Quality Start Leagues
With the recognition now of how wins is a terrible statistic for starting pitchers a lot of leagues have moved to quality starts. Wins for a pitcher has very little to do with their ability. Sure they can put themselves in a position for a win but they are then reliant on both their offense and bullpen to achieve that win. Just take a look at Jacob deGrom (you can probably hear me sigh from wherever you are reading this). Quality starts is something that is more so based on a pitcher’s performance alone. It’s simple, pitch at least six innings with no more than three earned runs and you earn a quality start. Let’s glance at the quality start leaders from last season.
| Rank | Player | QS | ERA |
|---|---|---|---|
| 1 | Lance Lynn | 10 | 3.32 |
| 2 | Yu Darvish | 10 | 2.01 |
| 3 | Shane Bieber | 10 | 1.63 |
| 4 | Carlos Carrasco | 9 | 2.91 |
| 5 | Trevor Bauer | 9 | 1.73 |
| 6 | Kyle Hendricks | 9 | 2.88 |
| 7 | Kyle Freeland | 9 | 4.33 |
| 8 | German Marquez | 9 | 3.75 |
| 9 | Zac Gallen | 9 | 2.75 |
| 10 | Gerrit Cole | 8 | 2.84 |
| 11 | Lucas Giolito | 8 | 3.48 |
| 12 | Jacob deGrom | 8 | 2.38 |
| 13 | Kenta Maeda | 8 | 2.70 |
| 14 | Adam Wainwright | 7 | 3.15 |
| 15 | Zack Wheeler | 7 | 2.92 |
| 16 | Hyun-Jin Ryu | 7 | 2.69 |
| 17 | Aaron Nola | 7 | 3.28 |
| 18 | Luis Castillo | 7 | 3.21 |
| 19 | Aaron Civale | 7 | 4.74 |
| 20 | Clayton Kershaw | 6 | 2.16 |
Of course, some of the best pitchers will have the most quality starts like Yu Darvish, Shane Bieber, and Trevor Bauer. What you might notice are some outliers like German Marquez, Kyle Freeland, and Aaron Civale. While performance level matters so do team philosophies. Pitchers have a better chance at acquiring a quality start if their respective teams let them pitch deep into games. Whereas if it’s the opposite like the Tampa Bay Rays, it won’t matter how good a pitcher is they likely won’t let them see the sixth inning. Let’s check out team’s quality start percentage (QS%) and their starting pitching ERA.
| Rank | Team | Team SP ERA | QS% |
|---|---|---|---|
| 1 | CLE | 3.17 | 62% |
| 2 | CHC | 3.77 | 50% |
| 3 | COL | 4.83 | 47% |
| 4 | CIN | 3.50 | 43% |
| 5 | HOU | 4.26 | 42% |
| 6 | SEA | 4.41 | 42% |
| 7 | SDP | 3.46 | 37% |
| 8 | PHI | 4.08 | 33% |
| 9 | CHW | 3.85 | 32% |
| 10 | OAK | 4.49 | 32% |
| 11 | STL | 3.86 | 31% |
| 12 | LAD | 3.29 | 30% |
| 13 | NYY | 4.24 | 30% |
| 14 | NYM | 5.37 | 29% |
| 15 | TEX | 5.32 | 28% |
| 16 | MIA | 4.31 | 27% |
| 17 | MIN | 3.54 | 27% |
| 18 | WSN | 5.38 | 27% |
| 19 | LAA | 5.52 | 25% |
| 20 | MIL | 4.18 | 25% |
| 21 | ARI | 5.04 | 22% |
| 22 | ATL | 5.51 | 22% |
| 23 | KCR | 4.70 | 18% |
| 24 | SFG | 4.99 | 18% |
| 25 | TOR | 4.55 | 18% |
| 26 | BAL | 5.09 | 17% |
| 27 | DET | 6.37 | 16% |
| 28 | BOS | 5.34 | 15% |
| 29 | PIT | 4.74 | 15% |
| 30 | TBR | 3.77 | 12% |
The Indians are way above everyone else and while that can be because of their stellar pitching staff they still let pitchers go deep into games no matter what. Aaron Civale popped up on the first list with a 4.74 ERA seeming like an outlier but that’s because the Indians love for their starters to accumulate innings. You can see a few outliers from the chart above, one is the Colorado Rockies.
Last season they actually let their starters pitch deep into games and while you have to battle the home/away splits they technically could be more valuable in this format. Another outlier is the Seattle Mariners. Marco Gonzales, Yusei Kikuchi, and Justus Sheffield all were top 45 in pitches per game. Unfortunately, the Mariners roll with a six-man rotation but you shouldn’t still discount them because it’s for that reason they are able to let their starters pitch deeper in their starts. Moving on to the major outlier which doesn’t surprise anyone is the Tampa Bay Rays. With one of the better ERA’s in the league they only produced a quality start in 12% of their games.
