Closer by Committee

Closer by committee, a highly used term, is a terrible misnomer.  I get that people people like alliteration, but an actual closer by committee would have to involve 3-15 pitchers standing on the mound at the same time, using Robert’s Rules of Order to determine which pitch to throw.  The ninth inning would take hours to play, as committee members jostle for positioning, form coalitions, file motions, second motions, debate, vote on cloture, file amendments, debate amendments, file motions to limit debate on amendments, etc.

I think every team should go closer by committee.

Link of the Day (Month? Year? Bueller?)

http://www.gocomics.com/feature_items/printable/564062?feature_id=13

Derek Jeter’s New Contract

Derek Jeter just signed a five 5-year, 450 million dollar contract extension with the Yankees, plus performance bonuses.  For each season Jeter fields at least fifteen ground balls, he earns an addition $20M.  If he reaches 50 hits in a season, he earns $1M, then an additional $1M for each hit beyond the 50th.  If selected by fans to appear in an all-star game, he earns $5B, plus another $5B if he plays in the all-star game.  If the Yankees make the playoffs, Jeter becomes owner of the Yankees, and also becomes owner of any team the Yankees defeat in the playoffs.  And if the Yankees win the World Series, Jeter becomes President of Earth.

When reached for comment, Yankees fans said, in unison, “It’s Derek Jeter.  Pay him whatever he wants.”

Football Endgames

Football coaches don’t think nearly enough about the amount of time they’re leaving on the clock when they score near the end of halves.  Time left on the clock is something they have a lot of control over.  And it has a considerable effect on the result of the game.  And yet, it seems like coaches bungle these decisions even more often than when deciding whether to go for it on 4th down.

For example, suppose a team trails by 5 with first and goal at their opponents’ 1-yard line and 0:50 left on the clock, their opponent having three timeouts.  Consider the result of three outcomes:

1. The team scores a touchdown on first down, leaving their opponent 0:47 and three timeouts.
2. The team scores a touchdown on second down, leaving their opponent 0:42 and two timeouts.
3. The team scores a touchdown on third down, leaving their opponent 0:35 and one timeout.
4. The team scores a touchdown on fourth down, leaving their opponent 0:30 and zero timeouts.
5. The team turns the ball over.

For each of these scenarios, you can estimate the team’s expectation of winning, based on the skill of their defense and the opponent’s offense.  For scenario 5, their expectation is zero, or very close to zero.  For scenario 1, their expectation may be as low as 0.3.  For scenarios 3 and 4, their expectation will be much higher, even assuming their opponent’s ability to move the ball.

Any yet, you never see a team in this situation call a running play that’s designed not to score.  They always try to score immediately, then seem overjoyed when they hand the ball–and the game–back to their opponent.  While the specific example above may not happen often, the thinking applies any time a game is close at the end of a half.  Teams should be thinking both about how many points they’re trying to score, and how much time they’re leaving on the clock, as both are significant determinants of the final score of the game

I can think of explanation why coaches do this.  There’s the classic risk aversion argument used to explain coaches’ timidity on fourth down.  Furthermore, there’s a certain gamesmanship going on; it looks bad when a coach is looking the game and appears to be not trying hard to score.  Players themselves may resist the idea of running plays designed not to score points but rather to run off the clock.  But there are a lot of game swung by teams leaving their opponents too much time on the clock.

Identity of Hurricane Kawasaki Revealed

I’ve written a fair bit about my mysterious fantasy teammate / co-manager Hurricane Kawasaki.  While I was careful to hide Hurricane’s true identity, it should now be obvious, based on statements made on this blog and off it, that Hurricane Kawasaki is indeed: Mr. Curt Schilling.

Drafts versus Auctions

I haven’t played a ton of auction fantasy leagues, but my feeling is they’re a considerable improvement over the standard snake draft.  However, I suspect auctions would make for an even greater improvement to actual drafts.  (Hat tip: football outsiders)

What’s the toughest division in baseball? (continued)

My co-manager Hurricane Kawasaki hit me over the head for my last post:

Based on that argument, you can argue that any pitcher coming up from the minors into the AL east would be facing an “easier” league and thus his poor performance should not be because he’s facing better hitters.

