MLB Strategy

Baseball gives you more ways to attack a game than any other sport. The moneyline, the run line, and totals are the foundation, but first-five-inning markets, team totals, NRFI and YRFI, and pitcher props all open up additional routes to the same angle.

That optionality is the whole point. Because you can isolate a starting pitcher, a single offense, or the early innings, you rarely have to bet a messy full-game number when a cleaner expression exists. The skill in MLB is choosing which market best captures the edge you actually have.

One thing to be honest about up front: building reliable systems is genuinely harder in baseball than in football or basketball. The day-to-day variance is enormous, the starting pitcher changes everything, and a clean two- or three-filter angle that would hold up in the NFL often falls apart over a full MLB season. The good baseball systems usually start with a strong narrative and then get tightened with extra conditions before we trust them. A great base example is Home Favorites Avoiding The Sweep, floated to the right. A home favorite that has dropped the first two games of a series, and was favored in that previous game, comes back in Game 3 facing a sweep, and quality home teams in that exact spot respond. It has gone 462-259 on the moneyline, just over 64 percent, and the story is as clean as they come: a good team, at home, backs against the wall, refusing to get broomed.

A close cousin of that system widens the net beyond the sweep, floated to the left. Here we simply back a home favorite that has lost its last two games and was favored in its previous game too, without requiring Game 3 of a series, as long as the moneyline price stays reasonable. Dropping the sweep condition costs a little accuracy but opens up far more games, and it has still run 634-408, just under 61 percent, across a massive sample. It is the same core idea, a good home team being doubted after a short skid, and it shows how loosening a single filter trades a bit of edge for a lot more volume.

The rest of this part covers how we think about pitchers and first-five markets, when team totals make the most sense, how we handle plus-money dogs and juice, and the first-inning markets that have become their own niche.

Starting pitching is huge in baseball. It is the single biggest factor we weigh, because the starter shapes the early innings more than anything else on the board. When our edge is built around a pitching matchup, the first-five-inning market is often the cleanest way to bet it, since it lets the starters decide the outcome before the bullpens and benches take over.

When the read depends on the full team profile rather than just the starter, the full-game moneyline makes more sense. Bullpens matter, but most of that information is already priced into the number by the time you see it, so we do not overweight it.

The decision between F5 and full game comes down to where your confidence lives. If you trust the starter and the early matchup, F5. If you trust the whole team to win the game, full game.

MLB props are their own discipline, and our capper Insider Mike treats them differently from the team markets. He does not build projection models for them; the brain is the model. The read always starts with the matchup in front of him and ends with the single most important input of all: the number itself. A great matchup attached to a bad line is still a pass, and a mediocre matchup attached to a soft line can be a bet. There are no traps in his mind, just spots where the price is right and spots where it is not.

A core lean carries across every prop market: unders are your friend. There are simply more ways to win them. In baseball a walk, a hit by pitch, a weather delay, an early hook, or a defensive gem can each erase one of the roughly four plate appearances a hitter gets, and any of those quietly kills an over. That is why we tend to avoid taking the over on an individual batter prop. If we like an offense, we would rather fade the opposing pitcher or back a team total than isolate one hitter on a total-bases or hits line and pray the sequencing cooperates. Nothing stings more than watching a batter rip a ball right at a defender and get nothing for it.

The batter angle we have leaned on most is the under on hit props. Modern lineups are full of strict platoon bats who may only get two plate appearances, and the threat of a pinch hitter or a weaker backup catcher later in a big spot is real. When one of those soft bats draws a stud starter, the hit prop often has not fully baked in how few quality swings the hitter is actually going to get, and that is a clean spot to fade.

Pitcher props vary by type. For strikeouts, we first find a line that feels a touch high or low, then check the starter’s CSW% (called strikes plus whiffs), K%, and SwStr%. From there we pull the projected lineup and look at their splits against that day’s handedness: their K%, chase rate, and whiff rate. For pitcher outs or innings, we look at the bullpen’s recent usage, who the team has lined up to start the next day, and whether they have a reason to give the starter a longer leash. Then we weigh how efficient the pitcher is (pitches per plate appearance) alongside his expected metrics like xERA, xFIP, and CSW%, and how he has done relative to that line lately. Finally we look at the offense he is facing: their wRC+, how patient they are, how other pitchers have fared against this line, and what the weather is doing, whether it is good hitting weather (hot, humid, wind blowing out) or good pitching weather (cold, wind blowing in).

