Archive for Meta Analysis

Necessity is the mother of invention. –Plato

I wanted to know how owners were valuing Michael Brantley’s playing time. Currently, at NFBC, he is going 233rd overall in NFBC drafts. Over a full season, he is projected to be more productive than the two outfielders going right before him, Carlos Beltran and Randal Grichuk. Owners, via calculations or their gut, are significantly downgrading a full season Brantley. But by how much? I needed to find the league replacement value.

I could go through all the whole league setting and final the values like I did for my Tout Wars league. While I recommend this detailed procedure for any league an owner takes seriously. I was just looking for a quick answer and stumbled upon one while looking over my Fantrax league.

Our friends at FanTrax.com have their players listed with projected stats and ADP. Having both downloadable made a projection sheet quickly come together.

Read the rest of this entry »

Over the past week, I have collected information on how spring training closers battles have worked out from 2013 to 2016. Today, I go over the results. It’s now time to release the tables.

The first set of data shows how the team’s initial closer fared.

Read the rest of this entry »

Today, I continue the comparison of my Pod Projections to Steamer in various fantasy categories, this time identifying players I believe have stolen base upside. My stolen base projections are calculated using a proprietary metric I developed that is revealed in Projecting X 2.0. Essentially, it’s a stolen base attempts per opportunities ratio and I use historical rates to guide my projected rate.

Read the rest of this entry »

The Pod Projections are back! My projections are based on the methodology shared in my eBook Projecting X 2.0, and the process continues to evolve and improve.

2017 Pod Projections Index:
Lance McCullers

Today, I’ll dive into the first hitter to receive the exciting 2017 Pod Projection treatment. David Dahl thrilled us during his debut last year, showing power (.185 ISO), speed (7.7 Spd score, five steals), and batting average ability (.315). That’s literally everything we want as fantasy owners. Of course, let’s not ignore the fact that he required an absurd .404 BABIP to reach that impressive batting average mark. That said, he plays half his games at Coors Field, so perhaps his average has some staying power. Early 2017 NFBC drafters are already falling over each other to roster him, selecting him as the 22nd outfielder off the board (just before Matt Kemp, Khris Davis, and Adam Jones), and 91st overall. WOWZERS! That’s some serious love.

Read the rest of this entry »

Playing time can make or break a baseball player’s fantasy value. An elite player may not finish above replacement level if he suffers an injury and plays only half the season, and a lackluster player could finish above replacement level simply by playing every single day. This is all intuitive, and the fantasy community generally approaches these kinds of things rationally. In other words, most players are appropriately valued, outside of the market inefficiencies that inevitably warp player values.

One-dimensional speedsters — dudes who steal a bunch of bases and do little else — are much harder to peg. Their value is tied up primarily in one category, as stolen bases (SBs) do not directly correlate with other categories the way home runs would with runs and RBI, for example. The issue becomes all the more confounding when one considers the contemporaneous scarcity of SBs relative to home runs. There’s more to value than just SBs and plate appearances (PAs), but the fact of the matter is the two statistics by themselves correlate very strongly with a player’s end-of-season (EOS) value (which, here, are informed by Razzball’s Player Rater).

In the last five years, baseball has seen 75 player-seasons of 30-plus SBs — 15 steals a year on average, a trend that didn’t fundamentally change in 2016 (although that doesn’t mean SBs aren’t scarce). A simple linear regression of SBs and PAs, the latter of which serves as a proxy for other counting stats such as runs and RBI, against EOS value produces a remarkable 0.71 adjusted R²:

Read the rest of this entry »

I love comparing my Pod Projections to Steamer, since mine are all done in a massively time consuming manner by hand, while Steamer is spit out by the almighty computer in mere seconds. It works as a sort of checks and balances system, as I am likely aware of various issues the computer doesn’t know about, while I might have some inherent biases I don’t even realize that the computer won’t suffer from.

So today, I continue my new series pitting my projections against Steamer in a specific fantasy relevant stat category. Last Thursday, I listed the hitters that I was most bullish on for home runs compared to Steamer, so today, I’ll look into the hitters Steamer is most bullish on compared to my projections.

Read the rest of this entry »

Over the last couple of years, I have run the “Steamer and I” series, pitting our hitter wOBA and pitcher ERA projections against each other and discussing the players our forecasts differ the most for. I’m going to do things a little bit differently this year by focusing on individual stat categories, identifying a group of players I’m significantly more bullish on compared to Steamer, and vice versa, in that metric.

We’ll start with home runs. I will be comparing my home run Pod Projections to Steamer, which have been extrapolated to the same number of at-bats I’m forecasting. Today, we’ll look at the hitters I’m most bullish on versus Steamer.

Read the rest of this entry »

Earlier this month, we took at a look at starting pitcher replacement levels for 2017. Today, we will continue this process for relief pitchers using the same methodology. Please refer back to that post as a primer on how I put my replacement levels together, though I’ll recap some of the methodology here.

2017 Replacement Levels: SP

There are two ways that replacement levels can be defined. This is either as a specific point per game (P/G) or point per inning total (P/IP) or as the nth player ranked at a position. For example, I could say, replacement level for RP is about 6.67 P/IP, or I could say that replacement level for RP is about the 70th RP.

Read the rest of this entry »

Nearly a month and a half ago, I shared the names of six starting pitchers who my old xK% metric suggested had the most strikeout rate upside this season, assuming their equation components remained unchanged. I then got sidetracked, introduced an updated version of the equation with new component coefficients and then even played around with incorporating CH% (changeup percentage) into an even newer version of the equation. So I never actually got around to the list of starting pitchers with strikeout rate downside. It’s now time to share those names with you very patient people.

Read the rest of this entry »

Yesterday, I participated in a 12-team NL-only auction league hosted by CBS (full analysis available soon from CBS). To add a wrinkle to the experience, I decided to try to construct an average team. No studs, but especially no scrubs. Just spend as close to $11.3 per player as possible. The main reason for this approach is that I wanted to stay away from the bottom feeders common in “Only” auctions. I was looking for regulars across the board. The strategy fell apart as my fortitude and simple rules failed.

First off, I wasn’t able to do much auction planning since I found out about it less than a week ago. Additionally, I didn’t want to use the traditional spread-the-risk approach of a bunch of $20 players. Mine idea was a No Scrubs approach. With $20 players, several $1 players enter the team. I wanted semi-talented players with jobs for every position.

After creating projections using the SGP method, I had to come up with an auction framework. In their book, Simple Rules, Donald Sull and Kathleen Eisenhardt go over how to create and utilize simple rules. Here their basic premise.

You want to make the rules as simple as possible to increase the odds that you will follow them. You can also limit your rules to two or three … to increase the odds that you will remember and follow them.

All right, I decided to go with just two rules.

1. Targets players between $5 and $17 ($11 +/- $6). I would not be able to get every player for exactly $11, so I was going to need some leeway.
2. Don’t overpay or reach for players.

Read the rest of this entry »