Major League Baseball

MLB Analytics & Game Predictions

A daily MLB board built around the starting pitcher: AI game breakdowns, market-derived win probabilities, run-line and total context, player prop lines and a graded record for the whole season.

Season note. The MLB regular season runs from late March to September, followed by the postseason and the World Series in October. Times on this page are UTC.

MLB Game Analysis

The starting pitcher is the single biggest input to a baseball line, so every MLB breakdown begins with both probable starters and reads them against each team’s recent form and its run scoring and pitching numbers, then the market: the read explains the moneyline and run line in terms of those inputs and the head-to-head history inside the series. Bullpen depth and the ballpark are part of how an analyst thinks about a baseball game, but they are not data Sportslyx ingests, and the breakdown does not claim otherwise. Where the public feed provides them, player prop lines are analysed alongside the game.

What the breakdown reads

  • probable starting pitchers
  • recent form
  • team run scoring and pitching statistics
  • market lines and movement
  • head-to-head history

Markets covered

  • moneyline
  • run line
  • game total
  • player props

Each breakdown carries a data-quality tier — rich, partial or weak — based on how much of the list on the left the feed actually delivered for the game. The tier is passed to the language model so the write-up hedges when the package is thin, and it is the first thing to check when a read looks off. Read more about how the AI analysis is produced.

AI MLB Predictions

The win probability on every MLB game page is derived from the market, not from the language model. The prices are converted into implied probabilities, the bookmaker’s margin is removed so the set sums to 100%, and the result is shown as each side’s fair win chance.

Baseball moneylines are short and change with the confirmed starter, so the probability is tied to the pitching matchup the market is pricing at refresh time. A late starter change moves both the price and the number on the page.

What the AI adds is the explanation, not the number. The breakdown reads the same data package the statistics section shows — form, availability, head-to-head, rest and the line, as far as the baseball feed provides them — and writes out why the market sits where it does and what it may be underweighting. A data-quality tier is attached — rich, partial or weak, depending on how much of that package the feed delivered — and the read is published before the game starts. It is graded from the official result afterwards, and nothing is edited after the fact. AI breakdowns are pre-generated for the upcoming slate and cached; the win probability on the card is derived from the current market line. The write-up is not rewritten when the line moves.

A probability is a statistical estimate. A 65% favourite is expected to lose roughly one game in three; a month in which it loses more than that is not evidence that the number was wrong, and a month in which it loses fewer is not evidence that it was right. The prediction-tracking table further down this page and the Pick Tracker exist so that the method can be judged on volume, not on memory.

The full board — every game on the slate with its probability and status — lives on the MLB predictions page.

MLB Statistics

The MLB statistics section is organised around the categories the breakdown actually reads. Every number comes from the public scoreboard and box-score feeds and the per-game data package built on game day; where a category is not provided for a competition, the page says so rather than showing a blank.

Runs scored and allowed

Run differential over the season and the recent window — the team-level number the moneyline is anchored to.

Starting pitching ERA and FIP

Earned-run average next to fielding-independent pitching for each probable starter, so a run of good luck is not mistaken for a run of good pitching.

Team OPS

On-base plus slugging for the team as a whole — the one-number summary of run production the breakdown reads against the opposing starter’s strikeout and walk profile. Shown for the season and the recent window, with the caveat that a hot fortnight at the plate changes the market more than it changes the hitters.

Recent form

Results over the last ten games, shown as context rather than as a predictor — a hot streak changes the market more than it changes the team.

Head-to-head record

Results between the two teams this season and inside the current series, shown because the same two teams meet on consecutive days with the same run-scoring profiles. Read as context — a sweep says less than the pitchers who produced it — and never as a predictor on its own.

MLB Matchups

A 162-game schedule makes recent form noisy, so the breakdown reads it through the pitcher rather than the team: a hot week means less than a starter’s last three outings. Head-to-head history inside a series is useful because the same two teams meet on consecutive days with the same run-scoring profiles, and rest is read from the schedule at the team level — bullpen usage is not part of the data the breakdown receives.

Form, head-to-head and rest are inputs to the breakdown, not the verdict. The market line is the anchor, and the matchup context is what the analysis uses to explain agreement or disagreement with it — a read that contradicts the line has to say which of these inputs the market appears to be missing, and by how much.

Every game page shows the matchup view in full: the participants’ recent results, the head-to-head record where one exists, the schedule context and the public line, drawn from the same data package the AI reads. Start from the MLB games list or from today’s board below.

