5 Lessons

Part 5: Post-tournament review

In this 5-part series, Max takes you through his exact process for researching and building his entry in this HUGE tournament.

The tournament is over and in this last installment in the series, you'll see Max deconstruct the slate to evaluate his overall process.


Hey guys, this is Max Steinberg. I am a partner at SaberSim and a daily fantasy professional and I’m back from Miami. I’m in California, and I’m here to bring you part five and the final installment of my road to the FFWC video series. And in this part I’m going to review my lineup, look at some of the top lineups, and then answer some questions that some of you have sent in about my video series, about my process that might’ve been left unanswered in some of the videos. So I’m going to start with the questions right off the bat. So we’ve got a few questions via email and one via Twitter. I just wanted to answer them and then I’ll get into reviewing my lineup. I think some of these will actually tie into my lineup review, so it will work out well.

First question, “Hello Max I have a couple of questions for you.” Okay, great. “I believe you maybe are more of a max entry guy versus single entry. How much different is your process for each of these? Do you think SaberSim’s strength is more suited for max entry events, or maybe coming closer to 20 max entry versus say one or three entries? And as a final question, being a numbers guy, how often do you find yourself making duck calls that maybe go outside the realm of the numbers or the game script?”

I think this is a good question. I think SaberSim obviously is built really well to make a larger amount of lineups, but I think this smart diversity feature is actually perfect when you’re making one lineup and you want to sort of give yourself some options based on the projections that you have, or some projections that you’ve made or adjusted off sabers.

And this is what I did in the Thanksgiving slate where I actually won the seat, is I only made one entry into the tournament, but that one entry was based off of building a bunch of lineups, looking at them and then choosing what I liked. And that’s because of the smart diversity feature, which I talk about in another video and I highly recommend you finding the smart city video on our YouTube page because I described in detail and it’s a really, really cool thing that we use in our lineup builder that leverages our simulation data in this very cool way that helps take into account upside of players and basically real outcomes of games that could happen. And if you set this to PI or very high, what’ll end up happening is you’re going to get lineups that are sort of based on things that are like high range of outcome possibilities.

And so that’s what these lineups really are. Like obviously the projected score hit here is the same amount, but the Saber score is sort of taking into account what are builders finding in the simulations of the possible upside of the players and it’s also incorporating correlation and ownership data. So what I actually did in the morning, which I didn’t get into in the video, is I had my build and I got my final build and I ended up actually just looking over some of these lineups. One thing I found is I was just getting a lot of Kyler Murray and I think it was for good reason. He was really cheap. Larry Fitzgerald is really cheap. Christian Kirk was really cheap. Kenyan Drake was really cheap. I thought they were going to do pretty well against this decimated Cleveland defense at home and they did and unfortunately Kyler Murray didn’t throw for that many touchdowns, which would have helped.

Oveerall I was happy with sort of that Arizona stack given I got so much out of Kenyan Drake, but I actually capped the max exposure to Arizona if I don’t do it, I think I end up getting about 30, 31% Arizona. So you can see scrolling down that there’s a lot of Kyler Murray in these lineups. And because of that I was like, okay, I definitely just have to go with Kyler Murray here. Obviously there’s like a Ryan Fitzpatrick one, I thought that was interesting, but given that overwhelming amount, despite the fact my gut initially said and here’s where I’m getting into the gut thing, maybe I’d want to go some other direction. This sort of is showing me, okay, clearly this is the best stack so then I can just use my gut instinct and say, okay, well is there one of these that I like more than the other because the Saber score isn’t that much different really.

So I think I can use my discretion here a little bit and sort of pick one. But my gut actually said that this was a pretty good lineup. I thought Miles Sanders, I just did not think was going to be high end and I liked pairing them with this Eagle’s D that I liked quite a bit. And then I just really liked the scam stack. I thought Odell Beckham Jr. also is going to get low ownership and he did. I mean if you look at the ownership maps here, I was one of two people who had Miles Sanders, if I’m calculating this correctly, or maybe three. Kyler Murray’s 18% that was actually kind of high, but, Blair Fitzgerald 7, Odell Beckham Jr 5, these are really low ownership numbers. My gut was sort of saying okay, this does seem like a pretty damn good lineup and I’m going to keep it.

