Peyton Marino
Footballguy
not really sure i'd say that dude looks like a slime ball. i think i agree with the initial categorization of #####.
Fantasy Football - Footballguys Forums
Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.
DFS data leak: employees cheating? (1 Viewer)
- Thread starter Willie Neslon
- Start date
Apple Jack
Footballguy
Taking all but 100 out for now.
Apple Jack
Footballguy
In related news, 300 new job postings just popped up on Monster.
karmarooster
Footballguy
Having one guy with inside info on DK and another guy with inside info on FD is infinitely more sketchy (and likely to result in +EV plays for both) than extrapolating ownership info for one site to the other (due to significant differences). Swapping info is much, much more akin to true insider trading than whatever calculus would be required to get a boost on FD by knowing DK's ownership percentages, and is the reason Regulation FD is in place (public companies would give out advance tips on their earnings to selected bankers with whom they had a working relationship in order to receive banking favors going forward). Most insider trading is not one person having inside info and profiting himself, but in providing that info to others.
In fact, this could be a basis for prosecuting the both of them (or simple wire/mail fraud would work):
http://www.jha.com/us/blog/?blogID=2242
To be held liable, a tipper (DK guy) must:
(1) tip -- provide DK's % to FD guy
(2) material non-public information -- DK ownership %
(3) in breach of a fiduciary duty of confidentiality owed to shareholders (classical theory) or the source of the information (misappropriation theory) -- in breach of duty owed to DK
(4) for personal benefit to the tipper -- in return for FD ownership %
Last edited by a moderator:
dgreen
Footballguy
I'm new to all of this. Just started playing DK this season. I'm sure this has been discussed elsewhere, but what's with the pricing of players? Why don't their prices adjust based on the market, like point spreads? Changing prices would then be a decent signal for ownership %, right? Price goes up on player X, then people are buying. Price goes down, then people aren't buying.
Chaos Commish
Footballguy
A temporary hold on employees playing. These guys need to make it permanent and shut down the boys club. Foxes in the hen house, not professional. I'm so close to pulling my funds from these sites and that link made it worse.
John Lee
Footballguy
Just to put your mind at ease, any member of the FSTA, of which the major sites are members, has to have their players' funds in a separate account from their operating funds. For that reason alone, there should never be a problem with the availability of your individual monies.
If you would rather be cautious, that's entirely up to you, but the only way I could see 'losing' our money is if the government intervenes and locks that money while they investigate and/or regulate...which I'm hoping against logic does not happen.
dgreen
Footballguy
Yep, they're looking forward to some regulation. They're big enough that they'll easily trade some extra costs on their part in exchange for keeping others out of the market.
RUSF18
Footballguy
Just out of curiosity, they "have to" have their funds separated but who is verifying that this is actually done? And what controls, if any, are in place that actually ringfences those funds and prevents them from being accessed by the people who have opened the accounts and deposited those funds in the first place?
It's easy to say "player funds are held here, operating funds are held there" until a simple ACH transfer can mingle those.
Last edited by a moderator:
(HULK)
(Smash)
I don't. I don't see the big deal at all.
Heck, I can spend 5 minutes looking at pricing and tell you the top 10 most commonly owned players with a strong degree of certainty.
(HULK)
(Smash)
Now this would be a huge deal.It's a start but they've been caught changing lineups before.I think the only way they can prevent this is to release owner percentage for all players Thursday night when the game starts.. if everyone has the same data then this type of "insider trading" would be squashed..
Condia changed his lineup after kickoff? How?!?!
RUSF18
Footballguy
I don't see how anyone can claim it's not a big deal when someone who puts a very large sum of money in play at DFS, has profited handsomely, and has a business relationship with these sites (aka all reasons someone would try to bury this and downplay the story), has come out and said it's a very big deal.
I give Dodds a lot of credit for that post.
Chaos Commish
Footballguy
Chaos Commish
Footballguy
never happenedNow this would be a huge deal.It's a start but they've been caught changing lineups before.I think the only way they can prevent this is to release owner percentage for all players Thursday night when the game starts.. if everyone has the same data then this type of "insider trading" would be squashed..
Condia changed his lineup after kickoff? How?!?!
karmarooster
Footballguy
I don't believe that ownership % on DK would be that much of a benefit on FD... and that's perhaps more concerning: the only way to get that kind of insane edge is to be trading the DK % info with a counterpart at FD (in return for the same).
(also as has been pointed out to me, this was partially on MLB which doesn't have the TNF game to give you a preview of % anyways).
Last edited by a moderator:
Chaos Commish
Footballguy
BassNBrew
Footballguy
While the FSTA has good intentions, do they have up to the minute access to the financials of these companies?
The the huge growth curve both companies are enjoying, I could easily see these companies only doing this for trailing numbers.
karmarooster
Footballguy
I just don't think he considered the possibility of info swapping. Yes, maybe extrapolating from one site to the other could be a slight benefit (but I don't think significantly more than the projected % anyways). Having the exact info provided to you from a counterpart in exchange for the same info on the other site is how this goes from "well I guess you could divide by the total salary available, account for scoring and roster differences, and voila there's your GPP LU!" to actually having precise data about the site in question.
