INDUSTRY FORUM

Comments

  • DFSx42

    I came here with popcorn in hand but was pretty shocked to find no discussion of it.

    Worst part about it is that DK has an obvious case of cheating and yet promotes the win.

    This is just horrible. This tweet here sums it perfectly.

    https://twitter.com/williambierman/status/1214015801721868288?s=19

    Vote with your wallets folks. Boycott Draft Kings. We can’t just continue to sit back, watch this and continue paying rake when there’s so many other options out there.

  • Messiah717

    @thedude404 said...

    The point is it’s against the rules. Your math is meaningless.

    The math is also wrong given the size of the slate.

  • Rasmus

    @bkamm1 said...

    Im not the math guy so someone else may have to jump in on the numbers of this but I believe you are wrong on your assumption. Yes, someone else would have filled the spots so your chance of winning based on the number of entrants is correct, its still 1 in 150,000 or whatever. The problem is by correlating lineups you have 150 more unique lineups to compete against meaning the chances of your score coming in first drops as more of the likely outcomes are covered as the lineups entered are all unique. If one were to look over an analysis of ownership and all 300 lineups fell for the most part in line with the overall field then there wouldnt be much of a problem. By having unique lineups they have made it harder as more possible combinations are now covered in a contest that has such a top heavy payout. If this were a 14 game slate it wouldnt be as big if an issue but in a 4 game it is. They essentially spread them out among all QB’s so that they had 45-50 lineups with the one QB who went off, generally speaking of course. Your chance of winning is still the same (based on # entries), the chance of you having the highest scoring lineup is different (based on combinations of players lineups cover using this strategy)

    If you enter a single lineup you do the same as entering 300 lineups. You try to get an advantage over the field. If you are very skilled and do it right you’re invested $20 with the expectation of winning a $24 on average (think 20% roi is very high for a pro). There is no secret synergy effect if your entering 300 lineups, it’s just 300 * your expected roi%. The size of the slate is meaningless.

    If he is indeed a pro and has an average roi of 20% then he has stolen 150 * ($4) = $600 from the community by his cheating.

  • lfn1992

    @bkamm1 said...

    Im not the math guy so someone else may have to jump in on the numbers of this but I believe you are wrong on your assumption. Yes, someone else would have filled the spots so your chance of winning based on the number of entrants is correct, its still 1 in 150,000 or whatever. The problem is by correlating lineups you have 150 more unique lineups to compete against meaning the chances of your score coming in first drops as more of the likely outcomes are covered as the lineups entered are all unique. If one were to look over an analysis of ownership and all 300 lineups fell for the most part in line with the overall field then there wouldnt be much of a problem. By having unique lineups they have made it harder as more possible combinations are now covered in a contest that has such a top heavy payout. If this were a 14 game slate it wouldnt be as big if an issue but in a 4 game it is. They essentially spread them out among all QB’s so that they had 45-50 lineups with the one QB who went off, generally speaking of course. Your chance of winning is still the same (based on # entries), the chance of you having the highest scoring lineup is different (based on combinations of players lineups cover using this strategy)

    To add to this, you can do the same thing with D/ST (only 8 possible) and stud RBs (this might have been more unique to this slate, but there were only Cook, Henry, and AK at high salary). If you coordinated QB and D/ST (i.e. you don’t pick Brees and Minnesota D/ST together), the possibilities go down further.

    Earlier I proposed changing the contest requiring people to pick 2 QBs and 2 D/ST. Would that help? I think it would, because rather than having 8 possibilities at QB you’d have 28 possible QB combinations with the 8 starters. Same with D/ST.

  • Sol88

    They’re right, each extra lineup is lower EV than the one before. However, 300 lineups lets you realize your EV quicker if you’re a +EV player (and thus take away from others), so in that way is still an unfair advantage.

    Either way, problem is you want a collision group capped at 150 total, but that assumes you know who the collusive groups are. A lot of users use the same set of projections and model but do not coordinate. That would look like collision. It’s actually very hard to duplicate a lineup if you’re smart about adding variance and forcing projected ownership mins or maxs.

    At that point you’re kind of looking for intent which you can’t find easily.

    That said, an absolute no brainier rule is ‘you can’t max enter the same contests as someone who shares your address’. Problem solved for a large number of cases.

  • superstars92

    • 189

      RG Overall Ranking

    • Ranked #32

      RG Tiered Ranking

    @Sol88 said...

