ERA, WHIP and FIP: The Pitching Stats UK Bettors Should Read Before Every MLB Game

The night I stopped trusting ERA on its own

It was a Sunday in May, a few seasons back. I’d staked a chunky position on the Texas under, mostly because their starter had a 2.41 ERA across his first eight outings. Felt safe. Eight innings later he’d given up six earned runs, the under was dust, and I sat there reviewing my notes wondering how I’d missed it. The FIP – fielding-independent pitching – had been sitting at 4.12 the whole time. The defence had been bailing him out. The clock was always going to run out.

That game taught me something I now repeat to anyone starting on baseball: ERA is the headline, but it lies regularly. WHIP and FIP are how you check whether the headline matches the underlying performance. None of these numbers replaces watching the game, but they tell you which pitchers the market is overrating and which it’s underrating, and that’s the entire job.

Below is how I read each metric, where the meaningful thresholds sit, and how I stitch them together for totals and runlines.

ERA – useful, but only as a starting point

Earned Run Average measures how many runs a pitcher gives up per nine innings, excluding runs scored on errors. The thresholds are well established and worth memorising because they shape every market price you’ll see.

An ERA under 3.00 is excellent – the territory of genuine staff aces. Between 3.00 and 3.75 is solid mid-rotation work, the type of starter who’ll keep you in most games. From 3.75 to 4.50 is the league-average band. Above 4.50 starts looking shaky, and 5.00-plus is where books generally lean toward the opposing total. The current MLB run-scoring environment puts the league mean somewhere near 4.10, but that drifts year to year as the rule package and the ball both change.

The rule package has shifted measurably. In 2024 MLB tightened the pitch clock to fifteen seconds with bases empty and eighteen with runners – down from twenty seconds the year before – and capped mound visits at four per nine innings. Faster games mean less time to escape jams and less time to settle when something goes wrong. ERA distributions across starters have compressed slightly since 2023 as a result. Don’t lean on five-year ERA averages as if the conditions haven’t changed.

The trap with ERA is that it absorbs everything that happens once a pitch is released – defensive range, bad bounces, the first baseman who didn’t stretch. A starter with a 2.90 ERA over twelve starts could genuinely be that good, or could be riding a glove-heavy infield and a low BABIP that’s about to regress. ERA tells you the story; WHIP and FIP tell you whether the story holds up.

WHIP – traffic on the bases

Walks Plus Hits per Inning Pitched is the metric I check first when I’m betting baseball, full stop. It tells you how often the pitcher puts somebody on. Runs come from baserunners. No traffic, no scoring.

The thresholds I use, refined over years of grinding through totals markets:

What WHIP catches that ERA misses is sequencing luck. A pitcher who allows ten baserunners but strands them all has a beautiful ERA on the night and an ugly WHIP. The ERA won’t last; the WHIP will. Over a half-season sample, sequencing luck washes out, and the pitcher’s results converge to what the WHIP suggested all along.

I want to flag a UK-specific habit I’ve noticed. Punters here will often quote a pitcher’s career WHIP because that’s what gets surfaced first in a search. Career figures are useful for veterans but misleading for anyone in their first three or four full seasons. Recent rolling samples – last six starts, last twelve starts – track the current shape of the pitcher far better than a five-year career average that might bury a recent mechanical adjustment.

FIP – the luck filter

Fielding-Independent Pitching is the metric that fixed my Texas under loss. FIP isolates what the pitcher actually controls: strikeouts, walks, hit-by-pitches, and home runs allowed. It excludes everything that happens when the ball is in play and a fielder gets involved. The output is scaled to look like an ERA, so you can read it directly.

If a starter’s ERA is 2.40 but his FIP is 3.90, the defence and a friendly BABIP have been doing heavy work. Regression is coming. If ERA and FIP align – both around 3.20, say – the surface number is real and the pitcher is performing as the underlying skills suggest.

The gap between ERA and FIP is what I write down first when I’m looking at a starter’s profile. A persistent ERA-minus-FIP gap of more than 0.50 in either direction is a signal. Pitchers who consistently outperform their FIP year over year do exist – they’re called “FIP-beaters” and the trait usually correlates with elite control of contact quality, like inducing weak ground balls. But across the league as a whole, the gap regresses and the FIP wins.

The thresholds for FIP run roughly parallel to ERA but with cleaner predictive power going forward: under 3.50 is excellent, 3.50 to 4.20 is solid mid-rotation, 4.20 to 4.80 is league average, above 4.80 is below average.

Stitching the three together for totals

How I actually use these in a totals market: I don’t look at one starter, I look at the matchup. Two ERAs, two WHIPs, two FIPs. Six numbers, plus the bullpen projection, plus the park, plus the wind.

The cleanest unders sit where both starters carry sub-1.10 WHIPs and FIPs aligned with their ERAs. That combination tells you neither pitcher is getting bailed out and neither is putting traffic on the bases. Even a high-scoring park, even a tailwind, struggles to produce a crooked number when both starters are limiting baserunners.

The cleanest overs are mirror images. Both WHIPs above 1.30, both FIPs running half a run worse than ERA – meaning regression is coming for the deceptive performances – and you’ve got a setup where traffic compounds and the bullpens are likely to be involved early. As a saying in the industry goes, integrity is the priority and the numbers behind the game are how you protect it. The same principle works for the punter: trust the underlying metrics over the surface ones, and the long-run results follow.

The 2024 rule changes also reshape totals analysis. Pitch-clock pace pushes pitchers into rhythm, and the data over the last two seasons suggests strikeout rates have edged up – small, but consistent. That nudges totals projections slightly downward in matchups where both starters are command-and-control types. Hard throwers benefit even more.

Pitfalls of small samples

Every April I watch the same mistake on UK forums: someone posts a 0.95 ERA over four starts and asks whether to back the under. The answer is almost always no. Twenty-five innings is nothing. The pitcher could be facing four below-average lineups, getting a generous strike zone, or simply riding a hot streak that the underlying batted-ball data doesn’t support.

The minimums I use before trusting a pitcher’s surface stats: roughly fifty innings for ERA to start carrying signal, eighty innings for WHIP, and a hundred-plus innings before I let FIP drive a strong opinion on its own. Below those thresholds I lean on prior-year baselines, scouting reports on stuff and command, and matchup-specific factors instead of regressing to the small-sample averages.

Park factors and bullpen quality both deserve their own treatment. If you want to dig into how the run environment changes with the venue, my breakdown of how MLB park factors shape betting markets covers the parks that distort starter ERAs the most.