Rewind: Quarterback Streaming

00:00:00
Speaker 1: Hey, everyone, it's jj with another late round podcast rewind. I hope you're all doing well, and I hope you're all staying safe. Quarterback streaming is one of my favorite fantasy football topics, partially because it's so controversial. There's a large number of people out there who think that it's not a long term sustainable solution in fantasy football. But I've been streaming for years with success, not just with my individual fantasy teams either, but through my other podcast called Living the Stream. I explain all of that on this rewind. Now you're gonna hear about some of the same concepts from the Superflex rewind last week. You're also gonna hear an advertisement at the top of the episode that will make you feel sad that the NBA playoffs aren't happening right now. I'm sorry about that, but you're gonna hear some of the same concepts from last week's Superflex show. That's just because there's some overlap and strategy here. I thought it would be good to put the episodes close together so everything is top of mind and makes more sense anyway. I hope you're able to walk away from this show with a little bit more confidence in quarterback streaming enjoy. One of my biggest passions within fantasy football is understanding how fantasy managers behave on a week week basis and using that knowledge to gain an edge over my competition. If we're being honest, the analysis that we do in the fantasy football world, it's usually hindsight driven. It's based on things that have already happened. In season long fantasy football, we usually aren't analyzing the decisions that we made and why we made them. We're looking at what happened and putting this giant puzzle together. We're saying this strategy worked based on these results, instead of can we actually apply this strategy. Now, there's nothing wrong with doing analysis the way that we do analysis. We've collectively come up with a lot of good strategies that work, and we've come up with ways to evaluate players by backtesting and looking at historical data. We just don't often see analysis that's solely based on predictions that we're made. It's usually based on things that have already happened, and that's where an edge can be found. You've got to sort of change the way you think about fantasy football for this to work, though, Like I said at the time, I love trying to understand how fantasy managers behave on a week to week basis, and that's how your mindset has to switch. You know. It's really easy to say that you can wait on wide receiver in your draft because we're going to see ninety or more wide receivers each year finished with at least one top twenty four performance. But that's all hind set analysis. It's based on descriptive data. How many of those ninety plus wide receivers had predictable performances. The answer to that is most definitely not ninety. Or take quarterback for example. If you're projecting Patrick Mahomes to outscore the twelfth ranked quarterback entering a season by one hundred points, then you're going to assume that Patrick Mahomes is one hundred points better than the worst starter in your twelve team league. That's not really true, though, because throughout a season, you're not starting that same player in your quarterback spot each week unless you drafted an elite passer, or at least you shouldn't be doing that. And this concept is different at quarterback than it is for other positions. Today, I'm going to talk about that. I'm going to talk about why you can effectively stream the quarterback position, and instead of giving you some analysis that's based soul on hindsight and how quarterbacks finished a season, I'm gonna provide information that's based on actual predictions. Now, before getting to all that, I wanted to remind you guys about a FanDuel offer that's running through the NBA Finals. Enter the featured NBA shot Last three Bonus contests during the Finals through June thirteenth to compete for a share of a special bonus on top of regular payouts. If the player in the MVP slot of your lineup scores the last three pointer of any game, you'll win a share of that night's bonus prize in addition to your regular contest payouts. For more info, visit fandle dot com slash Last three Bonus That's Last Hyphen the number three Hyphen Bonus or download the app today. Eligibility restrictions apply. Some of you may not be aware, but I've got another podcast that I do during the NFL season called Living the Stream. My co host Denny Carter and I have been doing the show for a handful of years now, and on the show each week we give quarterback defense and tight end recommendations for you to stream off the waiver Wire. It's a show about players you can get off the waiver wire. That's what streaming is. It's finding someone that you can plug and play off the waiver wire each week. Now, we've been tracking our picks since twenty fourteen, so we've now got five years worth of predictions, and on the show we tend to dictate whether a player is a usual, normal waiver wire play or if it's one that's a little bit deeper. Now, last season, when focusing only on our main plays for that given week, our quarterbacks averaged about twenty fantasy points per game. It was nineteen point nine to six eight. It was a good year for quarterback scoring and that got us into QB eight territory on a points per game basis. And to be clear, twenty points per game was only point seven points per game off of the QB four, which was Deshaun Watson. In twenty seventeen, the quarterback streamer averaged seventeen point nine points per game. When you include Aaron Rodgers and Deshaun Watson and your data set there who each played just seven games. That was a QB eight average, but if you remove them, that's a QB six points per game average. Through using only waiver