Monday, April 13, 2015

Daily DraftKings Matchups: Monday, April 13

Monday’s Optimal DraftKings Lineup

Note: Lineup subject to change due to weather and released MLB lineups. Check back before contest lineups lock for our final lineup.

C: Hidden
1B: Albert Pujols (LAA) @ R. Detwiler (TEX) — $5,400
2B: Robinson Cano (SEA) @ B. McCarthy (LAD) — $4,100
3B: David Freese (LAA) @ R. Detwiler (TEX) — $4,200
SS: Hidden
OF: Mike Trout (LAA) @ R. Detwiler (TEX) — $6,600
OF: Collin Cowgill (LAA) @ R. Detwiler (TEX) — $3,200
OF: Justin Upton (SD) vs. R. de la Rosa (ARI) — $4,300
SP: Shelby Miller (ATL) vs. M. Latos (MIA) — $7,100
SP: Hidden

Before I break down today’s picks, I want to take several minutes to discuss my general approach today (which I spent some time researching and refining over the weekend).

Unsurprisingly, batter versus pitcher (BvP) splits are heavily relied upon by those who play DFS. The logic is simple — not all matchups are made the same, and how a hitter performs against David Price is probably different than how he performs against Phil Hughes, which is probably different than how he performs against the league on the aggregate. The major problem with BvP splits, though, is there’s just not enough data to work with.

According to FanGraphs, it takes about 320 AB for SLG to stabilize for a batter (among the basic triple slash stats, I like SLG best for DFS analysis). While repeated AB against the same pitcher (as opposed to YTD data against all pitchers faced) probably reduces the number of AB needed for SLG to stabilize, that number doesn’t drop to anything resembling the typical sample size BvP data has to offer. That means we’re constantly left to guess how much weight to add to Anthony Rizzo’s impressive 7-for-20 (.350) with a homer against Mike Leake. Did Rizzo just catch Leake a couple times while he was running hot? Do those stats justify chancing it on Rizzo for $4,800 today?

My solution to this BvP sample size problem is to generalize the BvP matchup. Let’s take Rizzo versus Leake, for example. If we just look at those two in their specific matchups, we only have a 20 AB sample size — not much to work with. However, we can generalize Leake as a finesse, ground ball pitcher and then see how Rizzo does against other finesse, ground ball pitchers.

Quickly, a couple housekeeping notes. First, how did I classify Leake as a finesse, ground ball pitcher? I simply used the definitions provided by Baseball-Reference.com, who record batter and pitcher performance against a lot of very detailed splits such as these. Their definitions are as follows:

  • Finesse: Bottom-third of the league in K%+BB%
  • Ground Ball: Top-third of the league in GB/FB ratio
Second, where you can find these splits? They’re available on all players’ splits stats pages but you can download bulk splits using their splits stats finder.

By categorizing Leake as a finesse, ground ball pitcher, we’ve now created a bucket into which we can toss similar pitchers; according to Baseball-Reference.com’s definitions, this bucket would also include Hisashi Iwakuma, Tim Hudson, Alfredo Simon, Rick Porcello, Trevor Cahill, Tyler Chatwood, Jhoulys Chacin, Carlos Martinez, Jerome Williams, and plenty others. Suddenly the sample size isn’t just 20 AB. It’s a lot more. In fact, since the start of 2012 Rizzo has 601 AB against finesse pitchers and 470 AB against ground ball pitchers.

(Want to argue Martinez isn’t a finesse pitcher due to his blazing heater? That’s fine. Remember, these are general buckets that include dozens and dozens of pitchers, so on the average the pitchers typically resemble each other, though there are bound to be some ugly ducklings.)

Due to limitations in the way the data is organized we have to look at Rizzo’s performance against finesse pitchers and ground ball pitchers separately, and I kept it simple and just averaged his SLG against each type to yield a combined SLG against finesse, ground ballers. For Rizzo, this was a .556 SLG against finesse pitchers and a .500 SLG against ground ballers for an estimated .528 SLG against finesse ground ballers.

We also can’t split out finesse right-handed pitchers from finesse left-handed pitchers, however, we can make a loose assumption that if Rizzo’s SLG against right-handed pitchers is 5.6% higher than his SLG against all pitchers, which it is, then his SLG against right-handed finesse pitchers would also be about 5.6% higher than his SLG against all finesse pitchers. We have Rizzo pegged for about a .528 SLG against all finesse, ground ball pitchers, so a 5.6% increase to account for us only caring about the right-handed pitchers bumps our estimate for Rizzo’s SLG against finesse, ground ball, right-handed pitchers to .558.

