INDUSTRY FORUM

Comments

  • rausch180

    A bit different to the usual ones, but yesterday at NBA (18th Nov) they both managed to botch their GPP entries. Cash they had 2 common builds around the 5 core plays of Doncic, Sabonis, House, TT and Dinwiddie. Nothing to see there.

    Yet in the $55 NBA $200K FINGER WAG there were 4132 entries in the GPP, of which 111 had Kyrie Irving who had long been ruled out. ChipotleAddict had 104 with Irving, papagates 7 with him, no one else had any unsurprisingly. Same deal with Jeremy Lamb long ruled out, 29 lineups had him in the GPP, 27 were ChipotleAddict and 2 were papagates. Also in the $8 NBA $400K EXCELLENT 8’S the pattern was similar.

    Once ResultsDB is updated, you’ll be able to see for yourself.

    It is very uncommon for this to happen to a max entry reg. They both had a cash lineup that appears well reacted to all the news. The question is what is the chance of both players independently making a very rare mistake and the same mistake and only in GPP? Any one have a reasonable explanation?

  • Cocomilk

    @jjwd said...

    I actually never said that, and I don’t care to “justify” anything except logical thinking here. I think I’ve made my points. No need to monopolize the thread any further.

    haha bow down and back out just like every other nuthugger that nolimit and superstar has dragged through this thread

  • tahtinp

    It’s always interesting to compare the ownership both PG and CA had for a given tournament, assuming they do indeed rely on the same projection system on a daily basis. For example, in the Milly on Sunday, 11/17 CA had 66% and PG had 1.3% of Christian McCaffrey. I don’t get it.

  • Cocomilk

    @tahtinp said...

    It’s always interesting to compare the ownership both PG and CA had for a given tournament, assuming they do indeed rely on the same projection system on a daily basis. For example, in the Milly on Sunday, 11/17 CA had 66% and PG had 1.3% of Christian McCaffrey. I don’t get it.

    they had 33.65

  • Vdaddy8888

    I find it completely plausible that they did not overlap in this particular instance.

    Generally 150 max entry participants identify 1-3 NBA players that have 50 plus percent ownership in their lineup construction. This is a fair assumption based on data readily available. Let’s assume on given night two players are in 70% of their lineups. Is it possible for the remaining 6 players on 210/300 lineups to be different using the same tool and lineup build? Yes absolutely. If each player is at 10% ownership then instantly you are down to 21/300 and so on. Taking this out to its conclusion at 10% ownership on every player you arrive at 2/, .2,.02, .002, and .0002/300. So on a (lets use 10 game slate) since round numbers are easier there is a .0002 chance they arrive at the same lineup if 2 players are a 70% and the remaining 6 are at 10% ownership.

    Now multiply that likelihood over lets say 5000 unique tournaments that require 8 unique players and that are 150 max entries you arrive at 1/300 chance of sharing a lineup at some point holding everything else constant. Yes its that rare. Now, inserting the use of a common lineup construction tool will multiply the likelihood of a shared output, but to what extent, I can’t say. Is it 2X, 5X, 10X, but there is a coefficient here. Lets use 10X. again because its easier. So that would result in 1/500 unique tournaments at 150 max entries each where you would expect a minimum of 1 shared lineup.

    A better way to do it is to get their average total ownership percentage over an extended period of time. Lets say its 160% then you arrive at a 20% average on each unique spot in the nba roster construction.

    Awesemo has a tool that shows past roster construction (Lineup HQ). If these two players have never overlapped over (insert appropriate time frame that gets you above 150/300), then clearly rosters are being manipulated to avoid duplication. I don’t have the bandwidth for such an exercise, but it can be accomplished and if there are no duplicates, you can reach a conclusion.

    Separately, I’d like to see less steep payout scale tournament options available that would discourage this type of behavior. This type of mass unique entry “collusion”/behavior/conduct is more prevalent in tournaments which have steep payout structures that reward a disproportional share to 1st and 2nd verse 10-12. Finishing in the 99.94 percentile should not pay 30X less than the 99.99% percentile.

