In Japan, you bond with your coworkers by going to an enkai. You say it like “N-kai,” and the Kai is as in Cobra Kai, not “Kay” as in, “I’m so American I can’t envision speaking any language than N-glysh.” The enkai is usually where the office (not the TV show) gets together for some BBQ and beers. Everybody goes to the enkai, even if they don’t like beer and BBQ. Sometimes, Japanese managers conduct official office business at the enkai, which makes it somewhat awkward when they pass out documents to read while you’re navigating grilled meat. Then comes the nijikai, or “second meeting,” when the cool people leave the squares behind, and they go to a place where there’s more beer, some snacks, and some karaoke. But for the bold — and those who are truly initiated into the office — there’s the sanjikai, or the “third meeting.” This is the event where only the hardened drinkers, the Buddhist teetotalers, and those people trying to forget the horrors of 9-5 hyper-capitalism are found. You don’t talk about the sanjikai the next day, or the next week, because you’re likely not going to remember all of it. For the sanjikai, you’re out in the back alley bars not because you have to, but because you want to. Because you’re driven, by some supra-rational urge that verges on the paranormal, to see what’s hidden behind the curtain of reality. You want to suck the marrow out of life and maybe get a bone splinter in your gums. You know the world has made its mark on you, and you want to mark the world before you transit off this plane of existence into the cosmos, awaiting to be reborn on your ascent to Nirvana. You might not get home tonight if you go out to the sanjikai, but that’s OK because you’re willing to sit in the park and watch the stars until first train.
That’s what this third pre-season edition of the Top 100 Starting Pitchers is all about: You want to know what the others don’t know. You’ve probably drafted already, but you’re here to get prepped for who to pick up off the waiver wire and what to do for next year. You, my dear reader, are initiated into my office, and I invite you to the sanjikai of my weekly pitchers series.
This article is probably my greatest contribution to fantasy sports so far, and I hope you enjoy it.
We here at Razzball love making spreadsheets (see Room comma War) and sometimes we even like combining spreadsheets to make new data. Rudy created the Razzball Player Rater and has calculated the fantasy value of players historically back to 1903. Then Coolwhip (our weekly outfield writer and in-house meme creator) got the bright idea to combine the hitter/pitcher totals from the past five years to give an idea of just how valuable pitchers were in the overall scheme of things. That gave us a way of telling how valuable pitchers are compared to hitters for your typical fantasy league. Then, something fortuitous happened: the NFBC released their historical baseball ADP going back to 2003 just this week. Now we can look, with hyper-precision, at how pitchers performed relative to their ADP and the best practices for drafting them in our fantasy leagues.
And the answer is: they’re not worth it. Thank you for coming to my TED Talk.
I suppose I need a thesis statement. What would I have told my students to write? This is my paper about how to draft pitchers good and do other things good too. OK, I suppose that’s why I’m working in fantasy sports now instead of the ivory tower.
Or maybe, it’s time, now that I’ve been writing fantasy sports for a year instead of diving through archives, that I write my fantasy sports defense. Where I plant my stake in the ground and defend it, and maybe next year somebody submits it to the FSWA and I add another line to my CV. “Award winning guesstimator, EverywhereBlair!”
Nah, let’s just keep the punk aesthetic and DIY. Pitchers don’t matter as much as people tell you they do, and this is an article about it and I’m going to teach you how to do your own research. Nothing messes up “the man” like 7,000 words of teaching people how to achieve enlightenment on their own.
Buddha Blair has taken his seat under the tree, yet he denies the ascent into Nirvana, for he has decided to remain on this mortal plane in order to help others draft pitchers good and do other things good too. He has become a Lama. All he asks is that you consider the conditions necessary to have people perform at their optimum potential. Yes, I’m talking more than just pitchers.
Let’s put it this way: it’s March Madness right now. Did you know that there’s only a 1 in 9.2 quintillion chance that you get the bracket 100% correct? And there’s only 64 teams in March Madness.
There were 252 starting pitchers who played in MLB baseball in 2020. So, even in just making a Top 100 list, there are 1×10^72 combinations that could fill out the Top 100 list. Makes this ranking thing seem kind of meaningless, right? Why do we care about accuracy of rankings anyway? Why do we care about who goes above who? I could generate a new list every second until the sun goes black and still not get the rankings perfect. I mean, who had R.A. Dickey as SP1 in 2012? He was drafted 350th overall. But it happened.
Now! That’s not to say, “punt pitchers.” No, I’m not saying that at all. Saying that actually demonstrates a complete misunderstanding of what punting even is in football. Punting is just the transition from offense to defense and is a critical element in both stages of the game. What I’m saying, is to be smart about your pitcher choices for maximum return on investment.
