Happy Memorial Day. Thanks to all who have served this country, and to those who paid the ultimate price. My grandpa, who passed away in 2008, served as a machinist on a hospital ship during WWII. I'll never forget the anchor tattooed on his shoulder, and how he always told me war was no place to be. That's really all he said about his days in the Navy, at least to me.
I'm hitting pause temporarily on my 2020 pitcher rankings, until we know for sure if there will be baseball this summer. Goodness, I hope there is... In the meantime, I've overhauled the predictive version of DIGS, which now includes barrels (StatCast) dating back to 2015, flyball & hard-hit data (FanGraphs) dating back to 2002, and strike% breakdowns (Baseball-Reference) going back to 1988. The strike data, which includes % of strikes looking, swinging, and fouled, is used to compute expected strikeout and walk rates, the creation of FanGraphs contributor and fantasy baseball guru Mike Podhorzer. Incorporating all of this data has really helped to fine-tune "pDIGS," which, as mentioned in a previous post, is the expected version of DIGS. DIGS uses actual player outcomes like HR, BB, K, etc. to assess a player's value, while pDIGS essentially tries to come up with a predicted number for those same components. Instead of telling us what happened, it tries to tell us what could have happened, or what may be more likely to happen in the future.
0 Comments
I ran projections for nearly 170 different starters, but this series will focus on just the top 100 (ranked by projected WAR). In the first installment, we looked at players earning less than 1 WAR, or the "fringe average." Though nearly half of all starters fell into the sub-1 WAR bin, only 15 of them earned a spot in the top 100. The other 85 guys project out as average or better WAR producers, according to DIGS.
In this post, we'll look at pitchers earning 1.0-1.25 WAR, so 2-2.5 WAR over a full season. Traditionally, we fit these guys with a hat reading AVERAGE, although if you look carefully, you'll notice there are very few truly "average" pitchers to be had. Most of them, like most human beings in general, have a unique set of strengths and weaknesses. This is why I include the components that helped build their DIGS scores. This gives us a little more insight into just what average production really looks like. Remember, these projections are based on an 80-game season. For full-season equivalent, just double GS and WAR. Benchmarks:
Notes:
Projections are based on an 80-game season. For full-season equivalent, just double GS and WAR.
Benchmarks:
Notes:
While I ran projections for nearly 170 different starters, this series will focus on the top 100 (ranked by projected WAR). Among the 60+ who missed the cut, Michael Pineda (suspended), Domingo German (suspended), and Griffin Canning (elbow concerns) were the only players to score a better-than-average DIGS-. The first 5 off the list, #101-105, are Chris Bassitt, Rich Hill, Caleb Smith, Eric Lauer, and Pablo Lopez. Here is the first installment of the top 100, focusing on players earning less than 1 WAR. In a typical season, this is roughly 1.75 - 2 WAR, or fringe league average! In today's baseball analysis, there's a metric for every type of measurement you could imagine. Every major site has its own unique assortment, while aspiring analysts are creating dozens more as we speak. Good luck Keeping track of them all! Most of us identify a few "go-to's" and cling tightly to them.
I want to make DIGS one of your go-to's... but first, I need to instruct you on how to properly use it. The reason game score first caught my eye was simple. I appreciated that it was a single number you could use to evaluate a pitcher's performance. With game score, you didn't need to investigate all the individual components of a line score (IP, H, ER, etc.). The metric did it for you, and then packaged it up into a nice, clean number. In a sport overwhelmed with numbers, game score was refreshing to me.
In time, I learned a few things. While a game score is simple to its core, it, like anything else, still needs a context. If I told you a pitcher scored a 0 in his start, you'd know he pitched horribly. If he scored a 100, you'd know he was other-worldly on that day. Likewise, you probably wouldn't need to ask many questions if you saw a pitcher scored a 50. That's average, and baseball stats are literally built around the idea of "average." Pretty straightforward stuff. But what if a pitcher made 8 starts and averaged a game score of 52? How much better is that than someone who averaged 47? What if he averaged 43? The more time I spent with game score, the more I realized I needed to find ways to put it into context. When you are one of the only people working with a set of numbers (as I was, and likely still am with this metric), context is HUGE. Another thing I learned about game score early on was that people didn't really care. When Bill James, that stat's founder, called it a garbage stat, I think that was kind of a death sentence. But I also believe the baseball "numbers community" didn't take to it because it really isn't very sophisticated. In this sport & this age, people want sophistication! Maybe that's why Tom Tango built his own version of game score, tuning the values so they more closely aligned with what matters in pitching today. The more I dug in, the more my ideas for game score grew in complexity (creating game score WAR, adding park adjustments, etc.). Eventually, I'd come to the realization that this was a metric fans, analysts, & writers likely just wouldn't ever get behind. It was descriptive, but it would never be able to predict much of anything. But I loved the idea of it so much that instead of giving up, I set out to create my own. To create one that could do all the things I wanted! That's how DIGS was born, and for the past six months, I've been working on it with a couple major goals in mind; make it relevant, and make it easy to understand. And I believe I have done just that. My best ideas often come to me in the middle of the night. That was the case for my ERA conversion formula, which I simply called "digsERA." To be fair, the inspiration came from Baseball Savant, who converted their wOBA scores to ERA. If they could do it, so could I. That mantra has been the basis for a lot of my adjustments lately. Converting DIGS to an ERA paved the way for me to create "DIGS Minus" (DIGS-), another idea inspired by a major site, this time Baseball Prospectus. DIGS- is an index metric, which means it has context. A 100 is always average, no matter the year, the run environment, or the number of innings pitched. Always average. With this idea of context in mind, you'll find a lot of indexed stats on my leaderboards. All of the components of DIGS... batted ball%, HR% (or barrels), BB%, K%... they've all been converted to an index, where anything over 100 is above average. Why did I do this? You can find a pitcher's strikeout totals on any other site. But that number doesn't tell you everything. Even his K% doesn't tell it all, as a pitcher striking out 25% of batters in 2019 sure isn't the same as someone who did it in 2003. Things have changed, and index stats do all the contextual work for you. Oh, and while I don't possess all of the talents necessary to do wholesale regression analysis & predictability testing, I can tell you I've done enough number-crunching to confidently say that DIGS (and my newest creation, "Predictive DIGS," or pDIGS-) does just as good a job predicting itself into the future as the popular pitching metrics do (FIP, SIERA, DRA, etc.). I don't say that lightly. Ok, so the website is a huge change for me. I guess you could say I'm no longer a Giants blogger, which bums me out. I'll still be active on my Twitter account, and I'll always be around to talk Giants baseball. That won't change. But my heart is truly in this now, and with some absolutely badass design work (including an actual DIGS logo!) from Zach Ennis (@zachennis on Twitter), I'm ready to take my DIGS analysis to the next level. This site won't update a lot, but this seemed like the best way to share my projects with you. Browse around, and you'll find links to 2019 team profiles (complete with game logs for hundreds of pitchers), 30 years worth of individual leaderboards that I thoroughly enjoyed creating, and now thanks to a Baseball Cube membership, a growing list of Minor League leaderboards as well. If and when a 2020 season begins, you'll be able to find up to date data in the form of my custom leaderboards, again right here. DIGS is different, and I like that. If you're looking for a new & different way to look at pitching, I hope you'll take this journey with me. Thanks for reading, and here's to baseball returning to our lives again soon. ~ Kyle Goings |