Wednesday, March 18, 2009
THT Fantasy MailbagPosted by THT Fantasy Mailbag at 2:27am
Hey guys, I am in a weekly, 12-team, 5x5, head-to-head, mixed league with a rolling waiver system. We have put more emphasis on relievers this year, requiring each team to roster three. We also roster four SP, two P, five BN and two DL spots. Assuming all the closers are drafted first, do you feel I should look at guys with questionable stuff who are next in line to close or vulture saves, or should I concentrate on low-ERA/WHIP, high-K middle relievers? I am confident I will be able to snag some free agent closers during the season and the middle relievers could help with three other categories in the meantime. However, I don't want to lose too many games early on by not getting enough saves. They may not be mutually exclusive either, so are there some four-category value picks you would recommend?
If the price is right, I would strongly consider one of the low-ERA/WHIP, high-K middle relievers in place of a mediocre closer-in-waiting. Assuming all the true closers are gone, I don't see the value in starting a mediocre closer-in-waiting guy in until he is actually closing. You may want to bench such a player in case he becomes a closer, but I wouldn't start him. In a HTH league, you can also go on a match-up basis. If there's a week where one save is likely to be the difference in that category, you may want to press on the saves and pass on the ERA (for instance). But in general, I would go with the three-category guy rather than the zero-category guy. Some four-category guys, though, are JJ Putz and Jensen Lewis.
- Jonathan Halket
I’m a big fan of the site and I’m hoping someone can help with me in my first-ever keeper league. I adopted my friend’s team from last year. Although he knows nothing about sports, he somehow ended up with a decent team but still finished near dead last. So I’m left with five picks from a decent talent pool. I have three sure-fire keepers—Braun, Wright and Sabathia. With my last two keepers, I need help picking from a pool of Carl Crawford, Adrian Gonzalez, Adam Dunn, Joey Votto, John Lackey and Jose Valverde. As of now I am leaning towards keeping Lackey and Valverde.
Unless pitching is scarce in your league, I am going to disagree with your decision to keep Lackey and Valverde. In Lackey, I see a pitcher toward the end of his peak with a declining K rate. His ERA might be slightly lower this year than last assuming a regressed HR/FB rate (15.3% in 2008), but in a couple of seasons I can see Lackey as only a slightly above-average starting pitcher. Valverde is a good closer with a secure job, but my feeling is that the saves he would accumulate could easily be replaced by drafting a cheap closer late, someone like Joel Hanrahan or Chad Qualls.
On the other hand, a hitter's stats like Crawford's could not be easily replaced. At 28 years old, he should still be able to produce at a high level a few years down the road, and his 2009 numbers should more closely resemble his 2007 numbers than his disappointing 2008 ones. Then you are left with two players to consider: Votto and Dunn. I like both players a lot, but in my opinion Votto gets the nod, especially in a keeper league.
- Paul Singman
Hey guys, love the site and all the info. I have a question:
I'm trying to take raw projections and use them to calculate a single-number value. Can you give any guidance on how to calculate a value with the five different categories? For instance, how do you calculate the difference in value for two players below:
A: .280, 20/90/90/10
B: .300, 15/80/80/20
I seem to remember using stand deviation and other tools but can't for the life of me remember how to calculate this out. Any help would be appreciated.
Here's what I know some people do. Take the entire pool of hitters for your draft, so if you have 12 roster spots for hitters and 10 teams, you're looking at the top 120 hitters. Throw their projected stats into Excel, and calculate the standard deviation for each of the fantasy categories using the =STDEV function, as well as the average of each. You'll find something like an average of 16 HR, and a standard deviation of 8. You can use this to calculate a player's z-score in HRs, which is simply their projected HR minus the average, divided by the SD. So if you have Pujols projected for 36 HR, it would be (36-16)/8 = 2.5. Add up the z-scores across all categories and you'll get a sum for each player, all in like units. It works for pitchers too—a hitter with a sum of 5.0 should be worth the same as a pitcher with a sum of 5.0.
It gets tricky for the rate stats, however, because playing time is not factored into the stat; certainly, a 3.50 ERA for 200 innings pitched is more valuable than a 3.25 ERA for 20 innings pitched. So for BA, take the average of all your hitters again—let's say .275. Multiply each player's predicted at-bats by .275 to get the average number of hits he would get, and then subtract that from their actual projected number of hits. So if I have Ichiro at 220 hits in 650 AB (a .338 BA), his score would be 220 - (.275 * 650) = 41. That final number is often referred to as xH, $H, or something like that. It's that statistic that you'll want to find the average and standard deviation of, in order to calculate the z-scores for player BA's. Of course, a shortcut to all of this is simply to assume most top fantasy players get about the same number of ABs, and you can simply find the average and standard deviation of their batting averages themselves. But, this will slightly undervalue high batting average, high at-bat guys like Michael Young.
- Michael Lerra