Because last year, even I — an exiled professor hiding in the pandemic — managed to predict that Shane Bieber would be SP1. While there are countless iterations of who goes where in the rankings, there are easily predictable trends of how we can organize those players into meaningful categories based on where they could possibly land. And we can generally get a good idea of the kinds of statistics that give a player elite fantasy value. In other words, humans are pretty good at identifying the first round talent that will give you the best chance to win your fantasy league.
With one exception: pitchers. Pitching ranking and draft strategy is completely off-base. And part of the problem is because this whole “ranking” thing we do — that even I do — has become commodified behind the guise of “accuracy.” Somehow, we’ve come to venerate systems that mollify audiences with “consensus,” as if we [points to the community] can somehow agree on a specific hierarchy of pitchers when there are more than 1×10^72 possible combinations of rankings. In a community of 10 million fantasy baseball players, why are there only 50 rankers, and why do their ranks all look the same? And why are almost all of them wrong about pitchers?
In other words: ranking rankings is completely meaningless because we shouldn’t be looking at pitchers in a hierarchical manner in the first place. We should be looking at them in a probabilistic manner. And in order to do that, first we need to unlearn what the past decade of the commodification of the fantasy baseball field has taught us as “the accurate ranking.” Because what we’re being taught and sold about pitchers, isn’t accurate.
The next part is going to be a lot of data, and I won’t blame you if you want the TL;DR. Here’s the bullet point summary followed by the details:
- There is no “safe” SP1. There is no “safe” Top 3 SP. There is no “safe” Top 5 or Top 10 SP, either. You should be thinking of pitchers more in quantum theory terms than in high school algebra terms. It’s all about probability, not hierarchical ranks.
- You need to be prescient to pull off pocket aces successfully. When you do pull it off successfully, it is still usually less effective than drafting whatever hitter happens to be nearby in ADP.
- A number of the top pitchers do not meet their projections due to injury. Often, these major injuries are foreshadowed by injuries in the off-season or pre-season.
- Drafting multiple pitchers early in a draft serves only to diversify against injury at the cost of hitting. It is almost never a superior strategy to just drafting hitters.
- Drafting pitchers early (i.e., 1st round, pocket aces, acefecta) almost always puts teams at a suboptimal position. I asked Rudy about this, and he framed it thusly: you should only be playing suboptimal if you have a specific strategy that you are nearly certain will later compensate for your suboptimal start.
- This article is a critique of Average Draft Position (ADP), the aggregate ranking systems that replicate and reinforce ADP, and the industry that has developed around the commodification of rankings. It is my hope that this study begins your process of liberation from the constraints that ADP place upon fantasy players. I recommend using well-evidenced non-aggregate sources in order to perform better at fantasy baseball, such as the Razzball/Steamer projections, and The BAT/The BATX. ADP should be a guide only for when you can draft a player, not how valuable that player is.
- From doing this study, I reworked my own pitcher rankings based on a simpler formula that was more rooted in probabilistic yet aggressive metrics, and these are supplied at the end of the article.
ADP vs Results Study
This project was made possible by the recent appearance of two pieces of data: the first was Coolwhip synthesizing Rudy’s historical player rater, and the second was the NFBC providing historical ADP data going back to 2004. I downloaded each and compared the top 70-ish players by ADP from 2017-2020. This gave me an idea of the top 5 rounds of players going back 4 years. Want more data? Then sign up for the Razzball premium fantasy baseball tools to show support for the site and send a note to Grey that you love me and want me to do more stuff like this. :)
I’m breaking down the points from my study into a statement and the implications for fantasy managers.
1) Consensus SP1 from 2017-2020:
As ranked by ADP, the consensus SP1 in 2017-2020 never returned first round value. Clayton Kershaw came the closest in 2017, when he finished as SP2 with a second round value. Because Kershaw’s ADP was 4th overall, his actual value of 2nd round still represented 14 other players who would have been a better selection.
The consensus SP1 also has, within the confines of this study, never finished as SP1. The eventual SP1 was usually found in round 2 or 3, and one time in round 10.
This next part gets fuzzy, so let me state this for clarity: in recent history, ADP has not selected the proper SP1 in the first round and placed the proper value on them. It did happen a few times in the era prior to this study, so let’s elaborate on those circumstances for a sense of completeness.
In 2016, Max Scherzer was the consensus ADP SP1 and he finished with first round value. Except he had an ADP of round 2, so he’s a bit of an outlier to my thesis because I’m talking about pitchers in the first round. In 2015, consensus ADP SP1, Clayton Kershaw, finished with a first round value, but the actual SP1, Jake Arrieta, had an ADP of 100, and SP3 Zack Greinke had an ADP of 49. So, almost but not quite.
I found the smoking gun in 2014: Clayton Kershaw was consensus ADP SP1 AND he finished as overall #2 in fantasy baseball. So, the last time a consensus SP1 drafted in the first round AND returned first round SP1 value was: 2014. But, guess who was SP2 in 2014? Jonny Cueto, with an ADP of 171. Felix Hernandez finished SP3 with an ADP of 43. So, the point still remains: the ADP SP1 is almost never accurate, is almost never accurately valued, and is surrounded by other pitchers who return nearly the same value for dramatically reduced draft capital.
