November 22, 2013
Being Smart about Sits and Starts
Most of our writers didn't enter the world sporting an @baseballprospectus.com address; with a few exceptions, they started out somewhere else. In an effort to up your reading pleasure while tipping our caps to some of the most illuminating work being done elsewhere on the internet, we'll be yielding the stage once a week to the best and brightest baseball writers, researchers, and thinkers from outside of the BP umbrella. If you'd like to nominate a guest contributor (including yourself), please drop us a line.
Larry Schechter is a six-time winner of the renowned Tout Wars experts league and a winner of the USA Today-sponsored League of Alternative Baseball Reality (LABR). He is also a two-time winner of the CDM Sports national salary-cap challenge. In his new book, Winning Fantasy Baseball, Larry discloses the methods behind his success. The following is an excerpt from the In-Season Management chapter, and applies to all formats of fantasy baseball.
Winning Fantasy Baseball is now available for pre-order at Amazon.com and BarnesandNoble.com, and will be in bookstores in the U.S. and Canada on January 7th, 2014. More information, and early reviews, can be found at www.winningfantasybaseballthebook.com. You can follow Larry on twitter @LarrySchechter.
Weekly Lineup Decisions
Mono-league players sometimes keep an injured player in their lineup rather than replace him with someone they perceive will do more harm than good. And if a player’s dollar value is negative, that would appear to be the case. However, even players with a negative dollar value usually make a positive contribution toward gaining points in the standings, so you’re better off using them. (This is explained in chapter 6.)
I don’t like trying to time hot and cold streaks, since you never know when a player will suddenly turn around a bad streak or end a good streak. Trying to time this is like day-trading stocks: you might get lucky or unlucky. I’ve never done an analysis of all players’ weekly performances, but I suspect you would just find a lot of ups and downs, without much rhyme or reason. Sure, sometimes a player will have an extended hot or cold streak, but mostly the streaks are short and change quickly and without advance notice.
Having said that, I’m going to contradict myself and say that there are times when a player is obviously struggling and it’s probably best to bench him until he figures it out. If someone’s mechanics or timing are out of whack, they usually don’t just get it back the next day. But again, I could give you a million examples where this approach has backfired. Carlos Gomez began the first two weeks of 2013 hitting 6-for-37, a .162 average. I then benched him on my CDM teams and he went 11–for-18 the next week.
As always, you want to get the most value possible. Suppose you’re in a salary-cap game, and you have Brandon Phillips valued at 305 and Neil Walker at 284. They have identical salaries, therefore Phillips is the best choice. However, what if Walker is playing seven games and Phillips only six? In that case, Walker’s relative value is 7/6 × 284 = 331, and he is a better choice than Phillips.
I’m assuming here that Walker will play every game. There’s always a chance that he’ll get a day off, especially during the dog days of summer, so to be more precise you might want to multiply his games by 6.5 or 6.7 or so.
Some people might also take into account what parks Phillips and Walker are playing in and which pitchers they are facing. I usually don’t get that complex, unless it’s something obvious, such as playing at Coors Field. This doesn’t mean I automatically play a hitter just because he’ll be at Coors or bench him because he’ll be in San Diego. To be precise, you need to project the stats you think the hitter will produce, and the resulting value. For example, my 305 value for Phillips might be lowered to 270 for a week in which he’s playing half his games in San Diego.
Catchers are a little tricky. Even if they have seven games scheduled, they almost certainly won’t catch seven days in a row. You can assume that National League catchers will play only six. American League catchers may still play seven by being used at designated hitter. Also, keep in mind that for most catchers, all of these days off were factored into their original projection. Unlike full-time position players, who will get 550–600 at-bats, a full-time catcher would typically be projected for only 475–500 at-bats (possibly 525–550 if he’ll also be a designated hitter).
Early in the season, I do consider weather. A team may be scheduled for six games, but the weather forecast could be so threatening that a rainout or two is quite possible. Colder weather is also bad for hitters; a guy who typically hits .300 with 25 homers might be a .285 hitter with 18 homers if he always played in 50-degree weather.
For pitchers, taking matchups into account is, obviously, much more important. Some choices are easy. If a pitcher is similar to several other available arms and he has one upcoming start at Coors Field or in Texas, then bench him. If he has one start at a good pitcher’s park against a weak-hitting team, then start him.
I don’t think I’m saying anything brilliant here; this is all pretty obvious. My only words of wisdom would be to advise you not to overthink and overanalyze matchups. I know some fantasy players who go crazy looking at home and away splits, pitchers’ historical records against other teams, etc.
The ideal way to make a weekly lineup decision is to consider the likelihood of all possible outcomes, create projected stats, and convert that to a value.
You could take a lot of time each week analyzing all the pitching matchups and home/away splits and driving yourself crazy coming up with an exact figure, or you could make this process a little quicker. Here are a few examples…
No Designated Hitters at National League Parks
You calculate that if he starts two games, he’ll probably get eight at-bats. Most likely, he’ll get to pinch-hit a couple of times as well. So you take his current pace, but adjust it downward to reflect only 10 at-bats for a week (which would be approximately 260 for the year).
For this week, his projected value is $7.7. It’s interesting to note that because of the way value formulas work, if a player gets half the playing time and stats, it doesn’t mean his value will be cut in half. In this case, his value is less than half. If we want to get more complicated, we can consider how he hits on the road, what ballparks he’ll be in, what pitchers he’s likely to face, etc., and can further adjust the projections based on all of that. But we’ll keep it simple and just give him the $7.7 projection. If you’re in an AL-only league, you’re unlikely to have a better replacement. If you’re in a mixed league, you might.
Bad Pitching Matchup
In this case, your guess was six wins, leaving Niemann with a $6.9 projected value for the week. Again, if you’re in an AL-only league, you probably don’t have a better option. If you’re in a mixed league, you might.
If you think Niemann tends to pitch better at home than on the road, you could also adjust his ERA and WHIP downward, and perhaps his IP and strikeouts upward. You can also take into account how good a hitting team Detroit has.
Pitching at Coors Field
Minor’s original projection is shown here. Starting with that, you take a look at his home and road splits if you care to, and you consider how much better the Rockies’ hitters are at home. If you want to spend a fair amount of time analyzing this, you can. If you want to keep it pretty simple, you can just make an educated guess, which is what I’m going to do here. Considering Coors is the best hitter’s park in the league, I’ll estimate that Minor, if he pitched every game at Coors, would do this:
This is by no means precise. I’m just making an approximation.
Since San Diego is typically the best pitcher’s park in the league, you can make a similar guess for Davis. His ERA and WHIP will undoubtedly be better than normal. Since you expect his $7.2 value to rise, and Minor has been lowered to $3.9, you don’t need to bother estimating Davis’s exact numbers, because you’ve already concluded that he’s the one you’re going to start.