Streaming Models for Week 13 (The Deep Stream)

[Sith_Are_People_Too]

Week 14 of The Deep Stream is now available here.

The Deep Stream – 2019 Week 13

BY DAN KERBER / KERBERDAN89 / NOVEMBER 26, 2019

Welcome back to The Deep Stream where every week I share waiver recommendations for the three streaming positions (QB, TE, DST) based on the Deep Stream scoring projection models. Using the models, you can plan ahead and grab the best streamers before they become popular waiver-wire adds, saving you FAAB / waiver priority. The models can also help you identify players and defenses that are good season-long adds and see when you should get away from underperforming assets that are not likely to turn things around. You can find a more detailed introduction to the approach and models here.

Two great complementary tools to aid in your analysis are The Fantasy Footballer’s Stream Finder and Strength of Schedule Tool. For some examples of the Stream Finder in action, see my introductory post from a few weeks ago.

All scoring data is from FantasyPros with standard scoring settings. With those housekeeping items out of the way, let’s dive in.

Summary
A summary of the recommendations contained in the sections that follow is below.

Quarterbacks

Adds
Ryan Tannehill (79% available). Since taking over full-time in Week 7, Tannehill is the QB3 in full games played and his remaining schedule has him projected to be the QB2 rest-of-season (RoS). Unfortunately, he finished second only to Lamar “MVP” Jackson in scoring this week, which means you will probably have to pay up to get him.

Ryan Fitzpatrick (92% available). Fitzmagic is having fun playing QB regardless of how badly his team is doing, which translates to lots of points for your fantasy team. I would prefer not to play him this week against Philadelphia where he projects to be the QB21, but Weeks 14-16 he is the projected QB6 and a worthy add.

Streams
Sam Darnold (71% available). Two-week stream. Darnold’s next two opponents are the QB-friendly Bangles and Dolphins, making him the projected QB10 and QB7 in Weeks 13 and 14, respectively. Have other options to finish out your playoff run, though, as Darnold projects to be the QB28 Weeks 15-16.

Nick Foles (72% available). We still only have a two (full) game sample size with Foles, and his rough outing against Tennessee makes me reluctant to consider him viable RoS play. However, I will happily stream him this week against a Tampa Bay defense that is still giving up the most schedule-adjusted fantasy points to opposing QBs.

Stashes
Jeff Driskel (78% available). Weeks 14-15. In his three full games played Driskel is the QB12. I would avoid him this week, but he projects to be the QB2 in Week 15 thanks to a tilt with the coverage-adverse Buccaneers. He is also a reasonable play Week 14 vs. Minnesota.

Andy Dalton (95% available). Weeks 14,16. Dalton is a limited QB on a poor offense, but thanks to positive matchups he is projected to be the QB11 Week 14 vs. Cleveland and the QB6 Week 16 vs. the Dolphins.

Jacoby Brissett (48% available). Week 14. Brissett was a popular streaming option this past week and pooped right in his big boy pants. Despite that, he is still projected to be the QB4 when he lines up against the Tampa Bay Can’t-Covers and is a good stash. If he’s not currently available in your league, be sure to “drop it like it’s hot” and check to see if he is dropped when waivers are run.

Daniel Jones (70% available). Week 15. Another middling QB on a so-so offense in a great matchup, Jones gets a shot at returning to Danny Dimes form when he plays the not-tanking-but-still-bad Miami Dolphins in Week 15.

Drops
Jared Goff (65% owned), Philip Rivers (53% owned) and Jimmy Garoppolo (71% owned). One more time for those in the cheap seats: All three of these QBs have difficult remaining schedules, project as QB25 or worse RoS, and should be dropped. For Goff owners that have held on this long, “King” Goffrey projects to be the QB14 this week against Miami, so if you don’t have better options you can hang on for one more week. But he is projected to be the QB29 or worse every remaining week of the season after that and should only be played if you were made an offer you couldn’t refuse by local mafia-types to compel you to throw your fantasy playoff matches.

Josh Allen (84% owned). Since I first recommended trading him away four weeks ago, Josh Stallion has been the QB5 on a points-per-game basis. Super-sized ‘L’ for me. Despite that, I am quintupling-down and calling Allen a drop, but with a caveat. If we take an optimistic view and project him to match his performance level of the last four weeks for the rest of the season, he would go from being the projected QB24 to the QB13 RoS. So, if you believe that his last four weeks are more representative of his future performance than the rest of the season, AND you don’t have better options than a low-end QB 1, Allen is probably worth keeping.

Tight Ends

Adds
Jacob Hollister (40% available). Hollister is the projected TE7 RoS and the TE1 Week 16.