While I’m at it, I went back all the way to 2004 and found one more instance of consensus ADP SP1 being drafted in the first round and returning SP1 and first round value: Johan Santana in 2006. So, two times from 2004-2020 have “we” [points at all of us in the fantasy community] gotten the SP1 both correct in our projections and drafted at a proper value. One time in 2016, that SP1 returned surplus value.
1a) Conclusion and Implications for 2021:
Jacob deGrom is considered the consensus SP1 for 2021 fantasy baseball: he is both the SP1 by consensus ADP and by consensus rankings that use aggregate projections. If we take the historical scope of this study to be an indicator of a player’s certainty to not finish where they are expected (as opposed to their likelihood to finish), he is almost certain to not finish as SP1 and to not have first round value. We would expect deGrom to finish, at best, with a 2nd round value. Later in this article, I will point to innings pitched (IP) as a major factor in predicting a top finish for a pitcher, and deGrom was 21st in IP in 2020; he is projected for about 5% fewer innings than some other elite pitcher options in 2021. Jacob deGrom will very likely be a good pitcher, but he is unlikely to return first round value. Therefore, deGrom’s current ADP of 6 is not warranted. I would extend this to SP2, Gerrit Cole, who has a similar profile to deGrom and already has several concerning “knocks” against returning value on his draft capital, such as his known use of pine tar.
If you are insistent upon finding this year’s true SP1, you will likely find that player somewhere in round 2 or 3 based on ADP. That pitcher will likely throw a lot of innings with a reasonable K rate on a good team. I would look toward Yu Darvish, Brandon Woodruff, Jack Flaherty, or Aaron Nola.
1B) SP1 Obsessions
History tells us that we’re almost always wrong trusting consensus sentiment on the SP1. So…[travels over to a well-known site with rankings that ranks rankers] why does every ranker have Jacob deGrom as SP1? And why do almost all of them have deGrom in their top 10 overall fantasy value, with 2 rankers putting him #1 overall? Only one time in the last 20 years — Johan Santana in 2004 — has a pitcher been at the top of the fantasy charts. I mean, I did this study, ostensibly, in my free time. A bunch of those guys have full time jobs, some of them with multi-billion dollar networks and media revenue.
I can’t be the first person to recognize this. Can I? No, because Razzball’s been saying this for over a decade. Maybe they just needed a failed professor to come in and, you know, write it up like a punk Buddhist tract. Maybe I can say this because I’m not trying to sell you anything. I’m not trying to make rankings into a commodity. I’m asking you to leap and unlearn what you’ve been taught. I’m asking you to be uncertain with me, to ask questions, and to leave open the possibilities that the game we play has space for messiness.
That there is no perfect roster. There is no perfect ranking. There is no SP1.
There are just players who are more likely to wind up as “most valuable” given the 162-game span that we call a baseball season.
I would contend that it’s the increasing conformity mentality of the fantasy baseball community that’s reinforcing this notion of false pitcher value. After putting out my first set of rankings, people jumped to lambast me for my Gerrit Cole ranking. I have him ranked lowest of any ranker out there, at the equivalent of 6th overall. Because when I look at Cole’s 2020, I see a 1.73 HR/9, an FIP of nearly 4.00, a precipitous drop in K/9, a career-low BABIP, a decline in swinging strike rate, and career highs in launch angle and barrel rates. Grey wrote a whole post about pine tar and stats leading to a Gerrit Cole 2021 fantasy bust. So, is it that I’m a bad ranker that I have Cole so low, or is it that the rest of the community is so enamored with the conformity of their rankings that they don’t see the obvious warning signs?
That, sure, Gerrit Cole has the potential to be SP1, but he also has the potential to be SP20? And that the needle of probability on where Gerrit Cole finishes is wavering more toward SP10-SP20 than it is SP1? But you can’t neatly put that in an Excel sheet and number it, can you? So we put Gerrit Cole at the top of the rankings — where everybody else puts him — because if he makes it to SP1, you are all right. Not like, you’re OK. But, you’re literally all the same ranking distance away. And if he makes it to SP20, you all miss equally. You’re all equally wrong in the eyes of the aggregator.
Imagine you’re in one of those contests where you have to guess the number of jelly beans in the jar, and everybody can see each other’s guesses. And everybody willingly decides to put in a guess +1/-1 of each other, such that there’s 40 guesses that say either 1,001, 1,000, or 999 beans. When there’s actually 40 beans in the jar, does it matter that everybody was off by a mile? What if there were 1,002 beans in the jar? What if somebody told you they had the most accurate bean counting system because they were off by .1% when the competition was off by .2%? And that they came to that conclusion because they watched everybody else count beans first and then took all those answers as their own? How much more would you pay for the elite bean counting system, even if you knew the bean counting system was just as likely to overshoot the jar with 40 beans as it was to get the jar with 1,002 beans right?
