Evaluating Pitchers in Smaller Samples
I go through some of my own thought processes for evaluating pitchers when you don't have a ton of data to work with
I have had a crazy day and my back is sore so I’m laying down on the couch to take a bit of a break from the usual. But I can’t just lay down and do nothing, I have to use this time to write to all of you - one of my favorite things to do.
I’ve been studying the Bible pretty much daily for the last year now, and that has really gripped me. I think it would be enjoyable to a great use of my time to have some kind of Bible blog as well. The problem is that I don’t have much free time and nobody would read that. But maybe someday. For now, we’ll write about baseball statistics.
This post was inspired by reader Brandon K, who sent this in:
Hi Jon, I love the daily Reports that accompany your Daily Notes and you seem like a crazy smart impressive good looking man, I’d do anything to be more like you.
Do you think you could expand a little on the Last 3 Weeks leaders (CSW%, K%, K-BB%, Brl%, xwOBA, Contact%) and any other leaderboards you see fit for the benefit of seeking out potential waiver wire targets for deeper leagues?
Many of these players are universally owned. Perhaps something similar to OPS Leaders - Last 2 Weeks - Low Owned Hitters. Although with the fantasy season close to halfway over, maybe there are just fewer breakouts available that haven't been scooped yet. Thanks!
It’s possible that I did some editing at the beginning of that, but we’ll never know for sure.
You are right with that last point, breakout pitchers are pretty easy to find in April and May, but as the season progresses it gets much, much tougher as the field scoops up all of those names as they identify themselves. Most big time pitcher skill improvements happen over the off-season and not during the season, so by the time it’s June - the breakouts are spotted and added.
The daily notes indeed have a lot of recent data reports that are most of the time unhelpful as far as actually improving our fantasy teams goes.
For example, today we had this:
K-BB% Leaders - Last 3 Weeks
Garrett Crochet - 74 TBF, 35.1% K-BB%
Tyler Glasnow - 74 TBF, 35.1% K-BB%
Tanner Bibee - 69 TBF, 31.9% K-BB%
Paul Skenes - 92 TBF, 31.5% K-BB%
Reynaldo Lopez - 69 TBF, 27.5% K-BB%
Ryan Pepiot - 93 TBF, 25.8% K-BB%
George Kirby - 95 TBF, 25.3% K-BB%
Logan Gilbert - 103 TBF, 25.2% K-BB%
Hunter Brown - 73 TBF, 24.7% K-BB%
Ryan Feltner - 68 TBF, 23.5% K-BB%
But what are you going to do with that? Maybe add Hunter Brown, but in any kind of deep league he’s probably not been available all season and certainly isn’t now. Ryan Feltner would be, but it’s Ryan Feltner - you don’t really want Ryan Feltner.
So what can we do here mid-season when we’re trying to improve pitching staffs? Well I have some ideas.
It’s All About The Rookies
In the middle of the year, the best way to add on to your pitching staff is with the rookie call-ups. After April and May pass, you really start seeing a lot of pitcher call-ups, and in general terms this is the best group of pitchers to look to. Every once in awhile you’ll find a known commodity/veteran pitcher looking really good, but almost every time that happens it’s just a streak of good luck and that pitcher will go back to their older ways which put them on waivers in the first place.
The ceiling is much higher with the rookies. That comes with a much lower floor as well, of course. A good comparison might be Tyler Anderson and Christian Scott right now. You know what you have with Anderson. He’ll make a start every five games as long as he’s healthy, and he’ll have plenty of good starts and plenty of bad ones. The end of the season numbers won’t be good enough for fantasy purposes, but at least you know he won’t slap a 7.50 ERA on your fantasy team for very long. With a guy like Scott, you don’t have that same kind of assurance because he has never shown the ability to get Major League hitters out for any extended period of time.
So the young players have wide ranges of potential outcomes while the available veterans have much more narrow ranges.