    LU overlap, or lack thereof, is not a smoking gun, especially when their LU similarities are not actually that high. ~70% is a lot, but it’s not the top of the bell curve.

    Here’s a distribution of lineup overlap from week 17. It’s a natural bell curve, and they’re near the top but not at the top. If you saw them consistently at 90% to 95%+ overlap with no duplicates, then we could be concerned, but this isn’t over the top.

    The bigger issue to me is the QB swapping in Wildcard weekend that happened, but that behavior doesn’t exist in the week 17 lineup distribution.

    Wow this is great. How did you come up with this?

    Do you see a consisntant pattern between any two set of users?

  • dictator_teddy

    • 725

      RG Overall Ranking

    • 2017 DraftKings FHWC Finalist

    @spx333 said...

    from draftcheat twitter lol

    Hitler Reacts to His Wife Winning the Milly Maker

    https://youtu.be/zpjY8-AZLdI

    Absolutely hilarious, this is a must watch! Even if DK doesn’t do the right thing regarding this situation, at least I got to LOL for an entire 3 minute video and used google to take a look at some $1M meat curtains.

  • LegiaWwa

    @bhdevault said...

    This is a good listen. Dan makes so many solid points:

    https://rotogrinders.com/podcasts/daily-fantasy-fix-dan-back-analyzes-bachelor-milly-maker-controversy-3256271

    Is the podcast available on Spotify or Google Play?

  • fightingjohn

    Here are the facts:

    2 married people from the same household play on DK under 2 separate accounts.
    Both maxed out at 150 on the MM

    1 lineup hit for the 1 million

    ——————————————————————————————————————————————————————————————-

    Now the questions is how to prove wrong doing????? We see a spreadsheet with % owned on QBs across this 4 game slate. Okay big deal!!!!!! Yes it looks way out of wack, but now DK has to prove what went on in their household which is out of sight.

    To me there is no case here at all. I know people will think that is crazy, but it is the truth. DK did this to themselves and to the community. Unless they refine their TOS to something they can actually prove, then this will not end.

  • superstars92

    • 189

      RG Overall Ranking

    • Ranked #32

      RG Tiered Ranking

    @fightingjohn said...

    Now the questions is how to prove wrong doing????? We see a spreadsheet with % owned on QBs across this 4 game slate. Okay big deal!!!!!! Yes it looks way out of wack, but now DK has to prove what went on in their household which is out of sight.

    I said this in a couple of pages ago. It’s impossible to prove they did anything wrong. You can make huge assumptions involving math, involving their “reactions” on social media, involving the words they say, but it’s almost impossible to prove they actually colluded unless there is a camera showing everything. Ironically, a camera had been following their most important moments on Bachelor, but too bad there’s none in this situation.

  • fightingjohn

    @superstars92 said...

    I said this in a couple of pages ago. It’s impossible to prove they did anything wrong. You can make huge assumptions involving math, involving their “reactions” on social media, involving the words they say, but it’s almost impossible to prove they actually colluded unless there is a camera showing everything. Ironically, a camera had been following their most important moments on Bachelor, but too bad there’s none in this situation.

    Yes I saw a couple of people mention it. Bottom line is she is getting paid, because if they try to deny her the million based of some excel spreadsheet, they will be sued so fast and then this just gets bigger and bigger and not for the good.

    It comes done to this. If you don’t want to lose your money in a contest to these types of groups, then don’t play them.

  • superstars92

    • 189

      RG Overall Ranking

    • Ranked #32

      RG Tiered Ranking

    @depalma13 said...

    They didn’t collude. He played all 300 lineups.

    Everyone saying this could be right, but you can’t be 100% sure this is the case, and given the past situations, as long as there’s no camera actually showing what happened, the most likely outcome is nothing will get done.

    For everyone saying this, what is the actual logic that 100% proves he played all 300 lineups? It would be good if you guys can list that out because otherwise, it’s just going to be circumstantial evidence, and DK doesn’t care about circumstantial evidence.

  • superstars92

    • 189

      RG Overall Ranking

    • Ranked #32

      RG Tiered Ranking

    Also no one’s mentioned this, but it’s crazy to submit 300 or even 400 lineups into a 2.25 M dollar GPP that pays like 1 M to first.

    Even if I could submit 300 or 400 lineups, I would absolutely not choose to do that. You basically have to win the 1 M or at least like top 5 the GPP to be worth it, and yes in this case they won the 1 M, but most likely you are going to be down.