Because that’s the fantasy sports industry in a nutshell now. What would you say to the prescient ranker who had R.A. Dickey as SP1 in 2012? Would you shun them, or ask them for more information?
Support the DIY guys and gals and palz who are doing the research and math to make this game we love possible. Because otherwise, you’re going to find more and more results like I demonstrate in the next section.
2) Safety of Top Pitchers by ADP:
I see the claim that fantasy baseball players choose to draft pitchers early because they are the “safest bet to return value.” This is a false assumption that is not backed by the recent evidence.
In the table below, I demonstrate the “hit rate” of pitchers by ADP by year. I consider a “hit” to be a pitcher who performs either at or better than their draft capital, and a “miss” to be a player who performs at one round or worse than their draft capital.
|Year||Top 3||Top 5||Top 10|
As a reminder, this article is a critique of ADP and aggregate ranking systems that replicate ADP. The dismal return on investment of pitchers as supplied by “consensus thinking” — i.e., ADP — could be remedied by a proper ranking system that recognizes the failure of “groupthink” to produce successful pitcher draft picks.
However, because the recent movements to create aggregate ranking systems — which normalize outliers and actually revert aggressive rankings into passive rankings — fantasy baseball players are provided with rankings that actually recreate ADP rather than attempt to dissect and overcome the weaknesses of ADP.
I’m looking at an aggregate ranking system provided by a major provider right now, and it is identical to pitcher ADP across the major sites, with extremely minor differences (i.e. Snell and Woodruff swapped, Kershaw is the biggest deviation at 2 spots lower in ADP than in projections).
Now, far be it from me to tell you how to spend your money. Coming from an academic background, I have basically no idea how to get paid for my work, yet I love to do it. But you! Maybe you’re good at making money and you want to spend it wisely. Speaking just from my point of view, I couldn’t justify spending money on an aggregate ranking system that replicates and reinforces ADP because it normalizes “misses” instead of incentivizing properly valued “hits.” Making mistakes is fine; to err is human. No forecaster is perfect. But when the product being sold is more about normalizing misses (encompassed by the decline in pitcher prediction accuracy and simultaneous increase in pitcher ADP) instead of celebrating the outliers that provide maximum value, the consumer should question the intentions of the product that they are sold.
2a) Conclusions and Implications for 2021:
First, don’t draft a top 3 pitcher in the first round. Second, if you really want a top pitcher, you’re most likely to find them in the consensus top 5 (albeit, you’re still better tossing a coin on other, much more “affordable” pitchers).
You’ll notice that the trajectory of starting pitcher “hits” has declined precipitously in recent years. There are likely several reasons for this: 1) pitcher ADP inflation (there will be 4 SP with a first round ADP this year, compared to 1 in 2017); 2) short-season bias in 2020; 3) starting pitchers are throwing fewer and fewer IP recently (2020 proved to be the year where starting pitchers threw the fewest cumulative innings as a cohort, although this has been a trend for over a decade).
What am I doing with this information? I’m drafting SP even later than the consensus. If you don’t have the diamond hands for that, you’re best served rolling the dice on a Top 5 consensus ADP pitcher who represents the lowest ADP. In this case, that would be Yu Darvish or Aaron Nola.
3) Pocket Aces and Pitcher-Heavy Strategies
“Pocket Aces” is defined as starting the draft with two SP, and I’m seeing other options that have three or even four SP in the first 4-5 rounds. The mentality behind this is to “corner the market” on pitching stats. However, given the “miss rate” being 60-90% on the top 10 SP over the past few years, one would need to have immaculate foresight (or sheer luck) to “hit” on these strategies. The weakness of these strategies was evidenced by several tournament studies: no team in the 2020 RazzSlam with pocket aces finished in the top 20%; similar studies of various NFBC tournaments have revealed the same. The NFBC Main Event winner has drafted a hitter in the first round 17 years in a row, which matches this study’s findings that the top 3 SP return, at best, 2nd-3rd round value.
Even if you have immaculate foresight and nail pocket aces, you’re usually better served taking the hitters nearby. Using Rudy’s end-of-season valuations, here’s the spread of outcomes of “immaculate foresight” pocket aces compared to “immaculate foresight” hitters that were nearby in ADP.
|Year||Max Value Pocket Aces||Max Value 2 Hitters|
So, in 3 out of 4 years in the study, it would have been better to take the best two hitters rather than the best two pitchers.
But what if this is the year for pocket aces, you say? It very well could be. The problem is, it is so much harder to have immaculate foresight on those pocket aces due to the extremely high miss rate on pitchers when ranked by ADP.