Chaos Commish
Footballguy
sigh
(HULK)
(Smash)
It is ownership info for a different site than he was playing on. With different salaries, different lineups, and different scoring systems.
I'm sure it gives him a strong idea who is popular in general, but I don't need reams of data to tell you that.
I've barely looked this week, but I can tell you Gurley will be hugely owned at Yahoo this week. Its obvious to anyone with 1/2 a brain.
Chaos Commish
Footballguy
You don't understand the value of the data. Hint: It has NOTHING to do with who will be hugely owned.
BassNBrew
Footballguy
Really?
Rawls was only owned in 25% of my Sun-Mon lineups which I found very surprising. Ward was a no-brainer in college football and only owned at 20% clip. I certainly wouldn't have looked at other flyers if I'd known that fact.
Notice that you said strong degree of certainty? Not ever being wrong on this information gives you enough of an advantage to tilt the table your direction over time.
(HULK)
(Smash)
This would be a much bigger deal than what is reported imo.
Having ACTUAL data for the contest you're playing in does give an edge VS notional data that is back engineered from a significantly different contest.
The reverse engineered thing, I don't think it would make a huge difference as pricing and the difference between ppr and .5ppr are big factors impacting player selection differently at the two sites in question.
Tennessee_ATO
Footballguy
If we're assuming there is something other than variance at play here (and I am), knowing ownership %ages can't possibly be the explanation. The edge gained by that simply isn't enough to "explain" dude's run.
Last edited by a moderator:
John Lee
Footballguy
(HULK)
(Smash)
Look, the guy shouldn't have access to that data, as he probably did something shady with it. Him having the access is the big deal.
But reverse engineering Draft King's ownership percentage as a predicative factor for ownership percentage at Fanduel seems like it wouldn't give any significant edge.
Lets say I work for Yahoo (I don't). And I have insider access to know that Gurley is owned in 75% of teams this week. If I try to draw implications from that info for Fanduel, I'm going to hurt myself. Because Gurley is base priced on Yahoo and mid-priced on Fanduel. There will be zero significant correlation between the two ownership percentages this week.
(HULK)
(Smash)
I think he is mistaken in that Draft King's ownership percentage would be overly useful in predicting Fanduel's ownership percentage. Its too much of an apples to oranges comparison to be super useful.
(HULK)
(Smash)
Hint: A guy's price is the #1 factor in determining who is highly owner and barely owned. These sites price differently.
You don't understand the value of the data. Hint: It has NOTHING to do with who will be hugely owned.
ETA: And it definitely matters knowing who will be hugely owned. You need to fade those guys for an edge.
Last edited by a moderator:
RUSF18
Footballguy
I'm not anywhere near smart enough on this subject. I admit that.
I feel that Dodds is and I think he knows more about it than you.
(HULK)
(Smash)
Peyton Marino
Footballguy
you need to adjust for that player's pricing between sites. y=b0*x+b1 where y is FD ownership %, b1 is yahoo ownership %, x is the ratio of FD pricing to yahoo pricing, b0 is the relationship coefficient. just off the top of my head without even thinking about it. not that hard to glean some sort of relationship if you're a smart guy.
(HULK)
(Smash)
Dodds is a very smart guy and knows a lot about DFS. I just think he is mistaken that the ownership percentage from site A would be overly useful at site B. I can't see it generating a huge edge.
Unless the guy traded it. In which case, yes, then it is indeed a huge edge. And frankly, I'd bet that is what actually went down here.
karmarooster
Footballguy
OR, you could just swap the info with a guy at the other site. I don't think this will be limited to 1 guy when it all comes out.
Peyton Marino
Footballguy
the differences in pricing between DK and FD are very small for a lot of players. even if you don't figure out some perfect relationship, you can still use the info for similarly priced players.
(HULK)
(Smash)
I understand math very well. But there is a lot of differences at play between the sites. I bet if you pulled the ownership percentages for the first 4 weeks from the two sites, you'd have a very difficult time finding statistically significant correlations.
(HULK)
(Smash)
I think this is what probably went down. Given his counterpart @ FD also crushing it at Draftkings, and their link already found... it seems like this would be what happened.
And that would be a huge statistical advantage.
BassNBrew
Footballguy
Looks like they already have broken the charter.
Player Protection
The signatory company will ensure employees or other persons connected to the company with access to confidential player information (such as line-ups) will not:
• Play on their own games (apart from for testing purposes or in private leagues)
• Use confidential player information to gain an advantage playing against players on a different site
• Share confidential player information (such as win rate) to anyone outside of the company
(HULK)
(Smash)
Sure. Just like I can look at the pricing and tell you who will be heavily owned due to their being underpriced that week.
I just don't see this reverse engineering thing as an actual edge.
BassNBrew
Footballguy
Peyton Marino
Footballguy
fair enough HULK. I consider you to be a smart guy and didn't mean to imply you didn't understand/hadn't considered the mathematical possibilities of reverse engineering own %s. I guess when you think about it there are other differences that need to be considered between sites (different starting lineup requirements, for example). I still think you can find a decent relationship, it would certainly be hard though.
