Which Statcast Data Matters for Hitters in Daily Fantasy Baseball?

The Statcast era has brought about all sort of fun and exciting metrics, but which ones can truly help us evaluate hitters in DFS?

Since it debuted in 2015, Statcast data has been all the rage in fantasy baseball player evaluation. While it's technically no longer "new," it remains the shiniest toy in an analytics toolbox already filled to the brim with advanced stats.

And it's easy to understand the appeal when you head over to Baseball Savant and peruse yearly leaderboards littered with exit velocity, distance, and barrel metrics or click on player pages with eye-catching, color-coded percentile rankings that are bathed in a scorching red for household names like Mike Trout and Nelson Cruz or fast-rising superstars like Fernando Tatis Jr. and Juan Soto.

There's no question that having access to Statcast data is exciting, and one can only imagine what will come down the line next. But with new data comes a familiar conundrum in fantasy baseball: which of these new-fangled stats actually matter? While the abundance of data we have at our disposal in baseball is great, it's also easy to get overwhelmed and bogged down by information overload.

This is especially the case when it comes to daily fantasy. We don't have time to research every stat under the sun every day. We need to hone in on what's really useful.

It wasn't that long ago that Baseball Information Solutions' quality of contact stats on FanGraphs -- specifically hard-hit rate -- were the hot new thing in DFS. Anecdotally, it makes sense -- making hard contact should lead to more hits, and, most importantly, more extra-base hits and home runs.

Our own Jim Sannes detailed just how valuable hard-hit rate is to DFS in 2017 and 2016. In those studies, Sannes found a strong correlation between hard-hit rate and FanDuel points, confirming its usefulness as a measure of evaluating hitters.

Intuitively, Statcast's metrics should go hand-in-hand with hard-hit rate and other batted-ball data, and in theory, they should be a natural progression in precisely measuring quality of contact. Previous studies have demonstrated a strong correlation between power and certain Statcast metrics, as well, so it's logical to conclude that they should be a helpful tool when looking at daily fantasy matchups.

But which of those metrics stand out? Are any of them worthy successors to the old standby, hard-hit rate? Let's take a look.

Which Stats Actually Correlate With FanDuel Points?

For starters, anyone who's played MLB DFS knows that we're looking first and foremost for players who are most likely to do one thing: hit home runs.

Dingers are big-money plays. A solo shot alone nets you 18.7 FanDuel points. Add in some baserunners (hopefully from your stack!), and you're rolling in points on just one swing.

But we should confirm this through actual historical data. So to illustrate this, let's take a look at how some traditional rate stats and power metrics correlate with FanDuel points per plate appearance (FDP/PA).

The sample for all data in this piece reflects qualified hitters from 2015 through 2020, which is far back as Statcast goes. Between MLB reportedly "deadening" the ball this season, and evidence that they've been regularly altering the ball in recent years (including the infamous "juiced" ball of 2019), it's unclear how much hitting will be impacted in 2021, so it makes sense to expand our sample as far back as we can. That said, it's worth noting that these results saw minimal differences even when using more recent sample sizes.

The closer the correlation coefficient is to 1, the stronger the relationship. Conversely, if the number is closer to -1, it shows an inverse relationship.

Stat Correlation
with FDP/PA
SLG 0.932
OPS 0.924
wOBA 0.900
ISO 0.862
HR/PA 0.764
HR/FB 0.708
OBP 0.680
AVG 0.461

Wait, what? Slugging percentage? That old thing?

Yes, of all the statistics I examined for this exercise, slugging percentage showed the strongest relationship to FanDuel points per plate appearance. While that might seem odd at first glance, this reflects what Sannes also found, and it actually makes perfect sense when you look at FanDuel's scoring system, which closely mirrors the formula for slugging percentage when it comes to weighting extra-base hits.

The other highly-correlated stats that follow closely behind slugging percentage are also logical. OPS and ISO literally have slugging percentage within their respective formulas, while wOBA incorporates all the components of FanDuel scoring except stolen bases (extra-base hits, walks, HBP) but at weights more reflective of their worth in real life.

What it boils down to is that all these metrics incorporate power in some capacity. Specifically, they incorporate those sweet, sweet dingers, as both home run rate (HR/PA) and home-run-to-fly-ball rate (HR/FB) check in next.

On-base percentage has a solid correlation but lags behind by treating all hits as being equal. As one would expect, batting average performs even worse by removing walks and HBP.

Great! So, no need to dig any further, right?

Well, the problem is the aforementioned stats are results-based metrics that aren't necessarily as reliable in predicting future performance.

This is reflected by how long they take to "stabilize" in a given season -- i.e. when their samples start to become meaningful. That isn't to say a stabilization point is a magical in-season threshold where we can suddenly take a stat at face value -- a common misconception -- but it's roughly when we can partially attribute it to true skill and incorporate it as new data. Naturally, the smaller the stabilization point the better, and the larger the sample goes beyond that point over a campaign, the more we can trust it.

For instance, according to the original Baseball Prospectus study referenced on FanGraphs' sample size page, a hitter's strikeout rate stabilizes at just 60 plate appearances, which is approximately 15 games. That means that after one month, we've already doubled that threshold, which is why it's one of the first hitting metrics we can feel fairly confident in early on.

