MLB Stats That Actually Matter for Betting

If you know exactly two baseball stats, they're probably ERA and batting average. They are the stats on the back of the baseball card, the stats announcers use, the stats the box score puts in 18-point font.

They're also two of the noisiest, most misleading numbers in the sport - and a big reason new bettors lose money. Sportsbooks don't price games off ERA. The sharps don't bet off ERA. Once you understand why, the rest of the advanced-stat landscape gets a lot less intimidating.

Why ERA and AVG lie to you

ERA (earned run average) is the number of earned runs a pitcher gives up per 9 innings. Sounds clean. The problem: a pitcher only controls about three things - strikeouts, walks, and home runs. Everything else (a ground ball that finds a gap, a fly ball the outfielder lost in the sun, a one-hop liner that a great defender turns into a double play) is some mix of the defense behind him and pure luck.

Two pitchers with identical strikeout rates, walk rates, and home-run rates can have very different ERAs depending on the team behind them and the timing of the hits they allowed. That's not skill. That's noise.

Batting average has a similar problem on the hitter side. A hitter who crushes the ball at 95 mph but happens to hit it right at fielders can sport a batting average well below his actual quality of contact. A hitter blooping seventy-mile-an-hour singles can post a great batting average while contributing very little.

The pitcher stats that actually matter

FIP - Fielding Independent Pitching

Built from only the things a pitcher controls (strikeouts, walks, hit-by-pitches, home runs) and scaled to look like ERA. A 3.50 FIP "looks like" a 3.50 ERA. The crucial property: past FIP predicts future ERA better than past ERA does[FanGraphs]. If a pitcher's ERA is 3.00 but his FIP is 4.00, the market knows that his next 50 innings are likely to look more like 4.00. Bet the market, not the back of the card.

xFIP - Expected FIP

Like FIP, but it normalizes home-run rate to league average. A pitcher who's given up an unusually low or high HR rate so far gets corrected toward the mean. Useful early in the season; a sharper "what should we expect" number than FIP alone.

SIERA - Skill-Interactive ERA

FIP's more sophisticated cousin. Uses ground-ball / fly-ball data and accounts for the fact that strikeouts and walks interact differently with batted-ball type. Slightly better predictive power than FIP. If you're choosing one number, use FIP for simplicity. If you want the cleanest projection, use SIERA.

K% and BB%

Strikeouts as a percentage of batters faced; walks as a percentage of batters faced. A pitcher with a 30% K-rate and a 6% BB-rate is doing his job regardless of the ERA next to his name. These are the two most stable, most repeatable pitcher numbers - they "stabilize" (become predictive) faster than almost anything else.

The hitter stats that actually matter

wOBA - Weighted On-Base Average

Like OBP (on-base percentage), but weights each kind of hit by how many runs it actually produces. A home run counts more than a triple, which counts more than a double, which counts more than a single, which counts more than a walk. The single cleanest number for hitter quality.

The numerical scale is built to look like OBP - a wOBA of .320 is roughly average, .380 is excellent, .420 is MVP territory, .280 is rough.

wRC+ - Weighted Runs Created Plus

Takes wOBA and re-scales it to a league average of 100, then adjusts for park and era. A 130 wRC+ means the hitter created 30% more runs than league average. A 70 wRC+ means 30% fewer.

wRC+ is great for cross-comparison: it's the stat you reach for when you want to ask "is this Coors Field hitter actually as good as his stats suggest" or "is this Petco Park hitter being undersold by his raw numbers." Park-adjusted offense in one number.

Statcast metrics - exit velocity, barrel%, hard-hit%, xBA

Statcast is MLB's official ball-tracking system. The bettor-relevant outputs:

• Exit velocity - how hard the hitter hits the ball, in mph. Average EV is a stable, repeatable signal of contact quality. 90+ mph is good; 92+ is elite.
• Barrel % - the percentage of batted balls hit in the "perfect" combination of speed and angle that almost always becomes extra-base damage. The single best leading indicator for power.
• Hard-hit % - share of batted balls hit at 95+ mph.
• xBA / xSLG - "expected" batting average and slugging based on the quality of contact, ignoring whether the ball found a glove. If a hitter's xBA is much higher than his actual AVG, he's been unlucky and is likely to regress upward.

The team stat almost no one talks about: Pythagorean expectation

Pythagorean expectation is a formula that estimates a team's "true" winning percentage from runs scored and runs allowed[FanGraphs]. Teams whose actual record is well above their Pythagorean record have usually been lucky in close games and tend to regress. Teams below tend to bounce back.

Practical use: when a 25-15 team is on a hot streak and the lines are pricing them like a dynasty, check the Pythagorean record. If it says 22-18, the market is overrating them. The reverse is just as useful.

Putting these to work

You don't need to memorize the formulas. You just need to know which stat to look at for which question:

• Is a pitcher legit? Look at FIP, K%, BB%. Ignore W-L. Discount ERA.
• Is a hitter legit? Look at wOBA, wRC+, xBA, exit velocity. Discount batting average.
• Is this team's record real? Compare actual record to Pythagorean record. Big gap = regression coming.
• Is the player due to "bounce back"? Check xBA vs AVG, xSLG vs SLG. Big gap means the underlying quality has been better than the surface.

These five sources of truth - FIP, wOBA, wRC+, Statcast (exit velo / barrel%), and Pythagorean expectation - explain about 90% of what sharp MLB bettors look at when they evaluate a side. They're all freely available, and most of them surface directly in the Breakaway Stats player cards (next chapter).