The Barrel Ledger: Projecting Barrels for 2025

Author

Oliver Chang

Published

March 24, 2026

Introduction

Somewhere in Arizona and Florida, a groundskeeper wakes up every morning to tend to the fields, ensuring that the grass is perfectly manicured and the dirt is just right. When dawn turns to day, that field becomes a stage. Every swing, throw, and catch is meticulously recorded, added to the annals of baseball history.

As winter gives way to spring, the baseball world is abuzz with anticipation. Talk of over-unders, records, and player performance fills the air. But amid the chatter, one question looms: who will add the most barrels in 2026? A Barrel is a statcast metric that measures the quality of a batted ball, specifically those that are hit with a combination of exit velocity and launch angle that typically results in a high batting average and slugging percentage. Barrels are often used as a key indicator of a player’s power-hitting ability. By focusing on the exit velocity and launch angle, barrels try to isolate a player’s ability to make solid contact with the ball.

In this article, we investigate which MLB teams added and removed the most barrels from their rosters. However, a zealous statistician might wonder how we quantify “barrels” added/removed. We could just look at the 2025 season, but that would be too simplistic. We want to predict barrels for 2026, which means we need to look at the past few seasons for a better estimate. We will use a weighted average of the past three seasons (2025, 2024, and 2023) to project barrels for 2026. If a team adds an international free agent, we use their home run data to estimate their barrels.

The Data: Scraping, Cleaning, and Projecting

The data was collected from Fan Graphs. Specifically, we look at the offseason tracker for each team. For simplicity, we only consider free agents, trades, and prospective minor leaguers who are expected to make the major league roster.

Once we identify a key player, we grab their Barrel/PA data from Baseball Savant; we collect each player’s Barrel/PA for the past three seasons (2025, 2024, and 2023). If a player does not have a Barrel/PA for a given season, we just treat it as zero for the purpose of our weighted average calculation. In our dataset, we denote each player as an addition or subtraction for a team. Lastly, we get the project PA for each player from Fan Graph’s Zip projections.

Here’s what a data row looks like:

Code 
import pandas as pd
import numpy as np

# Set display options to show all columns and rows
pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', None)

# Read the CSV file
df = pd.read_csv("the-barrel-ledger.csv")
df[4:5]
Table 1
Team Name Transaction Type Projected PA Barrel/PA 2025 Barrel/PA 2024 Barrel/PA 2023 Team Change HR 2025 HR 2024 HR 2023
4 SDP Nick Castellanos Free Agent 357 5.4 5.8 6.9 Addition NaN NaN NaN

Sample Row: Nick Castellanos added to the Padres

Barrels in 2025 are weighted by 0.5, barrels in 2024 are weighted by 0.35, and barrels in 2023 are weighted by 0.15. If a player does not have Barrel/PA data, we estimate their barrels based on their home runs, using a rough estimate of 1.3 barrels per home run. For international players, we apply a 60% discount to the estimated barrels from home runs to account for the uncertainty in projecting their performance in MLB.

Thus, a player’s projected barrels are \text{Projected Barrels} = \text{Projected PA} \times (0.5 \times \text{Barrel/PA 2025} + \\ 0.35 \times \text{Barrel/PA 2024} + \\ 0.15 \times \text{Barrel/PA 2023}).

If Barrel/PA data is not available, then \text{Projected Barrels} = \text{Projected PA} \times 1.3 \times (0.5 \times \text{HR 2025} + \\ 0.35 \times \text{HR 2024} + \\ 0.15 \times \text{HR 2023}).

For international players, we apply a 60% discount to the estimated barrels from home runs, so \text{Projected Barrels} = \text{Projected PA} \times 1.3 \times (0.5 \times \text{HR 2025} + \\ 0.35 \times \text{HR 2024} + \\ 0.15 \times \text{HR 2023}) \times 0.4.

Let’s project Pete Alonso’s barrels for 2026, for example.

Code 
def compute_projected_barrels(row):
    """
    Compute projected barrels for a player.
    - If Barrel/PA data exists, use weighted average of available years (2025: 0.5, 2024: 0.35, 2023: 0.15)
    - Otherwise, estimate using HR data (assuming ~1.3 barrels per HR as a rough estimate), then apply 60% discount for international players
    """
    weights = {'Barrel/PA 2025': 0.5, 'Barrel/PA 2024': 0.35, 'Barrel/PA 2023': 0.15}
    
    barrel_pa_values = []
    barrel_pa_weights = []
    for col, weight in weights.items():
        if pd.notna(row[col]):
            barrel_pa_values.append(row[col])
            barrel_pa_weights.append(weight)
        else:
            barrel_pa_values.append(0)
            barrel_pa_weights.append(0)
    
    if sum(barrel_pa_weights) > 0:
        total_weight = sum(barrel_pa_weights)
        normalized_weights = [w / total_weight for w in barrel_pa_weights]
        weighted_barrel_pa = sum(v * w for v, w in zip(barrel_pa_values, normalized_weights))
        projected_barrels = row['Projected PA'] * (weighted_barrel_pa / 100)
    else:
        hr_cols = ['HR 2025', 'HR 2024', 'HR 2023']
        hr_weights_list = [0.5, 0.35, 0.15]
        