The best plays stack several of these together. A favorite spot for Mike is a starter whose pitch mix matches up well against a lineup that historically struggles with that arsenal, in cooler conditions with the wind blowing in, with friendly umpire data behind the plate, and recent form that lines up with the number that day. When all of that agrees, we are cooking. But the line still gets the deciding vote: if the price does not offer value, the read never becomes a bet.

The data stack behind this is specific. For a team’s offense we start with season-long rates and then layer recent windows, the last 30, 14, and 7 games, leaning on wRC+, K%, and BB% from FanGraphs. For an individual pitcher we look at Pitching+ from FanGraphs plus CSW%, SwStr%, K%, xERA, xFIP, BB%, and ICR (ideal contact rate) from Pitcher List. For an individual batter we pull the expected stats through PropsMadness, which surfaces the Baseball Savant data in one place. As always, hit rates are a reference we check at the end, not the foundation we build on.

Team totals are one of the most useful tools in baseball. They let you cut out the other team entirely and focus on a single angle: one offense that is either inconsistent, overpowered, or weak, or one pitching matchup you have a strong read on.

This is exactly the move from the daily-card example. When the moneyline is too expensive or too risky, like backing a big dog against an elite team, a team total lets you express the part of the game you actually believe in without betting straight up into a buzzsaw.

Any time you find yourself uneasy about involving both teams in a bet, ask whether a team total isolates the side you trust. Often it does, and it removes a variable you were never confident about in the first place.

Totals systems show the other side of how hard MLB can be to systematize, where a real edge usually hides behind a stack of filters. Take the Visiting Dogs Snapping An Over Streak under, broken out in the band below. The narrative is appealing: a visiting underdog that leans under on the season had been running a little hot to the over, just had that over streak snapped, and now plays in the middle of a series against a modest home favorite with a catchable total. On its own, half of that is noise, which is exactly the point. To get to a 336-209-26 record and an 18 percent ROI, the system needs every one of those conditions stacked together. The narrative gives you the idea, but in baseball you almost always have to layer filters before it becomes an edge, and even then we tread lightly.

Baseball is a high-variance sport, and plus-money underdogs are often best sized at a half unit. Backing a live dog at a plus price is a perfectly good bet, but the variance means you usually do not want full exposure on it.

Heavy juice is what pushes us off a play more than anything. A signal-backed favorite that is laying a big price often is not worth it straight up, and forcing it just because the signals fired is a mistake. That is the moment to look for a cleaner, cheaper route to the same idea.

Not every signal-backed dog deserves a moneyline bet, and not every favorite deserves the juice attached to it. The price always gets a vote.

First-inning markets, NRFI (no run first inning) and YRFI (yes run first inning), have grown into their own niche. They blend starting pitcher quality, the top of each lineup, park and weather conditions, umpire tendencies, and the early-inning market itself into a focused, single-inning bet.

A first-inning bet is really a bet on a small, specific set of inputs, which is exactly what makes it so cleanly handicappable. Start with the two starting pitchers and their first-inning history specifically, not their overall ERA. Some pitchers are notorious slow starters who give up early runs before settling in, while others are first-inning assassins who routinely post a clean frame. That split alone is often the whole bet, and it is the single biggest factor we weigh.

From there, look at the top of each lineup, since only the first three or four hitters are guaranteed to bat in the opening frame. A NRFI gets much stronger when both pitchers own their first inning and both lineups have soft tops; a YRFI gets interesting when high on-base bats lead off against a slow-starting arm. Park and weather matter too, with wind blowing out and hitter-friendly parks nudging toward YRFI. Umpire tendencies, specifically how wide or tight a plate umpire calls the zone, can tilt a borderline lean. Our First Innings builder pulls all of these factors into one view so you can filter the slate down to the handful of games where the inputs actually agree, instead of forcing a bet on every matchup.