earn an additional unit. A perfect example of this is Siena last tournament. Siena led #1 overall seed Duke for the majority of the game, before ultimately losing. Siena did not play in the First Four, but was a 16-seed. This meant that Siena’s conference, the Metro-Atlantic Athletic Conference (MAAC), only received one unit for the 2026 Tournament. In the new expanded tournament, Siena would have played a “play-in game” against another 16 seed, and if they played like they did against Duke, they would likely have won, earning the MAAC two units for the Tournament. Because there are now six play-in games for the lowest automatic bids, there are now six additional “attainable” units for low-major conferences. Attainable is in quotes because while it is possible for 15 and 16 seeds to win their first round games, history shows that it is extremely uncommon. Since the tournament expansion to 64 teams in 1985, 1 and 2-seeds are 307-13 in first round games, for a win percentage of 95.94%. While the mid-major conferences may not be pleased by this expansion, both the power conferences and the low-majors will see this as a good move.

    Now on to the negatives of this new format. This was not something that fans were clamoring for. In a quote to ESPN, John Calipari said, “I don’t know why you mess with something that’s working.” Agreed. The only people calling for this were the SEC and the Big Ten. They argue that their conferences are so superior to everyone else, that it is not fair that some of their teams knock each other out from making the tournament. And while the SEC and the Big Ten are both inarguably strong conferences, neither has had many champions in recent history. Since the tournament expanded to 68 teams in 2011, the SEC and the Big Ten have only combined for three champions!1 This is despite the fact that they routinely have the most teams in the tournament. There are many fans out there who have said that this is nothing other than a way for the NCAA to make more money. The NCAA admitted they will earn a profit on this expansion. However, they argue that the extra revenue will be minimal, and that this is in the best interest of the sport.

    The main issue that I have with the expansion is that it felt incredibly unnecessary. In my opinion, the biggest question I have is simply: “Why?” What is the point of adding eight more at-large teams? There was also no need to skip from 68 to 76 without first going to 72. This would have made much more sense, and would have made the bracket more aesthetically pleasing. There would be eight play-in games, four of them featuring the final eight automatic bids, and four of them featuring the final eight at-large bids. This would lead to play-ins for the 16-seeds and likely the 12-seeds.2 Instead, there are two extra play-ins for 15-seeds and two extra play-ins for 11-seeds.

    One additional thing to consider is the impact this will have on bracket pools and games. Will the ESPN Tournament Challenge choose to exclude twelve games from their competition? If not, then users likely will have less than 48 hours to create and submit their brackets. While this may not be a big deal to most people, it is big deal to me. Previously, I have excluded the First Four games from calculating tournament results. However, moving forward, can I exclude twelve games? My initial reaction is no, and the reasoning is sample size. Now including the “play-in” games, my sample size grows from 63 games per year to 75. This might allow me to make more informed conclusions per year, and try to conclude which models are better than others.

    Does the math support the move to include more at-large teams? Let’s look at the data. Since 2011, there have been 30 First Four at-large teams in the tournament (2 per year). Those 30 teams have won a combined 23 games, averaging less than a win per team. 13 of the 30 teams won their first round game, and 6 of the 30 won two games and made the Sweet Sixteen. 13 out of 30 winning their first round game is a win percentage of 43.33%, which is actually higher than the all-time 11-seed win percentage of 39.7%. Granted, the sample size is significantly smaller, but it is still important to note.

    Going into this, I will fully expecting to write about how awful this expansion is, and how it is killing the game, but there are undeniable benefits for low-major conferences that have to be acknowledged. If I were Charlie Baker, President of the NCAA, I would have pushed for a redistribution of First Four games, rather than expansion. The best part of the expansion is the addition of more attainable units for low-major conferences. Rather than expanding the tournament by allowing eight more at-large teams, keep the tournament at 68, and make all First Four games for the eight lowest seeded teams in the tournament, regardless of automatic or at-large bid type.3 In this format, there would still be four play-in games, but they would all feature 16-seeds. This way, low-major conferences have the opportunity to earn more units and get much more television exposure. Unfortunately, it is incredibly unlikely that the NCAA will ever decided to shrink the tournament, so 76 will have to be the new normal, at least until they decide to expand to 80… or 96… or eventually 128 teams.

