MLB Picks Using Sabermetrics

[11 d. drogba]

Hey guys, I'm fairly new to handicapping and I've always been interested in sabermetrics because baseball has such a great sample size. I had the idea of incorporating some of the sabermetrics statistics to project scores. I broke it down into batting and pitching. For batting, I used each team's 2011 wRAA and dividing by games played and shifting to match league average of runs scored. For pitching, I used each team's 2011 RP xFIP multiplied by 1/3 (assuming they pitch about 3 innings a game) and added to the projected starter's 2010-2011 xFIP multiplied by 2/3 (6 innings). Averaging one team's pitching with the opponent's offense, I came up with a projected run total. I did this for each of today's games, and I'll post the plays this little model I made came up with below. This is sort of a test run so I'll record how my picks do, and if this system works I'll continue posting. I would love some feedback on my model and any other advice. Before anyone starts criticizing my picks, I want to note, you should wait a few days before investing too heavily into my picks, so I can evaluate the model and make any necessary changes. My top plays will be in bold.

ML (I also made slight adjustments for Home Field Advantage):
Phi +109
Fla -110
SF -136
CWS -144
Cle +113
NYY -113

Totals:
Phi/Atl o7.5 -120
LAD/SD o7 -125
Oak/Min u8.5 -125
TB/CWS u9 -120
Tor/LAA o8 +102
Cle/Sea o7.5 +115
NYY/Bos u9.5 -120

For record keeping, I will assume these are all 1 unit plays for now.

[warriorfan707]

Interested to see the results, GL on your experiment

[11 d. drogba]

Thanks, I also just realized the scope of my plays is a little big. I'll leave those for now, but I'll narrow it down a bit for tomorrow's picks.

[GoggsViggs]

I like the concept. GL on your experiment, I'll be following.

[11 d. drogba]

Off to a pretty good start, also note: TB is HORRIBLE. Gotta feel for Price and Maddon. I can see picking against them everyday being a decent play at the very least until Longoria comes back and even then they're not working with a whole lot.

[IAMWINNER]

Do you read the newspaper articles on the teams also or just cap based on stats. Seattle is a team in meltdown right now.

[11 d. drogba]

Cap based on stats, but I try to keep up with headlines and what managers/players are saying to stay informed. Yes, Seattle is bad, but when have they not been bad? TB could go from a 96 win team to like a 60 win team. Even with all their offseason losses, I don't think anyone imagined them being that big of a dropoff. But maybe I'm wrong and Longoria will come back and make them respectable, we'll see.

[spargament]

Like the sabermetrics angle. Curious to see its success, BoL

[11 d. drogba]

Just tuned into Cards-Giants game, must say I'm a little disappointed Bochy decided to sit half his starters on one day. I know Torres had some issues, but to give Posey, Burrell, and Tejada the day off as well. They had an off day on Thursday so they should have been decently rested. (I'm a Giants fan btw) I'm pretty new to handicapping and I haven't pay this close attention to lineups in the past, but seeing all these starters out... I don't like it.

[11 d. drogba]

4/10/11
ML:
Phi +109 won
Fla -110 lost
SF -136 lost
CWS -144 won
Cle +113 won
NYY -113 lost

Totals:
Phi/Atl o7.5 -120 lost
LAD/SD o7 -125 won
Oak/Min u8.5 -125 won
TB/CWS u9 -120 won
Tor/LAA o8 +102 loss
Cle/Sea o7.5 +115 won
NYY/Bos u9.5 -120 won

Total:
8-5-0 +2.58 units
Top Plays (Bold):
2-0-0 +2.15 units

Okay start, Bochy resting starters and vintage Beckett hurt what could have been a very good day. I will be running numbers for tomorrow's games and narrowing my plays to only the best ones. Should have picks up in 2-3 hours.

[TheJames]

Good job on your picks

[11 d. drogba]

TB/Bos u9 was the top play, considered putting 2 units down, but I know at some point Boston's bats will wake up and Dice-K scares me. Also, the season is young and still working on improving my model, so tread carefully. Top picks in bold. All plays are for 1 unit.

