Ive been checking splits, L or R splits, check Vegas totals, look at the pitcher his splits as well, off the top of my head for what I do for hitters maybe something I'm forgetting

first I look at the size of the slate the larger the more margin for separation. I then begin to look at woba splits L and R. I start with the 3yr average for X player. After developing a list of 2 or 3 at each position that are appealing I typically look at the woba splits L/R for the last 14 days to see who's swinging a good bat. I put emphasis on guys batting 1-5th, and some more on guys batting first and 2nd with big power ala Manny Machado batting in the one hole.

I'm always looking for Gas Cans guys like Roberto Hernandez, Jerome Williams, and John Danks (pitches tonight).

I do consider park factors though I have had a horrible run @coors. I also consider totals though I seem to put more weight in the picking the pitcher favored by Vegas over batter with high totals from Vegas. Seems like I always have guys on teams with 8.0, 8.5, and 9 plus totals, but it's doesn't seem any less or more random. I usually get the same of better numbers from guys w/ 7-8 totals. ( maybe a sample size issue?)

I am very interested in putting in the time and learning DFS becoming a better player each day. Baseball season is long and I really hope to come out of it with a better understanding of the mechanics that will help me build better lineups.

I think most of the value in MLB is in GPPs especially single entry ones, but with maybe 170-180 games all year it's hard to get a gauge of how good you are if you're in 5,000+ fields given all the variance. Cash has been a lot tougher for me than basketball and I'm looking forward to a bigger market opening with hopefully lower rake.

^^ Coors games have been running pretty poorly for me as well. At this point in the season, games like those that were 'obvious' stacks that ran poorly could make or break your season, since stacking is so high variance.

For me I choose my hitters mostly based on pitching. That is, I try to identify the 3-5 pitchers who are in the worst spots (combination of being mediocre/bad to start with and/or pitching in a hitters park, etc.), then I stack against them. For awhile I was experimenting with 6-1-1 stacks but I've had more success doing something like a 3-2-2-1. When a team goes off for a lot of runs the scoring is usually concentrated in 2 or 3 guys anyway. And most often they are the usual suspects, guys in the 3, 4, 5 hitting positions, though I also like leadoff guys, particularly fast ones that can get you runs and SBs.

I don't put much stock in streaks, because I believe they are mostly just random noise. If anything I'm fine with hitters that are "cold" because they are usually better priced, though I don't necessarily go out of my way to pick them. The main focus for me is just to find pitchers in bad spots and then stack or mini-stack against them. Then the rest is sort of like a jigsaw puzzle trying to get as many of those guys as I can and still fit under the cap, which may involve punting a position especially if I've spent a lot on pitching.

can anyone share their results over a solid sample?

I find baseball a lot more enjoyable than I did NBA, granted I've put more effort into MLB for sure

I think I am also running cold on Coors but the silver lining may be that people probably aren't rostering as many Coors guys as they should be. Here is what the Coors totals distributions look like over the past several seasons. There are zero 1-0 games since 2012 which kind of blew my mind when I was looking at data. You actually have to go back to July 12th, 2010 for the last one, and there are only 9 in the history of Coors Field!



The other thing to note is what I was saying on the other account about RG run projections being wrong. Team medians (aka team totals) are marked by solid vertical lines (5 for teams, 11 for total), but the team means (+ markers) are 5.59 and 5.78, which sum to 11.37. The counterintuitive piece here is that the team totals sum to less than the game total, but the team means sum to more than the game total. So, If I gave you the game total, it would be wrong to force two means that sum to 11, but that is what RG is doing. Overall it would "only" be off by about 0.4 total runs in a hypothetical case where the above distributions are the true ones, but throw in some more error for vig or whatever and the bias could be fairly substantial.

Where did you get this data from? I did something similar manually in excel a few weeks back but wondering if there's a site that already did all this? Or did you do it yourself? Have the Coors numbers changed in recent years? I pulled every game log from baseball reference and got a median of 9 and mean of 9.71 (Graph below).

Ok, first thought is stop caring about L14 stats. As Tim said, streaks are often just noise. It's pretty common for the day's most popular "value play" to be a garbage hitter who happens to have homered recently. Let other people make the mistake of rostering Chris Parmelee while you take a better batter whose discounted because he's "slumping."

Second thought is that using wOBA by itself is dangerous. For example, someone on RG earlier this week (one of the founders I think actually) commented on how it was hard to ignore Brayan Pena against Trevor May. Really? Brayan Pena who has pretty much been a below average hitter his entire career? I mean Trevor May isn't a stud by any means but I wouldn't say he's a guy to load up against either. So why Brayan Pena? Cause the 2015 wOBA machine sees Pena's .340 vRHP and Mays .327 vLHB and says it's a good matchup. If you look at their respective BABIPs though you'll see that Pena is getting lucky and May is getting unlucky. Use BABIP, use ZiPS and Steamer projections, use more than just wOBA to assess hitters and pitchers. You don't need to know what a player IS doing, you need to know what he SHOULD be doing.

Try to be as objective as possible. If a minor leaguer comes up and is crushing in his first 10 games but you see that he was an average or below average hitter in the minors, don't make up some narrative to explain why he's crushing the majors. Trust yourself and your evaluation of a player, there is a lot of mediocre advice out there and it's easy to get swayed by it.

This was just a quick copy/paste from BRef but you could iterate through urls with a script changing team and year if you wanted.

http://www.baseball-reference.com/te...e-scores.shtml

Ah same thing I did then. I've got a spreadsheet of all 2014 game logs for this data.

I get the same numbers you have if I don't remove the Rockies away games.

Hmm let me check that I thought I took them out. Spreadsheets a couple of months own and just something I did for reference so could be a mistake in there. Yep. Put a wrong sign in the formula. I got mean 11.65 median 11 for them in 2014.

You didn't think that seemed off when you only saw 9 for coors?

I mean I looked at it for 5 seconds and closed it. Was just testing if I pulled all the info from bref and such. I didn't actually use the info for anything.

would be great if searching for games on FD website or android was as easy as the iPad app

and if making lineups on ipad was as easy as the website

Angels prices are broken

Stacking Boston Houston with Kershaw and Morgan. Maybe once a punt pitcher will work for me

Paulson wasn't an ideal last minute swap out on Victiv at near min price, yikes! never thought I'd have shares of Pennington tonight

It sucks how one tiny mistake can throw off a ton of tough-to-read, technical work. The way I had coded up the hist() bins caused R to throw an error when I tried to include the game total and I couldn't figure out why. Then I realized it was because I assumed that at least one game with a total of 1 existed in the data, which is how I figured out it did not.

gerlaod parra system in full effect tonight

One small (sign) mistakes throwing everything off...a classic issue in former engineering homework assignments and MATLAB code

Miggy blows out his hammy.

Parras good value against bad righties most of the time.

Carlos Gomez on track to be one of the rare hitters on DK to put up a negative number

edit : as soon as I posted that he got a hit nvm