Ryan Griffin (65% available). You will likely be competing for his services on waivers, but Griffin is the best widely-available TE add this week and projects to be the TE11 RoS.

David Njoku (75% available). Njoku’s projection assumes usage similar to his one full game in Week 1, so this is an extremely speculative add, but it is also one that could pay off nicely.

Jack Doyle (62% available). Doyle currently only projects to be the TE16 RoS, but with Ebron on IR, Doyle should be in line for a significant increase in target share, giving him mid-level TE 1 upside.

Delanie Walker (70% available). I am keeping him here one last week in the event he finally returns from injury. If he does, he projects to be the TE12 RoS.

Stashes
Vance McDonald (38% available), Week 14. McDonald is projected to be the TE3 when he plays Arizona and should be stashed. Though not currently widely available, his 0.6 point outing this last week means he will be dropped in many leagues when waivers run.

Streams
Nick O’Leary (0% owned). If Seth DeValve continues to be sidelined with an oblique injury, O’Leary becomes a sneaky streaming play this week against Tampa Bay, the second-best matchup in the league for fantasy tight ends.

DSTs

Adds
Carolina (58% available). Surprisingly available in over half of leagues, Carolina projects to be the DST6, DST6, and DST9 over the next three weeks. Week 16 they project to be the DST14, so it’s worth considering other options if you advance to the finals.

New York Jets (84% available). Sure to be a popular add this week, the Jets project to be the DST2 Weeks 13 & 14 and the DST10 Week 16. Be sure and have another plan for Week 15, however, when they play the Baltimore Points-Scorers and project to be the DST28.

Streams
Philadelphia (62% available), Green Bay (68% available). Two-week streams. Both teams have been listed as stashes in this column for the last four weeks, so hopefully you already have one of them on your roster. Philadelphia projects to be to be the DST4 this week and the DST3 in Week 14; Green Bay projects to be the DST3 this week and the DTS9 Week 14.

Stashes
New York Giants (94% available), Weeks 15-16. The Giant’s defense has been quite mediocre this year, but in their final two weeks of the fantasy season they play Miami and Washington, making them the projected DST5 over that time.

Kansas City (76% available). Weeks 15-16. The entirely-serviceable Chiefs defense projects to be the DST6 Weeks 15 & 16 combined.

Washington (86% available). Week 16. The DST22 on the year projects to be the DST4 championship week when they take on Daniel Jones who, as everyone knows, is happy to throw the occasional interception but gets downright giddy when given a chance to lose a fumble.

Houston (72% available). Week 16. Jameis Winston has committed a league-leading 24 turnovers through 12 weeks. He faces the Houston defense in Week 16 where they are the projected DST5.

Drops
Buffalo (87% owned, YTD rank: DST9, Projected rank Week 13-16: DST31). Despite being a top-10 defense on the year and putting up a combined 23 points the last two weeks, Buffalo’s remaining schedule is 31st-best against fantasy defenses, making them the projected DST31 RoS.

Chicago (96% owned, YTD rank: DST13, Projected rank Week 13-16: DST29). Another good defense with a bad schedule, Chicago has the worst remaining schedule in the league for fantasy defenses and projects to be the DST29 RoS.

That’s it for this week. As always please share your feedback in the comments section below or on Twitter. See you in the stream!

[beske187]

Ahh yess, been looking forward to this

[dbmville]

Curious do you have any data on how accurate your models have been? I mean I love the information and effort, very much appreciate it but just wondering?

[Sith_Are_People_Too]

Fair question. I have done accuracy comparisons against several fantasy analysts, including The Fantasy Footballers since they tend to rate very highly in accuracy, and they’ve compared very favorably. Doing the weekly ranking accuracy is a manual exercise and pretty time consuming, so I have not done it consistently, but over the full year in 2017 my rankings on QBs were within 10% of The Ballers in accuracy and on DSTs were 51% more accurate than the Ballers’ aggregate accuracy (I had not extended the model to TEs at that point). I’ve done spot-comparisons since and they’ve held their own, but haven’t analyzed a full year since then.

Although I use them for weekly start and stream decisions, the real power of them for me (and the reason I created them to start with) is for seeing trends to get breakout players before they are on my opponents’ radar, and identifying good streaming matchups a few weeks ahead and getting them before I have to compete with everyone for the “stream of the week”. They’ve worked very well for me for this purpose for a few years, so I decided to start sharing them this year. I am sure there are ways to empirically measure the accuracy of the future-looking projections a control or other analysts, but have not dug into that yet.

I appreciate the question and interest, and am open to your thoughts and feedback.