  • nam6641

    • x2

      2016 DraftKings FFWC Finalist

    The guy did it all but there will never be proof that is 100% by looking at their lineups. Statistically unlikely or rare but not a mathematical 0% chance that they did not enter their own. You’d need video or email or something like that to prove it for certain.

  • Dunzor

    @fightingjohn said...

    Here are the facts:

    2 married people from the same household play on DK under 2 separate accounts.
    Both maxed out at 150 on the MM

    1 lineup hit for the 1 million

    ——————————————————————————————————————————————————————————————-

    Now the questions is how to prove wrong doing????? We see a spreadsheet with % owned on QBs across this 4 game slate. Okay big deal!!!!!! Yes it looks way out of wack, but now DK has to prove what went on in their household which is out of sight.

    To me there is no case here at all. I know people will think that is crazy, but it is the truth. DK did this to themselves and to the community. Unless they refine their TOS to something they can actually prove, then this will not end.

    Normally you would be right and its the reason they haven’t done anything about other syndicates in the past because proving a violation of their guidelines is really subjective, but the one big different factor in this case compared to previous ones is that they literally broke like one of the only clearly defined and expressed rules within those community guidlines:

    “Unacceptable Behaviors:
    Team-building complementary lineups which serve to work together AND executing a strategy that may create any unfair advantage over individual play.

    Example A: You and 2 of your friends coordinate the makeup of the lineups you build AND coordinate which contests you enter using them.

    Example B: You and a group of friends collaborate in NFL contests to each draft different QBs and WRs, to guarantee you aren’t competing as directly with each other.

    Their play during the 17 week regular season probably doesn’t rise to a level that you can punish much like the other syndicates that are tough to catch, but for the wild card playoffs, they (read: he) literally did exactly what is deemed unacceptable under the guidelines so they have a very strong case to void the Wild Card winnings and it might be headed that way as the bad press builds up

  • Kingspark35

    @bhdevault said...

    This is a good listen. Dan makes so many solid points:

    https://rotogrinders.com/podcasts/daily-fantasy-fix-dan-back-analyzes-bachelor-milly-maker-controversy-3256271

    A good listen indeed. Some of the points he made are interesting.

    For me, the standard of proof should be ‘reasonable doubt’.

    But who decides without a jury ?

  • bigez952

    @bhdevault said...

    This is a good listen. Dan makes so many solid points:

    https://rotogrinders.com/podcasts/daily-fantasy-fix-dan-back-analyzes-bachelor-milly-maker-controversy-3256271

    I agree that this a good listen and Dan does a good job. I 100% agree with the end of this pleading to Draftkings to please not come out with a statement that we don’t comment on investigations with players. If they ever needed a time for transparency this is the issue if they want to come out of this with any integrity at all.

  • Sol88

    @superstars92 said...

    Wow this is great. How did you come up with this?

    Do you see a consisntant pattern between any two set of users?

    Hard to see patterns cuz I can only download 2 weeks of result CSVs. If anyone has older MM result CSVs please DM me and I can go back further.

    I can start examining NBA this week since that will have ~14 contests

  • fightingjohn

    @Dunzor said...

    Normally you would be right and its the reason they haven’t done anything about other syndicates in the past because proving a violation of their guidelines is really subjective, but the one big different factor in this case compared to previous ones is that they literally broke like one of the only clearly defined and expressed rules within those community guidlines:

    “Unacceptable Behaviors:
    Team-building complementary lineups which serve to work together AND executing a strategy that may create any unfair advantage over individual play.

    Example A: You and 2 of your friends coordinate the makeup of the lineups you build AND coordinate which contests you enter using them.

    Example B: You and a group of friends collaborate in NFL contests to each draft different QBs and WRs, to guarantee you aren’t competing as directly with each other.

    Their play during the 17 week regular season probably doesn’t rise to a level that you can punish much like the other syndicates that are tough to catch, but for the wild card playoffs, they (read: he) literally did exactly what is deemed unacceptable under the guidelines so they have a very strong case to void the Wild Card winnings and it might be headed that way as the bad press builds up

    Yes the example is clear, but can they actually prove “COLLABORATION” It could have been dumb luck it worked out like that, but probably not, but the fact remains is where is hard core proof they actually collaborated on this. You can take ANY 2 people in a contest and can try to say they did some wrong doing. If they void this, this will no end well at all and we no DK doesn’t want that.