    DK/FD have an implied interest in creating such tournaments as rewarding 1st place an $1M is sexier and sells more entries than 1-10th split 1.5M more evenly which I understand, but it promotes this type of behavior and conduct.

  • NoLimits0

    I’ll give you guys a primer on how to optimally collude as my last post here. If you follow this you pretty much will have a super low chance of ever getting caught, even if DK has an incentive to (which they obviously don’t). I don’t suggest anyone do this but I’ll give you guys on the idea of how it works.

    First you need to find someone or a group of people that make it plausible. So as a top pro, they could just sign up their mom or dad, but that’s not believable. Their mom or dad won’t have that great of sports knowledge and obviously aren’t the personalities to gamble 10k+ a week let alone 100k+. The brothers obviously both know their sports and come from a gambling background. Ideally actually a friend would work even better since that makes it less suspicious. The better example would be ragingphillip ans mawza who probably aren’t brothers but also collude.

    Second you have to set up some technology tools. Either you and your collusion partner both log in with your own accounts and submit on your own but just have it so one person generates the lineups and sends it to the other, or you both set up a VPN that is accessible to both of you. So that way I can log on to my partners account in that VPN and my partner can also log into that account in a VPN and from the hosts perspective they can’t tell it was logged on by two different people. VPNs will also allow you to play out of the country or like in a banned location like Vegas (who Im sure pros have gone to but still are able to play using VPNs). You absolutely cannot have both accounts being logged in from the same IP address. A one off occurance can be explained like my friend/brother was visiting me but you need to be careful not to always do it. You need to find a good enough VPN that goes around detection networks but plenty of those exist (assuming DK even checks logic tells me they don’t given they don’t care about collusion). If you don’t want to use a VPN it’s fine, as you just need to make sure you don’t screw up logging in using your IP addresses. VPNs are perfectly legal as that’s something you should be using at like a public hotel anyways so they won’t be violating anything.

    Now to the sports. The collusion tactic per sport isn’t the same. In MLB you want to increase variance so what you want is a whole bunch of different stacks and you don’t want similar exposures across the board on the stacks. Variance increases with more generic stacks. You absolutely want to avoid like having the same two stacks of 4×4 hitters but then having only a 1 off on the hitter. That absolutely does not increase variance. Variance means more in MLB because the distribution of outcomes is huge from 0 on a 0-4 day to 40+ on a 3 HR day, even if your predicted outcome was like 8 FPs. You still want to center your variance stacks around positive expected value plays but variance should play the largest role in MLB. So your weighting % for each team stack should still follow expectancy but your goal is to increase the # of these possible stacks across the board and not have 1-2 offs in the stacks.

    In NBA it’s the opposite. If you collude you want to maximize expectancy not variance so in that case you want to build very similar cores and have 1-3 offs. This is because the range of outcomes for NBA is less and you can realize your expected value much easier. In the NBA you should be going super heavy on a core of players over the field but then fill out the rest of your lineups with some 1-3 offs based on expectancy.

    In NFL it’s a cross between the two. You want to somewhat have a core of solid players but then you want to maximize variance in the rest of your lineup. This is either by going through QB – WR or TE stacks with that core or going severely underweight or overweight chalky players. So for example a way to increase variance would be to either have all McCaffrey or to fade him and instead choose the Carolina WRs. It covers more possibilities that way as long as you then can choose a similar core of the other player since it’s a combination of variance and expectancy.

    What I described above doesn’t even have to apply to collusion. If you are just doing 150 lineups it’s the same logic for GPPs.

    Finally you obviously would only want to collude when it’s beneficial. A beneficial time would be like you both want to do NBA and NFL on the same night but if you do it separately you are wasting time. Instead one should do NBA the other NFL and then help each other. NFL or golf would be the ones you probably don’t need time for since there’s no time crunch so there should be less collusion in those sports. NBA and MLB when there’s an overlap with another sport is where you should see the most likely collusion. Collusion should also occur the most when one partner/brother is busy (vacation, job, date) and you still want to maximize EV that night and not give up potential earnings so the other should do both lineups. In general you want both parties in the collusion to at least be knowledge about the sports and models so they can both fill in when necessary.