-Dan

[ydocgnortsmra]

I really appreciate the effort you put into the charts and have found them extremely useful the past few weeks. Keep it up and I will happily keep reading them!

[Sith_Are_People_Too]

Thanks @ydocgnortsmra! Glad you’ve found them useful!

[Kadders]

Awesome work been looking forward to these!

Do you do similar for wr and rbs our of interest!?

[Sith_Are_People_Too]

I don’t (yet anyway). There are more considerations for skill positions (e.g. WRs playing slot vs. outside, RBs that are pass-catchers vs. runners only, etc.). I think it’s possible, but I haven’t worked through it. The models I have I’ve used myself for a few years so I am confident in how effective they are. Will definitely consider it in the future though!

[Kadders]

Yeah course it’s a lot I’m just a bit greedy and love looking at them and make decisions etc

Basically you are/would be a 2nd team manager of my side

[dbmville]

I cant say it enough how much these are appreciated. They have seem pretty accurate honestly. Idk if its as easy as this but essentially you could add a second column for each of the weeks and have their finish position on the week or points on the week comparatively to what was projected through the model. That would also require the previous weeks to be shown and then thats a lot of information to sort through. I dont mean to add more work on it at all either just some thoughts while i dive deep into the information you gave
Honestly so great to have

[Mr_ButterFace]

If you aren’t making use of this post every week, you’re doing it wrong.

[Sith_Are_People_Too]

[beske187]

Thought you heard the last of me lol. But, Walker is on IR now, would slot in Jonnu in his spot in your projections?

[Sith_Are_People_Too]

It’s a slight downgrade based on how Jonnu’s four weeks. Average projected score Weeks 13-16 drops from 7.0 to 6.4, and projected ranking drops from TE12 to TE16 (small changes make a big difference in jumbled mid-tier TE rankings).

[Sith_Are_People_Too]

I don’t normally have time to do this each week since I have to focus on getting the model and article for the following week ready, but several players outperformed their projection yesterday and I had made several recommendations to fade some of them based on the projections, so I thought I would do an accounting on where it looks like the projections hit and missed on Thursday’s games.

We won’t have rankings or confirmed points using Fantasy Pros scoring until all the games are complete, but we can compare scores to projected points and infer if the projected ranking was in the ballpark or not. The one miss I am taking the biggest ‘L’ on is one Josh “Stallion” Allen. It’s clear his recent 5-6 weeks of improved results are sustainable, and future performance should be looked at through that lens. I’ll also take a small ‘W’ on the Atlanta D. Lots of respected analysts felt their two-game surge a couple of weeks back aligned nicely with coaching changes and was sustainable. I held firm that it wasn’t enough to override the season-long performance yet, and the last two weeks seem to bare that out.

Other results from Thursday seem like reasonable variance, but I’ll continue to look into the numbers to see what can be gleaned from any trends. Hoping for better results for the remaining slate of games this week, especially for the folks I’ve given recommendations to based on my model. Feedback or thoughts welcome.

Summary: QB 1 for 5 , DST 5 for 6 , TE 2 for 4

QB (projected ranking, likely hit/miss)
Trubisky – QB23, MISS (likely top-6)
Driskell – QB16, N/A, ruled out Thurs
J. Allen – QB20, MISS (likely top-10)
Dak – QB17, MISS (likely top-10)
Brees – QB9, MISS (likely bottom-12)
Ryan – QB11, HIT (likely QB10-14)

DSTs (projected ranking, likely hit/miss)
CHI – DST15, HIT (likely outside top-12)
DET – DST26, HIT (likely outside top-12)
BUF – DST31, MISS (likely top-6)
DAL – DST22, HIT (likely outside top-12)
NO – DST7, HIT (likely top-6)
ATL – DST32, HIT (likely bottom-6)

TE (projected ranking, likely hit/miss)
Shaheen – TE33, N/A, ruled out Weds
Hockenson – TE23, HIT (likely outside top-12)
Knox – TE22, HIT (likely outside top-12)
Witten – TE21, MISS (likely top-3)
Cook – TE20, MISS (likely top-12)
Hooper – TE2, N/A, ruled out Weds

[legacy_monarch]

Would you start Kyler over winston with those winds?
Or would you pivot to tannehill or goff?

[Sith_Are_People_Too]

I would try to get away from Winston if I could (projected QB7). I have Kyler projected to be QB13, Tannehill to be QB8 and Goff to be QB14. So if Tannehill is still an option I’d go with him. Unless the winds are forecast to be above 20 MPH, I would go with Kyler or Goff. If they’re over 20 before game time, I would take either over Winston.