Methodology
How the engine works
The loudest complaint about every simulation in this genre is the same one: you cannot see what decided the outcome, so you cannot tell whether it was fair. That seems like a solvable problem, so here is the whole thing.
Most baseball simulations roll a number against a rating and tell you what happened. This one plays out the confrontation. That difference is the entire design, and it is why a pitch here has thirty-three separate inputs rather than two.
Eighteen of them decide which pitch gets thrown. Thirteen decide what that pitch physically does. Two more get the last word.
Eighteen inputs choose the pitch
- The count
- Twelve count buckets, each with its own pitch mix. 3-0 is not 0-2.
- His actual repertoire
- Not a generic pitcher. The four pitch families he really throws, weighted how he really throws them.
- How this hitter handles that family
- A per-family skill, so a pitcher attacks what you cannot hit rather than what he likes.
- His fatigue
- A tired pitcher drifts toward fastballs, because that is what tired pitchers do.
- Times through the order
- The third time through, the family lean rotates and his stuff and movement take a penalty.
- His pitching coach
- Execution skill decides how faithfully the plan actually gets carried out.
- The leverage of the moment
- Itself computed from eight things: the inning, which half, the outs, a runner on first, on second, on third, and both scores.
The count and the family mix cross into a forty-eight cell grid, twelve counts by four pitch families, and the pitcher's own tendencies live inside it.
Thirteen decide what it does
- Stuff sets velocity
- The pitch type carries its own velocity band, and his stuff decides where in that band this one lands.
- Movement sets break
- Vertical and horizontal, per pitch type, so a sinker sinks and a knuckleball is genuinely unpredictable.
- Command sets scatter
- How far the ball lands from where he aimed. Pinpoint command scatters about a quarter as much as wild.
- Handedness mirrors it
- A lefty is not a righty with a flipped label. The whole break profile and location bias reverse.
- Natural zone bias
- Every pitch type has somewhere it wants to go.
- Where it is aimed
- In the zone or off the plate, which depends on what the pitcher is trying to do.
It lands in one of thirteen zones, the same ones Statcast uses, across twelve pitch types that each carry their own velocity band and their own vertical and horizontal break.
Two get the last word
- The catcher
- Framing can steal a strike that was a ball, or lose one that was a strike. It is tracked to the individual catcher.
- The ballpark
- Passed into the pitch itself, not bolted on afterwards.
And the pitch is the small part
Thirty-three inputs get you one pitch. Everything below decides what that pitch means.
Where the ball goes
Every ball in play gets an exit velocity, a launch angle, a spray angle, a distance and a hang time. Hard-hit means 95 miles an hour, because that is what hard-hit means. The whole contact model is calibrated against roughly 1.1 million real batted balls from 2015 to 2024, and a separate check refuses to build a league at all if the numbers that come out do not look like the numbers that went in.
Who catches it
Defense is not a team rating nudging a batting average. Nine fielders stand at real coordinates and each has his own range, arm and error. A ball in the 5.5 hole is an out to a rangy shortstop and a single to a statue. A gapper one center fielder cuts off for a double, another lets roll for a triple.
Who the player actually is
Sixty-six named ratings before you count his arsenal, including eight personality traits that move real outcomes rather than sitting in a profile screen. A smart pitcher and a smart catcher execute more of the game plan. A greedy one costs more in free agency. A loyal one re-signs at a discount. A high-controversy one drags the clubhouse.
How he feels about it
Morale runs off seven drivers and clubhouse chemistry rides on top. A player whose morale falls far enough asks out. None of it is a large multiplier, which is deliberate: mood should tilt a season, not decide one.
Where the game is played
Every park carries its own outfield distance and wall height in seven directions, its own surface, and whether it is open, domed or retractable. On top of that sits twelve months of real climate for that city, so the ball carries further in a July night game in Denver than an April afternoon in San Francisco. The park has hand-specific effects, so it does not treat a lefty and a righty the same.
What you are allowed to know
Scouting is fogged. You do not see true ratings, you see an estimate with a confidence band that tightens the longer and more closely you have watched a player. That applies to your own organisation and applies harder to everyone else. Trades are interesting because both sides are guessing, which is how it works in reality.
Why we publish this
Every simulation in this category asks you to trust it. Almost none of them show you why you should. That is how you end up with twenty-year-old forum threads arguing about whether the engine is broken, which nobody can settle, because there is nothing to check.
We would rather hand you the list and let you find the holes. If something in here does not match what you are seeing in your league, that is a bug report we want.
There is a deeper version of every section above on the feature pages, and the honest accounting of where we are and are not ahead is on the comparison page.