    1. 2026: Michigan / 2025: Florida / 2012: Kentucky ↩︎
    2. The seeding of the final eight at-large bids depends on the overall seeding of the automatic qualifiers. ↩︎
    3. It would always end up being the lowest eight automatic bids, due to weakness of the smaller conferences. The only way it would include an at-large team is if a lot of teams in weaker conferences were to put together undefeated, or close to it, seasons similar to what Miami (OH) did in this past season. ↩︎

  • A Review of my March Madness Models

    As many of you know, March Madness is my favorite time of the year. I pour countless hours into inputting data, creating models, and filling out brackets. This year was no different. I set a personal high in models created: 24, and I am happy to report, this was a pretty successful year and I am proud of the results. This article will be a review of my models as well as my general takeaways from following the tournament. Before beginning, below is the rankings of my models. There are two main ways to track a model’s success: correctly predicting games (regardless of the bracket), and accurately predicting the bracket. Each model’s final ranking was an average of their ranking in Total Wins and in Bracket Points.

    The #1 Model for the 2026 March Madness tournament was AFT. This model led the total wins category by correctly predicted 52 out of 63 tournament games (as well as all 4 First Four games, which are not included in this tally). In addition to its success correctly predicting matchups, it finished in third in the bracket points category. It is both encouraging and slightly disappointing that AFT was the best this year, because it is my oldest model and I had high hopes for many of my newer models. Although, it is nice to know that this model has stood the test of time and continues to be successful. It has been above 70th percentile in three of the last four years, and above the 95th percentile in two of the last four years. 52 tournament wins is also a record for any model I have ever created.

    Perhaps unsurprisingly, all of the top three finishers correctly predicted Michigan to win the championship, but what is surprising is that each of the top three models represent three of the four different categories of models:

    • AFT is what I like to call a Power Ranking. It uses a variety of team statistics and gives a single number as its result. When predicting matchups, the team with the higher number is projected to win the game. Power Rankings do not factor in opponent. Other examples of Power Ranking models are TOA Margin, Solidity Score (FF), and ZSCORE Sum.
    • 4 Way Game Score is a Resumé Ranking. This model uses a team’s results in the regular season and gives a single number as its result. This number represents how well a team performed in the regular season. All of the kWins models are also Resumé Rankings.
    • AFT Matchup is a Matchup Model. This type of model uses a variety of team statistics to predict the result in a given game. This type of model does factor in the opponent, and teams’ scores will change based on the opponent. Other examples of Matchup Models are XPoint, FourPoint, and FACTOR.
    • The models COMBO 1.0 and 3.0 are what I call Aggregate Models. These models use the results of other models to predict games. In this case, COMBO 1.0 uses AFT, ZSCORE SUM, and XPoint Standard SOS, to predict which team will win a game. For COMBO 1.0, whichever team is predicted to win in at least 2 of the 3 models is selected to advance. These models are successful because they tend to smooth out oddly predicted results from one model. Other examples of Aggregate Models are BC-Esque and Karn Composite.
    Defense Wins Championships… and Brings Success at Predicting Games

    This is perhaps the biggest takeaway I had from looking at this year’s results. Each of the models that factored in defense outperformed their solely offensive-minded counterparts. For example, XPoint ptSOS Def was the highest of three XPoints models, and the only one that used defense to predict games. The other example of this was FourPoint Def outperforming FourPoint. It is clear that using defense to help predict games gives a more complete picture of how the game is going to turn out. I will definitely be utilizing this moving forward, and ensure that all matchup models have at least some form of defense factored into their prediction.

    Farewell!

    It is time to say goodbye to these models. Whether they underperformed this year, or simply other models do essentially the same thing, these models will not return for next year.

    • ZSCORE Sum. This model will be retired after its fourth season, due to its lack of good results over its career. After performing well in an upset-riddled tournament in 2023, it has put up 3 straight years of under 50th percentile brackets. In its four-year run, it never reached 45 wins. The z-score framework may be back, but with different statistics, and a new name.
    • kWins 50/50 and Gradient. While both of these are solid models, I feel the kWins AL10N is a more encompassing version of them. It factors in the quality of a team’s last 10 games, and adjusts their wins accordingly. Additionally, it uses the same 25/50/75/100 framework for Quadrant 1-4 games as Gradient, and thus makes Gradient unnecessary.
    • Solidity Score (Fraudulent Formula). Unfortunately this model simply underperformed. After using the last 5 years worth of first round upsets to try to predict this year’s, it just fell flat. It is possible that this was due to the relative lack of first round upsets compared to years past. However, it just feels like this was a swing and a miss.
    • XPoint Standard SOS. After years of using this Strength of Schedule metric in my XPoint calculations, I feel that it is time to retire the “Standard SOS” method. This is back-to-back years now that the “ptSOS” method has dominated the Standard method. Thus it is time for me to move on from it.
    Requires Further Research
    • One big takeaway I had from looking at the Final Rankings Table was the massive discrepancy between TOA Margin and TOA Matchup. They use the same statistics, but TOA Margin is a Power Ranking model and TOA Matchup is a Matchup Model. This is the first time that two connected models have had such a wide gap between their results.
    • Another thing that always requires more research is how to properly account for the different strength of schedules. This is likely a never ending issue, as long as there are teams like Siena, High Point, and Miami (OH) that make the tournament, there will be a need to try to, as accurately as possible, adjust season statistics based on strength of schedule.
      • I may try to attack this at the game-by-game level next year, rather than adjust season statistics. This would require far more work and effort, but it might be worth it.
    Conclusion