  • yisman

    @WhiteyJones8 said...

    Based on their play history they likely did this the entire season. They pretty much maxed the Milly in both accounts almost every week. In some ways, a better excuse for them might just be that they didn’t know this wasn’t allowed.

    FYI, I was looking at some of the coverage of this and they had a quote from Tanner.

    He responded by saying (paraphrasing) “We did this all season and no one said anything. But now we win, and everyone is mad.”

  • Sol88

    @superstars92 said...

    Wow this is great. How did you come up with this?

    Do you see a consisntant pattern between any two set of users?

    Overlap % is basically the size (%) of intersection of sets of player picks between every two players. A really easy way to quickly identify which users are likely using the same set of projections or underlying model, but doesn’t tell you much about intent to collude or if they were coordinating which and how they made said lineups. Using the same projections and model isn’t wrong, but coordinating such use is.

  • thedude404

    • 2015 FanDuel NBA Playboy Mansion Finalist

    @slicktornewman said...

    Yes definitely a good listen. Dan admits many things most of us already knew, DK’s community guidelines are a farce and pros are definitely working together in every kind of contest, from single entry up to 150 entry contests. At least a respected industry insider (not sure that’s the right description but all I could think of) is speaking out about it. Good job Mr Back.

    Dan is definitely respected. I think the overwhelming themes through his podcast were 1. Yes many people are cheating and 2. It’s difficult to prove cheating. My response to all of that why did these companies develop DFS into a game where both 1 and 2 both coexist?

  • qatman

    For those who cannot see how the extra entries are unfair:

    Imagine the contest is winner take all (not too far off since 1st is ~40% of the entire prize pool). Having 300 unique lineups gives you 300 chances to win first prize. Having some duplicated lineups gives you fewer than 300 chances to win first prize. So if you can ensure (by colluding/multiaccounting) there is no overlap in any of your entries, you will have a better chance to win first prize.

    Personally I don’t think this is any different from other collusion/multiaccounting issues (Burrito Brothers) which should be banned but are not. Although rake is still the primary problem in DFS today.

  • Sol88

    @thedude404 said...

    Dan is definitely respected. I think the overwhelming themes through his podcast were 1. Yes many people are cheating and 2. It’s difficult to prove cheating. My response to all of that why did these companies develop DFS into a game where both 1 and 2 both coexist?

    Not everyone is cheating, that’s stupid, near sighted and bad for the industry. Bad take by Dan, he should know better.

    The data is public on DK. There’s no rampant cheating. It would show up in charts like the ones I posted.

    If people were rampantly colluding we’d see it and be able to point to it.

    We can’t. It’s not rampant.

  • fightingjohn

    DFS companies should get lawyers involved when coming up with their TOS and how can they stand up in court, because the proof of a violation will always fall with them.

    For example after looking at example b of the collusion violation different people draft different QBs AND WRs is a no no. Well the key word is “AND” what is their overlap in WRs then? I guarantee you that key word is everything in this case.

  • superstars92

    • 189

      RG Overall Ranking

    • Ranked #32

      RG Tiered Ranking

    @Sol88 said...

    Overlap % is basically the size (%) of intersection of sets of player picks between every two players. A really easy way to quickly identify which users are likely using the same set of projections or underlying model, but doesn’t tell you much about intent to collude or if they were coordinating which and how they made said lineups. Using the same projections and model isn’t wrong, but coordinating such use is.

    Oh yea in your post I see that moklovin is grouped in the really far right tail, which probably makes sense.

    I would say if they had like 17 weeks where somehow they were in the right tail (in terms of SD from mean) but never had an overlap, it would be highly suspicious. I have no idea if that’s the case with any users though.

    It would also be weird if they had like say 9 weeks of being randomly distributed and like 8 weeks in the far right tail together with no overlaps. That would imply like 8 weeks of 1 user playing and 9 weeks of two users playing (“users” can be substituted for “model” too).

  • X Unread Thread
  • X Thread with New Replies*
  • *Jumps to your first unread reply

Subforum Index

RotoGrinders.com is the home of the daily fantasy sports community. Our content, rankings, member blogs, promotions and forum discussion all cater to the players that like to create a new fantasy team every day of the week. Our goal is to help all of our members make more money playing daily fantasy sports!

Bet with your head, not over it!
Gambling Problem? Call 1-800-Gambler