So I think in user got, when you’re making a single entry, you don’t really have to use the top line up. In fact, if you really set smart diversity to very, very high, you really don’t, it sort of is going to be a great way to give yourself sort of a choice of what to do. I think all in all you can really use your got sort of picking the top line up out of the top 10 or so lineups that Sabersim builds you. I just happened to like this first line up so I hope that answers your question.

Question number two. In a relatively small tournament like the championship, does max use average percentile scores and for huge GBP’s would he suggest using 85 percentile since the winning lineup is going to be where all the players are hitting near their ceiling. He’s talking about our percentile projections and these are essentially based off our simulation data and you know if you want to take into account upside more, you can sort of use these projections, the percentile projections as your actual projections. While I think these numbers are helpful, I actually never go above the average. And that’s again because of the smart diversity feature. If you set it high enough it’s going to take the upside of the players into account and you really don’t need to worry about this. I suggest if you really do care about outside, which I think we all do, you can set smart diversity to very, very high and maybe not the top, top thing unless maybe you’re doing showdown, but one of these two, top two ones and then you’re going to really take into account upside in your lineups.

Given it was a 180 person tournament I didn’t have to care about like that much upside. So I did set smart diversity a little, little lower. But I think you can and when you want to take into account upside, just set the smart diversity feature to higher and you’ll take that into account.

Last question. So this is a question about one of the videos where I talked about the ownership fade slider, which is this thing here. So he says, “You had me confused on vid three with the ownership fade slider. You said as a contrarian player you have to set it to low, but I thought low settings didn’t fade the high end players, making sure I’m not misinterpreting at its use.” Let me just get into this a little more. So I set ownership fade to low and that being said, if you look at my lineup, wow! There’s some real low end plays, right?

What ended up happening is I didn’t really, quote unquote, need high ownership fade in order to have a pretty contrarian lineup. And the reason I was talking about in the video, how you have to sort of know yourself, is because of how basically we project. ‘Cause every single DFS player when they’re projecting players has flawed projections, they’re not perfect, right? So, if you’re working off completely perfect projections, you can just set this number to very, very high and you could just trust that you’re going to spit out like SaberSim’s going to give you really, really good lineups that are perfectly fading ownership, right?

The thing is that their projections aren’t perfect. They’re going to be very, very flawed. Even mine, like I work very hard on them and they’re very, very flawed and they’re usually flawed and this is where I’m talking about knowing yourself in the direction of I favor contrarian plays and I sort of like intuitively are low on plays that I know are going to be high end. And because of that, if I set ownership fade to really high, what happens is I end up sort of … I’m doubling that kind of mistake because what Saber’s going to do is going to favor those low end plays even more. Which I’m already doing. I actually set it to low because I know that I kind of just have the preconceived intention of doing this and I don’t need to do this. I think a perfect example is Leonard Fournette, like I projected him really high. I loved him. And I think part of the reason I loved him is I just thought he was going to be the lower owned play out of Saquon Barkley and Chris Carson and he really was, I mean, if you look at these lineups, let’s see, I’m not sure.

So Saquon Barkley, 42% owned. I’m sure there’s someone with … Chris Carson here is 50% owned. So I sort of was favoring him already. You know, in hindsight it actually would have been better to get Saquon Barkley because well A he would have done better, but I really didn’t need to. If I literally was just choosing between Fournette and Barkley, looking at my lineup and looking at the projected ownership and if I was projecting more reasonably, I would’ve just said, okay, I can use Saquon here. It’s totally fine, right? I’m so contraring already. It doesn’t matter. But because I’m already projecting kind of like a madman, and having Fournette over Saquon Barkley, which in hindsight was a little stupid, setting that ownership fade higher is just going to ensure that I’m just not going to get as much Saquon Barkley as I really should.

