Frequently Asked Questions

Are MLB home run picks actually accurate?

Home runs are rare events, so even the best matchup in baseball caps around a 15–18% single-game probability. That means the top daily home run pick will miss 80%+ of the time by design, not because the model is wrong. Accuracy only shows up over hundreds of predictions. We publish the receipts openly on our calibration page — predicted probabilities vs. observed outcomes, updated nightly.

Is The Dinger Almanac gambling advice?

No. We publish statistical forecasts — probabilities derived from public baseball data. We don’t sell picks, don’t guarantee outcomes, and don’t tell you what to wager. Any use you make of the numbers is your call. We call this stats analysis, not betting advice.

How is the home run probability calculated?

For every batter-pitcher-park matchup on the day’s slate, we estimate a per-plate-appearance home run rate by combining the batter’s HR rate, pitcher HR/9 allowed, Statcast barrel percentages, park factor by batter handedness, weather (temperature and wind), career batter-vs-pitcher history, pitch-arsenal matchups, and times-through-the-order effects. Each input is shrunk toward league average by a Bayesian prior sized to how quickly that stat stabilizes, then converted to a game probability via 1 − (1 − p)^expectedPA.

Where does the data come from?

MLB Stats API for schedules, rosters, lineups, weather, and probable pitchers. Baseball Savant for Statcast quality metrics — barrel rate, expected batting average, expected slugging, pitch-arsenal data. Both sources are public. The pipeline refreshes once each morning and publishes the top ten looks of the day before first pitch.

What’s the best MLB ballpark for home runs?

Coors Field in Denver is the clear leader on HR factor for both left- and right-handed batters, driven by thin air at a mile of elevation. Great American Ball Park (Cincinnati), Yankee Stadium (right-handed pull-side porch), and Oriole Park at Camden Yards also play HR-friendly. On the opposite end, Petco Park (San Diego), Oracle Park (San Francisco), and T-Mobile Park (Seattle) suppress home runs. Park factors used in our model are handedness-specific and updated based on recent park data.

Why do you publish a daily top ten?

Because that’s the honest ranking size. Everyone who reads the site should see both the strongest looks and how good the 10th-best is — usually not far behind the 3rd. Curating more would be cherry-picking; fewer would hide the meaningful range of probabilities. The top ten is the natural slice of the slate where the model has a discernible edge over random.

How does the strikeout probability model work?

Strikeout rate is the stickiest hitting stat in baseball — it stabilizes in about 60 plate appearances, which means we can trust it with lighter priors than HR or hit rates. The K model blends pitcher K/9 (shrunk by innings pitched), batter K% (shrunk by PA), recent K rate over the last 15 games, and heavily shrunk career batter-vs-pitcher K history. Park, weather, and handedness splits don’t meaningfully move K rates, so they’re omitted. Final probability is capped between 5% and 85%.

What’s the difference between your hit, HR, and strikeout probabilities?

All three estimate the chance of at least one of the relevant event in the game: one hit, one home run, one strikeout. Hit probabilities typically range from 35% to 85%, reflecting that even bad matchups get multiple plate appearances. HR probabilities sit between 1% and 20% because home runs are rare even for elite sluggers. Strikeout probabilities run 5% to 85% and depend more on the pitcher than the batter. We show all three on every forecast so you can see the full picture of what the matchup expects.

Can I unsubscribe from the daily email?

Yes, every email has an unsubscribe link in the footer. One click and you’re out — no confirmation step, no "are you sure." We don’t resend to verified addresses and we never sell or share subscriber data.

Question we didn't answer? Email hello@thedingeralmanac.com and we'll add it here.