Pre-Musings - CheeseIsGood's 2020 Daily Fantasy Baseball Prep: Part 2
To get ready for this MLB season, I wanted to walk through the stats that I am going to be referencing in the Million Dollar Musings as part of Premium this season. If you have a lot of experience with MLB DFS, or if you’ve been reading my Musings the past few seasons, most of this is going to be review. But I want to make sure we’re all on the same page so we can hit the ground running this spring.
- Build DFS Lineups Like a Pro!
- Access to Content and LineupHQ
- MMA, KBO, Soccer and PGA
The first thing I’ll say is that you do not need to be overwhelmed by the sheer volume of stats available in baseball. It is unlike any other sport in the amount of data that we have to consume. You do not have to be an expert in calculus or scouting to be successful in DFS. My goal in the Musings is to help you become a better DFS player by deepening your understanding of baseball stats and analysis and finding ways to use them in your game.
You can find Part 1 of this series here. Today, we’ll walk through pitching, which is where I’ll start each day.
Over the course of a full season, baseball is one of the most predictable sports, as we have so many stats and metrics to measure a player’s skill. However, due to the small sample size of one game, one at bat at a time, baseball is also the least predictable over the short term. Simply put, there is a lot of randomness in what happens once a ball is hit into the field of play. Trying to predict exactly the outcome of a ball in play is a fool’s errand. So, what I like to do is narrow things down to the things that we can project. You may have heard the term ‘Three True Outcomes’. These outcomes are a walk, a strikeout and a home run. There is no unpredictable randomness that can happen with these three outcomes. It doesn’t matter how the fielders are spread out, how big the outfield is, how slow the runner is, how quick the shortstop’s reflexes are, and on and on. Strikeouts will always be my starting point, so let’s start there. (I think that’s what a starting point means?)
Strikeouts, Strikeouts, Strikeouts. Walks, Walks. Innings.
There is simply no guesswork needed with a strikeout. The hitter is out and the pitcher gets a strikeout. Done. On top of the obviousness of that, strikeouts are also the single most projectable stat in the game. A player’s strikeout rate stabilizes quicker into the season than other stats and carries over from year to year more consistently than other stats. This is not at all to say that strikeouts don’t come with plenty of variance of their own, but it’s much less than anything having to do with batted balls. If I were only able to look at one metric every day, it would be the pitcher’s strikeout rate.
Walks are next in line to strikeouts in consistency and projectability. This is why plain old boring strikeouts and walks are the starting point of everything I do from a pitching angle. The very first thing I have done every day since sometime around 2009 is gather a list of the day’s starting pitchers and sort by strikeout and walk rate. So simple. Rather than re-hash all of this, I am going to link back to this article that I wrote last year. It takes a very basic look at how the top strikeout pitchers and the top DFS pitchers and how easy it is to find these pitchers based on nothing but strikeouts and walks:
So, that’s where it’ll start every day, with the strikeout and walk rates for the pitchers. For a basic guideline, the league average strikeout rate for starting pitchers is around 22%, with the average walk rate around 8%. In general, a 30% strikeout rate is elite, with just 13 starters topping that mark in 2019, while a total of 26 starters had a strikeout rate of 27% or higher. If you’re going to spend big money on a pitcher, you need to have that type of upside. On the flipside, as often as possible, I’ll try to avoid pitchers with strikeout rates below 20%. There will be exceptions to that based on salary and other factors, but when in doubt, I side with higher strikeout rates.
The next factor that becomes increasingly important every year is innings pitched. We’ve seen the average start get shorter and shorter over the past few seasons, which affects everything from strikeout upside to the potential to pick up wins. In DFS, not only do we get points for each additional inning, but more innings equals more strikeouts. I will talk about this in terms of both innings and number of pitches thrown. With some teams and pitchers, it’s not that they are on a specific innings limit as much as they are on a pitch limit. In general, 100 pitches in a start is the number I’m shooting for, though over the past couple seasons, that has become rare enough that 90 pitches is an acceptable number for a lot of starters. If we don’t regularly see a starter going 5+ innings with 90+ pitches, I am not going to be willing to pay a premium.
The Luck Factors – BABIP and HR/FB%
You will often hear me reference these two stats as “luck factors.” While that is not completely fair, these two stats have the most randomness and unpredictability for pitchers. BABIP stands for Batting Average on Balls in Play. While there are bits and pieces of BABIP that can be explained by hard hit rate and batted ball type, for the most part, all pitchers over time will regress towards league average BABIP. From season to season and start to start, there is basically no consistency whatsoever in BABIP. The league average BABIP is right around .295, and every pitcher over a long enough sample size can be expected to fall within a fairly tight range around that number, in the .270-.320 range. In general, fly ball pitchers will have lower BABIPs and some pitchers have shown a consistent enough ability to limit hard contact that they can often beat that league average, but when a pitcher has a stretch of unusually good or bad starts, it can often be attributed to nothing more than obvious good or bad luck on BABIP.