How to Choose a Rookie Pitcher
If you give me 200 somewhat recent Major League innings to evaluate a pitcher with, I’ll be able to give you a pretty good prediction of what the next 200 will look like. But we don’t have any Major League innings when a guy debuts, so what do we do?
We can look at the minor league numbers first. These are hard to trust because the competition is so wildly different than what they’re about to face in the Majors, so I can’t say it’s the most successful venture, but it’s better than nothing. What I look at primary from the minors is K-BB%.
Are you surprised? This is an easy one, because it’s easy to find. You can go to any player’s FanGraphs or Baseball Reference page and find this. But if you want to be way cooler, you can be a paid sub here and use my MLB & MiLB Stats Dashboard.
Let’s check some minor league K-BB% marks for some of the recent pitchers that have come up and had immediate, sustained success. You can find that right on my MLB & MiLB Stats Dashboard if you’re a paid sub and have access to the link.
Spencer Strider: 94 IP, 29% K-BB%
Bryce Miller: 167 IP, 21.1% K-BB%
George Kirby: 94 IP, 24.5% K-BB%
Grayson Rodriguez: 220 IP, 29.1% K-BB%
Tanner Bibee: 148 IP, 26.2% K-BB%
There are certainly more names that I’m missing there, but those are the guys I thought of that you could have picked up off of waivers the last few seasons as they got the call-up and would have gotten a lot of good stuff out of them right from the beginning. And you can see that they all have K-BB% above 20%, that 20% has kind of been the magic number for me for awhile now.
So I think it’s fair to say that most reliable rookie SPs will have high minor league K-BB%. It’s also certainly fair to say that the inverse is not true: not all minor league SPs with a high K-BB% will be reliable in the Major Leagues right away.
Here are some other pitchers with elite K-BB% in the minors that haven’t done it well enough at the Major League level for us to be interested in them for fantasy purposes:
JP Sears: 26.9%
Andrew Abbott: 26.6%
Ken Waldichuk: 25.1%
Kyle Harrison: 24.9%
Brayan Bello: 24.5%
Logan Allen: 24.0%
So the conclusion here is to use K-BB% as a filter. You can just ignore any rookie pitcher that comes up that had a poor K-BB% in the minors. But just because a guy had a high K-BB% down there does not mean they will have success at the highest level, so you don’t need to feel like you need to hold on if they have a bad first few starts.
I also really like to see swinging-strike rates from the minors. This is not an easy thing to find, unless you have my MiLB StatCast Dashboard, which shows that for all levels of minor league ball. Here are your leaders this year right from that dashboard, filtering to 150 batters faced in AA and AAA:
We’ve already seen Jack Leiter come up and fail (initially at least), but if David Festa (MIN) gets the call, you’d want to add him if you’re in need of a pitcher given this elite 18.3% SwStr% he’s posting in AAA. That’s a very good sign.
The other thing you can go on this dashboard for AAA pitchers is the pitch mix data.
What I personally look for here is this:
High SwStr%
Good numbers on the primary fastball
A deeper pitch mix
Ground ball rate
There are plenty of different ways to succeed at the Major League level. The only thing I would really require there is a high SwStr%. The ability to get whiffs is really important for pitching success, and if you can’t do it in the minors - you ain’t gonna do it in the Majors. The other stuff is less important. GB% holds very little weight to me but I’d still rather see a higher one rather than a lower one for a tiebreaker. A deep pitch mix can be irrelevant if the guy has two or three elite pitches, we’ve seen that plenty of times.
Not having a fastball that earns strikes in the minors is scary to me, and we have seen that hold back a ton of young pitchers - I talk about that all the time.
So that’s the minor league stuff I check on when looking at rookie pitchers.
Recent Data, Small Samples
This is really the heart of the question, so let’s talk about that. There’s not a pitcher report in the daily notes that filters out high-owned players. Maybe that’s a good thing to add, but even if I don’t do that - you can still find it yourself on the main MLB dashboard.