In the 2019 example, where Justin Verlander and Gerrit Cole made up the immaculate foresight pocket aces, it would have been impossible to draft them together unless you took Verlander (ADP 22) in round 1 and Gerrit Cole (ADP 25) in round 2. Had you taken the “traditional” pocket aces at the turn of round 1/ round 2 — Jacob deGrom and Chris Sale, ranked by ADP — your valuation would have been 31.8 compared to the 83.6 that Verlander/Cole returned. In other words, you would have needed an aggressive ranking system that refuted ADP. To pull off the ideal pocket aces that year, you would have needed be very lucky or very aggressive. The other years, you would have been better to just draft hitters. Of course, one could have done the hitter/pitcher/pitcher strategy from the 1 or 2 slot, but that deviates from the “pocket aces” discussion by means of introducing the third player. Without diving deeply into the threesomes and foursome starts (hehe!), I hope you trust me when I say the numbers generally reflect the importance of hitters in the early rounds in favor of pitchers.
But if you’re just insistent on getting the ideal pocket aces, it will take roughly [loads up internet calculator] 36 fantasy baseball teams each drafting a different set of pocket aces to land the possibility of the immaculate foresight pair. So, if you’ve got that time and money, go for it. For everybody else, draft hitters.
3a) Conclusions and Insight for 2021
Due to the foreknown IP limits on so many pitchers, pocket aces are almost guaranteed to be less valuable to fantasy baseballers in 2021. I’ve seen some drafts with teams starting with 3 or 4 SP. Don’t be intimidated by this. Don’t change your draft strategy. Don’t give in to “positional scarcity” FOMO. At best, only 2 of those pitchers will “hit,” and you’re taking hitters that are — statistically — more valuable than those pitchers anyway.
Pocket aces should only be on your draft strategy radar if you’re the kind of person that enters 50ish competitions. However, you should know in order to land that single successful pocket aces pair, you’re probably going to lose your other 49 competitions. And even if you land your immaculate pocket aces, you’re probably losing that tournament as well. If that’s your thing, you do you. Everybody else: take advantage of the inefficiency in draft rooms and keep drafting hitters.
4) There are so many misses…are hitters that much better?
In short, yes. Hitters in the first round generally return their value at a much better rate than pitchers. Again, I’m critiquing ADP here and those ranking systems that replicate ADP. There are many other ranking systems that do not aggregate and are therefore capable of “hit rates” that are much higher than the data seen below.
Here’s the outcome of how many hitters vs pitchers drafted in the first round finished in with first round value:
That sneaky pitcher in 2018 who returned first round value was Max Scherzer, who was actually SP2 by ADP behind Clayton Kershaw. Otherwise, all the other consensus first round pitchers by ADP failed to return their draft capital value.
Hitters by first round ADP have a better chance of returning their value. Even when they “miss” first round value, they generally return 3rd/4th round value.
4a) Conclusion and Insight for 2021
This synthesizes with the above points: you are better off drafting hitters early in your draft. You are more likely to “hit,” and even when you “miss,” the return on investment is generally better than when you miss on a pitcher. Historically, when you “hit” on two batters — which is much easier to do than with starting pitchers — you will have a higher return on investment than a successful pocket aces start.
5) So What Do I Do For My Draft?
My historical process that I’m conducting here is a twist on what the prognosticators like Rudy Gamble and Derek Carty are doing. Instead of forecasting into the future, I’m looking at what kinds of stats put pitchers in the top 10 for fantasy baseball.
And I took the numbers from 2017-2020 and checked them against multiple types of correlations, and the result was surprisingly simple: the top fantasy pitchers almost always have top 10 finishes in innings pitched, K-BB%, and SIERA rates. Additionally, we combine this information with incoming news about player injuries, and our rates of success on pitchers jumps dramatically.
Thankfully, Rudy’s rankings with Steamer do project SIERA. What a happy coincidence! Maybe that’s why Razzball had the best pitcher rankings in 2020? Could be!