In 2019, there were 113 starting pitchers that had 100+ innings. Only 13 had a BABIP above .320 and 20 had a BABIP below .270. If we go back to 2018, there were 128 starters over 100 innings, with 14 above .320 and 24 below .270. More convincing than those numbers are the fact that the list of pitchers who over or under perform the league average each year are completely different. There was only one pitcher above .320 both seasons and only two below .270 both seasons, but there were also two pitchers who showed up on opposite ends of the list:
Max Scherzer 2019 – .321 BABIP
Max Scherzer 2018 – .265 BABIP
Blake Snell 2019 – .343 BABIP
Blake Snell 2018 – .241 BABIP
If you think Scherzer and Snell were great pitchers in 2018 and jabronis in 2019, then I’ve got an amazing hair restoration product to sell you. Keep those Blake Snell numbers in mind, as we’ll be coming right back to those in a minute.
More so than the year to year BABIP, the start to start BABIP numbers are wildly unpredictable and all over the map. This is the main reason I don’t care at all when a pitcher has a few bad starts if those starts are just based on a high BABIP. As long as the strikeouts and walks are still lining up, the BABIP will come back down.
As for HR/FB%, this is a little harder to grasp, as there is a little bit of a wider bar between pitchers, in addition to the fact that the league average number has grown exponentially the past few seasons. But still, on the pitching side, there is much more randomness than you might think to the amount of home runs a pitcher allows. To get a baseline, here is the league average HR/FB% for starting pitchers over the past few seasons:
2016 – 13.3%
2017 – 14.2%
2018 – 13.1%
2019 – 15.5%
As I touched on in the Pre-Musings Part 1, the HR/FB% flew up in 2019 due to the actual baseball being used in MLB. Moving into this season, we don’t yet know which version of the ball will be used in 2020, but we’ll find out quickly, and the average HR/FB% is likely to land somewhere in the 13-15% range.
As we saw with BABIP, there is almost no consistency from year to year, much less start to start in which pitchers have high or low HR/FB%. Keeping in mind the league average increase in 2019: In 2019, 17 starters had a HR/FB% above 19%, with 13 starters below 11%. In 2018, 15 starters had a HR/FB% above 17% while 18 had a HR/FB% below 10%. Two pitchers appeared on the list for the high end and three on the low end for both seasons. It’s not quite accurate to say that a pitcher has zero input into HR/FB%, but it is an extremely variable stat from start to start.
It is much more about the plain old amount of fly balls a pitcher allows. The key to not allowing home runs is in inducing ground balls more than in limiting the percentage of fly balls that go over the fence. Here is a very interesting factoid that highlights that point. In 2019, the HIGHEST HR/FB% among all starting pitchers belonged to Dallas Keuchel at 23.9%. But, even with the highest HR/FB% in the league, Keuchel was all the way down at 64th in HR/9. Wow. Why do think that was? It’s very simple: Keuchel is a ground ball pitcher. He allowed so few fly balls that even with that unlucky HR/FB%, he was still not allowing very many homers.
The other factors we’ll discuss that affect HR/FB% are ballpark and weather factors. While the pitchers themselves don’t affect HR/FB% much, it is very affected by park factors. Every pitcher will have a higher HR/FB% in Cincinnati than they would in San Francisco. Overall, when looking at pitchers and the home runs they allow, I am not targeting pitchers with low HR/FB% or avoiding pitchers with high HR/FB%. I care more about their overall fly ball rate and ballpark factors, while expecting their HR/FB% to land around the league average over time.
ERA = Meaningless Nonsense?
ERA is one of the most used stats in the game. It is what is often first referenced by announcers or in articles talking about the effectiveness of a pitcher. However, ERA is much more about telling what happened in the past than what is going to happen in the future. As mentioned up top, there is a lot of randomness once a ball is hit into the field of play, and that causes ERA to be a poor predictor of future performance. Luckily for us, there have been several metrics created that are more accurate indicators of future performance. These are known as “ERA Estimators.” There are many different versions, using slightly different weights to categories, but all of them are a more accurate reflection of a pitcher’s skill than his actual ERA. The main reason is because they strip out the luck factors that we talked about above and show us what would happen to a pitcher’s ERA if their batted balls (BABIP, HR/FB%) were at league average. The three ERA estimators that I will reference in the Musings are FIP, xFIP and SIERA. Here is a quick summary of what they are. The actual equations for figuring out these metrics are quite in depth and can be found online if you care, but essentially this is what they boil down to:
FIP – ‘Fielding Independent Pitching’ – Regresses BABIP to league average.
xFIP – ‘Expected Fielding Independent Pitching’ – Regresses both BABIP and HR/FB% to league average.