MLB Team Performance

Team profiles hold the identity — crest, abbreviation and display name — and refresh daily. Record, runs scored and allowed, recent results and the probable starter are read on the game page, where the per-game package puts them beside the opponent’s numbers and the moneyline for that day.

Team profiles are rebuilt from the schedule feed every day, but they hold identity, not statistics: the crest, the abbreviation and the display name that keep a team consistent across game pages, the pick tracker and the breakdown. The performance numbers — record, recent results, the next scheduled game and the market’s number for it — are assembled on each game page from the per-game data package, because they only mean something next to an opponent and a date.

Two things the performance view does not do. It does not project a team’s next result; that is the breakdown’s job, and it is graded. And it does not backfill history the feed never delivered: a team with a short record is shown with a short record. The MLB team directory lists every team on the board with its profile, and each card opens that team’s upcoming games.

MLB Player Statistics

MLB is the only sport where Sportslyx covers player props: pitcher strikeouts, outs recorded, walks and earned runs, hits allowed; batter hits, walks, runs and RBIs. Lines come from the public core feed, and each prop read is graded from the official box score against a half-point line, so it never pushes; a pitcher scratched before first pitch is voided rather than counted as a loss.

Pitcher strikeouts, outs and earned runs

Strikeouts, outs recorded and earned runs per start for each probable starter — three of the pitcher props the public core feed prices and Sportslyx grades from the official box score. Per-outing rates sit next to the season number so a starter whose strikeout rate has moved in the last three outings is not read from a stale average. A starter scratched before first pitch is voided, not counted.

Hits and walks allowed

Hits and walks allowed per start for each probable starter, read together as the command side of the profile: a pitcher who misses bats but issues free passes is a different proposition from one who pitches to contact. Both are priced as props on the public feed and graded from the box score against a half-point line, so neither can push.

Batter hits, walks, runs and RBIs

Per-game production for regular hitters — hits, walks, runs scored and runs batted in — the four batter props the public feed prices for MLB. Read against the opposing starter’s profile and graded from the official box score; a hitter who did not appear is voided rather than counted as a loss.

Prop grading rule, stated once: a prop is settled from the official box score against the published half-point line, so it cannot push. A player who did not take part is voided — removed from the record, not counted as a loss — because a line on a player who never played was never a read.

MLB Prediction Tracking

Published MLB reads settled in each of the last 3 calendar months (UTC). The MLB moneyline read is graded on the final score. Player prop reads are graded from the official box score; a player who did not appear is voided.

MLB prediction record by month
MonthSettledW-L-PHit rateROI
September 202610173-28-072%-3.5%
August 2026380301-79-079%+1.8%
July 2026340268-65-780%-2.2%

Hit rate is wins divided by wins plus losses; pushes are excluded. ROI is measured at a flat stake per read, so one large result cannot dominate a month. 3 months is a small sample — read it as a window into the method, not as a forecast of the next month — and remember that past results do not predict future results.

Today’s MLB Games

Every MLB game on today’s board (UTC) in start-time order, with the public spread, total and moneyline where the feed provides them. The MLB regular season runs from late March to September, followed by the postseason and the World Series in October.

MLB Analytics FAQ

Why does the MLB analysis focus on starting pitchers?

The probable starter is the single biggest input to a baseball moneyline. Sportslyx reads both starters and both teams’ recent form before producing a win probability.

Does Sportslyx cover MLB player props?

Yes. Player prop lines — pitcher strikeouts, outs, walks and earned runs; batter hits, walks, runs and RBIs — come from the public feed and are graded from the official box score.

How is an MLB prediction graded?

Game reads are graded against the final score; prop reads are graded from the box score. All results are published on the Pick Tracker.

What data powers the MLB board?

Public scoreboard and odds feeds with probable pitchers and market lines, plus Sportslyx team profiles built from that data.

Is this betting advice?

Sportslyx produces statistical estimates and analytical breakdowns for information and entertainment. Nothing on this page is betting advice, and no outcome is certain.

MLB Breakdowns for Every Game — 7 Days on Us

Full Pro access: the AI breakdown and market-derived probability for every MLB game on the board, the daily reads, Combo Builder, Studio and the pick tracker. A card is required, nothing is charged today, and you can cancel before the trial ends.

7-day trial · $0 today · card required · cancel anytime before the trial ends

Sportslyx produces statistical estimates and analytical breakdowns for information and entertainment. Nothing on this page is betting advice, and no outcome is certain. 18+ only. Please play responsibly — see our Responsible Play page.