    This year’s March Madness was spectacular, as it always is. And, as was the case for every prior year, the attempt for a perfect bracket came up as empty as Tyler Tanner’s game-winning half-court heave against Nebraska in the Round of 32. While I may never get a perfect bracket, this year was definitely more encouraging than some of the previous years. 19 out of 24 models finished above the 50th percentile in ESPN. 21 out of 24 models correctly predicted the winner in 45 out of 63 games. Both of these are tremendous improvements. But what remains to be seen, is whether this year was simply a fluke, or the beginning of an upwards trend for the overall success of my models.

  • March Madness Pre-Final Four Update

    Before tomorrow’s Final Four games, I wanted to provide an update on how my models have performed so far. The link to see what each model predicted in every game is below.

    CLICK HERE FOR THE GAME-BY-GAME RECAP FOR EACH MODEL

    Model Tracking Format:

    There are two main ways to track a model’s success: bracket points and total wins.

    • The bracket points format that I use is the ESPN standard format: 10 points for 1st round games, 20 points for 2nd round, 40 points for the Sweet 16 games, 80 for the Elite Eight, 160 for Final Four, and 320 for picking the correct champion. It’s important to track a model’s success using this format because the end goal is to create a perfect bracket.
    • The way I track total wins is by running each game through each model, regardless of whether the matchup was originally predicted to occur in its bracket. A good example of this is the Nebraska-Iowa game in the Sweet 16. The best performing model in this format is AFT. AFT’s bracket had Florida vs Vanderbilt playing against each other in this game, but in order to track each model on the same sample size (63 games or 67 games with First Four included), I have AFT, and all models, predict the matchup that actually occurred. In this case, it was Nebraska vs Iowa.

    Takeaways:

    The best performing model in the bracket format is FourPoint Def. It currently sits at 900 points (97.1st percentile in ESPN). Unfortunately for this model, it predicted Duke to be the champion. Had Duke been able to close out a 19 point lead over Connecticut, this model would have predicted all Final Four teams correctly. If you want to read about Duke’s late-game collapse, you can click here.

    The best performing model in the total wins format is AFT. It currently has correctly predicted 50 of 60 tournament games (54 out of 64 if you include the First Four). I believe this is already a record for any of my models, and with three games to play, it has a chance to get to 53 wins. AFT predicts that Illinois and Michigan will play for the championship, and Michigan will cut down the nets.

    Only six models still have their champion remaining. Three have Michigan winning (AFT, AFT Matchup, and 4 Way Game Score), and three have Arizona winning (FACTOR, kWins tSOR, and ZSCORE Sum).

    The Connecticut victory over Duke was tied for the biggest upset of the tournament according to my models. Only 1 out of 24 models predicted Connecticut to beat Duke. The other biggest upset of the tournament according to my models was Texas A&M beating Saint Mary’s College in the First Round. Interestingly, there has not been a game this tournament in which no models correctly predicted the winner.

    There were 20 games this tournament in which all 24 models unanimously predicted a team to win. Unanimous teams were 20-0. The average margin of victory in these games was 23.1 points. The closest game was Duke vs Siena (71-65), and the biggest blowout, unsurprisingly, was Florida vs Prairie View A&M (114-55).

    Predictions:

    • If you want to see exactly who each model predicted, click the link at the top of this article.
    • Illinois vs Connecticut: 15 models predict Illinois, and 9 models predict Connecticut.
    • Arizona vs Michigan: 13 models predict Michigan, and 11 models predict Arizona.

  • Let’s Talk About the Duke Blue Devils

    This is second article of my newsletter series: “Let’s Talk About” with the first article being “Let’s Talk About Luis Arraez.” These articles will pop up anytime I feel there is something that has to be discussed as soon as it occurs.

    So Let’s Talk About the Duke Blue Devils. Where to begin? We could begin with Duke’s 19 point lead late in the first half. Or, we could begin with a 15 point lead at halftime. We could even begin with Duke’s 11 point lead with 8 minutes left in the game. But, I want to start at the end.

