…said absolutely no one ever. But feel free to post thoughts, questions, theories, code, data, bad beats, or whatever here. If anyone is interested in running or coding their own optimizer and CSV generator for DFS I can help with that.
4 Likes
Isn’t DFS the actual “no money, everyone’s solid” pursuit now?
not really, but it might be no money rake is 16% in tournaments.
I think MLB and NBA are still soft, but maybe it’s because I’m better at it. NFL there’s just so much data out there and a long time to process it that the edges are probably lower unless you’re actually good at it.
So yes
1 Like
how do you approach correlation in nfl gpps? My current thought is that your lineups should be as correlated as possible for mid-size GPPs (lets say in the 1000s entries) but the bigger they are the less correlation (relatively speaking, I’d still stack a receiver with my QB) as the odds of someone hitting perfect one-offs are much higher with 200k entries.
meh, i kinda hate that approach but i guess yeah, you won’t be printing money if you suck.
it’s just a fun game.
Seasonal FF definitely is
In a way, yes. The cash games in NFL are probably unbeatable. People who can’t tie their own shoes get an entire week to find someone who’s done the work and is sharing correct answers. MLB and NBA cash are still beatable but not for much I think. I finished at ~15% ROI in MLB this year but that was from game selecting super hard. You can’t just enter everything now like in the past. I’ve won every year in MLB but it’s way harder. My swongz in NFL are huge and I’ve lost last 2 years.
In tournaments, the edges come from exploiting ownership far more than anything else. That’s more interesting and complicated, but the odds of winning any single gigantic tournament are meager. I think today you’d probably need at least a six-figure bankroll to (1) play correctly and (2) get enough volume down. I’m not sure how many guys are actually max entering every GPP, but it’s probably a few dozen? Just in MLB alone you can get down about $20k in tournaments on one site. More for NBA and way more for NFL.
Compared to poker, DFS is way easier, the rake is twice as high, and the multitabling script regs are multiplied by an order of magnitude. Like basically if you could take all of the worst elements of poker and amplify them you get DFS. And I still don’t think any of the sites are in the black b/c of the vc money
I think correlation matters a bit less than people think and ownership matters way more, so the thing I think about with large field correlation is how it affects ownership. For example, some gamestack like Kyler + Kirk + Metcalf was high owned last week so I don’t think it’s worth much since it kills 3 lineup spots. Antonio Gibson + WAS D was a good one imo. I also think that high floor / high ceiling is way overblown.
yeah i think about it a lot in terms of QB ownership. Like if Rodgers is going to be popular, i tend to stay away more because it’s necessarily mean Adams is going to be very highly owned as well and there’s no valid stack without Adams. Otoh if Burrow is going to be popular this week, the ownership is going to be split on his 3 WR and Gio, so it’s easier to still play him at high ownership.
i’m not a big fan of the RB+D correlation for very large GPPs. You need defensive TD’s which means not only is your RB not scoring that TD but his team isn’t getting the ball.
1 Like
To expand on this a bit more, Lockett at 9% is so good. I had him on just about the same projection as Metcalf whose ownership I guessed would be in the 20% to 25% range. That’s what you should be looking for in tournaments.
The deep dive is a little more complicated. The reason Metcalf had a tough go of it is because he drew a shadow from Patrick Peterson. None of the sites I subscribe to were projecting that WR/CB matchup though. For instance, see:
However,
Dating back to Week 16 of last year, Metcalf has played six games against non-Peterson cornerbacks and averaged 4.7 catches and 96.2 yards per contest. However, against the Cardinals, he has only caught two passes for 23 yards combined in two games, and Peterson has mostly shadowed him during that time.
According to NextGenStats, Peterson tracked Metcalf on 42 of his 49 routes Sunday and allowed just one catch for six yards on four targets. Peterson’s toe-tapping interception in the end zone also came on a play in which he was guarding Metcalf.
Metcalf did not have a catch and was only targeted once when Peterson guarded him last year in Seattle.
The projected WR/CB matchups don’t always tell us much because the receivers switch around a lot:
Outside of shadow situations, LWRs and RWRs generally match up with both outside cornerbacks over the course of a game, as almost every wide receiver moves around the offensive formation to varying degrees, and almost every defense plays zone coverage on at least half of its snaps.
Lockett in the slot means avoids PP regardless of if you project the shadow coverage correctly. David Moore was probably a great play on a showdown slate but not sure he gets there enough on a full slate.
So, yeah, I liked Lockett and had a ton.
We can check your theory in here. Just need the data.
my DFS tale is a few years ago one week in my single entries I had four top performing players at their position and a fifth in the top 3
and still didn’t cash
that’s when I quit other than lol gambling a minute entry once a week if that. **** DFS
1 Like
Mission accomplished.
1 Like
Is there any decent free DFS option that could be used for a private group contest?
yeah. I expect them to be correlated a bit of course, but also to cap their ceiling, which i guess you can say is not as important.
If I do pay any attention to the D other than just spreading the ownership it’s to try and leverage a mediocre popular value play. So like last week I was planning on being really heavy on the Browns D vs Cin because of Gio’s ownership, but I switched off a bit when the Jamal Williams news broke and it was obvious Gio’s ownership % will go down.
1 Like
I used to have the data in a big database but can’t seem to locate it now. Would rather come up with a source where I can just download it all quickly because I really don’t feel like writing a scraper right now.
Not sure about free.
So are you also not playing RB+WR stacks from same team? I get that it’s subtractive scoring outside of QB/pass catchers and DST/kick returners but it still seems like the 2 players from same team comes up a lot.
I’m not playing RB+WR from the same team with very few exceptions.
And in a way my point about less correlation in super large tournaments is kinda about that. Like sure Ridley and Julio can both have good games, but it’s hard for them to be 2 of the top 4 WR on a Sunday slate, which you probably need to win the milly maker.
So my idea now is to run “skinny” stacks in the milly mostly (QB+WR/Opp WR preferably) and not a double stack or full game stack that you see often and is a very good strategy for low-mid size GPPs.
Also this is just me spitting out ideas I have in my head I’m not a good NFL DFS player by any means.