@Matt_barlowe@poodleknowsball That's what the chart shows. The lowest that curve ever went is -$658. Across 11,700 bets that's the most he was ever down, so no reload. Kelly stakes are a % of your current roll, they shrink during drawdowns, so ruin risk with a real edge is ~0. True PnL here is +$387K.
@Matt_barlowe@poodleknowsball Pro bettors can definitely cycle through their entire bankroll on one sport multiple times a year. Also annualized ROI in a sim is purely a function of stake size and bets per year. What did you assume? At quarter Kelly over 11,700 bets it compounds way past 29%.
@Matt_barlowe@poodleknowsball Even at 23% that’s what your capital earns in a year. The 9% is earned every time the bankroll cycles through a bet, and it cycles dozens of times a year. A $20K bankroll cycling $1M of handle at 9% makes 90K. That’s 450% annual
@Matt_barlowe@poodleknowsball ROI is on total handle, not capital. $387K at 9.38% means he wagered about $4.1M, but he never had 4.1M. It’s a bankroll recycled over 11,700 bets. An index fund pays 10% on your money once a year. A 9% edge on money you re-bet every week crushes that
@AndrewHeaters@BuffaloFreddy Mrtka is not slow + he’s skating is getting even better. He’s not the most gifted offensively but he makes a good first pass, certainly not “stone hands”.
@Sabremetrix@23sabres I don't know, Krebs was the 16th pick, Mrtka the 9th pick. Hague had a good D+1, but I doubt that closes the gap to a recent 9OA. In 21, Wheeler had Krebs as the 26th best NHL affiliated prospect. In 26, Mrtka was 22nd. You can nitpick the source, but it's close either way IMO.
@BillTCB 27-28 salary cap goes up a ton + Skinner’s buyout hit is 4m lower + Byrams raise will likely only be around $3m - can fit Tuch + Byram quite easily with a full roster. 26-27 is harder to fit Tuch into.