    Obviously blowing all of these leads is horrendous, but its March, and crazy comebacks happen all of the time. Throughout the entire second half, Duke could never quite put the Huskies away, allowing them to chip at their lead every so slowly until Connecticut cut the lead to 2 points with 3:22 remaining. But Duke being the juggernaut that they are (supposed to be), stretched the lead to 70-65 with just under two minutes remaining. Solo Ball made 1 of 2 free throws to make it 70-66 with 1:30 left. Here is error #1 for Duke at the end of this game. Probable Player of the Year, Cameron Boozer just loses control of the ball, Tarris Reed Jr of Connecticut recovers, and it leads an Alex Karaban three-pointer to cut the lead to 1 with 50 seconds left. Major credit goes to Karaban who seems like he never misses big shots. Boozer was able to settle in on the next possession and make a very difficult contested shot in the paint to extend the lead to three (72-69).

    Duke then makes error #2. They foul Silas Demary Jr with ten seconds remaining. Now I know everyone always says to foul up three points, and that is true, but not with ten seconds remaining. Connecticut had no offensive movement, and Demary was attempting to drive in the paint to try to create something. Had Cameron Boozer moved his feet to stay with Demary, he would have forced an uncomfortable situation for the much shorter UConn guard. If Demary doesn’t score and Duke is able to get the rebound, the game is likely over. Teams should foul when up three points only in the last five seconds of the game. This puts the trailing team in a much more difficult position, with purposely missing the second free throw and getting the offseason rebound likely being their best option. Instead, Demary goes to the line. He makes the second of two free throws. The score is 72-70 with 10 seconds left. To most everyone, it should feel like Duke just needs to dot the i’s and cross the t’s. But Duke does the unthinkable: they turn the ball over. Error #3.

    What makes this even more crazy is that Duke didn’t have a five-second call, or fumble the inbound, or get the ball stolen passing back to the inbounder like what most teams turning the ball over against the press do. Duke was able to get the ball to their point guard in the middle of the court. And instead of staying put and allowing them to foul him, Cayden Boozer tries to pass the ball over the top of the two players coming to trap him. The ball ricochets off of Silas Demary’s hands, Braylon Mullins recovers and kicks it ahead to Karaban who finds Mullins trailing the play wide open and he bangs home a 35-foot three-pointer with 0.3 seconds left to win it for Connecticut 73-72.

    Let’s dissect these three mistakes. First, Cameron Boozer does not just “lose control” of the basketball often. He averages 2.5 turnovers per game, and a 14.6% Turnover Rate. Both are impressive numbers for players who has such a high usage rate. Secondly, the Boozer foul on Demary with 10 seconds left. Connecticut was completely and uncharacteristically lost offensively. They had no motion, and all of their players were standing still outside the three point line. Demary was trying to take on Boozer and the Duke defense by himself. At that point in the possession, I probably would not have fouled at all. And finally, the biggest mistakes of all: Cayden Boozer’s turnover. I understand what Cayden was trying to do, if he is able to successfully complete that pass, there’s a chance that time runs out completely, or Duke gets an easy dunk to take a four point lead with less than three seconds remaining. However, you can not take that chance only up two points with less than ten seconds remaining. Simply let Connecticut foul, make at least one of two free throws, and Duke can’t lose in regulation. If Cayden can make both free throws, he’s an 81% free throw shooter, then Duke goes up 74-70, and then the game is over.

    Now I should be fair, both of the Boozer Twins have been exceptional all year long. Especially in the tournament. But, possibly for the first time all year, they showed their inexperience tonight on the biggest stage in the final moments. Cameron with the first turnover and the foul, and Cayden with the second turnover. As I said earlier, both are great players, and Cameron is certainly a top 5 pick in next year’s NBA Draft. Duke would not have gotten to this point of the tournament without them. It was an unfortunate end to the season for the Blue Devils who were the clear #1 team for most of this season. It was also an unfortunate result for most of my models, which you can check out on the 2026 March Madness page.

    One final note: Dan Hurley is an incredible coach, and he made a genius decision by not fouling immediately. He trusted his press defense to force a turnover, and they rewarded him. Braylon Mullins, a freshman himself, hit the biggest shot of his life to send Duke home. And credit again goes to Alex Karaban who gets the final assist, and unselfishly trusted the fate of the season to the freshman Mullins.

    To subscribe to the Let’s Talk About Newsletter, enter your email below!