BassNBrew
Footballguy
I don't think it's related to the heavily owned guys. Obviously Julio is heavily owned, it's about knowing whether the general public is lighter on Green, DT, or OBJ as a replacement options.
karmarooster
Footballguy
http://www.nytimes.com/2015/10/06/sports/fanduel-draftkings-fantasy-employees-bet-rivals.html?hp&action=click&pgtype=Homepage&module=second-column-region&region=top-news&WT.nav=top-news&_r=0
https://dfsreport.com/6840/fanduel-draftkings-mishaps/
These are two totally different stories, but I imagine they are actually related: (1) tweeting the info after kickoff but while DK rosters can be swapped (actually looks like this was a screw up), and (2) the same guy somehow crushing it in multiple GPPs (one reference below).
Player FanDuel % DraftKings %
Andy Dalton 2.3 2.5
Adrian Peterson 18.6 22.4
Devonta Freeman 6.7 8
Randall Cobb 8.2 12.3
A.J. Green 4.4 6
Allen Hurns 2 2.1
Greg Olsen 10.5 8.3
Stephen Gostkowski 10.8 N/A
Seattle D 22.6 30.7
The statement doesn't say that he "used" data, but that he "released" it... to whom? Apparently it was on Twitter, and it was after the games kicked off (but recall DK has a late swap option).
https://dfsreport.com/6840/fanduel-draftkings-mishaps/
These are two totally different stories, but I imagine they are actually related: (1) tweeting the info after kickoff but while DK rosters can be swapped (actually looks like this was a screw up), and (2) the same guy somehow crushing it in multiple GPPs (one reference below).
This was DK dude's FD roster (week 3):
Player FanDuel % DraftKings %
Andy Dalton 2.3 2.5
Adrian Peterson 18.6 22.4
Devonta Freeman 6.7 8
Randall Cobb 8.2 12.3
A.J. Green 4.4 6
Allen Hurns 2 2.1
Greg Olsen 10.5 8.3
Stephen Gostkowski 10.8 N/A
Seattle D 22.6 30.7
Last edited by a moderator:
Skoo
Footballguy
Player FanDuel % DraftKings %
Andy Dalton 2.3 2.5
Adrian Peterson 18.6 22.4
Devonta Freeman 6.7 8
Randall Cobb 8.2 12.3
A.J. Green 4.4 6
Allen Hurns 2 2.1
Greg Olsen 10.5 8.3
Stephen Gostkowski 10.8 N/A
Seattle D 22.6 30.7
How bout now?
32 Counter Pass
Footballguy
No question this is a black eye for the industry and gives me pause about investing more into this format. What is the solution to this issue? In part it may be to make the ownership % completely transparent because any other solution would be suspect, as some posters have already mentioned. Thoughts?
karmarooster
Footballguy
Went back to the FBGs content from Week 3 and compared to Thursday FD Ownership %:
Player FanDuel % DraftKings %
Andy Dalton 2.3 2.5
Adrian Peterson 18.6 22.4
Devonta Freeman 6.7 8
Randall Cobb 8.2 12.3
A.J. Green 4.4 6
Allen Hurns 2 2.1
Greg Olsen 10.5 8.3
Stephen Gostkowski 10.8 N/A
Seattle D 22.6 30.7
Thursday info:
Dalton - Not included
Peterson 10.9%
Freeman 5.4%
Cobb 4.2%
AJG 3.2%
Hurns 0.6%
Olsen 5%
Gostkowksi 11.1%
Seattle 14.4%
Player FanDuel % DraftKings %
Andy Dalton 2.3 2.5
Adrian Peterson 18.6 22.4
Devonta Freeman 6.7 8
Randall Cobb 8.2 12.3
A.J. Green 4.4 6
Allen Hurns 2 2.1
Greg Olsen 10.5 8.3
Stephen Gostkowski 10.8 N/A
Seattle D 22.6 30.7
Thursday info:
Dalton - Not included
Peterson 10.9%
Freeman 5.4%
Cobb 4.2%
AJG 3.2%
Hurns 0.6%
Olsen 5%
Gostkowksi 11.1%
Seattle 14.4%
Tennessee_ATO
Footballguy
Yeah, I but I have a very strong feeling that was a cherry-picked lineup.
If we are looking for an explanation for how a guy can become so +EV to explain this Ethan dude's run, ownership percentages simply ain't enough to explain it. We can debate how much of an edge it gives you, even if he's traded information with a FD employee (something for which there is no evidence whatsoever, so folks should probably calm down on those accusations). But there can't be any real question that the edge isn't THAT great. If his recent run isn't just variance from a top 0.001% player, then for those seeking an "explanation" it has to be something more nefarious.
I will say this: If the guy really has some inside information that he's using that gives him such a massive edge that he can dominate GPPs, then he's a complete #######. By (a) tipping the world to the issue with his Twitter post and (b) acting like Richard Pryor's character in Superman 3 by going really large he's pretty much destroyed a golden goose he could have ridden for years.