I’d like to clarify what I was trying to say, and how I think this analysis should be used.  Also, I think there was a key piece of information missing from the analysis.

For the analysis, I used baseballmonster‘s ease rankings to rate the difficulty of pitching in each division.  As Hurricane points out, the way ease rankings are calculated depends both on the quality of hitting in a league and the quality of pitching.  So if the AL East has the best hitters, but also has really good pitchers, it will appear as if it’s a pretty good division for pitchers, when really it’s a very tough division that happens to have some incredibly good pitchers.

In the post, I used the analysis to suggest that Javier Vazquez’s pitching performance shouldn’t be tied to his league change.  As Hurricane says by email, “Javy might have been the best pitcher in the NL east, but he’s at best the fifth best on his own team right now.”  My analysis was highly dubious.  Of course, I find Hurricane’s assessment that Vazquez is pitching on par with his AL track record to be equally dubious.  Vazquez’s ERA is currently 9.82.  His worst ERA over the course of a season in the AL is 4.91.  I think it’s perfectly fair to call that struggling.

The better application of this analysis is in thinking about where to find pitchers to add to your fantasy roster.  If the divisional ease ranking of the AL West and NL Central are favorable to pitchers, as I suggested last post, then there are presumably pitchers getting good results in those divisions.  Whether they’re getting good results because they’re strong pitchers or because they’re facing weak hitters is somewhat immaterial.  It’s not completely immaterial since there’s plenty of inter-divisional play, but it’s nonetheless valuable to know which divisions have weak hitting relative to their pitching.  That seems to be the case for the AL West and NL Central.

One question worth addressing, however, is how many starters are owned in each league?  The 10th best starter in the AL West might be more valuable than the 10th best starter in the NL West, but they may not both be available.  Who’s owned differs by league, but here’s the breakdown for mine:

• AL West: 0.48 ease rating, 13 SPs owned.
• NL Central: 0.42, 12 SPs owned.
• AL East: 0.10, 18 SPs owned.
• AL Central: -0.23, 14 SPs owned.
• NL West: -0.37, 13 SPs owned.
• NL East: -0.39 , 11 SPs owned.

(Note: injured players who haven’t yet pitched and relief pitchers who qualify as a starter are excluded from these counts.)

This information tells a different story.  The AL East has a more favorable ease rating for pitchers, but 18 starters are owned, far more than any other division.  This suggests that the starting pitching is indeed stronger in the AL East, skewing the ease rating.  If you’re looking for a spot start, chances are the 15th best starter in the AL Central, or the 12th best in the NL East is a better bet than the 19nd best starter in the AL East.

This analysis probably needs some further hammering out, but I’d nonetheless look to the AL West and NL Central as the best divisions for pitchers.

What’s the toughest division in baseball?

I’ve been playing around with the Ease Rankings feature on Baseball Monster, trying to figure out which divisions are toughest to pitch in.  I’m focused on starting pitching, so I customized league settings to look only at wins, strikeouts, ERA and WHIP.  Baseball teams play divisional opponents more than anyone else, so I averaged the Ease Rankings by division:

• AL West: 0.48 (most favorable to pitchers)
• NL Central: 0.42
• AL East: 0.10
• AL Central: -0.23
• NL West: -0.37
• NL East: -0.39 (least favorable to pitchers)

Going a step further, I created pitcher ease ranking for each team by excluding the team they pitch for from the average.  The analysis shows that the easiest teams to pitch for are Oakland, Texas, and St. Louis.  The toughest teams to pitch for are San Diego, Atlanta, and Cleveland.

It’s early, and these rankings are likely to change, but they’re pretty interesting and have interesting implications both for fantasy and real baseball.  For one, don’t blame Javier Vazquez’s early struggles on his league change; thus far, the Yankees’ divisional rivals haven’t been hitting the ball nearly as well as those of Vazquez’s former team, the Braves.

