October Trends: How the 2020 Postseason Stacks Up
We need your support to make FanGraphs sustainable so you can continue to enjoy all of the articles you’ve come to rely on and enjoy, like this one.
We’re asking for your help to make it through the offseason and ensure that we’re here to cover baseball in the 2021 season and beyond. You can read more about the state of the site here.
The ALCS between the Rays and Astros and the NLCS between the Dodgers and Braves have both produced tight, dramatic contests thus far, full of home runs and low scoring. As such, it’s a good time to check in on some of the trends that defined the brief 2020 season, and how they compare to what we’ve seen in the postseason, and how this October compares to recent regular and postseasons.
For starters, well, there are the starters. As I noted just two weeks into the abbreviated regular season, the length of the average start had fallen below five innings, and while it rose slightly over the remainder of the 2020 campaign, it still finished below five. Updating the table I included with that piece:
| Season | IP/GS | Change | K% | BB% | HR/9 | ERA | ERA- | FIP | FIP- |
|---|---|---|---|---|---|---|---|---|---|
| 2015 | 5.81 | -2.6% | 19.5% | 7.1% | 1.06 | 4.10 | 103 | 4.03 | 102 |
| 2016 | 5.65 | -2.8% | 20.2% | 7.7% | 1.24 | 4.34 | 104 | 4.3 | 103 |
| 2017 | 5.51 | -2.4% | 20.6% | 8.1% | 1.34 | 4.49 | 103 | 4.48 | 103 |
| 2018 | 5.36 | -2.8% | 21.6% | 8.0% | 1.21 | 4.19 | 101 | 4.21 | 101 |
| 2019 | 5.18 | -3.4% | 22.3% | 7.7% | 1.44 | 4.54 | 101 | 4.51 | 100 |
| 2020 | 4.78 | -7.7% | 22.9% | 8.3% | 1.30 | 4.46 | 100 | 4.46 | 100 |
The year-to-year drop in innings per start was the largest we’ve seen in this span, and indeed the largest we’ve seen in the Wild Card era, if not longer. Granted, it was a short season, with a short ramp-up, as well as a ton of seven-inning doubleheaders, something we’ve never seen at the major league level before, but that wasn’t really factor when I checked in on the first two weeks; I used data through August 5, at which point only one seven-inning doubleheader had been played.
Anyway, through this year’s Wild Card and Division Series, starting pitcher workloads decreased even further:
Starters have averaged fewer than five innings per turn in three of the past four postseasons; with the likes of Gerrit Cole, Stephen Strasburg, and Justin Verlander frequently pitching deep into games — six or more innings a total of 14 times between them — last year was the exception. While the average through the first two rounds fell to its lowest level this year, the drop-off relative to the regular season is about on par with the period in question:
| Year | Reg IP/GS | Post IP/GS | Dif |
|---|---|---|---|
| 2015 | 5.81 | 5.51 | -5.2% |
| 2016 | 5.65 | 5.11 | -9.6% |
| 2017 | 5.51 | 4.73 | -14.1% |
| 2018 | 5.36 | 4.68 | -12.7% |
| 2019 | 5.18 | 5.14 | -0.9% |
| 2020 | 4.78 | 4.37 | -8.5% |
| Avg | 5.45 | 4.93 | -9.5% |
The particularly short starts from this postseason, I believe, reflect both the urgency of the short series — including the unprecedented best-of-three round (ugh) — as well as the increased use of openers and bullpen starts. On that subject:
| Year | GS | <= 3 IP | % | >= 6IP | % |
|---|---|---|---|---|---|
| 2015 | 72 | 9 | 12.5% | 36 | 50.0% |
| 2016 | 70 | 7 | 10.0% | 24 | 34.3% |
| 2017 | 76 | 15 | 19.7% | 25 | 32.9% |
| 2018 | 66 | 15 | 22.7% | 19 | 28.8% |
| 2019 | 74 | 10 | 13.5% | 33 | 44.6% |
| 2020 | 66 | 16 | 24.2% | 18 | 27.3% |
Nearly one-quarter of the starts from the first two rounds lasted three innings or less. Admittedly, that’s lumping together some quick hooks of ineffective starters (Hyun Jin Ryu, Chris Paddack, Zach Davies, and Carlos Carrasco, for example) with bona fide openers (Deivi Garcia, Dustin May, Ryan Thompson) who were followed by primary pitchers, and also bullpen game starters (including, inadvertently, Mike Clevinger). While it may be something of an apples to oranges comparison with any year before 2018, the year that the opener came into vogue, this year’s best-of-three round has created its own type of fruit salad anyway.