So that’s what I mean by just knowing yourself when using ownership fade flutter. Obviously it’s very good, but you have to make sure that you’re not already sort of bias to making projections that are contraring already. Otherwise you’re just sort of going to make your mistakes worse. So I think that’s really important.

Here, let’s just look at my lineup right now. This is what I ended up with and I just want to look at it and see if there’s any regrets. I think in terms of my game stack, I really don’t have any regrets. I think still in hindsight, this felt like overwhelmingly the best value. And looking at these ownerships, I think there’s not really any regrets there. I think Larry Fitzgerald and Christian Kirk and Kyler Murray were low end enough that that seems like a really, really profitable stack. And if I were to do it all over again, I’d probably keep them.

Then Odell Beckham Jr. Coming back, he was only 5% owned. I’m not sure if I can find a lineup with Jarvis Landry. I might try to do that real fast. Jarvis Landry is 25% owned. I think in stacking this game, it was really great to get OBJ and stab because I ended up being really contraring there and I think OBJ, he’s a higher [inaudible 00:12:00] he’s a player who can get big plays. So I think it makes sense that Odell Beckham Jr. gets a big score and there’s more place for Kyler Murray and Larry Fitzgerald and Christian Kirks. I think that made sense. And it’s not surprising to me that Sabersim actually gave me an Odell Beckham Jr. lineup instead of a Jarvis Landry lineup. So really liked that play and you know it, it worked out fine.

I think Odell Beckham Jr. does not look totally healthy, but I’m still fine with that plan. I’m okay taking that risk. George Kittle obviously worked out really well. He was just one tight end that I was really fine with. Kenyan Drake also worked out really well. Miles Sanders worked out really well. There’s a lot of stuff that worked out really well here. But I think the main thing that’s my big regret, which I just touched on earlier, is I really feel like I over projected Leonard Fournette and in reality I think Saquon was probably a one or two points better than Fournette.

I think Chris Carson, I just don’t in the long term think he was a better play than Fournette and because of that I wish that I would’ve projected them more reasonably and it’s possible that Sabersim would have given me a Saquon lineup instead of a Fournette lineup. I’m sad it didn’t work out that way cause I think I would have placed in the top 10 if that was the case. So it’s too bad that happened. This is the type of mistake that I know that I make. So that’s why in general I sort of keep that ownership fade low because I usually end up with these contrarian lineups.

One thing you also might’ve noticed on part four of my video is I was going to use a Patrick Mahomes, Tyreek Hill stack and it’s really too bad that that weather was so bad cause I was really on it. And I think in hindsight I don’t regret getting off of it because the weather was really bad. You know it was really slick. I think passing in general is depressed and even for Kansas City I think it was a little depressed and Tyreek Hill obviously still had a really big game but I think my analysis was right there so I’m happy I did get off of it but it is kind of sad to see that the winning lineup actually had Mahomes and Tyreek Hill cause that would have been really, really nice.

I don’t know that I would have actually ended up with the same lineup cause I don’t think I would have had Greg Ward or Boston Scott. I probably would have had Chris Conley, but my lineup probably would have been a little different. I don’t know if I would have won but it’s a little too bad.

I’m really happy with my process at least that I was sort of on that and I think this lineup in general was interesting. It was a good amount of shockyness with a non shocky stack and I think that’s a smart thing to do. I think if you have a quarterback receiver stack in these tournaments and that quarterback receiver sack is not shock, I think you can feel pretty safe just going shock the rest of the way. I think Boston Scott was an interesting play.

Philadelphia was going to use a lot of the running backs and Patriots D obviously was a good play and ended up kind of under the radar at 4,000 they only got 8% ownership so I think that was a really savvy play by this guy who was actually a poker player. His name’s MazeOrBowie on Twitter. I think his name’s Jesse Martin, so congrats to him, Mr. Good seeds. I thinks this guy’s a pro.

