Hi. I'm Harshil. I'm new here. I'm currently a junior at Wake Forest University double-majoring in Economics and Philosophy, which means I'm learning to ask big questions about why things cost money and whether anything truly exists — which is useful preparation, as it turns out, for watching the Orioles bullpen in 2025 (and seemingly 2026). I grew up in Clarksburg, Maryland, going to Camden Yards with a regularity that probably should have been classified as a dependency. I have never hit above .200 in Little League, but I have spent enough time on FanGraphs to be in their top 3% of visitors for 2025. I'm also studying to be a Chartered Financial Analyst, which has absolutely nothing to do with baseball but does explain why I look at a batting average and instinctively want to annualize it. Zach was kind enough to let me write something. Here's what I noticed.
The Birdland Metrics Core Player Benchmarks dashboard tracks ten Orioles players across thirty individual statistical benchmarks, updated daily. If you've ever wanted a daily reason to feel something about baseball before noon, this is your tool. Let's walk through those stats. All of them.
The Batter Stats
xWOBA (Expected Weighted On-base Average): Essentially wOBA, but based on how hard and at what angle you hit the ball and not what actually happened to it afterward. Hit a 103-mph laser directly at the third baseman? Your wOBA gets nothing. Your xwOBA gets credit. It's the stat that separates a hitter who is struggling from a hitter who is struggling and also getting robbed. When xwOBA runs significantly higher than wOBA, the results are probably about to get better. When they match, what you're seeing is real.
OBP (On-Base Percentage): How often a batter reaches base by any means. Hits, walks, hit-by-pitches. OBP is what happens when you let batting average grow up and get a real job. A .340 OBP is the Birdland Metrics benchmark for Adley Rutschman, and for good reason: the most valuable thing a hitter can do is not make an out.
SLG (Slugging Percentage): Total bases divided by at-bats. A single is worth one base, a double two, and so on. The problem is that SLG ignores walks entirely. A player who walks 100 times gets zero credit. SLG has strong opinions about power and no opinions about patience.
K% (Strikeout Rate): Strikeouts divided by plate appearances. The beauty of K% is that it stabilizes fast — after just 60 plate appearances, it starts telling you something real. This means K% is one of the first stats worth trusting in April. League average is around 22%.
BB% (Walk Rate): Walks divided by plate appearances. Takes a bit longer to stabilize (about 120 PA) but it measures the single most repeatable offensive skill a hitter has: the discipline to not swing at garbage. Adley's benchmark is 11%, and he's sitting at 17.6% through the first five games.
Barrel%, Hard Hit%, Exit Velocity, and Launch Angle: These are the Statcast metrics. Sort of biomechanical X-rays of a player's swing. A "barrel" is a batted ball struck at an ideal combination of exit velocity (95+ mph) and launch angle (roughly 26–30 degrees), which is the sweet spot where batting average exceeds .500 and slugging exceeds 1.500. Hard Hit% measures how often a player hits the ball at 95 mph or harder. Exit Velocity is the average speed off the bat. Launch Angle is the vertical angle at which the ball leaves the bat. Together, they tell you how well a hitter is hitting the ball regardless of what happens afterward. They're the quality-of-contact stats, and they stabilize relatively quickly, about 50 balls in play for EV and barrel rate. These are what you should look at in April.
The Pitcher Stats
ERA (Earned Run Average): Earned runs allowed per nine innings. A pitcher's April ERA is roughly as predictive as a fortune cookie.
FIP (Fielding Independent Pitching): An ERA-like metric built only from strikeouts, walks, hit-by-pitches, and home runs — the outcomes a pitcher actually controls. FIP removes defense and luck. The Birdland Metrics dashboard tracks both ERA and FIP for Shane Baz specifically because the gap between them tells you whether his results are real. If ERA is 4.50 and FIP is 3.20, the pitcher is better than his results suggest. If it's the reverse, duck.
WHIP (Walks + Hits per Inning Pitched): Baserunners per inning. Simple and intuitive but slightly incomplete (it counts a bloop single the same as a line-drive double), but useful as a quick-glance traffic indicator: how many runners is this pitcher putting on? A WHIP of 1.15 or lower throughout a full season is considered elite.
K/9, BB/9, and HR/9: Strikeouts, walks, and home runs per nine innings. K/9 tells you how dominant a pitcher is. BB/9 tells you how much he's struggling with the strike zone. HR/9 is the single most volatile pitching stat. A pitcher can give up zero home runs for a month and then surrender three in an inning on a windy Tuesday at Wrigley. Kyle Bradish's K/9 benchmark is 9.5, and the fact that he posted 7.7 in his first start is the kind of thing worth watching closely.
SV (Saves): The dashboard tracks saves for Ryan Helsley, because when you trade for a closer, you want him to close.
How the Projection Model Works
The Birdland Metrics projections use a Marcel + Statcast model. Marcel is named after Marcel the Monkey, because the system is supposed to be so simple a monkey could run it. It weights the last three seasons in a 5/4/3 ratio: the most recent year gets about 42% of the weight, two years ago 33%, and three years ago 25%. Within each season, stats are further weighted by plate appearances, so a season where a player had 600 PA counts more than one with 300 PA.
After the Marcel baseline, the model layers in Statcast adjustments for batters: exit velocity, barrel rate, and hard-hit rate each nudge the projection. If a hitter's exit velocity is well above league average (88 mph), his wOBA projection gets a small bump, capped at plus-or-minus .030 wOBA. Barrel rate adjustments flow through to home run projections. These adjustments are batters-only. Pitcher batted-ball data is too noisy at the individual level.
Then age curves: hitters peak at 25–27, pitchers at 25–26 and decline faster. Finally, everything gets regressed to the mean based on sample-size reliability. A player with 600 PA of data gets a projection that's mostly his own stats; a player with 200 PA gets pulled harder toward league average.
How to Actually Read the Dashboard
Look at Statcast numbers first. Exit velocity, barrel rate, and hard-hit rate stabilize fastest. These are the earliest signals that a player's underlying quality has changed. Look at K% next (60 PA to stabilize). Then BB% (120 PA).
The benchmark counter at the top of the dashboard (currently reading 11 of 30 met) is your season pulse check.
When are you allowed to panic? Not until June. Seriously. If a player is missing their benchmarks by mid-June with 200+ plate appearances, then it's time for a conversation. Until then, stay put.
How frequently does it update? The Birdland Metrics dashboard updates daily. As of this morning, it reads 11 of 30 benchmarks met. Rutschman is clearing all three of his. Bradish is clearing none. The season is seven games old. It is, in every measurable way, too early to know anything.
And yet here we are, already deep in a spreadsheet, already worried about exit velocity and launch angles and the precise velocity of a sinker because this is what we do. This is what the stats are for. Not to predict the future, but to watch the present more carefully than anyone else, and to know, in the middle of all the noise, which numbers to trust.
The Orioles are 3-4. Adley is hitting the ball harder than he has in two years. Bradish's slider is still filthy.
Check back tomorrow. The numbers will have changed. That's the beauty of it.