  • Michael Jordan Makes History Again… This Time in NASCAR

    Yes, this is the same Michael Jordan that won six NBA Championships. What many may not know is that Michael Jordan co-owns the NASCAR team, 23XI (Twenty Three Eleven) with Denny Hamlin, a fantastic current racer for another team. Hamlin is a longtime veteran winning 60 races, although he has never won a championship. The two founded the team in 2020 signing Bubba Wallace to be its only driver starting in 2021. Wallace drives the #23 car, an homage to Jordan’s jersey number. In 2022, 23XI Racing added their second car to the Cup Series: the #45 car, another homage to Jordan’s other, less popular, jersey number that he wore with the Wizards at the end of his career. Kurt Busch and Ty Gibbs both drove the #45 car, before finally bringing in Tyler Reddick for the 2023 season and beyond. And that, is where this story begins.

    Arguably, there is no bigger race in motorsport than the Daytona 500, with respect to the Monaco Grand Prix, Indianapolis 500, and 24 Hours of Le Mans. It is certainly NASCAR’s biggest race of the season, and it is the first race on the calendar. Reddick began the race in 26th position, and finished Stage 1 in 20th place. The end of this race was chaos, but Reddick somehow found his way into the lead and converted for his, and Michael Jordan’s, first Daytona 500 victory. Reddick lead only 1 lap… the last lap.

    The following week at EchoPark Speedway in Atlanta, GA, Reddick was looking to join a small list of only 5 other drivers that have won the first two races of the season. Reddick started far better than he did at Daytona, starting on pole position. He finished in the top 6 in both stages. This race was a trainwreck. Over half of the field was involved in a wreck at some point throughout this race. But in the first overtime, it was Bubba Wallace in 1st, Carson “Hurricane” Hocevar in 2nd, and Tyler Reddick behind Bubba in 3rd. Reddick was pushing Bubba and the two looked to be working well together when all of the sudden Bubba goes high to block Hocevar. Not only did this not work, as Hocevar slid under him to the middle, but it gave Reddick free air on the bottom to gain the lead of the race. A lead that Reddick would not relinquish, as he won back-to-back races to start the season.

    Next up, Circuit of the Americas. Typically a Formula 1 track, Circuit of the Americas (COTA) is a road course. And in today’s NASCAR, that means that Shane Van Gisbergen is the favorite. Going into this race, SVG had won the last 5 road courses, and was looking to tie Jeff Gordon for longest road course win streak in NASCAR history. But there was also another driver looking to make history at COTA: Tyler Reddick. No driver had ever won the first three races of the season in NASCAR history. It appeared unlikely, although Reddick does have a previous victory at this track. Going into qualifying, Van Gisbergen was +120 to win the race, while Reddick was +1600. Even after qualifying, and Reddick took a second straight pole, SVG was still +120 despite qualifying in 13th, though Reddick was up to +400. It was looking more and more likely that one of the two drivers would make history. It was a fun race, filled with lots of Green Flag pit stops and passes, but a caution came out with 19 laps to go. Reddick was in the lead, but SVG was not far behind. It felt like SVG was just taking his time before picking the right spot to make the pass for the lead. About 10 laps after the restart and 10 laps before the finish, SVG had closed in significantly, and was less than a half second behind Reddick, but that proved to be as close as he would get. Reddick just flipped a switch and started to pull away. When the White Flag came out, Reddick was four seconds ahead of Van Gisbergen. Michael Jordan had made his way down to pit lane, and was ready to celebrate with his team. When Reddick crossed the finish line, he became the first driver to win the first three races of the season in NASCAR history, and Michael Jordan was the owner of the first car to win the first three races of the season in NASCAR history.

    Even two and half decades removed from his dominance on the court, Jordan has now brought his dominance to the track. Jordan’s first three-peat took place from 1992-1994, and his second took place from 1996-1998. This three race run by Tyler Reddick has brought Jordan one more three-peat, albeit as a team owner. Reddick now has a commanding 70 point lead in the standings after only three races. It is very possible that he can bring Jordan and 23XI Racing their first Championship. Well, I guess it would be Jordan’s 7th total Championship.

  • A True Golden Goal for the United States

    A True Golden Goal for the United States

    I was not planning on writing an article today. But after watching the Men’s Hockey Gold Medal Game in Milan, there was no way I couldn’t write down my thoughts and emotions. Fair warning, lots of superlatives ahead.

    I will start off by saying that this was the greatest hockey game I have ever seen. Now I will be the first to admit, I do not watch a ton of hockey. I watch occasionally in the regular season, and a good bit in the playoffs. With that being said, I have never seen a game with such speed, physicality, and most of all, emotion. It was clear from the opening puck drop how much this game meant to each team. The game began with a furious blitz by the Americans before the first line change. Canada thwarted the initial attack until 6 minutes in, Matt Boldy made a spectacular play splitting two Canadien defenders while flipping the puck off of the ice to himself, before beating Jordan Binnington to put the Americans on the scoreboard. After that, however, it was all Canada for essentially the next 50 minutes. The ice appeared to be tilted towards towards the American net, but Connor Hellebuyck was fantastic, only allowing one goal, despite a barrage of shots on goal. That lone goal coming on a Cale Makar shot with less than two minutes to go in the 2nd Period.