So let’s dig into that and see if we can’t find any names to recommend right here! Let’s look at the last 30 days, pitchers under 30% owned. If you’re in a truly deep, competitive league, you’ll need to get under 30% or even lower to find guys available in your league - because these ownership numbers use Yahoo and ESPN leagues which can be huge jokes.
Here’s what it looks like:
So I set these dials
Game Date: 5/19 through today
GS: 3 (this sets a minimum of 3 starts to qualify)
Ownership%: 0-40%
And then I sorted by K-BB%.
We see Ryan Weathers there at 29.1% in four starts. Not any old joe could make four Major League starts with that high of a rate, but there is still a great deal of randomness that goes into a one-month sample. Given the mediocrity we’ve seen from Weathers in the much bigger sample (11.2% K-BB% if you go the whole way back to 2023), I’d pass on him.
DJ Herz could be a potentially interesting name here. We also see an elite 17.1% SwStr% there, so he’s doing plenty of things right. The downside is that it’s just three starts and 246 pitches. That’s a really small sample. But this is the world we’re operating in, right? To be the first man to the table in a 20-team league, you have to make some guesses on small samples. And a 17% SwStr% in a small sample is still pretty impressive. You could then check the arsenals tab on the guy. What we’re hoping to see there is good marks on the fastball and good marks on at least one other secondary pitch (breaking or offspeed).
In this case, we see good marks on the fastball with a 19% SwStr% and a 53% Strike%, however the velo on the pitch is concerning and so is how easy it is to get into the air (53% FB%). You could check pitch modeling stuff on the fastball as well if you wanted to see if the movement and whatnot grade out well. In this case, they don’t. I’d also be scared that he’s throwing that fastball at such a high clip. That’s fine if you’re a Bryce Miller or a Paul Skenes, but this fastball doesn’t appear to be anywhere near that good.
One more thing to check then is his minor league numbers, so back to that dashboard, this time the “Stats by Level” tab:
The 14.9% K-BB% in AA is discouraging, but the 31% K% being a part of it makes me feel a bit better. He’s posted high strikeout rates at every levle, so that’s a positive.
My final judgement on Herz isn’t the point here, but to show the full process through to the end - I wouldn’t be confident that Herz will be very good at all moving forward. More red flags than … green, or whatever color the flag would be in this analogy.
I’m not going to go through more names, that isn’t the point here. Another thing I like to check, as you may know, is the combination of SwStr% and Ball%. This is one of my “Magic Formulas”. This is a filter we can put on the dashboard like this:
I don’t have a Ball% slider on there, I ran out of room, but you can just sort it. So we have seven names here under 40% owned with a SwStr% above 14%. What we’re looking for is something like a Ball% under 34%, so you can mentally filter out the bottom two or three there and be left with:
Spencer Schwellenbach
Jared Koenig
Sean Manaea
Carlos Carrasco
You can eliminate Koenig because he’s an opener, and then you have veterans that are just recently on the positive side of variance in Manaea and Carrasco.
But at least we get one hit here in Schwellenbach. Then you can dive into more on him and find a bunch of the stuff out that I already wrote about today in the daily notes.
One other tab that I’ve been using a bit more is the Stuff by Game tab. This will give you those advanced pitch modeling numbers at the start level so you can see trends. Maybe a veteran pitcher will make a tweak to a pitch mid-season or change the pitch mix to favor his better pitches, and that kind of stuff can lead to improvements in these advanced metrics that do correlate with success.
So to take a couple of the names we shrugged off earlier (the blue line is Stuff+, the orange line is Location+):
So if we saw blue lines shooting up in the last few starts, then maybe that would be reason to look further - but in these two cases we don’t see that. We just see some better command with the Location+, which isn’t very interesting since that’s a pretty random stat at the start-to-start level.
Alright, well that was a ton of words to answer one question, but I think this will be helpful to some of you out there who have access to my tools and want to be a bit better at using them.
Talk to you next time.