In fact, I looked back at Grey’s 2020 pitcher performances, and if he had amended them aggressively in response to the injuries heard about in the pre-season, you’d be surprised at the outcome. Already, Grey was the top starting pitcher projector in the country last year. If he had been a bit more aggressive, his numbers would have been astounding. So, I actively cut out the pitchers that we had injury reports on — Stephen Strasburg, Max Scherzer, Walker Buehler, Mike Clevinger — and here’s the table that results. Grey’s Top 100 returned proper value or surplus on 4/5 pitchers. Also, Jack Flaherty was on the Cardinals and the Cards lost 30% of their season to Covid at one point, and they played a ton of 7-inning double-headers to catch up, so don’t blame the system on that one. If we really want to play “What If,” you can take a look at Flaherty’s K-BB% and SIERA and project him for another, say, 3 starts, which would put him at proper value. In that hypothetical scenario, Grey would have been 5/5 on proper/surplus value on the healthy SP in his 2020 Top 10 pitchers. How ’bout them apples? [smashes a sticky note with his phone number against your window]
|2020 Projection||2019 IP Rank||2019 K-BB% Rank||2019 SIERA Rank||2020 Player Rater Finish|
So, what we have are some pretty nice “hits.” But it’s important to note where Grey placed these pitchers in his overall rankings, too. Gerrit Cole was 16th overall, and deGrom was 24th. Even though they were his top pitchers, he placed them outside of first round value. And, they returned second and third round value, respectively. Of these pitchers, only Bieber returned first round value, and he was drafted in the fourth round on average. So even though Grey was more-or-less accurate as he could be given the circumstances, a player who drafted Cole in the first round hoping for “certainty” would still have been missed out on a hitter who returned a higher overall value. The ideal start to a draft last year would have been hitter/hitter/hitter/Bieber.
As Grey and I noted many times last year, Shane Bieber was likely to be the pitcher to return the most value in 2020 due to his low ADP (4th round) and high performance (indeed, scoring as well if not better than deGrom in 2019). Those who listened to us would have been on track to receive the best possible start to their fantasy draft in 2020.
Overall Study Conclusion:
This study in pitcher valuation demonstrates that pitchers almost never return their draft value early in the draft. I recommended all fantasy baseball players — from casual to high stakes tournament players — to draft two hitters to start their draft. If you are using a trustworthy ranking system that is unconcerned with ADP, you may certainly have an effective selection of pitchers in the second or third rounds. However, the evidence does not endorse the selection of a pitcher in the first round in any circumstance. There is no historical evidence whatsoever that pocket aces should be an optimal start, and that draft methodology should only be used as a suboptimal beginning when the player is certain that they can get superior hitters later in the draft.
I’ve said it many times and I want to reiterate: this study is a critique of ADP and those ranking systems that replicate ADP and/or minimize deviations from the norm. High stakes and tournament players would likely benefit from ranking systems that deviate from the norm, because successful projection systems enable a more efficient draft that has a higher likelihood of winning against a strong field. Casual players would also benefit from this approach because it liberates them from following groupthink, thus making their drafts more individualized, customizable, and enjoyable.
The aggregate ranking systems that replicate and/or reinforce ADP may be getting traction in their “accuracy,” but my study in this article demonstrates that such systems that rely on “groupthink” for their validity are also — as time passes — becoming worse at predicting starting pitcher outcomes. Last year, you would have been better just closing your eyes and throwing darts at a board rather than following ADP or aggregate drafting systems for your top starting pitchers. Other years, you would have been just as well suited flipping a coin. The fantasy sports industry is driven by rankings. Yet, as consumers of rankings, we should not be accustomed to increasing instances of “misses” as the normalization of our game. Forecasting is indeed an art and a science and a deeply personal excursion through data, and we should not let it be replaced with aggregation and groupthink that normalizes conformity in the name of “accuracy awards.”
Of course, this whole study is “small” in scope. I looked at only 4 years of nearly 18 years of available data, and I looked at only the first five rounds of the draft. However, I think we have enough data in the sample size to begin making significant statements about the start of your fantasy baseball drafts. If you want more, well, [stares at PayPal].
Whew! So! It’s time to put my money where my pretty little mouth is. I already gave you one set of rankings through week 1 and week 2 of the pre-season. You can find my first fantasy baseball pitcher draft strategy here, and my second fantasy baseball draft strategy here. Now, I’m giving you a completely different set of rankings based on IP, K-BB%, and SIERA, as best as I can do on my yeoman’s salary. If this works out, I’ll be reporting to you next year from the sunny beaches of Grey’s backyard, where he’ll undoubtedly have chained me up to protect me from fleeing to ESPN and becoming their new ranker.
Now, because I’m gung-ho in this punk DIY thing, I’m not going to take the easy way out and just aggregate The Bat, Steamer, and Fangraphs Depth Charts. Instead, I’m taking Rudy’s in-house rankings (developed in consultation with Steamer), where I hand-calculated the K-BB% using his stats, included his projected SIERA, and also factored my own custom ERA projection using K-BB% weights from 2021 projections and 2020 results. I used this combination because some of the 2020 results were problematic due to pitchers like Shane Bieber striking out a ridiculously high number of batters — much more in line with ratios seen by elite relievers (given the 70-80 IP in 2020) rather than the 180ish IP projected for 2021. I then factored in cumulative IP from 2019-2020 and compared to ADP. After much sorting, I had my results.