    The turning point of the game came with around 7 minutes left in the 3rd Period. Quinn Hughes weaved through the Canadien defense and fired a shot on net, but Binnington turned it away. It was the first real scoring opportunity for the Red, White, and Blue in a long time. It led to a Sam Bennett high sticking penalty on future American hero Jack Hughes. The 4-minute double minor penalty gave the USA team an unbelievable opportunity to score before regulation ended, although it would cost Hughes parts of two of his teeth. He left the game briefly before returning with around a minute left in the power play. Hughes would then give a high stick of his own to Bo Horvat, leading to 4-on-4 hockey for 49 seconds, before giving Canada a 5-on-4 power play for over a minute. Hellebuyck and the American defense were gritty and were able to kill the power play and essentially ice the clock to end regulation.

    After regulation, so much was going through my head. Visions of 2010 came to mind, when Sidney Crosby scored the game-winner in the Gold Medal Game, mere seconds after I foolishly asked, “Why does everyone say Sidney Crosby is so good if he never scores?” I thought of the Miracle on Ice, 46 years ago, to the day. And finally, thanks to the broadcast, I relived Quinn Hughes waving off the bench and scoring the game-winner to defeat Sweden in the Quarterfinals only days ago.

    When it comes to 3-on-3 hockey, the game could be over at any moment. I, along with everyone else watching I’m sure, was on the edge of my seat. There might not be a better goal-scoring combination of 2 attackers and 1 defender in the world than the Canadiens’ 1st Line: Connor McDavid, Nathan MacKinnon, and Cale Makar. The two teams traded shots, and possession over the 1st minute, and time seemed to slow down. Both teams had great scoring opportunities only to turned away. But then, as Hellebuyck turned away a Connor McDavid shot, Jack Hughes took the puck, passed to Zach Werenski, who gave it back to Hughes, and Hughes sent it up the ice. Werenski gained control, spotted a trailing Jack Hughes, sent him the puck, and Hughes beat Binnington and found the back of the net. And just like that, the celebration was on. Less than two minutes into Overtime, Jack Hughes had scored the Golden Goal and etched his name into United States Men’s Hockey History.

    The celebration was a beautiful sight to see. Gloves, sticks, and helmets thrown into the air and an army of white and blue jerseys stormed the ice. As the initial celebration began to wind down, we got to witness some really special moments.

    • Connor Hellebuyck and Jack Hughes, the two brightest stars of the game, hugging and sharing a moment of true joy.
    • Zach Werenski, Auston Matthews, and Matthew Tkachuk retrieved the jersey of Johnny Gaudreau, a member of the Men’s USA Hockey Team who tragically died after being struck by a drunk driver in 2024. They then skated around the ice holding his jersey high, as if giving Johnny his own Olympic Gold moment.
    • The Jack Hughes Interview. I truly have no words about this interview. It was perfect from start to finish. The bloodied mouth and broken teeth. The American pride. The humility and love for his teammates. Please, just take the time to go watch the interview.
    • Finally, Werenski and Dylan Larkin went to the stands and brought out Gaudreau’s two children to join them in the team on-ice picture with their gold medals.

    These were just a few of my favorites from the postgame festivities. This game felt different from start to finish, and this group of players was truly special. Not only did they bring home the first Gold Medal in Men’s Olympic Hockey in 46 years, but they did so beating Canada. The very team that had defeated them in the 2010 Olympic Final, and in 2025’s Four Nations event. To say this Canadian team was stacked, is an understatement. And yet, it was the United States who prevailed. The grit, toughness, and passion that this team played with, will never be forgotten.

    1960, 1980, and now 2026. Congratulations to the US Hockey Men’s National Team.

  • The NFL has a WR1 Problem

    No, the problem is not that there are too few or too many Wide Receiver 1’s. The problem is that teams are over-valuing their star wide receivers, and it can set them back for years. I’ve thought this for a long time, but when watching Colin Cowherd this week, he put up a list of the 12 highest-paid wide receivers in Average Salary per Year. For the 2025-2026 season, those 12 highest-paid wide receivers had a combined 0 playoff wins. To take it a step further, only 4 of the 12 made the playoffs this season. George Pickens is looking for a big contract this offseason, but should the Cowboys give it to him? Or, should the Cowboys invest in their defense and offensive line instead of allocating even more Salary Cap space to their Wide Receiver room? He is just one example, but teams across the NFL continuously make the same mistake: spending big on wide receivers.

