Week 1 review

So my fantasy teams are doing ok, my head to head team more than my roto league team. The Dodgers, meanwhile, are 4-2. Let's see how.

Game1: Penny vs Zito, W 5-0.
Gamescores: Penny 67, Zito 35
Zito got destroyed

Game 2: Lowe vs Cain, W 3-2
Gamescores: Lowe 50, Cain 66
Furcal scores from 2nd on a grounder to Ray Durham to win the game.

Game 3: Kuo vs Valdez, L 1-2
Gamescores: Kuo 57, Valdez 60
Rain delays scratch Billingsley vs Lincecum, although both appear in relief.

Game 4: Kuroda vs Germano, W 7-1
Gamescores: Kuroda 71, Germano 67
Kuroda goes 7 on 77 pitches, Dodger offense comes alive

Game 5: Penny vs Peavy, L 1-4
Gamescores: Penny 38, Peavy 86
Doesn't that say it all? Peavy goes the distance allowing two hits.

Game 6: Lowe vs Young, W 3-2
Lowe 61, Young 64
Guess Hu beat Hoffman?

The Dodger bullpen was really the unsung hero. The offense didn't do a whole lot, and the starters were outdone for the most part.

April Report

Time for a bunch of useless summary statistics, including tonight's game since the season started April 2.

Overall Record
16-11
Record with 3 or fewer runs scored:
4-8
Record with 3 or fewer runs allowed:
11-1
Record with 3 0r fewer runs scored by both teams:
4-1
Record in 1 run games
6-3
Extra Innings Record
2-0
Double digit run games, allowed:
2, 0
Shutouts
1
Times Shut out
0

Wins
Average Game Score: 58.8
Minimum: 36
Median: 61.5
Maximum: 76
Average IP: 6.15
Average Runs Scored: 5.81
Average Runs Allowed: 2.25

Losses
Average Game Score: 37.7
Minimum: 20
Median: 45
Maximum: 50
Average IP: 5.24
Average Runs Scored: 2.73
Average Runs Allowed: 6

Pitchers by Mean Game Score
Hendrickson (2 starts): 58.5
Penny (6 starts): 57.5
Tomko (4 starts): 51.5
Wolf (6 starts): 49.7
Lowe (6 starts): 45
Schmidt (3 starts): 39.7

Jason Schmidt - A cut above?

The first thing to notice is that I did these on my desktop in Excel rather than on my laptop in Minitab. The second thing to notice is that Schmidt is a much better pitcher than most of the Dodger Starters. For one thing, he's got a much better strikeout rate, even after last year featured his worst numbers in that category. As a result, he peaks a lot higher. His greater durability will help this area as well. What is particularly good is that in the time series chart, it is evident that he does not stay below 40 for long, meaning he does not slump much, but in fact he tends to stay mostly in above average territory.

Schmidt has some health concerns, and his injury would hurt the Dodgers the most. Also, Schmidt has to adjust to his new team. This will mean more offense (no difference here unless he pitches differently), but a new defense (better? we'll see), along with different park factors for home games (which may prove unfavorable). However, if he is healthy and maintains this level of performance, he should be considered the staff ace.

Here are the descriptive stats:

Starts: 32
Mean: 55.22
SE: 2.76
Min: 20
Q1: 48
Med: 53
Q3: 65.25
Max: 84

Brett Tomko

Ladies and Gentlemen, Brett Tomko will not be starting for us.

Tomko could be a very good starter, but only if you get him on the upswing. As the time series chart shows, you could run a sine function on when he'll be good. As such, he's really better suited to the bullpen; at times, he's a setup man or fill-in closer. Otherwise, he should be behind Elmer Dessens, and you can do that with a bullpen. The histogram just looks silly on him, as he's just all over the place in terms of performance. He can be pretty good, but for the most part, he's not.

Here are his descriptive stats:

Starts: 15

Mean: 46.20

SE: 4.61

Min: 21

Q1: 28

Q2: 47

Q3: 61

Max: 74

Brad Penny

Penny's histogram shows some bimodal data, and with a slight skewing to the right.

The time series plot is a bit more telling, as it reminds us that Penny performed worse in the second half of the season, when he was playing through some pain. As a result, he pitched basically without a breaking ball, and did not peak nearly as often as he did with one.

Penny showed some excellent potential, though, particularly early in the season. In August and September, though, he demonstrated his fallability as well as his poor health. Could Brad Penny lose 30 lbs and become a health nut, he could likely be one of the best pitchers in the game. As it is now, though, he's that frustrating C+/B- gifted kid.