    That’s a primer. Don’t do it obviously but it’s pretty easy explanation of all the details. As long as you stick to the technology plan, even if people had an incentive to catch you, you shouldn’t get caught.

  • stv1313

    • 932

      RG Overall Ranking

    Two years ago, Cal made a constructive post regarding the ChipotleAddict/Papagates scandal. I think this particular section is relevant today:

    “After multiple calls with Martin Crowley (papagates) and Matt Kalish (Co-Founder of Draftkings), I am further convinced that the two brothers are not violating DraftKings terms. The two brothers built a model together. They work from the same source data. Some days they interpret and act on the data very similarly, other days very differently. While this produces the optics problem, it is not collusion, even if they are splitting funds. It only becomes collusion if they are working together to align their lineups to circumvent entry limits.”

    After 13 pages of back and forth from all of us, I don’t believe that their is a smoking gun that proves the brothers violated DK’s TOS on Monday.

    Here’s my final take on how I think this ultimately went down:

    As stated two years ago, the brothers share a projection database. Monday evening, Brother #1 updates the database (after lock) and mistakenly includes “old” projections for Kyrie and Lamb – ignoring the fact that they’ve been declared out. Apparently, FantasyCruncher may have played a role. Who knows.

    Brother #1 then builds 150 lineups (making a few clicks here and there to tweak ownership percentages for players he really likes) and he quickly uploads them to DK – just like he’s done hundreds of times in the past. He never looks at the lineups after the upload and, therefore, doesn’t notice the smattering of red O’s that appear next to Kyrie and Jeremy Lamb. Ooops. His bad.

    On a personal note: A similar event once happened to me and I kicked myself for not taking the extra minute to verify that my upload looked “good”. Although I’ve stopped MME’ing, I have always promised myself that I’d look at the lineups within DK’s system after I do a massive upload if/when I go back to MME’ing.

    Brother #2 has access to the same database. He doesn’t need to check the projections because Brother #1 is responsible for updating it. So, he doesn’t notice the Kyrie/Lamb issue. Making a few clicks here and there to tweak ownership percentages for players he really likes, Brother #2 now builds 150 lineups and uploads them to DK – just like he’s done hundreds of times in the past. Similar to Brother #1, he never looks at the lineups after the upload. Ooops! His bad. That’s what he gets for not auditing his brother’s update of the database.

    I also think that it’s important to note that there is some overlap in the brother’s lineups. As best I can remember, there are about 10 identical lineups that they used Monday night. They did not post 300 unique lineups between them.

    An hour or two into the night, Brother #2 contacts Brother #1 and asks why the projections were screwed up. Brother #1 apologizes for the screw-up and they move on, until our OP makes this post and things blow up.

    Without more of a smoking gun, I don’t think that I can really conclude anything more. Personally, I wouldn’t mind throwing these guys under the bus and pointing to obvious collusion, but it just doesn’t seem to be there. As I see it, with the system they are employing, they aren’t violating DK’s TOS and we’re all just gonna hafta live with the system these two brothers are using.

    Personally, I wouldn’t be surprised to learn that Brother #1 has occasionally posted the lineups for Brother #2 (or vice versa), thereby violating the TOS, but I certainly don’t believe that it’s a regular occurrence and there is no proof that it’s actually happened. If it was a regular event, I trust that DK would have uncovered it by now.

    Call me stupid and naive, but this is what I’m going with.

  • blenderhd

    • 638

      RG Overall Ranking

    @stv1313 said...

    Here’s my final take on how I think this ultimately went down:

    As stated two years ago, the brothers share a projection database. Monday evening, Brother #1 updates the database (after lock) and mistakenly includes “old” projections for Kyrie and Lamb – ignoring the fact that they’ve been declared out. Apparently, FantasyCruncher may have played a role. Who knows.

    Brother #1 then builds 150 lineups (making a few clicks here and there to tweak ownership percentages for players he really likes) and he quickly uploads them to DK – just like he’s done hundreds of times in the past. He never looks at the lineups after the upload and, therefore, doesn’t notice the smattering of red O’s that appear next to Kyrie and Jeremy Lamb. Ooops. His bad.