To quickly recap, in quality start leagues you should move up starting pitchers who pitch for the Indians, Rockies, Cubs, Astros, and Mariners (slightly). Move down any starter who plays for the Rays.
As for individual performance here were your 2020 innings pitched per game leaders.
| Rank | Name | Age | Team | IP/Game |
|---|---|---|---|---|
| 1 | Zach Plesac | 25 | CLE | 6.9 |
| 2 | Kyle Hendricks | 30 | CHC | 6.8 |
| 3 | Trevor Bauer | 29 | CIN | 6.6 |
| 4 | Adam Wainwright | 38 | STL | 6.6 |
| 5 | Lance Lynn | 33 | TEX | 6.5 |
| 6 | Zack Wheeler | 30 | PHI | 6.5 |
| 7 | Shane Bieber | 25 | CLE | 6.4 |
| 8 | Framber Valdez | 26 | HOU | 6.4 |
| 9 | Yu Darvish | 33 | CHC | 6.3 |
| 10 | Marco Gonzales | 28 | SEA | 6.3 |
| 11 | German Marquez | 25 | COL | 6.3 |
| 12 | Aaron Civale | 25 | CLE | 6.2 |
| 13 | Gerrit Cole | 29 | NYY | 6.1 |
| 14 | Brad Keller | 24 | KCR | 6.1 |
| 15 | Kenta Maeda | 32 | MIN | 6.1 |
| 16 | Antonio Senzatela | 25 | COL | 6.1 |
| 17 | Sandy Alcantara | 24 | MIA | 6 |
| 18 | Dylan Bundy | 27 | LAA | 6 |
| 19 | Patrick Corbin | 30 | WSN | 6 |
| 20 | Zac Gallen | 24 | ARI | 6 |
| 21 | Lucas Giolito | 25 | CHW | 6 |
| 22 | Aaron Nola | 27 | PHI | 5.9 |
| 23 | Luis Castillo | 27 | CIN | 5.8 |
| 24 | Zach Davies | 27 | SDP | 5.8 |
| 25 | Clayton Kershaw | 32 | LAD | 5.8 |
| 26 | Dallas Keuchel | 32 | CHW | 5.8 |
| 27 | Dinelson Lamet | 27 | SDP | 5.8 |
| 28 | Chris Bassitt | 31 | OAK | 5.7 |
| 29 | Carlos Carrasco | 33 | CLE | 5.7 |
| 30 | Jacob deGrom | 32 | NYM | 5.7 |
| 31 | Alec Mills | 28 | CHC | 5.7 |
| 32 | Brandon Woodruff | 27 | MIL | 5.7 |
| 33 | Zach Eflin | 26 | PHI | 5.6 |
| 34 | Kyle Gibson | 32 | TEX | 5.6 |
| 35 | Zack Greinke | 36 | HOU | 5.6 |
| 36 | Andrew Heaney | 29 | LAA | 5.6 |
| 37 | Hyun Jin Ryu | 33 | TOR | 5.6 |
| 38 | Sixto Sanchez | 21 | MIA | 5.6 |
| 39 | Max Scherzer | 35 | WSN | 5.6 |
| 40 | Kevin Gausman | 29 | SFG | 5.5 |
| 41 | J.A. Happ | 37 | NYY | 5.5 |
| 42 | Jesus Luzardo | 22 | OAK | 5.5 |
| 43 | Justus Sheffield | 24 | SEA | 5.5 |
| 44 | Nathan Eovaldi | 30 | BOS | 5.4 |
| 45 | Mike Fiers | 35 | OAK | 5.4 |
| 46 | Kyle Freeland | 27 | COL | 5.4 |
| 47 | Kwang Hyun Kim | 31 | STL | 5.4 |
| 48 | Brady Singer | 23 | KCR | 5.4 |
| 49 | Jose Berrios | 26 | MIN | 5.3 |
| 50 | Johnny Cueto | 34 | SFG | 5.3 |
So here we are, and thus proving our previous short recap, you see four Indians pitchers in the top 30. Out of all of the Indians pitchers, one who sticks out is Zach Plesac. He lead the league in innings pitched per game while also being 11th overall in pitches per game. He actually didn’t have a single start where he didn’t reach six innings. I quickly want to note Aaron Civale as well because even though he didn’t have a solid ERA he still was 12th on this list and was actually 7th in pitches per game last season. He is not only being extremely undervalued in quality start formats but all formats as well.
We also see the Rockies starters on this list with German Marquez, Antonio Senzatela, and Kyle Freeland. Both Senzatela and Marquez are top 25 in pitches per game and seem like decent values for this format. Keep in mind skill of course matters as well but finding people like Senzatela late in drafts makes sense because you are chasing upside anyway.
