(Photo by George Walker/Icon Sportswire)
If you’re a consistent reader at Pitcher List, you may know that I pick streamers every day inside my morning SP Roundup articles. I’ve been doing it since 2014 and for the last two years, All-Star reader David McNickle has compiled the results for full transparency because he’s awesome.
A few notes before we dive into the numbers:
- Every streamer had to be owned in 20% or fewer leagues according to Fantasy Pros’ consolidated ownage metrics. I can’t tell you how upset I was when Zack Wheeler and German Marquez went above the threshold.
- I was forced to give a streamer every day. No exceptions.
- Fine, there were exceptions as streamers were often changed after the piece was finished. This was due to rainouts, late scratches, injuries, and sometimes flat-out misinformation.
- 156 pitchers were streamed across the entirety of the season. I assigned a Win or Loss to each start based on overall performance as I saw fit. I leaned toward a Loss when possible and turned to the comments to ultimately decide if it were debatable.
Here are the results from 2018:
All Streams
Total Streams | IP | Wins | ERA | WHIP | Strikeouts | IPS | K/9 | BB/9 |
156 | 921 | 65 | 3.54 | 1.20 | 808 | 5.9 | 7.90 | 2.80 |
Successful Streams
Total Streaming Wins | IP | ERA | WHIP | Strikeouts | IPS | K/9 | BB/9 |
103 | 713 | 2.22 | 0.97 | 646 | 6.15 | 8.20 | 2.34 |
Unsuccessful Streams
Total Streaming Losses | IP | ERA | WHIP | Strikeouts | IPS | K/9 | BB/9 |
53 | 208 | 8.05 | 2.00 | 162 | 4.08 | 7.00 | 4.46 |
I’m satisfied, but most of all shocked at the results. There is an inherent floor when picking streamers and I would side with upside over being conservative often. The reasoning is simple: if you are in a position of streaming, it means that you playing catchup, not trying to maintain the status quo. Aim for the heavy swings each week at the risk of some terrible plays that ultimately don’t hurt given you’re already tough situation.
I would try to balance this, of course, and I think the Wins vs. Losses is reflected strongly here in that approach. Successful streams produced a legit fantasy ace across 100+ starts and 900+ innings this year, while unsuccessful starts were detrimental to your week. A 66% chance of picking an ace is a gamble I will gladly take, and across all 900+ innings, streaming blindly would have produced for any team.
For reference, among all qualified starters in 2018, our “Mega Pitcher” of streamers would have returned the 25th best ERA and 30th best WHIP in the majors. That’s a #3/4 starter ratio wise, not to mention the 800+ strikeouts to go with it. In roto leagues, taking this stream once per day, would have secured a #1 spot in Wins and Strikeouts, while not losing ratios. Obviously not all streamers are available and there’s a lot more to it than that, but I’m a bit startled to see how well this worked out.
For H2H this gets a bit trickier. You don’t have the sample of the full season to wipe out the bad ones, and as seen above, even getting three losses in seven days could have been a death sentence for your week. It’s still a risky proposition with each stream and if you aren’t given the luxury of “in the long term”, it can hurt very badly.
I really should talk more about those losses because they really hurt. An 8.05 ERA should showcase the innate risk your taking with these streamers as they could kill your week in a H2H league if you’re spinning the wheel a just once or twice. The overall value is still there if you do it often enough, and there were plenty of streams I had to choose when there were no solid options on a given day. There are ways to still raise the floor even if you’re streaming.
All of that being said, it’s hard to say this will work again. Streaming shouldn’t be this successful. It really shouldn’t. This wasn’t purely me throwing darts at the wall, though, and let’s take a look at the most streamed pitchers I chose + the teams I streamed against the most and how they did.
Most Streamed Opponents
Opponent | Games | W | L | Pct |
---|---|---|---|---|
NYM | 19.0 | 12.0 | 7.0 | 0.632 |
SD | 16.0 | 13.0 | 3.0 | 0.812 |
DET | 11.0 | 10.0 | 1.0 | 0.909 |
BAL | 10.0 | 7.0 | 3.0 | 0.700 |
CIN | 9.0 | 6.0 | 3.0 | 0.667 |
PHI | 9.0 | 5.0 | 4.0 | 0.556 |
PIT | 8.0 | 4.0 | 4.0 | 0.500 |
LAD | 8.0 | 4.0 | 4.0 | 0.500 |
SF | 7.0 | 6.0 | 1.0 | 0.857 |
TEX | 7.0 | 6.0 | 1.0 | 0.857 |
I can’t say I’m shocked about most of these teams, especially that the Dodgers and Pirates didn’t return quite as well as the others. The Mets getting 19 starts makes sense when considering I started plenty of Miami Marlins arms, a few Braves arms, and members of the Phils. No, not the Nationals. They have aces and…nothing close to aces in that rotation.
