I at the last minute picked up Jimmy G and played him over Stafford.
2 Likes
Wence 53 / 88
Fulgham 7 / 48
Scott 13 / 68
Gallup 1 / 15
Lamb 1 / 23
Elliott 12 / 50
DiNucci 6 / 27
Ward 1 / 27
Zeurlein 0 / 16
Right about Wentz, a little off on the RBs, a lot off on DiNucci. The biggest mistakes aren’t if you project a guy at 60% and he’s 40%. It’s if you project a guy at 25% and he’s 5%. Nonlinearity in ownership to profitability. Think about coin flipping contests: if 60% pick heads, you don’t win much picking tails; if 95% pick heads you make a ton with tails.
Both QBs fumbling is a great setup for a weird MVP slot.
Mostly daily talk here where I make a weekly donation.
But for a standard year long league, I was offered Zeke for Aaron Jones. Jones has had a calf injury keeping him out for the last two weeks with an uncertain future. Zeke now plays on an offense that can stack the box because they don’t have to worry about Dallas’ QB doing anything despite being loaded with options. But Zeke does play in the worst division in football history.
Thoughts on this?
1 Like
Both those players might benefit from a change of scenery.
Snap take this trade, imo. Zeke is a clear upgrade and will get his numbers.
Meh, I played DFS few years ago. I just gave up when I realized it’s about as beatable as roulette or craps.
I play season-long fantasy, but really no one here posts about that because I assume interest is way down this year because of covid. And I don’t because my team is 0-7 (about to be 0-8) because of the most ridiculous string of injuries I’ve ever seen in all my years of playing this stupid game. (In this way I will concede that DFS is vastly superior to yearly fantasy, but it’s not as fun for me because I’m playing against strangers, and millions of them at that.)
I almost certainly will not play next season if covid hasn’t been solved by the time it starts.
Auto-reject this trade. Dallas O sucks big balls and Jones should be back next week. Even if Dalton returns, with their patchwork line, Dallas is going to struggle running the ball.
Only reason I might consider this is if you’re in must win every game mode and can’t afford the risk that Jones misses another game - per Packers he would have been limited today if they had practice - so odds are decent he should play.
That’s 100% wrong. I’m profitable over a fairly large sample. It’s just not very profitable unless you’re extremely good at it. Which is fine for me.
A few years ago it was absurdly beatable. You were just not good at it, which is okay, but lol at DFS not being beatable a few years ago.
Also the idea that a hobby should be profitable is silly. If you enjoy the sweat you can get from like a 100 bucks a week worth of sports action, it’s totally worth it even if you’re -EV.
That being said, it’s obvious no one here cares about DFS and the thread is at best a private message convo (and even that seems shaky as Lawn appears to be mostly annoyed with DFS at this point) it’s probably a good idea for me to stop posting about it.
I mean, okay, if you say so. I distinctly recall reading about and discussing on 22 articles stating how “90% of the money won in DFS is won by 10% of the players” back when Fanduel and DK were getting popular. I’m not sure I’d call something like that “absurdly beatable” but good on you if you were part of that 10%; obv I was not, I didn’t have access to nor the knowledge to create the tools to have that level of success. I won a few weeks here and there, but over time it wasn’t profitable.
I do know I’ve cashed way more than 10% of the time in traditional fantasy games than in DFS, for a variety of reasons, not the last of which is playing with the same leagues year after year and learning the other players’ tendencies. But even then I’m subject to the whims of the injury gods, like this year where most of my original starters have been lost for the season.
Of course, it was the same in online Poker. 90% of the people were terrible. The only difference today is that the 90% that are losing aren’t that bad as they used to be.
I’m not saying it’s “easy” to win, but it takes very similar skill set to that of online poker.
Poker used to be fun too but then I played a million hands of it. Same thing here, and the only reason I’m still playing is because I’ve already done most of the work and my marginal cost of entering is low. Also like poker, the sites have taken continuous steps to remove skill from the game. I still enjoy playing MLB but I’m biased because it’s the sport I follow and know the most about.
I know you like NBA but it’s impossible for me to get deep into watching NBA reg season, so I don’t have any particularly great insights. The good thing about NBA is there’s a ton of granular data and it’s free–you can pull all of it with JSON directly from NBA.com’s API. If you’re looking for a project, I don’t think anyone has created a great automated minutes model. Do you code at all?
PGA and NFL are both awful. All of the good PGA data is hoarded by ShotLink and NFL has too much info available with more time for the fish to figure it out. Also, you aren’t realistically going to take down any of the NFL tournaments in a given year because the fields are so huge, even the singles and 3max. It’s basically an anti-sweat. In MLB I win at least a few of those every season and it would be pretty hard not to given the field sizes and number of slates.
1 Like
Does ShotLink keep a lot of the data from the public? I always enjoy looking at golf stats and figured all of the strokes gained data was public, is there some cool shit us plebs don’t see?
None of the ShotLink data is public.
Fanduel has introduced fantasy dating and has a free play contest for Reality Bachelorette in the lobby.
All of the stats on pgatour.com are from ShotLink though? datagolf.com has a bunch of strokes gained data which I believe is calculated using ShotLink numbers
None of this is raw data. One row of the ShotLink data would at minimum contain something like:
Golfer ID | Course | Hole | Shot # | GPS coord
That’s one row for every shot by every golfer for every tournament. In other words, the level of analysis is the shot level. The most granular data I’ve seen on the web is at the hole level.
By the way, ShotLink data is collected by “volunteers” who pay to do the job.
Not sure if I can run simulations tonight and maybe not even this week at all. This game is insane anyway with all of the players on IR.
I’m digging through some old folders and found a bunch of work I did on NBA a few years ago. Here are plots I generated that shows how margin of victory relates to fantasy scoring (Fanduel). First minutes then fantasy points:
And then also points / minute which appears to have a much more diluted relationship:
The superficial takeaway from this is that game stacking seems like a good idea in tournaments. And indeed, I think it does, but a really important question is how much of a game stack and which players. I did a study on that as well, but I can’t seem to locate the folder for it. I’m working from memory, but the general results were that the number of players to stack from the same game depends a lot on the size of the slate. For larger NBA slates, most of the optimals were 2+1 and 2+2. Medium and smaller slates had more 2+2 and some 3+2, respectively. I seem to recall 3+1, 4+1, and 3+3 being pretty rare but not impossible.
A tougher question and one I don’t think has a particularly great answer is which players to stack. In a tight game with important crunch time minutes, it makes sense that high USG would be king, but I think there are too many other factors in play like correlation and salary. Also, correlation within teams seems to vary. In football, QBs will obv be + correlated to pass catchers. In NBA, I didn’t really find any reliable overall effects and think you’d need to deep dive correlations among players within teams, and probably only attach low to moderate weight to those.