MLB Totals Betting: Over/Under Strategy Using Weather, Ballpark and Pitcher Data

My best month betting MLB totals came not from picking winners but from understanding something deceptively simple: the run-scoring environment shifts from year to year, and the market is slow to catch up. In April 2024, I noticed that league-wide scoring had jumped from the previous season, but opening totals were still anchored to the prior year’s averages. I hit overs at a 58% clip for three weeks before the lines adjusted. That lag — the gap between what the league is actually producing and what the books expect it to produce — is where totals bettors make their money.

MLB totals betting, also called over/under betting, asks a single question: will the combined runs scored by both teams in a game be over or under the number set by the bookmaker? A typical MLB total sits between 7.0 and 10.0, depending on the pitchers, the park and the conditions. It sounds straightforward, but the variables feeding into that number are layered enough that even small analytical advantages compound across a 2,430-game season. This guide covers the three pillars that shape every total — league environment, ballpark factors and pitching matchups — and shows you how to combine them into a framework you can apply to any game on the slate.

Understanding the Run-Scoring Environment: League Averages and Year-to-Year Shifts

Every totals model starts with the same number: league-wide runs per game. In 2023, MLB averaged roughly 4.6 runs per team per game after the pitch-timer and shift-ban changes boosted offence. In prior years, that figure sat closer to 4.1 to 4.3. The difference between a 4.1 and a 4.6 environment is massive for totals bettors — it shifts the break-even point on over/under bets and changes which games qualify as value.

I track league-wide runs per game on a rolling 14-day basis during the season rather than relying on static preseason projections. The reason is that run-scoring is seasonal within the season itself. April games tend to have lower scoring because of colder temperatures and pitchers who are relatively fresh. June and July see scoring spike as heat helps the ball carry and pitcher fatigue accumulates. September introduces another wrinkle with expanded rosters and teams at different motivation levels.

The practical application is this: if the 14-day league average is 9.4 combined runs per game and the bookmaker posts a total of 8.5 on a neutral-park matchup with average pitching, the over has structural value. If the 14-day average has dipped to 8.2 because of a cold snap across the Midwest and Northeast, that same 8.5 total shifts toward the under. Most recreational bettors set their totals thesis at the start of the season and never update it. That’s where the edge is — in the constant recalibration.

Why the Same Pitching Matchup Prices Differently at Coors and Tropicana

I once tracked a pair of pitchers who faced each other twice in a ten-day stretch — once at Coors Field in Denver and once at Tropicana Field in St. Petersburg. The total for the Coors game opened at 11.5; the Tropicana game opened at 7.5. Same two arms, same two lineups, four full runs of difference. The park was doing all the work.

Coors Field sits at 5,280 feet above sea level, where thinner air reduces drag on the baseball and allows it to travel roughly 5% further than at sea level. That altitude effect inflates home runs and extra-base hits, which is why Coors consistently produces totals in the 11-13 range. Wind of just one mile per hour adds approximately three feet of ball flight, and a temperature increase of ten degrees Fahrenheit boosts distance by about 1%. When you combine altitude, dry air and the cavernous outfield dimensions, Coors becomes an outlier that demands its own model.

Tropicana Field, by contrast, is a domed stadium with climate-controlled conditions and a playing surface that suppresses offence. The total there rarely climbs above 8.0 unless both starters have ERAs north of 5.00. The dome eliminates wind, temperature and humidity as variables, which means the total is almost entirely a function of the two pitchers and the lineups.

For totals bettors, the lesson is that park factor is not a nice-to-have adjustment — it’s a foundational variable. I apply a park-factor multiplier to every total I model. FanGraphs and Baseball Reference both publish park factors indexed to 100 (where 100 is neutral). A park factor of 115 means that park inflates offence by 15% relative to the league average; a factor of 88 means it suppresses offence by 12%. Multiplying the expected combined runs by the park factor gives a more accurate baseline than the raw pitcher data alone, and it’s the single quickest way to identify when a bookmaker’s total is off by half a run or more.

How Pitching Matchups Shape the Total: Combining ERA, FIP and Opponent wOBA

After the ballpark sets the baseline, the pitching matchup determines the deviation. Here’s where I part ways with most casual bettors: I don’t use ERA as my primary pitching metric for totals. ERA is a backward-looking number that includes sequencing luck, defensive quality and bullpen inheritance. For totals, I need a metric that isolates what the pitcher himself controls, and that means FIP — Fielding Independent Pitching.

FIP strips out the defence behind the pitcher and focuses on strikeouts, walks, hit batters and home runs. A pitcher with a 3.20 ERA but a 4.10 FIP has been getting lucky with his defence; a pitcher with a 4.00 ERA but a 3.30 FIP has been getting unlucky. For totals betting, the FIP tells me the true run-prevention talent more accurately than the ERA, and I weight it accordingly in my model.

The second metric I layer in is opponent wOBA — weighted on-base average. Instead of looking at the opposing team’s batting average or OPS in a vacuum, I check their wOBA specifically against the starter’s handedness. A left-handed pitcher facing a lineup whose right-handed hitters carry a .340 wOBA against lefties is facing a harder task than his overall numbers suggest. That platoon-adjusted opponent wOBA is the best single number for estimating how many runs the lineup will produce against this particular arm.

My process for any given game: start with the weather impact data and park factor, then layer in each starter’s FIP and opponent wOBA. The total I estimate from those inputs is my “true total.” I compare that to the posted line. If the gap is 0.5 runs or more and the direction is consistent across all three inputs — park, weather and pitching — I have a play. If the inputs conflict (great pitching in a hitter’s park on a hot day), I pass. Totals betting rewards conviction built on convergence, not hunches built on one variable.