Some interesting late round shots you might want to take would be with the Royals pitching staff. They let their pitchers throw as well but the quality starts haven’t come due to the skill set. Their pitchers are young and growth could easily come. Brad Keller was 14th in innings pitched per game and 17th in pitches per start. Kris Bubic was 25th in pitches per start. Brady Singer was 48th in innings per start and 40th in pitches per game.
Yes, a top-end skill set matters most for quality starts leagues, but team tendencies can help you find success. Pitchers like Aaron Civale, German Marquez, and Brad Keller could be overlooked due to their lack of either ERA, ballpark, or strikeout upside. When in fact they bring more value to quality start leagues than we think.
Mike Minor …
So one of my leagues has gone to quality starts, but most projection systems don’t project quality starts. However, after looking through historical data, I’ve actually found a pretty good equation for approximating QS%.
QS% = IP/Game * 1/3 – ER/Game * 1/4 + ER/IP – 1.2
I was having trouble by only including the first two variables, but adding that ER/IP variable better accounts for pitchers who may go deeper into starts but allow few enough ER to still qualify for the QS. For example, 3 ER in 9 IP is the same as 0 ER in 6 IP as far as quality starts go, but if you only include IP and ER in the equation, it will drastically penalize the first pitcher.
I’ve found it to be pretty good. The nature of projection systems means the QS numbers don’t vary that much, but it can help separating low IP/high pitch count guys like Glasnow from the high IP/low K guys like Marco Gonzales.
Very interesting and thank you for reading.
Interesting but what’s the concept? Is it just tinkering for a good r^2? I see two constants being multiplied by data, then another constant at the end being subtracted. What is the reason for those numbers? Hunting correlation backwards is dangerous. When you have 3 constants, 2 multiplicative, in a short formula, the potential error is now off by at least a couple orders of magnitude based on the precision of those guesses.
P.S. At the risk of denigrating an entire field of academics that considers itself credible, economics makes insane post-fitted models all the time (besides the fact that only a psychopath would EVER! just decide to put the dependent variable on the horizontal axis as a convention! Argh what is wrong with these people!?). “Well, you see, when we simply multiply our (laughably simplistic) assumption by a wild self-serving guess and then shift it again by another random number to fit the data, it somehow reinforces precisely what we set out to prove. QED!”
That came off too harsh at your equation – my bone to pick is more broad with models that get really precise and then multiply the whole thing by a guess (thus killing the precision part). Sort of like how projections are good at rates and then multiply by a wild guess of IP or PAs. I like that you’re getting closer to a result and it’s something I’ve been annoyed by forever.
Nah, I get it. It’s worthwhile to have a discussion on procedure/methodology. I feel like this equation is fairly straightforward, however, considering the two variables that go into a quality start are IP and ER. I felt the ER/IP variable was useful in helping to mitigate the effects of more IP likely yielding more ER. It uses the basic elements of the definition of the statistic to come up with an approximation.
Ultimately, behind it all is an oversimplified linear regression. That being said, I only claimed it was “pretty good” at estimating QS given no other information. Since this is a fantasy baseball blog, it seemed relevant to provide readers with a rough equation to get to QS if their league uses it.
Thing is QS are certainly going to decrease in the future, and some people also appreciate streaming bulk guys. I do play plenty of both, though. Nice article.
Thank you!
Yes indeed, good article that I’ll have to come back to when draft season starts up. I will say, though, that it’s a bit sad that only 21 pitchers were throwing enough innings on average last season to even qualify for a quality start. Unless the stat changes to reflect only 5 innings of work, QS might decrease enough to make it almost as obsolete as the complete game category in leagues.
I would love to see QS added to the Auction Calculator. Otherwise (when dealing with QA leagues, which I often am) you have to leave it out and just mentally bump up the guys who tend towards deeper starts.
Good article, but raises the question whether QS is really the right stat either. I took my league a couple years ago to IP. Seems weird, but it is literally the number of outs your pitchers get. Suddenly, 5.2 IP with good ratios isn’t worse than 6 IP with meh ratios. Let the ratios and # of Ks measure quality of outing. If I’m stacking up W, QS and IP for combining pitcher length and quality of outing within his control, I’m going IP every time.
My long-time league started with W, then shifted for QS for last 5-6 years and finally this off-season made the move to IP. As noted above, QS is trending toward becoming obsolete. I think in particular in reduced workloads in 2021 it might be very scarce. With pitching already two rate stats (compared to 1 for batting) I think it appropriately mitigates the risk of just streaming/stacking innings as the penalty is severe enough to be a proper deterrent. But with that said, I don’t love IP and was interested if anyone had come across anything else.
Note> league uses yahoo so somewhat curtailed about their available options.