On the other hand, the stabilization points for slugging percentage (320 at-bats), on-base percentage (460 plate appearances), and batting average (910 at-bats) are all incredibly high. This is due to the astronomical marks for doubles-plus-triples rate (1610 plate appearances) and BABIP (820 balls in play) to stabilize, which makes perfect sense -- there's a ton of yearly variance on balls in play. OPS and wOBA aren't specifically listed in the study, but this clearly puts them in the same bucket.

ISO (160 at-bats), home run rate (170 plate appearances), and HR/FB rate (50 fly balls) fare a bit better. ISO strips out some of the luck by subtracting batting average -- thereby giving home runs more weight -- which explains why all three stabilize in roughly the same timeframe of 40 to 60 games, per FreezeStats. That's not bad! Home runs are far less noisy than balls put in play. But we also know that not all home runs are created equal, so we're still at the mercy of outside elements like park factors and weather. It's certainly not perfect.

We can do better, which is where batted-ball data can help us.

The Immense Value of Hard-Hit Rate

So, we know we want home runs, and stats like slugging percentage, wOBA, and ISO are closely tied to FanDuel points. But we need to find a more reliable means of predicting those stats rather than just relying on the descriptive stats that overlap significantly.

All of which brings us back to good ol' hard-hit rate. I straight-up jacked the headline from my colleague's 2016 piece for this header, but it sums things up perfectly.

Let's quickly show out how batted-ball stats correlate with the power metrics I referenced in the previous section.

Hard% 0.685 0.659 0.646
Fly-ball% 0.583 0.571 0.234
Soft% -0.369 -0.320 -0.354
Ground-ball% -0.524 -0.476 -0.171
Medium% -0.595 -0.598 -0.557

It's easy to see why batted-ball data has been a long staple of DFS research -- that's pretty good! Hard-hit rate used in conjunction with fly-ball rate gives us a strong indication of a given player's power. The value of hard contact is further shown by medium-hit rate actually having the strongest negative correlation in all three metrics -- even worse than ground-ball rate, which is pretty wild.

On top of all this, batted-ball stats also stabilize quickly. According to FreezeStats, hard-hit rate, fly-ball rate, and ground-ball rate all stabilize around 80 balls in play, which shakes out to just under 30 games. That's the best we've seen so far.

Logically, this makes sense. Hitters have more control over quality of contact and whether they tend to hit fly balls or ground balls. We're also eliminating the end results of these batted-ball events, making this a stronger indication of skill. We don't need to worry about whether a ball was robbed on the warning track or barely snuck over the fence for a home run. It's simply reflecting how they hit the ball.

Now that we've revisited the value of hard-hit rate, it's time to see whether anything from Statcast can dethrone it.

Like Shooting Fish in a Barrel

I mentioned at the start that some studies have suggested a strong relationship between certain Statcast metrics and power. So, let's put our sample to the test and see how a variety of Statcast measurements compare with hard-hit rate and fly-ball rate when it comes to correlating with power.

"BBE" is a batted ball event, while "EV" stands for exit velocity. "95+ MPH%" refers to Statcast's version of hard-hit rate, which measures the percentage of times a batter hits the ball at 95 miles per hour or faster.

Barrel/PA 0.842 0.853 0.823
Barrel/BBE 0.824 0.837 0.846
EV on LD/FB 0.702 0.716 0.790
Hard% 0.685 0.659 0.646
Avg Dist 0.672 0.616 0.364
95+ MPH% 0.642 0.636 0.702
Avg EV 0.603 0.593 0.645
FB% 0.583 0.571 0.234
Max Dist 0.554 0.579 0.607
Max EV 0.446 0.476 0.508

We may have a new sheriff in town.

Both barrels per plate appearance and barrels per batted ball event top the chart in all three metrics, while average exit velocity on line drives and fly balls (EV on LD/FB) also hits the podium. Hard-hit rate still holds up well, though, placing top-five in each category.

Now, in fairness to our old friend hard-hit rate, that metric applies to all types of batted-balls, including grounders, whereas barrels will almost never be a ground ball (searching Baseball Savant's database will dig up only a handful of outliers), and EV on LD/FB specifically excludes grounders.

What's interesting is that FanGraphs' hard-hit rate maintains a higher correlation over Statcast's version (95+ MPH%) in two of the three metrics, although they're fairly close across the board. You'll occasionally find confusing discrepancies between the two versions with certain players, and this doesn't really help pick a preferred choice in that regard.

Likewise, average exit velocity and distance pop up with solid marks, but they don't appear to have any meaningful advantage, either.

Regardless, the main takeaway is the eye-opening correlation between barrels and home run power. In a sense, this shouldn't be surprising considering the very definition of a barrel. According to, barrels are "batted-ball events whose comparable hit types (in terms of exit velocity and launch angle) have led to a minimum .500 batting average and 1.500 slugging percentage."

Remember how hard-hit balls and fly balls work well in conjunction? Here are the league averages on hard-hit fly balls from 2015-2020.

