fantasy strategy: a case study in lineup construction

[NT!]

warning: long. skip to the bottom for a case example if you like.

I split this off from the day-to-day thread hoping that it would attract some more general thoughts from both serious and more recreational fantasy players.

I am trying to look at which of these factors you weigh most heavily in deciding which hitters to sit or start on a given day:

1. batting vs. SP (over what sample size?). or batting vs. a team in general.
2. l/r split, home/away split, pre-ASB/post-ASB, or monthly splits
3. last 7 / last 15 / last 30
4. long term performance (season, career)
5. opposing pitcher's splits
6. need for a particular stat

Obviously if there is one split that is particularly huge, you go with that (i.e. Chad Tracy sucks huge balls against lefties). But if it's close, is there one stat that you lean towards? Which one do you look at first?

I tend to look first at how the hitter has done over the last week or two, and I try to know the L/R splits for my hitters. Then I look at who the opposing pitcher is, how he has done against the hitter's team, and how he's done lately.

What I tend not to check out enough, I think, are home/away splits for hitters and pitchers. I also sometimes don't check a SP's splits vs. L/R which I think is important, and I don't know how good some bullpens are.

What I really don't have a strong idea of is when I need a particular stat badly enough to sacrifice some other areas. Generally, I have been going for the players who bring the most value to the table on the whole, because I am generally pretty close to the people ahead of and behind me in most statistical categories. I have never found myself in a situation where I would sit a guy who is a four-tool threat for a guy who is just an average/steals guy.

Here's one where I've left the names off, three guys I am thinking of starting at UTIL tonight. I am in third, am close in all statistical categories, but the two guys ahead of me are beating me on HR, RBI, R, OPS and one is punting steals. (6x6, also includes AVG).

Last 7:
Player 1: 3R 2HR 2RBI 0SB .316 1.000
Player 2: 5R 0HR 3RBI 1SB .391 1.005
Player 3: 4R 3HR 9RBI 0SB .348 1.206

Last 15:
Player 1: 8R 2HR 6RBI 0SB .262 .795
Player 2: 13R 3HR 8RBI 1SB .413 1.147
Player 3: 6R 4HR 12RBI 0SB .324 1.158

Batting vs. Pitcher:

Player 1: 3/12, 2HR, 5RBI, 1.083 OPS
Player 2: No ABs
Player 3: No ABs

Opposing pitcher, this season:

Player 1: 1-6, 8.56 ERA, 1.85 WHIP, LHB: .940 OPS against
Player 2: 2-2, 5.88 ERA, 1.35 WHIP, RHB: .583 OPS against
Player 3: 2-0, 4.43 ERA, 1.61 WHIP, RHB: 1.063 OPS against

Season:

Player 1: 54/217 40R 6HR 30RBI 1SB .249 .735
Player 2: 74/283 39R 8HR 34RBI 8SB .261 .749
Player 3: 56/231 30R 12HR 34RBI 0SB .242 .818

Other info:

Player 1: Performing well beneath career numbers, heating up over the last week, may be slightly hurt, should be hitting leadoff in a very strong lineup. No major split vs. L/R pitchers. Opposing team has a strong bullpen. Playing in a hitters' park.

Player 2: Long history of being a slow starter, typically heats up this time of year, usually hits 4th or 5th in a below-average lineup. Opposing team has a crappy bullpen. Playing in a hitters' park.

Player 3: Aging slugger, slumping until recently, no major monthly splits or 1st/2nd half. .949 career OPS vs righties. Hits 4th in fairly average lineup.

I was going with Player 1, but am now thinking I should go with Player 3.

[Vyse]

ESPN's Daily Notes [img]/images/graemlins/cool.gif[/img]

[Benholio]

1. Pretty much never. Hard to get a high enough sample size to draw any conclusions from batter vs. hitter, and batter vs. team is pretty useless.

2. L/R split is totally relevant if there is enough difference between the 2. Pre-ASP/post-ASP might be marginally useful, but I don't buy into that much. Monthly splits are pretty useless.

3. I try to avoid playing streaks and slumps. Maybe if a player is in a really bad slump I will take that into account.

4. Season/career: Most of the decision comes from this.

5. Opposing pitcher's splits: The rest of the decision comes from how good the pitcher is.

6. The league I've been in for the last 5 years is points based, so I can't speak on roto-stats.

Basically, how good is the hitter (career/projections for this year), how good is the opposing pitcher, then righty/lefty splits. The rest is a waste of time to me.

[Vyse]

Why would you not play a solid player having a great hot streak for minimal investment? Silly.

[Kneel B4 Zod]

I pay a lot of attention to L/R matchups, b/c they are relevant and b/c they will sometimes affect playing time decisions.

now I won't site Prince Fielder against a lefty, but I'll sit a Chad Tracy

[Hoya]

Do you guys usually have big enough benches for this to be particularly relevant?

[NT!]

usually only comes up for me at UTIL or in the OF, otherwise no

[mbillie1]

I think overall skill is #1 (eg I will always start Beltran and Ordonez over say, Chris Duffy or someone). If skill is close (I have Atkins and Zimmerman both at 3B) I will play the hot hand / better season production. I don't worry about the day to day splits... I usually carry a ton of pitchers and only 1 (or 0) bench players anyway. I am adding Sosa to the mix so he'll be my bench OF... and am pretty much only going to start him if one of my regular of (Rios, Beltran, Ordonez) is hurt, slumping terribly or has a day off.

[NT!]

btw, in case anybody cares, the players were jd drew, mike cameron, and frank thomas. i went with drew because weaver (the seattle starter) sucks and the sox usually mash him, so drew should have some run scoring / run producing chances. also he has two HR in 12 at-bats, which is a small sample but i put a little more stock in it if it's power numbers rather than just hits falling in.

[THAY3R]

L/R matchups and opposing pitchers ability.


Recent performance aka slumps streaks and batter v pitcher stats is very bad, do not rely on that.