The fact that we have to make these assumptions prevents the data from being perfectly representative of the specific matchup at hand, but it’s fairly close, and any error introduced by making these assumptions is probably less than the error incurred by using a 20 AB sample.

Are you lost? I hope not because there’s two final steps left.

First final step — adjusting for the specific opposing pitcher. The .558 estimate we’ve thus far generated is our projected SLG for Rizzo against the average finesse, ground ball, right-handed pitcher. However, over the last three years Leake has held the average left-handed batter to a .446 SLG whereas the average finesse, ground ball pitcher has held their batters to a .416 SLG. This means Leake has allowed a SLG to left-handed batters that’s 7.2% higher than the league average for pitchers of his type, so let’s increase Rizzo’s projected .558 SLG by 7.2% — now we’re at .598.

Last final step — adjusting for ballpark. The game between Rizzo and Leake will take place at Wrigley Field, which has inflated SLG by 3.6% over the last three years. Take Rizzo’s .598 SLG and increase it 3.6% and we’re now at .620.

To summarize, what is that .620? That’s Rizzo’s expected SLG against all finesse, ground ball, right-handed pitchers adjusted for both Leake’s performance relative to the league average finesse ground baller and ballpark. So, when we see Rizzo has slugged .600 against Leake in 20 AB against him, we can feel confident that’s about the type of production we’d expect from him in this specific matchup.

Now, that was a super-lengthy explanation of my general method. I have a file set up that allows me to quickly calculate these expected slugging percentages based on generalized BvP matchups, and by comparing expected SLG to a player’s projected (or actual) season-long SLG we can identify good hitters to target.

And that’s how I landed on all of today’s Angels!

Adjusting for everything — power/finesse, ground ball/fly ball, specific strength of pitcher relative to others of his type, and ballpark, I generated the following Expected SLG for their game tonight against Ross Detwiler (Steamer projected SLG also shown for reference):

  • Pujols — Projected: .539 | 2015 Steamer: .462
  • Freese — Projected: .561 | 2015 Steamer: .387
  • Trout — Projected: .566 | 2015 Steamer: .535
  • Cowgill — Projected: .529 | 2015 Steamer: .336
These projected SLG are extra valuable in this case because the Angels as an entire team have just 16 AB combined against Detwiler. We can figure Detwiler isn’t a great pitcher so an Angels stack should be safe, but I’d prefer to quantify how good each matchup is and then make my selections based off that. Based on these estimates, Cowgill and Freese look to be great buys today, while Pujols is clearly better against the Detwilers of the world as well, and Trout is only marginally better than his usual, otherworldly self.

Also in the lineup today is Cano, who squares off against Brandon McCarthy. Now, this isn’t the same McCarthy as we’ve seen in past seasons as he’s improved a good amount since the start of last season, but Cano is still 3-for-11 (.273) with a homer and just one strikeout in his career against the Dodgers’ new right-hander. Using the same analysis above, I project Cano for a .503 SLG in this particular matchup, which is higher than his .460 Steamer projection for the season and well worth his modest $4,100 price tag.

With Upton it’s the same thing. Rubby de la Rosa has little experience in the majors, and Upton is just 0-for-1 with a walk in limited time against the Diamondbacks’ starter. Using the method above and bucketing de la Rosa as a ground ball pitcher with walk and strikeout rates that bucket him as average (in the middle-third between finesse and power), we can project Upton for a .538 SLG in this specific matchup, including ballpark, which compares favorably to his .442 Steamer projection and makes him worth the price tag as well.

As for Miller (let’s not forget the pitcher!) he held Miami scoreless in his first start of the season, and I don’t think it’s a fluke. He’s developed a nice two-seamer/sinker that he’s relying upon a lot. He actually unveiled the pitcher last August and while the improvement wasn’t immediate (4.35 August ERA), he pitched like an ace in September (1.48 ERA) and then again in his first start of the 2015 season. In fact, in that first start against the Marlins the former fly ball pitcher didn’t allow a single fly ball. Combine that with Miami’s struggling offense and Marlins Park’s spacious dimensions, and I’ll roll the dice with Miller again.

Baseball Professor is sabermetric-slanted fantasy baseball blog with fresh content updated daily. Their fantasy analysts provide you with everything you need to run your team on a daily basis, including their SP Matchup Ratings, Fantasy News and Notes, and a no-fluff daily podcast (or as they call it, a Profcast). Get the fantasy info you want at www.baseballprof.com.

No comments:

Post a Comment