Where starts of at least six innings were three to four times as common as such short starts in 2015, ’16, and even ’19, the rates of those types have been converging, last year aside; those “long” starts have become about half as common as they were just five postseasons ago. Meanwhile, for the first time in this span, starts of two innings or fewer outnumbered those of seven innings or more, 11 to seven, a count that does not include Lance McCullers Jr. going seven in Monday’s ALCS Game 2 (the three LCS games have boosted the average to 4.48 innings per start, in case you’re wondering).
We haven’t even discussed actual pitcher performance, but as you might expect from the lower innings per start, postseason starters aren’t doing great:
| Year | SP ERA | SP FIP | SP ERA- | SP FIP- | RP ERA | RP FIP | RP ERA- | RP FIP- |
|---|---|---|---|---|---|---|---|---|
| 2015 | 4.33 | 4.17 | 108 | 108 | 3.55 | 3.40 | 88 | 88 |
| 2016 | 3.88 | 3.86 | 113 | 109 | 2.88 | 3.16 | 84 | 89 |
| 2017 | 4.08 | 4.32 | 101 | 101 | 3.97 | 4.20 | 99 | 98 |
| 2018 | 3.90 | 3.96 | 104 | 103 | 3.60 | 3.75 | 96 | 97 |
| 2019 | 3.36 | 3.91 | 90 | 97 | 4.24 | 4.23 | 113 | 105 |
| 2020 | 4.52 | 4.55 | 110 | 106 | 3.74 | 4.02 | 91 | 94 |
A couple of caveats are necessary here. First, I’m using the regular season FIP constants, which are calculated so that a league’s ERA and FIP reconcile for that season, but that isn’t the case within the limited samples of games here. The ERAs have generally been lower than the FIPs, which is because scoring is lower in the postseason relative to the regular season, a fact you probably knew already; this year, teams scored 4.65 runs per game during the regular season, and 4.44 during the first two rounds (more on that subject shortly). Second, for the ERA- and FIP- metrics, I’m not using any park factors; these are just the split ERAs (or FIPs) relative to the overall league ERAs (or FIPs).
Still, what stands out is that last year was a significant aberration in that starters far outperformed relievers. Every other year within this span, it’s the other way around, and thus far this year, the gap has been wider than any year since 2016.
Turning to the offenses, first it’s worth noting that the scoring drop-off we saw through the first two rounds was actually less than usual, even with the stifling performances of the Braves, who entered the NLCS having allowed five runs in five games, and the Dodgers, who likewise allowed just 11 runs over the same stretch:
| Year | Reg R/G | Post R/G | Dif |
|---|---|---|---|
| 2015 | 4.25 | 4.36 | 2.6% |
| 2016 | 4.48 | 3.70 | -17.4% |
| 2017 | 4.65 | 4.29 | -7.8% |
| 2018 | 4.45 | 4.00 | -10.1% |
| 2019 | 4.83 | 4.03 | -16.6% |
| 2020 | 4.65 | 4.44 | -4.5% |
| Avg | 4.54 | 4.14 | -8.9% |
If we include the games to date from the ALCS and NLCS, this year’s level drops to 4.28 runs, a dip of 8.0%, but it should be noted that the participating teams have had the chance to line up their rotations to put some of their top starters out there thus far, and we shouldn’t expect those series to continue producing just 2.5 runs per team per game.