SIERA – ‘Skills-Interactive ERA’ – Regresses both BABIP and HR/FB% while also factoring in ground ball/line drive/fly ball data.
In the simplest terms, if a pitcher has a low ERA and a high FIP, xFIP, SIERA, that pitcher has been lucky and is likely to be worse moving forward. If a pitcher has a high ERA but low ERA estimators, he is likely to improve moving forward.
Remember a few sections ago, we saw that Blake Snell went from one end of the BABIP luck spectrum to the opposite end from 2018 to 2019. This is a very good example of how all these stats work together to show us how to find the true talent level of a pitcher.
Blake Snell 2018 – 31.6% K, 9.1% BB, .241 BABIP, 10.7% HR/FB%, 1.89 ERA, 3.30 SIERA
Blake Snell 2019 – 33.3% K, 9.1% BB, .343 BABIP, 15.4% HR/FB%, 4.29 ERA, 3.56 SIERA
Just from the ERA, we would assume Snell was a superstar in 2018 and a scrub in 2019. But his strikeout rate was slightly better in 2019 and his walk rate was the same. The big difference in the two seasons was just the crazy swing in BABIP and a more moderate swing in HR/FB%. You can see that his SIERA shows that he had fairly close to the same skills in both seasons, and moving forward, he is more likely to land somewhere around that 3.30-3.65 ERA. With swings this big in full seasons, it’s easier to see how the swings can be even bigger from start to start. Some days, a pitcher is going to allow 15 ground balls and 10 of them find a hole between two fielders, and the end result will look like he had a bad start, when really he did absolutely nothing wrong or out of the ordinary.
I will reference SIERA most often, and will always lean to these ERA estimators ahead of actual ERA.
The Tangled Mess of Batted Balls
GROUND BALLS / FLY BALLS
Strikeouts, walks, innings. These are what they are, and they are somewhat predictable. But once that ball is hit, yikes! We discussed the biggest cause of variance, BABIP, but that doesn’t mean that there is nothing we can learn from batted balls. There is one thing that is very much under a pitcher’s control, and that is ground balls vs fly balls. As wide open and random as the leaderboards are from one season to another with BABIP and HR/FB%, they are remarkably consistent in the highest GB% and FB% for pitchers. After strikeouts and walks, ground ball rate and fly ball rate are the next thing to stabilize in a pitcher’s skill set. Let’s take a look at league averages to get a starting point:
2017 – 44.2% GB, 35.5% FB
2018 – 43.2% GB, 35.4% FB
2019 – 42.9% GB, 35.7% FB
As the launch angle revolution and obsession with home runs has continued, the average ground ball rate has crept down slightly over the past few seasons, but you can see from these numbers that there isn’t much movement here. Anything in the 40-45% range is an average ground ball rate, while fly balls hold steady around 35-36%.
I will refer to anything around 49% or higher as a high ground ball pitcher. 60% is an extreme mark that only one or two starting pitchers hit in a season. 2019 saw 20 starters above a 49% ground ball rate, with 22 on that list in 2018. One very important note is that while ground balls are a skill based on pitch type and approach, major league pitchers do have the ability to change their style at any time. When a pitcher sees an extreme change in ground ball or fly ball rate, it is usually not random, but rather a change in their actual pitching style. For example:
Stephen Strasburg 2018 – 43.6% GB
Stephen Strasburg 2019 – 51.1% GB
Strasburg is one of the most talented pitchers in the game, and he can pitch any way he darn well pleases. If he wanted to strike out 35% of batters and allow 50% fly balls with more homers, he could do it. If he wanted to get 60% ground balls with average strikeouts, he could do that. Using data from brooksbaseball, in 2018, Strasburg threw 45% 4-seam fastballs and 7% sinkers. In 2019, he threw 29% 4-seamers and 19% sinkers. It’s really just that simple. He got more ground balls intentionally in 2019. There is no need to get overwhelmed with all this pitch type data if this is not something you enjoy digging into or have time for; that is why people like me are around. I’ll keep you alerted if there is a noticeable change in someone’s pitch type, but if you find this interesting, you can learn a lot from digging into these numbers.
