Auction Calculator Results: Starting Pitcher Tiers
The Auction Calculator is now loaded with the 2017 end-of-season data along with some 2018 projections. I’m sure the preceding comment will be sufficient to keep many readers busy for a while. I’m glad some came back. While its output can lead down several different discussion paths, I’m going to analyze what I consider to be the third starting pitcher tier. I feel many 2018 leagues will be won or lost by navigating this minefield.
So far this offseason, fantasy owners have placed four starters (Kluber, Kershaw, Sale, and Scherzer) in the top tier, After those four, I believe there are a dozen or so pitchers who would make acceptable aces, especially if they can be doubled up with another pitcher from this tier.
After the second tier, there seems to be a tier where it is hard to differentiate the talent. In the #2EarlyMocks, the Tier runs from about 72nd overall (Arrieta) to 191st (Bundy). I really noticed the difference while watching to NFBC style drafts at First Pitch Arizona. Here is one league’s draft board.
The third tier starts with Gerrit Cole (#74) and seemed to end with Drew Pomeranz (178th). What is the real difference between these two pitchers? Here are the Steamer projections for the pair.
| Name | K/9 | ERA | WHIP |
|---|---|---|---|
| Cole | 8.5 | 3.90 | 1.25 |
| Pomeranz | 8.8 | 4.27 | 1.35 |
Close with Cole having a nice ERA advantage. Additionally, I ran the auction calculator (Steamer projections) using my normal setup of 15 teams, no bench, and OBP. Cole ended up 24th overall with Pomeranz at 46th with a range of $9 ($15.4 to $6.5). The main difference is Cole is projected for ~20 more innings pitched. Is it worth waiting to get some arms once a team has an ace or two?
It’s time to dig a little further into this tier. First, here is one of my favorite graph to make to find various tiers. It is the action values ranked in order along with the actual values.
Additionally, here are the top 50 pitchers for reference.
The top tier is the first five pitchers since Syndergaard is projected for 191 IP. There is another tier at Carrasco ($25.5) and Bumgarner ($22.1). The talent drop is steady from there until 29th where Salazar ($13.3) drops to Lance McCullers ($11.6). So using the projections, it seems that an extra tier exists. The third tier gets split and become tiers three and four.
For the end of these two tiers, I was around the right spot with the $4 decline (rare drop near the end) happening from the 42nd to the 50th ranked starter. On average, an owner will get three of these top starters in this 15-team league.
Using a normal distribution of talent, teams will spend $30.8 on their ace (Strasburg), $16.5 on their #2 (Lester), and $9.5 on their #3 (Stroman) for a total of $56.8. A team could pair Strasburg and deGrom right off the start for $58 in value and wait on starters for a while and pick up two starters rank from 46th to 60th and still come out ahead. There are several ways to navigate these pitchers, just don’t fall too far behind because it’s becoming harder to find good pitcher during the season.
I feel like I am rambling a bit but need to get back and focused on this third tier which is now two tiers (some relievers snuck their way into the list, it happens).
- Has the dirt bike accident dropped Bumgarner down to #13 overall?
- Where is Darvish going to end up throwing?
- Can I trust Jeff Samardzija as my #2?
- Is Luke Weaver now a #2?
- Is it worth taking a chance on Rich Hill and see if he can throw more than 141 innings?
- What is Luiz Gohara doing this high?
- How healthy is Garrett Richards?
- Could I take Jon Gray as my #3?
- Or Rick Porcello?
A couple first round talents are in these tiers, it’s just figuring out the right pitcher. Basically, who is this season’s Robbie Ray and Luis Severino? Beyond finding an ace, teams can start figuring out which pitcher are likely to produce surplus value as industry rankings and more ADP values begin to roll in.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
Tyler Lyons hasn’t made a start since 2015…?
And it’s not likely Chad Green starts for the Yankees.
Thanks. I’m not sure why each one is getting some game starts. I’ll let those in charge of depth charts know.
I noticed them and added this disclaimer:
some relievers snuck their way into the list, it happens.
Greinke at 15 is….curious.
I would NOT put Noah in tier 1. injury risk too high
Agree. Same with Kershaw getting all his starts.
No Zack Godley?
Only projected for 113 ip for some reason.
We are working out the playing kinks for sure.
This is the first time I’ve seen Samardzija valued properly. I picked up on his curveball usage late in ‘16 and forecasted 200K because he was using it in the same way he did his old slider (when it was dominant).