4/11/11
ML:
Ari +105
Cin +104
CWS -120

Totals:
StL/Ari u9.5 -101
Cin/SD o6.5 -125
LAD/SF o6.5 -115
Tex/Det o8 +105
TB/Bos u9 +110
Cle/LAA o8 +100
Tor/Sea o6.5 -110

Previous results:
Total:
8-5-0 +2.58 units
Top Plays (Bold):
2-0-0 +2.15 units

[LLXC]

Question: Why did you decide to look at pitchers' FIP, because that stat is great for ranking pitchers but I don't think leaving out a team's defense entirely is the right idea. Just my thoughts.

[11 d. drogba]

"- xFIP has the highest correlation with future ERA of all the pitching metrics."

http://www.fangraphs.com/library/index.php/pitching/xfip/

I know defense is important and I am missing that in my model. I don't really have/know a way to quantify team defense in terms of runs, as there is so much to consider (range, arm, fielding...). I know UZR is used for individual players in the sabermetric world, but I couldn't find a team UZR. I'll look for other ways, and any suggestions are welcome.

[11 d. drogba]

I looked into defense a little more and found a stat called Defensive Runs Saved (DRS), which looks promising. Still not sure how I would work it into my model and what kind of weight it deserves compared to hitting and pitching. I'll continue thinking about it and would appreciate any input.

[11 d. drogba]

Still working on system, I did add defense, but realized I way overstated it, sticking with earlier picks for today.

[11 d. drogba]

4/11/11
ML:
Ari +105 loss
Cin +104 win
CWS -120 loss

Totals:
StL/Ari u9.5 -101 loss
Cin/SD o6.5 -125 loss
LAD/SF o6.5 -115 win
Tex/Det o8 +105 loss
TB/Bos u9 +110 loss
Cle/LAA o8 +100 loss
Tor/Sea o6.5 -110 win

Total:
11-12-0 -1.84 units
Top Plays (Bold):
2-3-0 -0.85 units

Really poor day. Juan Pierre lost the game for CWS and Dice-K blew. On the bright side, I figured out how to add defense using defensive runs saved/games (DRS/G) for 2011 and scaling it down because the sample size is only like 10 games for each team and then I subtracted this from the projected run totals. Hopefully, this makes the system more accurate.

[warriorfan707]

MLB is following in the footsteps of NBA when it comes to amazing these days

[11 d. drogba]

Seriously, Tor/Sea... Felix gives up 7 and Sea comes back from down 8 with 5 outs left using 3 runs off walks. Toronto's bullpen prior to this game wasn't bad either.

[11 d. drogba]

I feel pretty good about these picks. The system now incorporates my interpretation of the most important sabermetric statistics for hitting, pitching, and defense, wRAA, xFIP, and DRS, respectively, using 2010-2011 data for projected starting pitchers and 2011 data for everything else. Sample size is still a little small for 2011, but will improve with a few more weeks of games. All 1 unit, pick of day bold.

4/12/11
Pick followed with some notes
Phi -118 (Livan xFIP 4.61, Phi RP xFIP 3.27, Phi proj hitting 6.33 runs, projected score: Phi 5, Was 4)
KC +135 (Francis xFIP 3.73, Min proj hitting 2.53 runs, projected score: KC 5, Min 3)
CWS -120 (Jackson xFIP 3.61, CWS proj hitting 5.68 runs, Oak proj hitting 3.24 runs, projected score: CWS 5, Oak 4)

Cin/SD o7 +103 (Cin proj hitting 5.74 runs, projected runs: 9)
LAD/SF o6.5 +105 (projected runs: 7)
Cle/LAA o7 -115 (LAA RP xFIP 5.23, projected runs: 9)
Tor/Sea o6.5 +105 (projected runs: 8)

Previous results:
Total:
11-12-0 -1.84 units
Pick of Day (Bold):
2-3-0 -0.85 units

[11 d. drogba]

4/12/11

Phi -118 loss
KC +135 loss
CWS -120 won

Cin/SD o7 +103 won
LAD/SF o6.5 +105 won
Cle/LAA o7 -115 loss
Tor/Sea o6.5 +105 loss

Total:
14-16-0 -3.09 units
Pick of Day (Bold):
2-4-0 -2 units