wire quarterbacks, twenty sixteen, we had an eighteen point eight points per game average that tide Kirk Cousins for the QB six. In twenty fifteen, it was nineteen point one points per game that was the QB seven, And in twenty fourteen we averaged seventeen point eight points per game. That was the equivalent of streaming the QB seven as well. So basically, what these results are showing you is that we've effectively streamed between a QB six and a QB eight every single year that we've been doing this, and we've got receipts. We've got over one hundred and seventy episodes that we've done where our listeners are using the picks that we're providing. And here's the other thing to all of this, we're not benefiting from one of our quarterback streamers becoming an every week starter. We had to stop streaming Deshaun Watson once his ownership got too high when he was a rookie. We had to do the same with Ryan Fitzpatrick and Jamis Winston last year. There are a lot of examples of players that we streamed who then went on to be pretty good starters, and that would have made these numbers look even better. Now. I could stop the podcast there and say, see, this is why quarterback streaming works. But there's reasoning and math behind all of this too. I've talked about some of these concepts in the past, but they play into why quarterback streaming can work. The first idea is replaceability. This is really what drives the idea of streaming and why you can stream positions that are in low demand in your lineup, or what I call onesie positions. So we're talking quarterback tight end defense here. Because you're only using one of those positions in your lineup each week. What that does is it forces usable players at those positions onto your waiver wire. It's really just a simple supply and demand equation. So let me just give you an example of how this works. I've got a ten month old daughter. She does her business in diapers like any other ten month old would. If someone were to come up to me and say, just this week, you can either have and use five diapers with unlimited toys for your child to play from, or you can have five different toys with unlimited diapers. What would you choose as a parent, Well, I'm not trying to clean up mess after mess or give my kid a diaper rash, So the answer is to obviously just have a few different toys for her to play with while having unlimited diapers, because diapers are more important. So if you ask that question to one hundred parents, what's gonna happen is that we're gonna see all these toys just hanging out there with no home. And no, this isn't the plot line to Toy story four. But when everyone chooses diapers, they're not choosing toys, and the reason is because of the importance of those things. And even though this analogy was weird as hell, that's the same thing that goes on in fantasy football. Teams don't need quarterbacks, tight ends, and defenses like they do wide receivers and running backs. So in most leagues, if you hit your waiver wire in any given week, you'll have an easier time finding usable players at those positions. You'll at least have an easier time finding usable players at those positions that no one else cares about because clearly there will be weeks where obvious running backs and wide receivers are being added off the waiver wire. But everyone in your league once and needs those players. No one needs Sam Darnold in two quarterback leagues. This is all different. When you increase the demand of the position, it limits the supply. So when teams now need multiple quarterbacks to complete their roster, there aren't as many usable players on the waiver wire. So I want to be clear here, no one is advocating you stream quarterbacks in two quarterback leagues. Now. The other reason that streaming can work, or the other reason streaming has worked, is because of the idea of predictability, and this is a little bit more complex than the basic notion of replaceability. With predictability, I'm really asking how predictable was this performance by this quarterback on both a season long and weekly level. Let's start with the season long level, which isn't necessarily as important for the idea of streaming, but it still helps wrap your brain around why streaming can work. Really, with season long predictability, we're just looking at a player's average draft position versus the postseason result. So how we viewed a player entering the year versus how we viewed him once the season was over. There's this misunderstanding in fantasy football circles that early round quarterback picks are safe, that you're getting Patrick Mahomes in the third round because he comes with a lot of safety, But that's not really the case. The correlation between where a top twenty four quarterback was drafted and where those top twenty four quarterbacks ended a season since twenty eleven has actually been weaker when compared to top forty eight running backs and wide receivers. And I'm using top forty eight there because you're obviously starting at least double the amount of running backs and wide receivers in your typical league. And this is the case when you move up to top eighteen versus Top thirty six, top twelve versus top twenty four, and even top six quarterbacks versus the top twelve running backs and wide receivers. That may sound insane, but just to give you some particular insights, since twenty thirteen, the number one quarterback drafted according to Myfantasy league dot COM's ADP has an average finish on the year of QB seventeen and that's thanks to injury. That's the same thing that happens with early round running backs and wide receivers. But even on a points per game basis, the average points per game is just eighteen point six. You're not gaining any sort of edge with that. So the reason I'm bringing up season long predictability