Before we go any further, I want to stop and dispel a common misperception. While it is generally true that when all else is equal, a ground ball pitcher is preferred due to lack of home runs, there is nothing inherently bad about being a fly ball pitcher. In my opinion, any pitcher who is intentionally getting the type of contact he wants is doing it right. Justin Verlander is one of the highest fly ball pitchers in the league. I assure you that is no accident, and it is not bad that he allows fly balls.
I consider anyone over 40% FB to be a fly-ball pitcher, with 45% and higher being an extreme fly baller. We also need to pause here and point out that all this batted ball data is separated by handedness of the opposing batter. Some pitchers throw a different pitch mix against righties and lefties, so you will see guys who are fly ball pitchers to righties and ground ball pitchers to lefties. This data is already going to be nicely sorted out for you to see on Plate IQ, and if you don’t have the time or interest in digging into it, this is another area where we’ll have you covered here in the Musings.
HARD HITS
The next step to batted ball data after ground balls/fly balls is hard hit rate. This is also where things start to get quite tricky and with a lot of unknown. There are pitchers who have shown a consistent ability to limit hard contact, but it is not nearly as much of a bankable skill as it is on the hitting side. Essentially, while the pitcher has the most control over any at bat as far as strikeouts, walks, ground balls and fly balls, the batter is the one with more control over hard hit rate. Like HR/FB%, this is another stat that has been increasing at an alarming rate the past few seasons. Making matters more difficult, there is more than one way to measure hard hit rate. I will most often reference the numbers from fangraphs. Here are the league average numbers:
2017 – 31.8%
2018 – 35.3%
2019 – 38.0%
That is a clear and obvious increase year over year and there’s no reason to think it’s any accident. Players are stronger, pitchers are throwing harder, and batters are being coached to swing like it’s going out of style. While we don’t know if these numbers will continue to climb, we’ll always be comparing one pitcher to the next, so regardless of the league average, a lower number is better for a pitcher. But the real bug-a-boo here (is bugaboo one word? I like the hyphens. Speaking of that, did you know a dash and a hyphen are not the same thing? Sorry, I digress). Again I say, the real bug-a-boo is the lack of consistency for the majority of pitchers in hard hit rate. Unless a pitcher has shown similar skill year over year, I am hesitant to overvalue hard hit rate, because again, the hitters have more say in that. However, there is no question at all that certain pitchers have a real, projectable skill in this area, and I will talk about this in the Musings when sorting between pitchers with similar skills in other areas.
Overall, when it comes to batted balls, I am most interested in matching up ground balls or fly balls vs opposing teams in a favorable way. For example, a ground ball pitcher against a low power ground ball hitting team doesn’t have much power downside. Even better if that pitcher has a low hard hit rate. We’ll talk about this more in the next Pre-Musings with the batters, but fly ball vs fly ball is actually generally a favorable situation for the pitcher. The more extreme fly balls become, yes home run rate goes up, but fly balls that are not home runs are almost always outs and fly ball pitchers against fly ball hitters is what leads to the most easy pop ups. The thing I want to avoid is pitchers with no discernible skill in inducing a certain type of batted ball against a team with hard hitting batters with either line drive or fly ball ability.
- Build DFS Lineups Like a Pro!
- Access to Content and LineupHQ
- MMA, KBO, Soccer and PGA
Let’s Do Some Cliff Notes!
It’s never too early in the year for Cliff Notes. If you’re new here, you may think that I am shortchanging world famous Nebraska native Clifton Hillegass, the creator of CliffsNotes. He agreed that CliffsNotes was a dumb way to say it, and he is all on board with Cliff Notes. He is the real hero.
There is a good chance that your eyes glazed over in some of my rambling, so let’s boil down these pitching stats to a simple list of what you can expect to see in the Musings this year. In order of importance:
1) K% – Strikeout Rate
2) BB% – Walk Rate
3) Innings/Pitch Count
4) ERA Estimators – SIERA, FIP, xFIP
5) BABIP and HR/FB% – The ‘luck factors’
6) GB%/FB% – Ground Balls / Fly Balls
7) Hard Hit %
Again, you do not to become an expert in all these areas, but I do hope to be able to help you gain a better understanding of what makes certain pitchers great, and other pitchers, well, for lack of a better term, Chris Tillman.
I’ll be back next week with a look at the hitting side of things as we continue to prepare for Opening Day 2020!
Pre-Musings Part 1
Pre-Musings Part 2
Pre-Musings Part 3
Pre-Musings Part 4