        hr_values = []
        hr_weights = []
        for col, weight in zip(hr_cols, hr_weights_list):
            if pd.notna(row[col]):
                hr_values.append(row[col])
                hr_weights.append(weight)
        if hr_values:
            total_weight = sum(hr_weights)
            normalized_weights = [w / total_weight for w in hr_weights]
            weighted_avg_hr = sum(v * w for v, w in zip(hr_values, normalized_weights))
            projected_barrels = weighted_avg_hr * 1.3
            projected_barrels = projected_barrels * 0.4
        else:
            projected_barrels = np.nan
    
    return projected_barrels

df['Projected Barrels'] = df.apply(compute_projected_barrels, axis=1)
result_df = df[['Team', 'Name', 'Projected PA', 'Barrel/PA 2025', 'Barrel/PA 2024', 'Barrel/PA 2023', 'HR 2025', 'HR 2024', 'HR 2023', 'Projected Barrels']]

result_df = result_df.sort_values(by='Projected Barrels', ascending=False)
result_df[0:1][["Name", "Projected Barrels"]]
Table 2
Name Projected Barrels
70 Pete Alonso 72.07585

Sample Row: Pete Alonso projected barrels

72 barrels for Pete Alonso. A reasonable projection since his barrels for 2025, 2024, and 2023 were 89, 58, and 62 respectively.

The Ledger: Who Added Most Barrels?

Now that we have projected barrels for each player, we can calculate the net barrels added or removed for each team. Net barrels is as follows:

\text{Net Barrels} = \text{Incoming Player Brl/PA} \times \text{Projected PA} \\ - \text{Departing Player Brl/PA} \times \text{Projected PA}.

Here’s the list of MLB teams who added the most barrels.

Code 
# Calculate net barrels by team
# Add projected barrels for incoming players, subtract for departing players
from IPython.display import Markdown
from tabulate import tabulate

net_barrels_by_team = []

for team in df['Team'].unique():
    team_df = df[df['Team'] == team]
    
    # Sum projected barrels for additions (incoming players)
    additions = round(team_df[team_df['Team Change'] == 'Addition']['Projected Barrels'].sum(), 2)
    
    # Sum projected barrels for subtractions (departing players)
    subtractions = team_df[team_df['Team Change'] == 'Subtraction']['Projected Barrels'].sum()
    
    # Calculate net barrels
    net_barrels = additions - subtractions
    
    net_barrels_by_team.append({
        'Team': team,
        'Additions (Barrels)': round(additions, 2),
        'Subtractions (Barrels)': round(subtractions, 2),
        'Net Barrels': round(net_barrels, 2)
    })

# Create a dataframe with net barrels
net_df = pd.DataFrame(net_barrels_by_team).sort_values('Additions (Barrels)', ascending=False).reset_index(drop=True)
net_df.index = net_df.index + 1
Markdown(tabulate(net_df[["Team", "Additions (Barrels)"]]))
Table 3
1 NYM 146.66
2 BAL 142.6
3 PIT 125.6
4 COL 80.26
5 CIN 78.3
6 MIN 74.89
7 TOR 67.77
8 BOS 67.32
9 CHW 61.58
10 TEX 56.22
11 MIA 55.07
12 TBR 53.39
13 ATL 52.89
14 LAD 51.53
15 SDP 51.49
16 SEA 50.79
17 PHI 50.36
18 SFG 39.37
19 CHC 39.33
20 LAA 37.19
21 ARI 34.86
22 MIL 32.48
23 ATH 26.15
24 KCR 25.32
25 HOU 13.5
26 WSN 7.38
27 NYY 0.36
28 STL 0
29 CLE 0
30 DET 0

Net Barrels Added/Removed by Team

The New York Mets lead MLB with 147 barrels added. While they lose Pete Alonso, they add Bo Bichette, Jorge Polanco, and Marcus Semien, among others. Not surprisingly, the Baltimore Orioles trail in second with 142 barrels. The Orioles add Pete Alonso! Overlooked amid the excitement, Taylor Ward also joins the Orioles from the Angels. The Pittsburgh Pirates round out the top three with 125 barrels. They are going to be an interesting team to watch in 2026. Marcel Ozuna would give the Pirates much-needed plate discipline and power. Perhaps he can instill more plate discipline in Cruz’s approach to at-bats.

Who net the most barrels?

More importantly, who net the most barrels?