    On a personal note: A similar event once happened to me and I kicked myself for not taking the extra minute to verify that my upload looked “good”. Although I’ve stopped MME’ing, I have always promised myself that I’d look at the lineups within DK’s system after I do a massive upload if/when I go back to MME’ing.

    Brother #2 has access to the same database. He doesn’t need to check the projections because Brother #1 is responsible for updating it. So, he doesn’t notice the Kyrie/Lamb issue. Making a few clicks here and there to tweak ownership percentages for players he really likes, Brother #2 now builds 150 lineups and uploads them to DK – just like he’s done hundreds of times in the past. Similar to Brother #1, he never looks at the lineups after the upload. Ooops! His bad. That’s what he gets for not auditing his brother’s update of the database.

    I’ve been repeating this very point several times that they use the same projection model and source data, just like the same way you’d share a file on Dropbox between two people. If one is uploading a projection set that mistakenly has a value for Irving and Lamb, the likelihood of the other one doing the same is not a statistical improbability, it’s in fact the most probable outcome by far. That’s why the “what are the chances of both” types of arguments are without merit.

  • vino24

    • 2019 FanDuel WFBC Finalist

    @stv1313 said...

    I also think that it’s important to note that there is some overlap in the brother’s lineups. As best I can remember, there are about 10 identical lineups that they used Monday night. They did not post 300 unique lineups between them.

    I mean, this is how they were caught last time, because they did end up with 300 unique lineups. It would be in their best interest to have a few overlapping lineups in so that they can say “see, we didn’t break any rules.” Just seems like a silly point to bring up IMO.

  • stv1313

    • 932

      RG Overall Ranking

    @vino24 said...

    I mean, this is how they were caught last time, because they did end up with 300 unique lineups. It would be in their best interest to have a few overlapping lineups in so that they can say “see, we didn’t break any rules.” Just seems like a silly point to bring up IMO.

    I don’t disagree, vino24. I only really mentioned it because some other posts in this thread implied there was no overlap.

  • NoLimits0

    @blenderhd said...

    I’ve been repeating this very point several times that they use the same projection model and source data, just like the same way you’d share a file on Dropbox between two people. If one is uploading a projection set that mistakenly has a value for Irving and Lamb, the likelihood of the other one doing the same is not a statistical improbability, it’s in fact the most probable outcome by far. That’s why the “what are the chances of both” types of arguments are without merit.

    As I also mentioned plenty of times previously even if we do give them the complete benefit of the doubt and take their word for it, in the best case, they are still liars.

    I linked papagates response years ago before, you can find it in his profile. He specifically states they only use “similar methods”, no different than him vs other pros and every night they do it 100% of the time on their own. If you are using the same database, same projections, even as stv suggested both then logging onto the same FC with the same data and methodology to generate lineups, that’s NOT doing it 100% on their own every night.

    So in the best case they are liars and scummy. Since they clearly lied before, I can’t take their word this time either.

  • slicktornewman

    “For example, in the Milly on Sunday, 11/17 CA had 66% and PG had 1.3% of Christian McCaffrey. I don’t get it”

    This stinks bigtime. Obviously CA contest lineups were chosen to use McCaffrey and PG not. How anyone can say they weren’t working together in this instance at the very least, is beyond me. It really doesn’t matter though because these boys generate way to much rake for DK/FD to want to expose them. This is why I won’t enter any contest higher then 3 entry max.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    Wait is the CMC argument for real here? It took me 1 second to realize what happened.

    Obviously, if you looked at Results DB more (actually I really don’t even need to), I bet you CA had like very little DJ Moore and Papagates had like a ton of DJ Moore.

    If you had 150 lineups, you obviously would want to split your CMC and Moore on two different sets of lineups, you build the first set for CMC and you build the second set for Moore. Here, it’s just 300 lineups.