I just don’t see much difference between The ? and Chris Archer but people love to throw money at Archer and let Samardzija slide.
Archer picked up a heavy HR problem during the HR explosion and I don’t see it going away even if the BABIP regresses him closer to his FIP. Gimme The Shark and his park and the NL rate bonuses.
Samardzija looks like a massive bargain. Although I definitely wouldn’t take him over Archer. Really only had a HR problem in 2016 and we all know how unreliable those numbers can be in small samples.
Probably not 250K. Archer won’t approach the 212IP mark again anytime soon and that 11.5ish K/9 screams outlier (I’ll have to look and see). 230K is a better bet and probably with a worse WHIP & ERA than Samardzija who had a hard luck year. Archers HR issue looks larger than Samardzija’s. JS looks like a true talent of around 11/12% HR/FB and Archer looks to have gone 13%+
Just my take. I just wouldn’t take him for anything more than a #2 with a pile of Ks, and I’d rather spend far less to get the same thing from the ?
And if you didn’t start him in COL, his overall ERA would be 0.40 lower
I would argue that Shark also has a HR problem. In fact his HR/9 was higher than Archer’s in 2017. Tough to do in San Fran’s park.
I can’t argue that Archer should be as high as 8th overall but I’ll take his 250 Ks over Shark’s 200 without a doubt.
The inflation for catchers is huge with the auction calculator. I’ve even moved catcher to the lowest priority and still have crazy results. Sanchez valued as the #2 behind only Trout?
Default is 2 catcher slots, change it to 1
I had already done that to get those results.
The catcher pool drops off hard at ~20. Usually, I have to sift through the players’ pool to find those guys I’m willing to roster and then find a plan to get them.
https://www.fangraphs.com/fantasy/deep-diving-into-the-catcher-lagoon/
Thanks Jeff! Depending on your league (say 1 catcher league) are you spending $30+ on a Gary Sanchez?
I don’t have a problem doing it. Most of the time, catchers go for less. He’s was the 30th best hitter according to wRC+. Add in he is a catcher, he could move into the top 15. He won’t so buy low.
The catcher inflation is huge because the calculator’s positional adjustments use a faulty logic.
Thanks. This is great stuff. A question about the positional adjustments: Is there any way to adjust them manually, beyond shifting the order of the valuation, or to turn them off, to get at raw value, normalized for a full $260 budget? I have sorted by the PTS column, but that doesn’t leave a usable position-neutral auction value.
Additionally, I’m sure there is some math behind valuing DH-only players positional adjustment at $0.4 when the next lowest adjustment is $12.7 for 1B/DH, but it strikes me as incorrect (or at least not reflective of real world auction values).
The 0.4 is the lowest value, but because of the program’s nature (assumes a perfect league with no biases), the 1B adjustments will stay.
Thanks for this amazing tool, Jeff! Could you add RP (R+RBI-HR) as a category?
Jeff – I have some serious misgivings about the Last Player Approach used within the calculator. It mostly has to do with value versus cost and the adjustments made i.e. the aPOS column. Who do I contact about discussing the model?
Couldn’t you just ignore the positional adjustments, export the results, and apply your own positional adjustments?
You could. And that brings up a secondary issue. The PTS column (ignoring positional adjustments) should be a “value” value not a cost (i.e. $) value. It is more labeling/algebra, but the mAVG/mRBI/mR etc. columns are independent of cost; you can always apply the dollar value at the end of the calculation when those positional adjustments are made.
The fundamental flaw of the positional adjustments (aPOS) are that you are forcing the last player picked’s value to equal cost, i.e. $1 and adjusting everyone else based on that adjustment.
There is a more optimal way of doing this, but it is based on a different assumption.
It’s not going to change and I never use as a final valuation for even my leagues (I use SGP).
The adjustment from value to cost is independent of the valuation method (whether its SGP, Z-scores, etc.). My issue isn’t what method is used but adjusting from the PTS column to the aPOS column is currently a wrong approach in the calculator.
Simple example: based on the calculator using Steamer projections, in a 15 team mixed league Gary Sanchez is the 8th best hitter. Based on the photo that was attached he went 34th in that mock draft. Do you really think he’s the 8th best hitter?
Not related to the SP tiers, but more to the early mocks recently posted:
Trea Turner at 3-4?! Sweet merciful Christ.
You hate the two scarcest categories? I’ll take the 75SB and .330 average upside ?
correct me if im wrong but isn’t steamers weakness projecting SB’s?