As for homers, after record levels during the 2019 regular season, we saw something of a drop-off this year, but it was still the second most homer-happy season in history. Lo and behold, even more balls flew over the fences in the first two rounds, that despite the shift to the cavernous neutral zone of Globe Life Field for the Dodgers-Padres series, which — gasp — featured just three homers in three games:
| Year | Reg HR/G | Post HR/G | Change | Reg HR/PA | Post HR/PA | Change |
|---|---|---|---|---|---|---|
| 2015 | 1.01 | 1.26 | 25.1% | 2.7% | 3.4% | 26.2% |
| 2016 | 1.16 | 1.01 | -12.2% | 3.0% | 2.7% | -9.7% |
| 2017 | 1.26 | 1.37 | 8.9% | 3.3% | 3.7% | 11.7% |
| 2018 | 1.15 | 1.08 | -6.4% | 3.0% | 2.8% | -8.4% |
| 2019 | 1.39 | 1.28 | -8.0% | 3.6% | 3.4% | -6.3% |
| 2020 | 1.28 | 1.41 | 9.8% | 3.5% | 3.7% | 6.7% |
| Avg | 1.20 | 1.24 | 3.2% | 3.2% | 3.3% | 4.1% |
Through the first two rounds, we saw a higher rate of home runs per game and per plate appearance than we’ve seen in any regular season — an increase of nearly 10% relative to this year’s regular season per-game rate, though a bump of just 6.7% if we go by plate appearances (note that I’m not referring to the percentage point gains here, which are much smaller). By either measure we’ve seen some rather drastic swings relative to regular season rates, sometimes positive, sometimes negative, and while most of it comes out in the wash, the reality is that we’re seeing more homers. The increase stands out more if we factor in the higher strikeout rates, which are a product of both better pitchers taking a greater share of innings, and relievers taking a greater share than usual:
| Year | HR/Con | HR/Con | HR/Con Dif | Reg K% | Post K% | K% Dif |
|---|---|---|---|---|---|---|
| 2015 | 3.8% | 5.1% | 33.1% | 20.4% | 24.2% | 18.7% |
| 2016 | 4.4% | 4.1% | -5.6% | 21.1% | 24.4% | 15.3% |
| 2017 | 4.8% | 5.7% | 18.9% | 21.6% | 24.7% | 14.3% |
| 2018 | 4.5% | 4.3% | -2.7% | 22.3% | 24.7% | 11.0% |
| 2019 | 5.4% | 5.4% | -1.1% | 23.0% | 26.1% | 13.9% |
| 2020 | 5.3% | 5.8% | 9.7% | 23.4% | 25.1% | 7.3% |
| Avg | 4.7% | 5.1% | 8.8% | 21.8% | 24.9% | 14.2% |
Again, I’m working with percentage gains rather than percentage point gains in order to emphasize the relative increases and keep them on the scale of the previous table. Where by home runs per plate appearance this year’s first two rounds saw a 6.7% increase, it’s up to 9.8% if we factor in contact rates, because pitchers are missing more bats — though the extent of the increase there is less than it’s usually been.
Naturally, given what we’ve witnessed in recent years, this leads to speculation about whether the ball is juiced, or at least has different characteristic from what we saw in the regular season. On that subject, I’ll defer to Rob Arthur, who earlier this season observed that the 2020 baseball was less aerodynamic than last year’s ball and that its drag was inconsistent (which isn’t to say that 2020 hasn’t been a consistent drag in the grand scheme of things). Writing on October 9, he noted that while there hasn’t been any “playoff surprise” like last year, when it appeared that the postseason ball had been dejuiced, we’ve seen a more juiced ball this postseason, albeit via an as-yet-unexplained spike in exit velocity rather than drag.
That trend will certainly take more data to iron out, and indeed, likewise for the trends above. Thus far we’ve seen downturns in starting pitcher usage and overall scoring rates, but higher home run and strikeout rates. As a result, a greater share of the runs scored are coming via home run:
| Year | Reg R/HR% | Post R/HR% | Dif |
|---|---|---|---|
| 2015 | 37.2% | 41.7% | 4.5% |
| 2016 | 40.2% | 42.5% | 2.3% |
| 2017 | 42.3% | 50.9% | 8.7% |
| 2018 | 40.2% | 42.4% | 2.3% |
| 2019 | 45.2% | 46.6% | 1.4% |
| 2020 | 43.7% | 50.2% | 6.5% |
| Avg | 41.3% | 45.9% | 4.6% |
Already in the two LCS we’ve seen 11 out of the 15 runs scored come via home run, so it’s possible this trend will set a record in that regard. And while it probably shouldn’t need to be said except as a corrective to some of the handwringing in the broadcast booths, teams that out-homer their opponents are now 24-1 including the two LCS. I strongly suspect that will remain the case through the two Championship Series and the World Series, though given the lack of off days in the LCS, we’re likely to see the innings per start rise, as it’s untenable for teams to overburden their top bullpen arms night after night. We’ll check back in on some of these trends as October continues.