Then, because it’s basically impossible to forecast with any sense of precision the exact finish a player will have, I put everybody in a tier based on their likelihood to finish in the Top 3 / Top 5 / Top 10 / Top 20 based on their projections and their ranking in the qualities aforementioned: IP, K-BB%, and SIERA, while demoting players with known injuries or issues. I’ll say it every week: don’t look at this as hierarchical. Instead, look at it as descriptive of the pitchers who have the qualities to have the best possible outcome. When you start to view players as bearers of certain qualities of stats that are complementary, you no longer worry about sniping or “getting your guy,” because you know that there are just guys who are “more or less likely” to reach their projections and you adjust your draft strategy according to that.
So, this is a new set of rankings that I’m going to start the MLB year with because I feel it’s even better than my previous rankings. If you want my previous rankings, you can find them under my Top 100 Starting Pitchers 2021 week two article. I’ll note in the comments where major changes occur between the sets.
But wait! Didn’t I say that I was gonna teach you how to do this? Isn’t that the last part of liberation? Sure! It’s super easy. Start by scrolling way up and clicking on “Stats/Projections>Steamer/Razzball Projections>Pitchers”. Rudy offers all his projections in .csv form for free. How punk is that? Guy does all this work and just gives it to you for nothing. Then, you can calculate K% and BB% by dividing K and BB by Total Batters Faced, respectively. You calculate K-BB% by subtracting BB% from K%. Rudy provides SIERA in his rankings. If you want to do historical stuff, download past data from Fangraphs>Leaderboards. I did some of my own magic in my own rankings, but play with your results as you like. Add in weighting of categories, historical data, other SABRmetric stuff, whatever you like. Have fun, and be creative!
Tier 1: Top 5-ish
|Rank||Name||2021 K-BB%||2021 SIERA|
Comments: If there’s one pitcher who’s most likely to finish in the Top 3, it’s Shane Bieber. He’s elite in IP, K-BB%, and SIERA. If you’re looking for a dynasty pitcher to acquire, it’s him. If you’re looking for an SP1 with the highest chance of returning first round value, it’s him. I still wouldn’t draft him in the first round. Bieber is league average in hard hit % and Cleveland is clearly jettisoning talent. Bieber is just 25 and could honestly be one of the greatest pitchers of our generation, but the Spiders are a mess internally and we could see some of that internal drama spill out on to the field with negative results.
The surprise of is bracket is probably Brandon Woodruff. I know! When we’re looking at checking all the boxes of IP, K-BB%, and SIERA, he’s right up there. As the consensus SP13, he’s already considered by many rankers to be “very good,” and I would suspect that the last category holding him back from top 10 status would be “wins.” Wins proved to be wonky in my study and not entirely descriptive of a top finish. In fact, in many circumstances, players with dismal win counts finished in the top 3 fantasy starting pitchers. Dinelson Lamet had 3 wins last year and finished as SP7. Devin Williams had more wins than Lamet. Trevor Bauer had 5 wins and was SP3. Pete Fairbanks, the nutso setup guy for the Rays, had more wins than Trevor Bauer in 2020. See? Don’t get all flustered over wins. ENYWHEY. With Woodruff going around the fourth round, I’m getting the Shane Bieber vibe I had in 2020. You could start hitter/hitter/hitter/Woodruff and that could be the maximum return on value this year. We’ll find out in October! In any case, at his current ADP and given his ceiling, Woodruff should be a target for all drafters.
Tier 2: Shinelackers
|Rank||Name||2021 K-BB%||2021 SIERA|
Comments: You know that part of the study where I talked about omitting the people who had obvious problems? We’ve got a mix of the good and the bad in here. These pitchers simply lack “the shine,” but not the Shining. Bauer gets hit hard, and he’s basically admitted to pine tar. Now that he’s got his contract, do we see a change in his performance? Same thing goes for Cole except we have even more evidence about the pine tar, and we know he went through an awful period last year of a 6+ FIP. Flaherty and Castillo are bright spots who are good but just not as good as the previous tier. Scherzer’s numbers look stunning but he’s struggled in IP recently. Buehler has had a high IP load before, and if he reaches 170/180 IP this year, he could be in the top 5. These second tier pitchers should be considered as “likely” to reach the Top 10 SP, and most likely to return a value of 2nd-4th round.
Tier 3: The Affordable Aces
|Rank||Name||2021 K-BB%||2021 SIERA|
Comments: This tier was basically my SP1/SP2 tier from the last iteration. I could make a case for every pitcher here to make the top 10, and they’d all be equally valid. I see the likely outcomes for these pitchers to be anywhere from SP 5 to SP 40. If Corbin Burnes makes it to 170-180 IP (he came close in the minors once), you’ve got a top 10 SP. Zack Wheeler and Zach Eflin’s projections are likely a bit wonky due to 2020. When you look at their peripherals, they are thrilling and you’ll want them everywhere. Obviously the steal of the draft here is Kevin Gausman, who I have ranked nearly 30 spots above consensus. He has had 180 IP seasons multiple times, and now that he’s got his craft figured out, you might be Corbin Burnes-ing him in 2022 if he gets enough innings. Meanwhile, Snell and Glasnow probably won’t see enough IP to be a top 10 threat, but their stuff is certainly worthy of this tier. They have the ability to reach elite IP but lack the track record. German Marquez checks all the boxes on the stats and I can’t ignore him simply because of his home field. Sonny Gray would be in Tier 2 based on stats alone if he wasn’t already dealing with back spasms. He’s dealt with them all his career and will probably be fine this year, meaning he could be an extraordinary value.