HH FB Results by Year AVG SLG wOBA ISO
2020 0.510 1.756 0.883 1.246
2019 0.523 1.801 0.887 1.279
2018 0.493 1.662 0.857 1.169
2017 0.525 1.767 0.891 1.242
2016 0.503 1.671 0.858 1.168
2015 0.499 1.630 0.860 1.131

Well, will you look at that! The batting averages are practically spot on with the minimum for barrels and the slugging percentages even exceed it. Hard-hit fly balls sure look like a good proxy for barrels, right?

Except there's more. Although a .500 average and 1.500 slugging percentage are the bare minimum criteria for barrels, they actually tend to well surpass those thresholds in real life.

Let's now look at how barrels performed in those same seasons.

Barrel Results by Year AVG SLG wOBA ISO
2020 0.797 2.713 1.364 1.916
2019 0.814 2.820 1.408 2.006
2018 0.772 2.627 1.371 1.855
2017 0.826 2.885 1.468 2.059
2016 0.810 2.765 1.434 1.955
2015 0.794 2.721 1.442 1.927

Okay, then.

It's official. Barrels are unquestionably the "S-Rank" of batted balls. Not even hard-hit fly balls can touch them.

Finally, to top it all off, Statcast metrics like average exit velocity and barrels stabilize after 40-50 balls in play, which typically comes at roughly 18 games. Not only is that slightly faster than hard-hit rate and fly-ball rate, but that's awfully close to the gold standard of strikeout rate (15 games).

How Barrels Apply to Daily Fantasy

We've confirmed that barrels have a whole lot going for them as a metric, so let's circle back to where we started.

We already established that barrels correlate strongly with ISO, and they clearly boast some ridiculous marks in slugging percentage and wOBA, too, so it stands to reason that barrels should also have a strong relationship with those stats and, presumably, FanDuel points.

Barrel/PA 0.683 0.726 0.590 0.842
Barrel/BBE 0.654 0.671 0.555 0.824
Hard% 0.572 0.601 0.514 0.685
EV on LD/FB 0.562 0.583 0.495 0.702
95+ MPH% 0.543 0.573 0.516 0.642
Avg EV 0.517 0.553 0.499 0.603
Avg Dist 0.489 0.529 0.413 0.672
Max Dist 0.445 0.464 0.357 0.554
Max EV 0.352 0.377 0.301 0.446
FB% 0.352 0.367 0.213 0.583

Once again, both barrel metrics outclass hard-hit rate across the board, though it's a testament to the latter's utility that it still hangs on to the bronze when it comes to correlating with FanDuel points per plate appearance.

EV on LD/FB and 95+ MPH% fall in the same vicinity as hard-hit rate. But what's interesting is that if we narrow the sample down to just 2019, EV and LD/FB had the strongest relationship with FanDuel points of the three, while the two hard-hit rates were identical, which reflects the higher league-average exit velocity produced from the juiced ball that season -- the highest in the Statcast era. In 2015 -- the lowest league-average exit velocity -- the inverse was true, with hard-hit rate correlating most strongly with FanDuel points (the barrel metrics are still tops in both outlier seasons).

Ultimately, all three metrics are probably somewhat interchangeable in evaluating quality of contact for our purposes, but considering the ball is expected to be deadened this season, giving the edge to hard-hit rate over the other two metrics might be wise in 2021.

That said, the main takeaway here is the fantastic showing by barrels. Specifically, barrels per plate appearance demonstrates the strongest relationship with FanDuel points, slugging percentage, wOBA, ISO, and home run rate. Barrels per batted ball edges it out in HR/FB rate, which makes sense given the two metrics only measure batted balls.

Closing Thoughts

It's no secret that much more goes into creating successful FanDuel tournament lineups than just individual hitting talent -- team context is just as vital. Pitching matchups, lineup strength, platoon splits, park factors, and weather are all important, which is why targeting the right teams to stack is such a proven and successful strategy.

But at the end of the day, we're still looking for batters who can hit for power and score points in bunches, and that's reflected in the strong relationship between slugging percentage and FanDuel points.

Slugging is too reliant on the end results that lead to FanDuel points, though, and we should be trying to find what leads to the opportunities to generate those points and stabilizes quickly: that's barrels.

Barrels have the strongest correlation with FanDuel points amongst all Statcast metrics, and they even exceed that of hard-hit rate, a proven metric in DFS. Barrels also stabilize in a smaller timeframe than most other hitting stats.

At the same time, it's worth noting that hard-hit rate still holds up against Statcast, and it's a reminder that the latest thing doesn't always mean we have to throw out what came before.

Outside of barrels, no other Statcast metric showed a decisive advantage over hard-hit rate as a way of evaluating hitters. EV on LD/FB correlates better than hard-hit rate in power metrics like ISO and home run rate, so there's a strong argument for incorporating it into your research, but it isn't necessarily a step up over the tried and true combination of hard-hit rate and fly-ball rate.

That said, there's no question the big winner here is the barreled ball, and barrels per plate appearance makes a strong case for being the metric you want to focus your hitters around. If you haven't been incorporating barrels into your FanDuel research, it's time to join the party.