    This doesn’t prove collusion either necessarily, but it absolutely does not disprove it. It defnitely learns more towards collusion if I had to pick one. Heck, that’s the first thing people do when they make like X lineups, they split around 1/2 into the CMC core and 1/2 into the DJ Moore core…here X just happens to be 300 whereas most people X is 150 max…

    I don’t even need to look at ResultsDB to know that Papagates with his 0% McCaffrey probably had a ton of DJ Moore + Greg Olsen + Curtis Samuel (probably mostly Moore since he was the best play) and that Chipotle with his 60% McCaffrey probably had very limited DJ Moore + Greg Olsen + Curtis Samuel, despite them being chalky. Now, if Papagate with his around 0% McCaffrey had like very little DJ Moore to, now that’s something…but I bet you it’s not the case. It’s never the case when you split builds into sets.

    It’s the ideal strategy in GPPs, for whatever lineups you do. It’s actually a strategy I hope people learn more of, since it’s a GPP winning strategy. You build your lineups to maximize “sets” of players who together won’t do well.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @tahtinp said...

    It’s always interesting to compare the ownership both PG and CA had for a given tournament, assuming they do indeed rely on the same projection system on a daily basis. For example, in the Milly on Sunday, 11/17 CA had 66% and PG had 1.3% of Christian McCaffrey. I don’t get it.

    Read my response above and check your resultsDB to see if I was right. I don’t even need to look to probably be right.

    It’s how you actually build GPP lineups, you split into sets. If you only had a maximum of 150, that’s how you would do it (split into two sets of around 75). Here, obviously, they have 300, so their sets would be size 150 and 150.

    Am I right in saying CA’s McCaffrey exposure + Papa’s DJ Moore exposure will be around 100%?

  • bhdevault

    • Lead Moderator

    • Blogger of the Month

    Admit it, how many of you are like:

  • BerkeleyBoss

    @blenderhd said...

    I’ve been repeating this very point several times that they use the same projection model and source data, just like the same way you’d share a file on Dropbox between two people. If one is uploading a projection set that mistakenly has a value for Irving and Lamb, the likelihood of the other one doing the same is not a statistical improbability, it’s in fact the most probable outcome by far. That’s why the “what are the chances of both” types of arguments are without merit.

    I’m not sure if there’s much value in me continuing to post the same thing here, but…

    The mistake doesn’t appear to have anything to do with their shared source data. PG explicitly said in this thread that he relies on FC excluding a projection for players that are ruled out (so I assume PG/CA projections simply exclude those players altogether, not uploading a 0 projection for them — if they uploaded a projection for them, he wouldn’t expect FC to give them a 0, because his data overrides that). If that’s not done, he needs to notice and manually adjust it himself. It wasn’t done this time and he didn’t notice. Neither did CA. That’s a bigger coincidence than you mention, though perhaps not as big as others are implying. I’ve made the mistake myself in the past on FD with players that are Doubtful.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @NoLimits: I have no idea about baseball because I don’t play, but the strategy in NBA is you build cores (like VDaddy pointed out, me and him talked about 100% Dinwiddie yesterday and hopefully he also had 100% Nance) and the strategy in NFL is you build sets around cores. Baseball is probably more complicated because I am not sure how the SP is handled (you definitely don’t want stacks vs SP I know that at least).

    I’m a pretty crappy player in NBA and NFL though so maybe I need to rethink my strategy HA.

    100% Carmelo Anthony it is tonight! Subject to change watching Syracuse highlights and then Rockets highlights…

    @VDaddy I have to re-read your post, so much math at once.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @BerkeleyBoss said...

    The mistake doesn’t appear to have anything to do with their shared source data. PG explicitly said in this thread that he relies on FC excluding a projection for players that are ruled out (so I assume PG/CA projections simply exclude those players altogether, not uploading a 0 projection for them — if they uploaded a projection for them, he wouldn’t expect FC to give them a 0, because his data overrides that). If that’s not done, he needs to notice and manually adjust it himself. It wasn’t done this time and he didn’t notice. Neither did CA. That’s a bigger coincidence than you mention, though perhaps not as big as others are implying. I’ve made the mistake myself in the past on FD with players that are Doubtful.

    Also like I mentioned this FC “glitch” that’s being talked about isn’t even easy to experience.

    Try generating some lineups with Dame Lillard for example. It’s not easy. You have to actually try to make a mistake to generate that lineup.