Tier 4: Something’s Always Wrong
|Rank||Name||2021 K-BB%||2021 SIERA|
|Lance McCullers Jr.||16.24||3.71|
Comments: I know this will reveal just how old and white and 90s I am, but Toad the Wet Sprocket were a really good band. “Something’s Always Wrong” will always be a banger to me, and I’m a guy that tunes his headphones to Periphery. These are the pitchers that had the stats to be a tier above, except for something’s always wrong. Again, I’m ranking by likelihood to achieve premier outcomes, so when we start seeing question marks, we pay attention to those. Greinke’s K-BB% from 2020 seems unnatural. Bundy’s SIERA is concerning. Morton and Paddack struggled with IP compared to other pitchers. Mahle lacks a track record. Gonzales’ K-BB% was stunning in 2020; but was it more noise than signal? Joe Musgrove, Lance McCullers, and Jesus Luzardo could be tier 2 players if they were more consistent. The pitchers you’re looking at in this tier have a range of SP5-SP60, making them more volatile, but nonetheless upside options who would nicely fill out your rotation. But the takeaway is this is the last tier of players who have the qualities to be in the top 10 overall; they just have some hangup that’s lacking among the other players.
Tier 5: End of Rotation and Streamers
|Rank||Name||2021 K-BB%||2021 SIERA|
Comments: Here’s where the deviations from my previous rankings become obvious. The above tiers took you to SP40 or so, which is fine because they were really aggressive and you can see a bunch of players that ADP has ranked high are down in this tier (hello Julio Urias!). These are the players you draft as SP 4-7 to fill out your standard team, and they have an outcome range of SP10-SP90. Huge, right? But that’s why we’re ranking in a different manner. You’ll notice the K-BB% of these players are markedly lower than the players in the previous tiers, and that their SIERA is markedly higher. Also there’s Dinelson Lamet, who has an elbow injury and you should just stay away from because there’s a ton of players who aren’t injured at his ADP. I like a lot of players in this tier: you know I’m here for Robbie Ray and Yusei Kikuchi. Elieser Hernandez stuns on paper and could be crucial to high stakes teams pulling out a big win. I don’t believe in Drew Smyly’s small sample size but his numbers can’t be ignored. I have no idea why people are going all in on Sandy Alcantara when his numbers are nearly identical to Dallas Keuchel and slightly better than Zach Davies. Sixto could be a small sample size casualty but unless he cranks up his K-BB% he simply has a lower ceiling than many are projecting for him. Dustin May looks nice on paper but the Dodgers rotation will constrain his IP.
Tier 6: Dart Throws
Comments: These guys have a range of SP20-SP150. They’re the kind of pitchers you draft in best ball, deep leagues, or tournaments. If Tony Gonsolin gets a job, roster him everywhere. If Freddy Peralta gets a job, roster him everywhere for the next decade. Alex Wood looks nice on paper but the guy’s getting spinal work done. Carlos Carrasco is injured in many ways. Stephen Strasburg isn’t healthy. Framber’s finger is busted but he’s healing like the predator (true quote from Dusty). Luke Weaver looks nice on paper and could be a comeback player if he regains form. Randy Dobnak has a new slider and people are losing their mind over him in spring training but he doesn’t have an actual roster spot. Corey Kluber could be a value but he’s got a lot of history working against him. Shohei Ohtani is nice if you combine him as a pitcher/hitter but he’s not likely to reach the IP levels needed to be an elite fantasy pitcher. He’s more floor than ceiling. Compare Ohtani’s projections to Kohei Arihara, who is going like 200 spots later. Anybody who isn’t listed above is basically a dart throw. Blame omissions on my editor, who is me.
Well, now that I’m reaching a solid 7,850 words, I’ll take my leave and say thank you to the people who made it to the end. Thanks for staying for the sankjikai. I told you that you would learn something new! I hope this post is effective for you, and if you have questions, leave them in the comments. Keep your hate mail in your desk. Support your content creators and be nice to your league mates. Have an awesome week and see you for the regular season soon!
Aye, you made it this far, didn’t ya. EverywhereBlair is, well, located at home right now. He’s a historian and lover of prog-metal. He enjoys a good sipping rum. When he’s not churning data and making fan fiction about Grey and Donkey Teeth, you can find him dreaming of shirtless pictures of Lance Lynn on Twitter @Everywhereblair.