    According to Spotrac, the Cincinnati Bengals were the highest spenders in the league on their wide receiver room at a $51.4 million cap hit.1 The Bengals Wide Receiver Room accounts for just over 20% of their total spending. The Bengals are probably the best example of why spending big on wide receivers might not be the best idea. The Bengals were 31st in Total Yards Allowed per game, 32nd in Touchdowns Allowed, 32nd in Rushing Yards Allowed Per Game, and 30th in Points Allowed Per Game.2 These are horrific statistics, and one could argue it is a direct result of not having the cap space to spend more money on the defense. The Bengals ranked 20th in Defensive Spending in 2025, but some of the teams that ranked lower, spent less on their defense because many of their best defensive players are still on their rookie contracts, such as the Eagles, Rams, Jaguars, and 49ers. All of these were strong defenses in 2025, despite low defensive spending. Unfortunately, the Bengals did not draft as well as those teams, and thus did not have the same production from players on their rookie contracts.

    I now want to take a different approach on why spending big on a star wide receiver can be detrimental to a team’s future. When looking at the highest paid wide receiver for the teams that made the Super Bowl in the last 8 years, there is a clear and obvious trend:

    The only teams to spend over 10% of their cap on their wide receiver rooms were the 2020 and 2019 Kansas City Chiefs, and they had a Hall of Fame Coach, Quarterback, and arguably Defensive Coordinator.3 Another thing I noticed when creating this table was how teams got better after shedding massive wide receiver contracts, or got worse after taking new ones on. For example, in 2024, the Seattle Seahawks led the NFL in WR Room Cap % at 15.70%. After releasing Tyler Lockett and trading DK Metcalf, the Seahawks brought their WR Room Cap % down to 6.50% on their way to winning the Super Bowl in the 2025-2026. That might be an oversimplification of their tremendous turn-around, but its actually a part of a fascinating trend. After winning the Super Bowl in the 2020-2021 season with the 6th lowest WR Room Cap %, the Tampa Bay Bucs decided to reward their veteran wide receivers. Over the next 5 years, the Bucs were never outside the top 6 in WR Room Spending, and they only have 2 playoff wins to show for it, despite being in one of the worst divisions in the NFL the past 5 years. Here’s another example, the Los Angeles Rams won the Super Bowl in 2021 with a WR Room Cap % of 7.60%. Over the next three years, the Rams spent big: over a 10% WR Room Cap % all three years, getting as high as 15.37% in 2024. Their reward? Missing the playoffs in 2022, losing in the Wild Card Round in 2023, and losing in the Divisional Round in 2024. In 2025, the Rams got under the magical 10% threshold, and went on a run in the playoffs, losing to the Seahawks in the NFC Championship.

    The data is there, teams that spend over 10% on their wide receiver room have a tendency to underperform. There are great wide receivers all over the league that teams can get at a good value. Additionally, there is never a shortage of wide receivers in the draft. If I was an NFL General Manager, I would draft a wide receiver every season, and target veteran receivers who are good leaders in free agency. And I would definitely stay under 10% on wide receiver spending.

    1. While Chase and Higgins combine for more than $51.4 million in Average Annual Value, their cap hit is actually much lower. For the purposes of this article, I will be looking at cap hit. ↩︎
    2. https://www.foxsports.com/articles/nfl/2025-nfl-defense-rankings-team-pass-and-rush-stats ↩︎
    3. 2022 Kansas City Chiefs WR Room Cap % is listed as less than 5% because Spotrac was missing Juju Smith-Schuster’s 2022 contract. It had only 2 receivers listed, and the Cap % was 1.73%. When factoring in JuJu’s contract, the actual Cap % is likely to still be under 5%. ↩︎

  • Let’s Talk About Luis Arraez

    Thirty years ago, a player with a career batting average of .317, with an average of over 600 at bats a season, would have been considered one of the top players in the sport. And yet in today’s era, that same player is continuously overlooked and possibly under-valued. Luis Arraez just signed a one-year deal with the San Francisco Giants for $12 million. That ranks him 8th among 2nd Basemen in Average Annual Value. If we were to consider Arraez as a 1st baseman (the position he has played the most over the last two seasons), he would rank tied for 14th in AAV. Why is a career .300+ hitter not getting paid like one? The simple answer is that Arraez does not do much else besides getting base hits. He’s a below-average defender, he’s slow, and he does not hit the ball hard.