Code 
# Create a dataframe with net barrels
net_df = pd.DataFrame(net_barrels_by_team).sort_values('Net Barrels', ascending=False).reset_index(drop=True)
net_df.index = net_df.index + 1
net_df['Net Barrels'] = net_df['Net Barrels'].map(lambda x: f"{x:8.2f}")
Markdown(tabulate(net_df[["Team", "Net Barrels"]]))
Table 4
1 BAL 126.41
2 PIT 87.94
3 COL 68.22
4 TOR 50.03
5 CIN 43.48
6 MIN 42.93
7 MIA 28.27
8 LAD 25.26
9 CHW 23.19
10 SFG 13.94
11 ARI 9.6
12 SDP 9.23
13 ATH 8.72
14 ATL 7.31
15 BOS 7.18
16 NYY 0.36
17 CLE -1.23
18 PHI -8.65
19 DET -16.81
20 LAA -29.9
21 KCR -30.59
22 MIL -31.74
23 WSN -36.31
24 SEA -41.88
25 NYM -45.88
26 TBR -48.05
27 TEX -52.75
28 HOU -58.73
29 CHC -82.45
30 STL -98.15

Net Barrels Lost by Team

The O’s net the most barrels this offseason, garnering 126. That’s a huge boost for a team that had a poor 2025 season. The biggest additions are Pete Alonso and Taylor Ward; those two will be a fearsome pair in the lineup. More importantly, the players they lose include Gary Sanchez, Jorge Mateo, and Alex Jackson. All of that said, the Orioles should be a formidable team in 2026.

Following the Orioles are the Pittsburgh Pirates with 88 net barrels added. In addition to Ozuna, Ryan O’Hearn joins the Pirates. They lose Andrew McCutchen and Ke’Bryan Hayes, but, quite frankly, their heyday is behind them. Expect the Pirates to be in the mix for the NL Central two or three spot.

Before discussing the Colorado Rockies, it should be noted that they add Paul DePodesta to their front office staff as President of Baseball Operations. DePodesta, portrayed by Jonah Hill in the movie Moneyball, is a pioneer in the use of sabermetrics in baseball. The Rockies, with their 43 wins last season, offer a clean opportunity; a chance for DePodesta to implement his novel vision for the team. With that, the Rockies add Willi Castro, Eduoard Julien, and Jack McCarthy, among others. These players are not household names, but, in the aggregate, they roughly add 80 barrels. It’s the Moneyball philosophy in action. If their main loss is Thairo Estrada, from their embarrassing season, then the Rockies should be in a much better position in 2026.

The figure below shows the net barrels added or removed by each team. The blue bars indicate net gains, while the red bars indicate net losses. The teams are sorted by net barrels, with the team that added the most barrels at the top and the team that lost the most barrels at the bottom.

Code 
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches

plot_df = pd.DataFrame(net_barrels_by_team).sort_values('Net Barrels', ascending=True).reset_index(drop=True)

colors = ['#d73027' if x < 0 else '#4575b4' for x in plot_df['Net Barrels']]

fig, ax = plt.subplots(figsize=(10, 9))

bars = ax.barh(plot_df['Team'], plot_df['Net Barrels'], color=colors, edgecolor='white', height=0.7)

for bar, val in zip(bars, plot_df['Net Barrels']):
    offset = 1.5
    ax.text(
        val + (offset if val >= 0 else -offset),
        bar.get_y() + bar.get_height() / 2,
        f'{val:+.0f}',
        va='center',
        ha='left' if val >= 0 else 'right',
        fontsize=9,
        color='#333333'
    )

ax.axvline(0, color='#333333', linewidth=0.8)
ax.set_xlabel('Net Barrels', fontsize=12)
ax.set_title('Net Barrel Changes by Team — 2026 Offseason', fontsize=14, fontweight='bold', pad=15)

gain_patch = mpatches.Patch(color='#4575b4', label='Net Gain')
loss_patch = mpatches.Patch(color='#d73027', label='Net Loss')
ax.legend(handles=[gain_patch, loss_patch], loc='lower right', fontsize=10)

ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.tick_params(axis='y', labelsize=10)
ax.tick_params(axis='x', labelsize=10)

plt.tight_layout()
plt.show()
Horizontal diverging bar chart showing net barrel changes by MLB team for the 2026 offseason. Blue bars indicate net gains; red bars indicate net losses.
Figure 1: Net Barrel Changes by Team - 2026 Offseason

Conclusion

Barrels are not the whole story. A team can lead the league in barrels and still lose 90 games if the rotation crumbles or the bullpen hemorrhages leads. What this ledger offers is a narrow but useful lens: which organizations made deliberate, quantifiable investments in hard contact?

By that measure, three teams stand out. The Orioles are the clearest winner, pairing Pete Alonso’s prodigious barrel rate with a roster that shed dead weight. After a dismal 2025, Baltimore has the ingredients for a genuine bounce-back. The Pirates are the most intriguing dark horse; Ozuna is a legitimate middle-of-the-order threat, and if Cruz matures at the plate, Pittsburgh could surprise the NL Central. The Rockies are the longest of long shots, but DePodesta’s arrival signals that the organization is finally thinking rigorously about roster construction. Forty-three wins is a low bar. Clearing it will not require much.

On the other side of the ledger, teams that shed barrels without replenishing them face an uphill climb. Power is difficult to manufacture mid-season. The teams in the red on the chart above have bet that pitching, defense, or depth will compensate. Some will be right. Most will not.

Spring training will soon give way to the real thing. The groundskeeper’s work will be done, and the numbers will start to speak for themselves. Check back in October.

Note

Past articles:

Github: - Running on Numbers