MLB Daily Games Strategy: Adding ISO To Your Arsenal

Written by

Updated on July 22, 2014 6:46PM EST

Adding ISO to your DFS arsenal at CI

I could start every single strategy article I write about MLB DFS the exact same way...frustration with the day to day variance inherent in the game. From the first week of this season, when I committed to playing daily MLB, every day, I've worked hard on refining my personal strategy for building a successful lineup. I've added in various tools, dropped others that were too time consuming and yielded too little reward. I've written about all of this before. The underlying theme is that baseball is just a lot harder to predict than basketball or even football.

Actually, it's not harder to predict. Baseball has the best data available of any of the major sports, and there are several algorithms that argue that they weight all the variables optimally to spit out the best plays of the day, some even digging down to batter and pitcher success against/using specific pitches. That's incredible. So, with tools like those, it's not hard to predict baseball. It's hard to be right.

How often does your most logical, most researched, most justifiable play go 0/4? Half the time? Seems about right to me. And that's crazy. It doesn't matter how much salary cap you expended on said player, how good the circumstances are, he might not score you a single fantasy point. Now that is nearly impossible in NBA, and pretty darn rare in NFL. MLB is the only sport where money doesn't buy you any

Adding ISO to your DFS arsenal at CI

I've tried to let a lot of the frustration go, take it in stride, know that everyone else is in the same boat, etc. That hasn't stopped me from trying to figure out the best way to predict DFS success for hitters. Today I want to focus on the corner infield spots. You typically are hoping for power in the form of homeruns and extra base hits from these two positions. I usually pay up here while saving in middle infield and outfield. So how to make that investment a little more sound?

I took a look at the top 15 DFS 1B and 3B according to fantasy points per game (FPT/G):

POS	PLAYER	TEAM	DD Salary	FPT/G
1B	Jose Abreu	White Sox	12800	18.3
1B	Edwin Encarnacion	Blue Jays	11650	18.3
1B	Paul Goldschmidt	Diamondbacks	12500	18.2
1B	Miguel Cabrera	Tigers	12000	16.5
1B	Anthony Rizzo	Cubs	11300	16.4
1B	Brandon Moss	Athletics	10400	15.5
1B	Albert Pujols	Angels	10900	15.2
1B	Freddie Freeman	Braves	10200	15.2
1B	David Ortiz	Red Sox	10100	14.7
1B	Steve Pearce	Orioles	8450	14.1
1B	Michael Cuddyer	Rockies	9650	14.1
1B	Adam LaRoche	Nationals	8500	14.1
1B	Mike Napoli	Red Sox	9350	13.9
1B	Mark Teixeira	Yankees	9350	13.6
1B	Joey Votto	Reds	9400	13.2

____________________________________________________________

POS	PLAYER	TEAM	DD Salary	FPT/G
3B	Josh Donaldson	Athletics	10100	15.6
3B	Anthony Rendon	Nationals	10350	15.6
3B	Adrian Beltre	Rangers	10200	15.4
3B	Todd Frazier	Reds	10250	15.3
3B	Kyle Seager	Mariners	9800	14.2
3B	Ryan Zimmerman	Nationals	10450	13.6
3B	Brock Holt	Red Sox	9350	13.5
3B	Neil Walker	Pirates	9550	13.3
3B	Lonnie Chisenhall	Indians	7600	13.1
3B	Conor Gillaspie	White Sox	9050	12.8
3B	Trevor Plouffe	Twins	8250	12.7
3B	Aramis Ramirez	Brewers	8550	12.6
3B	Carlos Santana	Indians	9000	12.4
3B	Matt Carpenter	Cardinals	8650	12.4
3B	Casey McGehee	Marlins	7350	12.3

Notables who don't make the cut include Adrian Gonzalez, Adam Dunn, Justin Morneau, and Chris Davis at first, David Wright and Evan Longoria at third. As you can see, salary follows DD pts closely.

Next, I went to FanGraphs to get the 2014 season data for these positions (no splits, just full season data for this year) to see what statistics most reliably sorted players into the lists shown above. In other words, I want the player who is going to succeed best in my DFS scoring system, so I want to know what actual baseball statistic of his predicts that. So I sorted first and third basemen on things like HR, SB, RBI, R, wOBA, OPS, and ISO.