Most Streamed Pitchers
Pitcher | GS | W | L | Pct |
---|---|---|---|---|
Mike Minor | 6.0 | 5.0 | 1.0 | 0.833 |
Andrew Suarez | 6.0 | 5.0 | 1.0 | 0.833 |
Tyler Mahle | 6.0 | 5.0 | 1.0 | 0.833 |
Nick Kingham | 6.0 | 4.0 | 2.0 | 0.666 |
Chris Stratton | 6.0 | 2.0 | 4.0 | 0.333 |
Derek Holland | 5.0 | 4.0 | 1.0 | 0.800 |
Jaime Barria | 5.0 | 4.0 | 1.0 | 0.800 |
Homer Bailey | 5.0 | 2.0 | 3.0 | 0.400 |
Dereck Rodriguez | 5.0 | 5.0 | 0.0 | 1.000 |
Dan Straily | 4.0 | 3.0 | 1.0 | 0.750 |
Considering the amount of discussion about my “underserved” love for Nick Kingham, he definitely helped out as a streamer when he was cruising, it’s too bad it didn’t work out in the long term, as one start on July 26th ended his three successful streams and ended him as a starter for the season (with Chris Archer joining the rotation).
I’m a bit shocked how I kept going back to Chris Stratton, which was mostly at the beginning of the season save for his time against the Padres in September. He just isn’t good.
Same goes for Homer Bailey, though my instinct is that these were desperate streams rather than strong ones. It’s hard believing in Homer Bale.
Conclusion
2018 was a successful year for streaming. There is a science to it gauging ceiling and floor, recent trends in team performance and if pitchers had recent success due to a fortuitous evening or legitimate changes in their approach. Examples of the latter include Derek Holland and Mike Minor, while there’s often an opportunity to get easy starts from no-names that have a strong matchup and occasional strong stuff, ala Tyler Mahle, Jaime Barria, and Dereck Rodriguez.
I don’t expect this to go that well in 2019, though this should show there is still opportunity to be had picking streamers even in this massive “information age” of fantasy, where “sleepers don’t exist anymore.”
I hope to also improve the methods of which we judge the offenses that we stream against. There are popular metrics like wOBA (over what range? Full season? Month? Two Weeks?), BAA vs LHP/RHP, Home/Road, etc. but I think there is still more work to be done here to nail down a more consistent return. There’s a chance that we’ll have something ready to go for the 2019 season to help with this.
We’ll see. Let’s do it again next year.
Nick,
Question: Why is Framber Valdez not getting ANY love? Guy flew through the Houston system and showed chops in MLB, despite very little minor league seasoning.
I think it’s because his stuff isn’t all too great. I’d quantify his fastball and curveball as both “good enough” but neither are excellent + he doesn’t have a decent third pitch.
It makes him worthy of being a #5 option for an MLB squad, but with Josh James more likely to take a rotation spot in 2019 and McHugh working back in the mix, it’s suddenly getting very crowded for Framber to find room.
A decent last round option, but not something I’d bank on consistent production.
Thanks! Seems coming up his curve was unhittable.
Did you happen to break it down by month? At the beginning of the season you said you didn’t like streaming in April but I’m guessing you probably did pretty well. I’ve always thought there was opportunity in April due to the fact that offensive numbers are historically lower plus there are usually more streamer worthy pitchers available.
I know one year doesn’t prove anything but if you can look back or keep track of streaming by month I wonder if anything would pop out (maybe September as well).
Good work! In terms of if this performance will regress, I would be curious to see the full range of peripheral stats for the streamers (xFIP, xwOBA, etc.). Although I’m not sure if that’s possible to put together.
I suspect the next step in streaming analysis is really focusing on the strengths and weaknesses of different teams. How is Oakland at hitting 95+ MPH fastballs? Mets on fastballs up in the zone? Yankees on sliders down and away?
The issue with this detail is that your sample size for the data underlying it (xwOBA, swtk%) gets really small, so maybe you have to use multiple years.