We talked about Ty Johnson in part four I think he was an interesting play. It did not really work out for Mr. Good Seeds and that’s too bad. I think you’d easily have a 12 to 15 point game, especially given he had four receptions. He somehow only had three rushing yards because Math Patricia used this very backup running back from the practice squad, but I think this is a cool line up as well.

This is one that I would have considered Jamis Winston if it weren’t for the thumb injury that he was sort of nursing all week and Mr. Good Seeds also got kind of unlucky here too with Chris Galvin getting injured. He might’ve won this tournament without that. This might’ve been like the savviest lineup of the week really with Winston and Godwin, bringing it back with Amendola and Ty Johnson and just a little… A couple of things didn’t go his way otherwise this could have been the winner. I’m sure he’s feeling pretty sad about it even though he did win a million dollars. And then Peter Gibbons, this was interesting to, so both the top lineups did have Greg Ward who sort of caught a last second touchdown and was close to Stoneman I think or he was Stoneman.

But this one was a fun sweat given, he had a huge Dallas Rams game stack and the Rams were keeping their starters zone at the end, but he really needed Robert Woods to do something and that just didn’t happen. But otherwise, some really savvy plays here. Kenyan Drake, Saquon Barkley, Ezekiel Elliott all scored over 30 points. Pretty crazy. Jared Goff just could not get himself to have a big game, but maybe passing a last second touchdown to Robert Woods would have been different for Peter Gibbons. So congrats to him. Then Mossimo bring up the rear. Derek Carr and Darren Waller I think that makes sense. He also had Ty Johnson, so this actually seemed like sort of the savvy play of some of the good players. I know I’m not using them, but I do think he was a good play.

Then Julio Jones, that’s a really interesting play for Mossimo that ended up really working out. Obviously there’s a lot of entires. He wasn’t going to get shadowed in this game and I think he probably had six entries and this was one of the ways he went contrarion. I’m kind of surprised he didn’t use, if he was going to use, Julio, he didn’t use someone like Matt Ryan. But I’m seeing now that Derek Carr was only 5,600 so now that does make sense.

But again, some core plays, you know I mentioned Christian McCaffrey liking him, I mentioned liking Danny Amendola, sadly I only had one entry, so I didn’t have a lot of variety in my lineups. But you know, these are just all solid plays that people run. It’s just all about making a good game stack, having the right combination and SaberSim really does that for you.

So all in all. It was a really great trip. Miami was beautiful and I had a great time. I’m happy with my 20th place finish. Wish it was a little better. I was definitely sitting pretty and after the morning games having a Miles Sanders Eagles D stack that had 46 points already and having all this [inaudible 00:18:39] exposure but it didn’t work out for a top 10 finish and that’s okay.

I really hope you enjoyed this whole video series. I personally really, really enjoyed it. It was really fun. I hope it let you understand what a professional’s process is throughout the week and how it’s actually pretty simple and it’s something that there’s no secret sauce or anything like that. A lot of the stuff I do is stuff even my proprietary models, even my expect fantasy point model, that’s something you can just find online from someone else. It’s really not something that is so secret or whatever.

It’s just really all about using SaberSim and honestly just trusting what is popping out. Using a little bit of your intuition to pick whatever lineup that you really feel is fast and you’ve got instinct. And then going with it and trusting that your projections are good and trusting that our simulation data is good.

SaberSim is a really cool tool. I really, really recommend it. Luckily for you we are offering a three day free trial and even if you tried us before, we’ve wiped the slate clean so you can actually try us again for free if you want for three days. This weekend is a great time, drafting’s had some really big tournaments, Vandal has some really big tournaments. Really, really recommend trying us out. We have a lot of new features that are really, really cool and it makes it really easy to make twenty, twenty plus months of lineups and do mass multi entry or it’s really good for even just picking one lineup like I did here. So I hope you enjoyed this video series. Good luck to you for the rest of the daily fantasy season, and thank you so much for watching.