Some of my instincts going in:

Most DFS sites weight HR extremely heavy in their scoring.
Sometimes it's SB that push otherwise similar players apart.
A lot of experts and professional DFS players rely on wOBA.

None of these produce a ranked list of corner infielders that approximates the goal, however. I found that the best metric to rank these players on to match the DFS FPT/G ranks was ISO, or isolated power. Here's a link to the FanGraphs excellent description.

First Base:

Name	Team	HR	R	RBI	SB	ISO	wOBA
Jose Abreu	White Sox	29	51	74	1	0.327	0.405
Edwin Encarnacion	Blue Jays	26	57	70	2	0.314	0.411
Brandon Moss	Athletics	22	48	67	1	0.261	0.377
Paul Goldschmidt	Diamondbacks	18	71	65	8	0.253	0.414
Anthony Rizzo	Cubs	23	65	53	2	0.242	0.392
Miguel Cabrera	Tigers	14	58	76	1	0.22	0.383
Lucas Duda	Mets	14	38	49	2	0.217	0.361
Michael Morse	Giants	14	36	47	0	0.213	0.358
Mark Teixeira	Yankees	17	39	48	1	0.212	0.339
Adam Dunn	White Sox	14	34	39	1	0.209	0.351
Albert Pujols	Angels	20	59	65	4	0.205	0.344
Matt Adams	Cardinals	12	32	43	3	0.204	0.371
Chris Davis	Orioles	16	39	50	2	0.197	0.31
Freddie Freeman	Braves	13	65	53	0	0.191	0.373
Justin Morneau	Rockies	13	38	60	0	0.189	0.366

Third Base:

Name	Team	HR	R	RBI	SB	ISO	wOBA
Kyle Seager	Mariners	16	40	64	4	0.216	0.366
Josh Donaldson	Athletics	21	65	70	3	0.215	0.338
Todd Frazier	Reds	20	58	54	15	0.211	0.371
Adrian Beltre	Rangers	14	52	52	1	0.199	0.391
Anthony Rendon	Nationals	13	69	53	8	0.196	0.352
Mark Reynolds	Brewers	14	34	33	5	0.179	0.303
Lonnie Chisenhall	Indians	9	41	41	2	0.179	0.395
Carlos Santana	Indians	14	40	40	2	0.176	0.332
Trevor Plouffe	Twins	7	43	44	0	0.172	0.323
Pedro Alvarez	Pirates	15	41	47	6	0.169	0.322
Luis Valbuena	Cubs	5	34	25	0	0.162	0.327
Pablo Sandoval	Giants	12	44	43	0	0.161	0.33
Nick Castellanos	Tigers	6	30	35	2	0.137	0.314
Matt Dominguez	Astros	12	37	43	0	0.136	0.291
Evan Longoria	Rays	11	52	48	4	0.132	0.318

As you can see, the top 6 1B are identical when sorted by ISO. For 3B, we get the top 5 correct. While it's certainly not perfect at predicting fantasy points, ISO is a metric I'll be looking at some more. As FanGraphs notes, ISO is not considered predictable for small samples, yet it performed better in this setting than HR, wOBA, or anything else I checked. I also haven't applied any rigorous test of which stat is a better predictor, I'm simply matching lists by hand. Remember that DFS is not the typical context in which sabermetric stats are used. Most people are using these numbers to predict seasonal fantasy output or actual baseball career output.

One way I will try to take advantage of this ISO correlation is to look for players with great pitching matchups, that perhaps have really high (favorable) ISO splits. For example, tonight Jose Abreu vs LHP Chen (ISO vLHP .310) and Adam Jones vs LHP Santiago (ISO vLHP .340) look like strong plays on this statistic alone. Facing RHP tonight, Trout vs Gonzalez (ISO .289), Seager vs deGrom (ISO .272), Smith vs Hendricks (ISO .239), and Duda vs Ramirez (ISO .249) look like good plays. I happen to believe in deGrom, so I will be avoiding Smith and all the Padres as usual.

Predicting one game is always going to be a nightmare, but by adhering to some logical process (see many of my previous MLB DFS Strategy articles), refined over time with new information, we can rest a little bit easier. ISO is one of the new factors I'll be considering when deciding between similar players. Particularly when you're going to be spending top dollar, having one more piece of data to justify a selection is nice. Good luck this week! As always, feel free to comment or question below!