    Let’s start with Arraez’s defense. When he broke into the league in 2019 with the Minnesota Twins, he played mostly 2nd Base, but he was used as a Utility player, also playing games at 3rd Base and Left Field. This continued until 2022, where he was a primary 1st Baseman, but was still utilized in both the infield and outfield. In 2023, he was traded to the Marlins where he played almost exclusively 2nd Base during his year and a half with the team. A little over a month into the 2024 season, he was traded to the San Diego Padres, where he has played 1st Base primarily with a few games at 2nd Base sprinkled in here and there. Many expect Arraez to play 2nd Base for the Giants. So even though Arraez can play multiple positions, and has done so throughout his career, he has never really excelled at any of them. His range is poor, ranking in the bottom 15th percentile in Outs Above Average in every single year of his career.1 Additionally, since his career began in 2019, his Fielding Run Value, a statistic that measures how many runs above or below average a player is, is 4th worst in the MLB, only behind Nick Castellanos, Kyle Schwarber, and Andrew Benintendi. Despite his ability to play multiple positions, he is a defensive liability. As a 2nd Baseman, he would be tied for last with Jonathan India, and as a 1st Baseman, he would be the 8th worst. These are really bad defensive metrics that should only continue to decline with age.

    Now onto his speed and baserunning (in)ability. After his rookie season, in which he ranked in the 52nd Percentile in sprint speed, Arraez has never gotten back above average, and the last three years in particular have all been in the bottom 30th percentile in Major League Baseball. While he does rank slightly above average in his career on taking extra bases when the ball is in play, he is ranked dead last since 2019 in Baserunning Run Value only using stolen base opportunities. And again, these baserunning and sprint speed metrics are expected to decline as Arraez continues to get older.

    Finally, let’s get into Arraez’s bat: his supposed specialty. It is true that he gets a ton of base hits. However, the ball is rarely barreled. Arraez’s hard hit rate has never been above the 8th percentile in the league, and his barrel rate has never been higher than the 11th percentile. His bat speed has been the slowest in the Major Leagues all three years that bat speed has been measured (2023-2025). Yet, Arraez continues to get hits. He is a 3 time batting champ, winning it three straight years with three different teams. He does not whiff at pitches, and he does not strikeout, never being lower than the 99th percentile in either of those statistics. His expected on-base percentage had been in the top 20th percentile every year of his career until 2025 when it fell to 53rd percentile (still above-average).

    So then, let’s get into how valuable is Luis Arraez really? His Baseball Savant page is full of bright reds and blues, signifying the sharp contrast between his strengths and weaknesses. As a hitter, he has provided above average Hitting Run Value for the majority of his career (only negative values in 2021 and 2025). As a fielder, he is a liability, no matter where he plays. And finally, as a baserunner, his Baserunning Value is so low because he steals so few bases despite having the 9th most opportunities in the league since 2019. I would argue that Arraez can still be a valuable piece to many MLB teams. Despite the obvious differences in their hitting style, I would compare Luis Arraez to Kyle Schwarber. Both are below average defenders and runners, despite being excellent hitters, in their own ways. Yet, Schwarber signed a 5-year, $150 million deal this offseason, and Arraez signed a 1-year, $12 million deal, despite Schwarber being 4 years older. Granted, Arraez cannot hit for power the same way that Schwarber can, but when comapring the two players’ last four seasons, the stats are similar:2

    Arraez: .318/.359/.417 , 30 HR , 225 RBI , 115 OPS+ , 10.8 WAR

    Schwarber: .226/.349/.507 , 187 HR , 434 RBI , 134 OPS+ , 11.1 WAR

    Yes, Schwarber obviously hits for power far better than Arraez, but despite having over 150 more home runs, and over 200 more RBI’s, their Wins Above Replacement are essentially equal. Schwarber is making $38 million more than Arraez for the 2026 season, and yet they have both averaged around 2.7 WAR per season. In my opinion, the Giants are getting a lot more value for their money with Arraez than the Phillies are with Schwarber for 2026. But hopefully the Giants will follow the Phillies lead, and simply make Arraez the DH. They will be a better baseball team because of it.

    1. Baseball Savant: https://baseballsavant.mlb.com/savant-player/luis-arraez-650333?stats=statcast-r-hitting-mlb ↩︎
    2. Baseball Reference: https://www.sports-reference.com/stathead/baseball/versus-finder.fcgi?request=1&seasons_type=perchoice&player_id1=arraez000lui&p1yrfrom=2022&p1yrto=2025&player_id2=schwar001kyl&p2yrfrom=2022&p2yrto=2025 ↩︎