  • tahtinp

    @superstars92 said...

    Wait is the CMC argument for real here? It took me 1 second to realize what happened.

    Obviously, if you looked at Results DB more (actually I really don’t even need to), I bet you CA had like very little DJ Moore and Papagates had like a ton of DJ Moore.

    If you had 150 lineups, you obviously would want to split your CMC and Moore on two different sets of lineups, you build the first set for CMC and you build the second set for Moore. Here, it’s just 300 lineups.

    This doesn’t prove collusion either necessarily, but it absolutely does not disprove it. It defnitely learns more towards collusion if I had to pick one. Heck, that’s the first thing people do when they make like X lineups, they split around 1/2 into the CMC core and 1/2 into the DJ Moore core…here X just happens to be 300 whereas most people X is 150 max…

    I don’t even need to look at ResultsDB to know that Papagates with his 0% McCaffrey probably had a ton of DJ Moore + Greg Olsen + Curtis Samuel (probably mostly Moore since he was the best play) and that Chipotle with his 60% McCaffrey probably had very limited DJ Moore + Greg Olsen + Curtis Samuel, despite them being chalky. Now, if Papagate with his around 0% McCaffrey had like very little DJ Moore to, now that’s something…but I bet you it’s not the case. It’s never the case when you split builds into sets.

    It’s the ideal strategy in GPPs, for whatever lineups you do. It’s actually a strategy I hope people learn more of, since it’s a GPP winning strategy. You build your lineups to maximize “sets” of players who together won’t do well.

    I should have been clearer when I said “I don’t get it”. What I meant was, if they are indeed independently created 150 lineups apiece, this would seem close to impossible going off of the same or similar model. The question is, did they work together to get an effective ownership at around 1/3 of CMC? To me, if they were independently building lineups with the same or similar model, each of their exposures to CMC would be closer to 1/3. I agree that this proves nothing for certain, but I wanted to point it out nonetheless.

  • willyvonka

    @slicktornewman said...

    “For example, in the Milly on Sunday, 11/17 CA had 66% and PG had 1.3% of Christian McCaffrey. I don’t get it”

    This stinks bigtime. Obviously CA contest lineups were chosen to use McCaffrey and PG not. How anyone can say they weren’t working together in this instance at the very least, is beyond me. It really doesn’t matter though because these boys generate way to much rake for DK/FD to want to expose them. This is why I won’t enter any contest higher then 3 entry max.

    It’s pretty questionable to draw conclusions like that from one day. Just look at their CMC ownership from the previous three Millys in which he played:

    11/10: CA – 0.67%
    PG – 2.67%

    11/3: CA – 49.33%
    PG – 20.67%

    10/27: CA – 17.33%
    PG – 10.67%

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @tahtinp said...

    I should have been clearer when I said “I don’t get it”. What I meant was, if they are indeed independently created 150 lineups apiece, this would seem close to impossible going off of the same or similar model. The question is, did they work together to get an effective ownership at around 1/3 of CMC? To me, if they were independently building lineups with the same or similar model, each of their exposures to CMC would be closer to 1/3. I agree that this proves nothing for certain, but I wanted to point it out nonetheless.

    Oh no, I think the key is you have to understand how football lineups are built. They are absolutely not built using an optimizer the same way NBA lineups are built. In the NBA, you will see very similar distributed teams from across all users (not just them, just out the other pros, there’s a large discussion about this in another thread) because you would indeed optimize from a similar model.

    In the NFL, you build lineups using “sets” in the way I described. So that’s why if I see someone with 0% CMC, I bet they have a ton of DJ Moore or Greg Olsen. If someone has a ton of CMC, they won’t have much DJ Moore nor Greg Olsen. You found the lineups on ResultsDB right? You can check yourself. The 0% brother probably has a lot more DJ Moore + Greg Olsen than the 60% brother.

    Basically, if you use RG’s optimizer for basketball with no input, it’s going to be pretty good. If you use RG’s optimizer for NFL GPPs, it won’t be good unless you override it with human input using sets.

    Like if it was just 1 person creating 20 lineups, he absolutely does not want to choose the top 20 optimizer lineups in NFL. You would in NBA. In NFL, you want to split up your lineups into sets of players.