is because it's a very quick way to tell you to stop drafting your damn quarterbacks early. When you draft them late, you have less of an attachment to that player. And when you have less of an attachment, you have an easier time letting go of that player when he doesn't perform. And when you let go of that player, you can start to stream. This is why I say that streaming the quarterback position shouldn't be your go to strategy, and most li you should be looking to find a late round quarterback who can become a weekly starter, like Patrick Mahomes last year or Matt Ryan in twenty sixteen. If your late round quarterback doesn't work out, that's when you feel okay to stream. So that's predictability on the season long level. What's honestly more important is predictability on a weekly level, because that's what we're talking about with quarterback streaming. Like I said at the top, we often do analysis in hindsight. We can say, here's how many times this semi obscure player finished with this many top twelve performances, But how often are we actually using that player when he blows up with wide receiver and running back? That number of times is a lot less predictable than it is at quarterback. And really you can use simple logic to understand this as opposed to math. In a given game, a quarterback is dropping back to pass thirty or more times, and we know who those quarterbacks are going to be. Meanwhile, you've got split backfields with running backs, and then a double digit target game is awesome for a wide receiver, and we don't always know who those players are going to be because they're not touching the ball like quarterbacks are. That's the simple logical reason that quarterbacks are more predictable week to week, but you can see it in their scoring as well. There's a statistic called the coefficient of variation or CV, which takes a data set standard deviation and it divides it by the data sets. Mean. What this allows you to do is compare variability across data sets, since standard deviation can look different based on a particular data set, meaning quarterbacks score more points than other positions, so the standard deviation or the amount of variation in that data set is going to be higher. But when you compare that standard deviation to a data set's average, it becomes easier to compare across data sets. You can compare a quarterback versus a running back versus a wide receiver. Naturally, if a coefficient of variation is really high, there's more variance. If it's low, that means there's less variance. That's just how the fraction and equation works out. As you'd guess, quarterback coefficient of variations are a lot lower than the other positions in fantasy football. If you take the QB one to the QB six over the last eight years, and if you look at their week to week scoring, you'll see a coefficient of variation of about thirty eight percent. That's forty seven percent for RB ones, fifty two percent for wide receiver ones, and forty eight percent for high end tight end ones. The QB seven through the QB twelve group also has a CV of thirty eight percent, but the RB twos have one of about sixty three percent. Wide receiver twos are at fifty nine percent, and low end tight end ones are at sixty five percent. Essentially, what I want to show here is what happens when you get into the QB nineteen to QB twenty four range. These are non starters in your twelve team league, and in some leagues a lot of these players are just hanging out in your waiver wire. Over the last near decade, the CV for week to week scoring among this group has been fifty seven percent. So they've shown more variants in Fantasy points score than QB ones, but they've got less variants than RB twos and wide receiver twos. What this means is that it's just as easy to predict a performance from a random low end QB two a legitimately mediocre quarterback, as it is a legitimate starter in your RB two spot. So what can you do with that information? Well, if we know quarterbacks are going to give us more predictable output, then our goal should be to aim for the games where that output is hitting its higher range of outcomes, and there are ways to then find those higher range of outcome games. It's why we look at matchups and things like over unders. We're increasing our odds that these players will hit a ceiling, all while knowing that there's some sort of floor because of their predictability. Now, I'm not selling quarterbacks streaming as something that will always work. I'm not gonna tell you that you'll match the qb ones production. I'm not saying that every week will be a great one from your quarterback. Even last season we had multiple single digit fantasy point performances from our streamers. But what gets lost is that you're still gonna get those types of performances from your non streamers. And what also gets lost is what I talked about earlier, and that's that early round quarterbacks bust at a much higher rate than the fantasy football world seems to think. So with absolutely no investment at the quarterback position in most leagues, you can effectively stream the quarterback position. And the reason you can isn't based on straight up luck. It's not even based on crazy research that takes our hours to do. It's just based on how fantasy football is formatted, and the incredible depth of the quarterback position has just made it all even easier to deploy. That's gonna do it for today's show. Thanks to all of you guys for listening. If you get to subscribe to the Late Round podcast, make sure you are by searching Forward on iTunes, Google Play, a SoundCloud, tune in, Stitcher, and Spotify. And as always, if you're on iTunes, please do me a favor and leave a rating or review, and you can follow me on Twitter at Late Round QB. Thanks everyone, I'll talk to you later in the week.