Hockey is also much less of a binary sport (try predicting how Joe Quinville will structure his lines for a typical Chicago Blackhawk's game -lol). Sure, baseball has platooning and there is enough variance to keep fish interested, but sabermetrically speaking... It seems to me, that there is always an optimal batter vs. pitcher match up for cash games. No?

Not necessarily true for GPPs, because you're not trying to get the best "average" score.

.

I don't want to derail thread, but just curious if you're talking about cash or GPPs. Why would it be hard to come up with a perfect projection model for cash games? Not saying I could do it, but why not some sabermetric savvy programming or Excel expert? Feel free to PM me if you're knowledgeable about this. Thanks.

Only if you have a perfect projection system that had no error or noise in the calculations. In reality every projection you have for each player with have enough error that you won't be able to distinguish between several players projected similarly. However there are no perfect projection systems. Just take a look at the few of the public ones out there. There are pretty big differences in their projections on some players.

What is the "perfect" time scale? Do I use a guys last 3 games, last 10 games, 2015 splits, 2014-15 splits, 2013-2015 splits, entire career splits?

Seems like it's relatively important for Odorizzi not to have a big game.

Many DFS players share ideas with each other either over private chat, in public via Twitter, or through reading each other's newsletters. It results in a lot of group think and herding. In some cases, though, players explicitly join forces and use identical lineups or identical lineups that intentionally differ by only a single player or two.

Re: projections, we can quibble over the definition of "perfect" but I don't see any reason why a near perfect projection system cannot exist. Note that perfect doesn't mean non-random outcomes; just that it accurately projects the range and probabilities of each potential outcome.

Touche! That's the million dollar question (at least for me -lol). But like poker, you're talking about ranges, not absolutes. Surely, there is an ideal range, age, etc. that zeroes in on optimal.

Again, I'm not trying to start an argument. Just curious if it's worth me putting a bunch time into DFS cash games. I find it a very enjoyable way to follow my love of baseball and if I can make some $$ at it, all the better. But I don't want to just give $$ away to people who have ridiculous advantages using Excel solvers, etc.

Trying to remember the last time Reyes lived up to his DK salary. If he has this year, it must've been the day I didn't roster him.

Because it's daily wouldn't this be impossible?

How can you solve any game with infinite outcomes?

"Perfect" projection obviously doesn't mean there is zero variability between the predicted and actual outcomes. I mean that is true of basically all predictive modeling in all cases ever, in any discipline, including physics. Some things are more predictable than others, i.e. a higher signal:noise ratio. Just because there is a lot of noise in dfs outcomes doesn't mean it isn't virtually solvable, at least to the point where you wouldn't play against it at 10% rake. And of course there are qualitative factors to consider externally of any model since there are things models don't do particularly well, but those are probably the dfs analog of physical tells in poker: i.e., you wouldn't base your fundamental strategy on sikk readz as opposed to just having very good, mathematically-sound poker strategy.

Fair enough, then again I can say the same about most regs, advanced algo through somewhat decent gomer that does everything by hand, like me. The games are beatable because most regs aren't all that good, the part timers that get picks from experts are mostly clueless, and recs are total fish, and there's a lot of them.

I can see the cash games drying up and soon as growth in the last group slows. Then it's all GPP's, or lower rake.

BvP guys last night:

Trout
Pujols

I doubt there are infinite outcomes. Chess and backgammon have a ridiculous number of outcomes and they both have been mostly solved, afaik. Also, make sure you're not confusing variance with infinite outcomes. I agree baseball is high variance.

Wandy Rodriguez doesn't seem as bad as the general public is making him out to be.

Angels did get 1 hit last night too, but that was vs a righty. Most of the team has much better numbers vs lefties though.

Oh yeah I'm definitely on the Angels and am by no means scared of Wandy. It's just funny how hard people are ragging on him. His 2014 numbers v RHB are pretty bad but it's a 22 IP sample size. His prior year numbers aren't as bad, though they aren't stellar either.

I'm gonna use Trout a lot but it's not like there aren't other guys to attack on the mound tonight, like Miguel Gonzalez, the Harangatang, 2 kids at Coors. I'll be going after Leake and both guys at Tampa as well, Garza too.

c. gonzalez seems so due but he lets me down every time

how much better do we think king felix is than the rest of the SPs tonight? i originally had him but if i go with kaz i can upgrade to trout.

I like Felix a lot but I could only
Afford him in 1 out of 3 lineups

I love Cole and Kazmir tonight

His upside (1) at home (2) vs MIN (3) is redic...that said i'm more all in on Kaz for value with some shares of King Felix, Cole, and Cashner for a bit of a contrarian pick.

Billy Hamilton might steal 6 bases, tonight.