    Hence if you had 300 lineups between two users and they were colluding, you would do the same. However, this example doesn’t prove much either way without more detailed data.

    Looking at a single exposure in NFL is definitely not the right way to approach it.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @willyvonka said...

    It’s pretty questionable to draw conclusions like that from one day. Just look at their CMC ownership from the previous three Millys in which he played:

    11/10: CA – 0.67%
    PG – 2.67%

    11/3: CA – 49.33%
    PG – 20.67%

    10/27: CA – 17.33%
    PG – 10.67%

    You have to understand my post on how NFL GPPs are built before commenting on individual ownership of NFL players.

    No top pro builds NFL lineups optimizing say the top X lineups. A lot of top pros would do that in NBA but almost none in NFL.

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    For the record, there’s no way anyone can possibly prove or dis-prove collusion by looking at ownership % and lineup builds between two users. The reason this post is more fascinating is because it’s a more solid piece of evidence which although you cannot 100% prove anything, as others have argued, you can prove it in a statistical manner which probably could win you a civil case.

    It’s like Erin Brokovich if you ever watched the movie. What are the chances all the people who got sick all happened to be in a similar location using the same water. Yes, it’s possible they all independently got sick, but more likely the water had an issue. What are the chances that the only two screw ups were brothers who have a checkered past and the screw ups were also so egregious that they played two O players. Yes, it’s possible they made the mistake on their own, but unlikely. Erin Brokovich won her case.

  • BerkeleyBoss

    @superstars92 said...

    Also like I mentioned this FC “glitch” that’s being talked about isn’t even easy to experience.

    Try generating some lineups with Dame Lillard for example. It’s not easy. You have to actually try to make a mistake to generate that lineup.

    I’m not sure that you understand what the issue with FC is. What you’re referring to is when FC does zero out an inactive player. FC has correctly given Lillard a 0. That’s exactly what PG relies on and what happens an overwhelmingly high % of the time. The problem is that every now and then, it doesn’t happen. Every now and then, you’ll see Lillard projected for 45.6, which is obviously an oversight (I wouldn’t call it a glitch). That’s what PG said happened in this case.

  • jayzee666

    @bhdevault said...

    Admit it, how many of you are like:

    RG really needs to just save this thread and repost every 2-3 weeks. It would save a lot of typing AND refreshing!!
    Like watching Family Guy ruruns. You know all the lines and jokes but can’t stop watching anyway!

  • superstars92

    • 365

      RG Overall Ranking

    • Ranked #63

      RG Tiered Ranking

    @BerkeleyBoss said...

    I’m not sure that you understand what the issue with FC is. What you’re referring to is when FC does zero out an inactive player. FC has correctly given Lillard a 0. That’s exactly what PG relies on and what happens an overwhelmingly high % of the time. The problem is that every now and then, it doesn’t happen. Every now and then, you’ll see Lillard projected for 45.6, which is obviously an oversight (I wouldn’t call it a glitch). That’s what PG said happened in this case.

    No I get that. FC is correctly giving Lillard the 0, just like it gave Draymond the 0 yesterday. PG is saying it DID NOT correctly give Kyrie or Lamb the 0. But that’s odd because first it normally does give the players 0 and second what are the chances it glitched and messed up for both him and his brother in two seperate times and only for those 2 so that no one else was affected by it. So the glitch itself is super rare, but yet he’s claiming it happened to both him and his brother and only to them. Furthermore, even after the glitch happened, they both could have seperately caught it in the upload to DK but both didn’t and the reason given was because both were working on Monday Night Football? I work on NBA and NFL on the same night too and I can tell you, the first thing I do in the Kawhi situation is to check if Kawhi was indeed taken out of all my lineups in the late swap before working on NFL (which is a showdown slate in an hour, showdowns take like 10 mins to do), and if you just check for that, you would inevitably find Kyrie in those lineups.

    It would make sense since this glitch is super rare that it only affected 1 user and he didn’t see Kyrie or Lamb. Like I said, coicidences do happen. It’s just too much of a coincidence.

  • 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.

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