Here are his numbers:

Starts: 33

Mean: 49.76

SE: 2.80

Min: 24

Q1: 35

Q2: 53

Q3: 63

Max: 76

The histogram shows the skewness, but what is really telling is that the mean and median are more than a standard error apart.

More: The LA Times has a bit more after talking to Brad Penny. This I found particularly insightful:

A split-finger changeup Penny developed last spring gave him a devastating third pitch to go with his 96 mph fastball and sharp curve. But he stopped using the splitter at mid-season because he was afraid it might lead to arm problems.

"It's hard on the arm and I shied away from it," he said. "Then I had trouble with my curve for a while, so I was out there throwing one pitch. It's tough to get by just with your fastball."

The rest of the article is here. Also encouraging is that Penny lost 10 lbs in the offseason. A step in the right direction, certainly. I wonder if he'll start throwing a cutter with that fastball and curveball. . .

Billingsley's Charts

At a mere glance, Billingsley did not appear to have skewness, but a somewhat more symmetrical distribution. Further, he did not fare as well with several shorter starts toward the end of the season. Nonetheless, Billingsley put together a decent performance, and for the most part displayed a general upward trend in game scores.

Starts: 16
Mean: 50.13
Min: 26
Q1: 40.75
Q2: 50
Q3: 58.5
Max: 79
P-Value: 0.686

Billingsley had the distinction of pitching the best game, by game score of any Dodger starter this season, topping Greg Maddux's masterful 8-inning performance on 68 pitches. Being 22 years old suggests that he will get better, as well. Especially considering how late in the season it was before his control finally started to look respectable.

The Low-Down on D-Lowe

The above graphs are a representation of Derek Lowe's starts over the 2006 season. The graph on the left, the histogram, shows a clear left-skewed tendency, demonstrating that while the majority of Lowe's starts were good, he also demonstrated the ability to make very bad starts. The graph on the right is simply a time series plot, which shows his game score by start number. The significant dip in the graph is where Lowe faltered in July. A bit more quantitatively, here's how he did:

Starts: 34
Mean: 51.59
SE Mean: 3.04
Min: 9
Q1: 41.25
Q2: 52.5
Q3: 67
Max: 75

Additionally, I ran a two-sample, two-tailed t-test against the Dodgers overall, and the observed p-value was .378, so Lowe's starts were not significantly different from an average Dodger start, which one would expect from the guy who made the most starts for the game.

But apart from that awful July, which was keyed by that nightmarish start against the Twins, Lowe had a pretty good year. Looks like a 16 game winner (17 starts with game scores over 52.5), and without that July he might have been a Cy Young winner. He turned in 215 innings, as well.

Lowe also has fairly consistent drop-off patterns, and it looks like the worst obstacle he could face would be a repeat of July. Apart from that, he's one of the top starters in the game.

Game Scores From this Past Season

Bill James' Game Score statistic is a nifty little metric to use for pitchers to judge their dominance. ESPN typically includes it with their box scores, but you can compute it yourself fairly easily. The way it works is the formula

Gamescore = 50 + (number of outs recorded) + 2(innings completed after the 4th) +
(strikeouts) - 2(hit) - 4(earned runs) - 2(unearned runs) - (walks)

Seems simple enough. So how did last year's Dodgers stack up? Well, here were the basic stats for the team as a whole:

Descriptive Statistics: Dodgers Overall
Games: 162
Mean: 48.64
SE: 1.32
St Dev: 16.86
Min: 9 (Derek Lowe)
Q1: 34
Med: 51
Q3: 63
Max: 79 (Chad Billingsley)

And here are the means for the individual starters:
Kuo: 58.00
Stults: 56.50
Maddux: 53.42
Lowe: 51.59
Billingsley: 50.125
Penny: 49.76
Sele: 47.93
Tomko: 46.20
Seo: 45.20
Hendrickson: 42.00
Odalis: 33.25

Despite having a better ERA for his limited starts than Kuo, Stults did not come out as dominant as the Taiwanese power-pitcher. My intentions are to discuss players in depth more for season preview purposes as time passes, and I'll focus on those players who could still be Dodger starters. So that means no Maddux, no Sele, no Seo, no Odalis, and since I want to use relevant defensive info and park factors, no Schmidt and no Wolf either. I'll have something to say about them, though.

The Superbowl is Sunday, so it's definitely just about time to start thinking about baseball.