Loss Aversion in Fantasy Baseball (continued)

I was already planning to follow up yesterday’s post with some reasons why loss aversion might exist in fantasy sports, when reader Joe makes the following comment:

The goals of fantasy sports isn’t just winning; there are two: winning and maximizing bragging rights.

Picking up some new dudes and saying, “I KNEW that Whosthat McRandom was going to be awesome” is much more awesome than bragging, “I KNEW that Oldnewsy von Fartington still had another year in him.”

This is a decent summary of a casual attitude towards fantasy sports. “I probably won’t win anyway, so I might as well have some fun along the way.” It’s not my attitude, but I’m not a casual player–I’m an extremely competitive one. In my view, there’s no point in bragging if you’re losing. If you want to brag, play a strategy most likely to win, then brag when you win. Find creative ways to brag if need be. Back in 2003-2004, I KNEW that Kerry Kittles still had another year in him when he helped me win a 20 team fantasy basketball league.*

*I didn’t actually know this, so much as strongly suspect. The following season, Kittles played 11 games off the bench, then retired.

However, there actually is a case to made for picking cagey veterans over hotshot sleepers, from a bragging perspective, which is that no one else is doing it. Over at Baseball Analysts, Sky Andrecheck argues that experts should make recommendations based on how their opinions set them apart from other experts. It’s not hard to find an expert who’s big on Jason Heyward or Gordon Beckham; it is hard to find an expert who likes Johnny Damon and Chipper Jones over those two.

Throughout the season, I’ll check in on the veterans I praised and prospects I questioned in yesterday’s post and compare their stats. If my picks pan out, no doubt I’ll brag relentlessly.

Finally, I’d like to generalize the point I’m trying to make about young and old, and, just for fun, I’ll do it in the form of a delta-epsilon definition:

In fantasy baseball or basketball, for every set of players without proven track records ε, there exists a set of players δ who play the same positions, are older, and are deemed worse by fantasy experts, such that the expected value of δ exceeds the expected value of ε.

Loss Aversion in Fantasy Baseball

In behavioral economics, loss aversion refers to a bias where people prefer a chance of gain over the combination of locked-in gain and chance of loss, even though they have the exact same results. For example, consider the following cases:

Case 1: Choose between receiving $100, or having a 50% chance of receiving $200 dollars.
Case 2: You receive $200. Then, you choose between losing $100, or having a 50% chance of losing $200 dollars.

Mathematically, the two cases are identical. Yet experimentally, when presented with offers as written, people tend to prefer the second option in case 1, and the first option in case two. Basically, people like the feeling of upside risk, knowing they stand to gain extra, but don’t like the feeling of downside risk, or knowing they might lose extra.

In fantasy sports, this bias pops up in the strong preference for younger players over older players. For instance, consider these two options:

Option 1: A young outfielder, with little major league experience, has a 50% chance of becoming a very good player this year, but otherwise will have little or no value.
Option 2: An old outfielder, who has been very good for many years, has a 50% chance of providing little or no value this year, but otherwise will remain a very good player.

Again, these two options are mathematically identical, but there’s a significant preference for Option 1 among fantasy players. Spotting these opportunities is a key way to gain edge in fantasy sports, and a key reason why my teams tend to include players at the tail-end of their career.

For instance, this year I have one team with an outfield of Bobby Abreu (36 years young), Raul Ibanez (38) and Johnny Damon (37), with Chipper Jones (38) and a banged-up Lance Berkman (34) slated to share my utility spot. Some of those guys won’t pan out. But I’m just as confident in this group as I would be in an outfield of Jason Heyward (20), Hunter Pence (27), and Chris Coghlan (25), with Gordan Beckham and Billy Butler (both 24) at utility. And my team came much more cheaply, allowing me to slot more reliable players elsewhere.

I’ll check back on this question throughout the season, but right now I feel pretty good.