Mark Atwater

(Long Post) I know a lot of ppl out there can't or don't want to pay for DFS projections - well a lot of ppl are building different kinds of dfs sites w AI right now, many are free - I've been testing a couple myself Wanted to pass this one along, because @TwoGunFF who created it was a big baller successful dfs guy back in the day - on the site twogungpt.com/ he pledges that it will remain free indefinitely, because it basically is a hobby to him One of the commercial sites just uploaded their early run of projections and powns - and obv there will be a lot of changes etc in the coming days (theirs weren't even complete just yet) - TwoGun's projs and powns were already up and re-run nightly I think I compared about the top 50 highest proj players and the difference in proj between the sites was +3.1 and -2.5 at the most (the avg diff was 0.30) - I think there were 9 players with proj +2/-2 among 50 guys (which is 18% of the highly proj pool I looked at) - I think that seems reasonable for an early run comparison on a Wednesday, but that's up to you to judge really - prob need to check it on Sat or Sun morn to see how they line up 3 & 2 pts does make a lot of diff when you get into optimizers, sims etc, but it's possible his might actually be better at the current time (like I said it is a very early run from the commercial site - and I haven't actually tested or used the TwoGun projs or optimizer lineups myself yet) - and also that they may converge more in the coming days - it's also possible his projs may actually be better overall than some sites idk Just example, the diff on Gibbs dk proj was only .53 pts - Puka is off by 3.12 pts (the most of any non-injury type situation), but TwoGun's Puka projs look more accurate to me currently from the eyeball test - and I'm guessing those will prob converge and narrow the difference gap Another thing I want to look at in the coming days is to see if the chalk cheap value plays line up somewhat between the sites in terms of proj and pown - because those are often key to how ppl play a slate The powns are way off currently, but that may fall in line today or in the coming days (the other site hasn't really fixed theirs yet I think), but I would take those w a grain of salt until you see how they line up. And accurate pown estimates are very important for how I'm playing gpps currently It also does have a fancy, pretty slick (and a bit complex) optimizer on it with a lot nice options for builds - one part was a little tricky when I was trying to generate a pool of lineups - he also is actively working on other upgrades and new features for the site - I think the AI is actually teaching itself as well and building, which is wild to me (but maybe that is getting normalized) So I would say check it out - but again I'm not guaranteeing anything, just passing it along w my thoughts so far - writing stuff like this actually helps me to organize and think through things, and I generally just post em here instead of in personal notes or a journal or whatever