Monday, December 29, 2008

Rox Talk - Tale of Two Pitchers?

Tale of Two Pitchers?

Okay so the column on the left is 2007 and the column on the right is 2008. The top line is the number of strikes thrown by the pitcher and the percentage of the total number of pitches thrown by pitcher (i.e. 63% of pitches thrown were strikes). The next three lines describe the type of strikes and the percentage of the type by total strikes (yeah don't ask me why they are greater than 100%, 43 extra pitches of some sort!). The final five rows deal with what happened to the batter or the ball (Groundball, Flyball, Line Drive, Strike Out, or Other [Walks]) with the percentage based on total batters faced. So basically the percentages are the same except a bit more flyballs and little less strikeouts in 2008.

So what about these next numbers. Again the ratios (ratios used since the 2008 pitcher didn't pitch as many innings) are pretty much the same from 2007 and 2008.

Moving on we start to see a big giant difference in the next series of numbers and that is HR allowed. 5 percent difference in Home Runs in 2008! The percentage with these is simply the percent of hits allowed (runs allowed, HR allowed, BB allowed) per batter faced.

And so we have a pitcher who pretty much had the same basic pitching stats with some more flyballs, little less strikeouts, and some more home runs which led to the following bottom lines (see final tally below). Pretty fascinating that difference between a 17 win season and a 4 win season is so small...was Jeff Francis strikeout pitch in 2007 turning into a homerun pitch in 2008? Did batter's adjust to Francis and Francis not readjust? What did the Red Sox's find during the 2007 World Series that had the rest of the NL looking for?

Monday, December 22, 2008

Rox Talk - 2008 Pitching

This is my crazy pitching blog post, stop if you hate numbers ;-)

Basically Cook and Jimenez were the work horses for the team. De la Rosa was the surprise and savior. The first four pitchers accounted for 75% (2007 had 5 pitchers to account for 75% of the IP) of the starts and overall the starting pitching accounted for 61.7% (2007, 64.1%) of the innings pitched for the entire team. If a quality start is defined as a game score of 50 then only Cook and Jimenez averaged a quality start during the season. Cook's 81 game score was the highest for the team (2007, high game score was 77). Hernandez had the distinction of having the lowest game score of 5 (beating out Fogg's game score of 4 last year)!

In addition to the the above stats, I also looked at some other wacky stuff. One question I have always had was what would be a starter's record if he won (or loss) those games because of relievers (or was bailed out by the offensive or relievers and got the no decision)? Also which pitcher actually pitches with the lead?

So in normal speak the Roxs starters were 54 - 62 with 46 no decisions (versus 55 - 47 and 61 no decisions in 2007). In my revised look the starters were 76 - 74 with 12 ties (versus 84 - 60 with 19 ties in 2007). Cook would have won 18 games. Once again Jimenez was the unluckiest as he left 16 games winning and the bullpen could hang on only 12 times. As I mentioned last year I don't think this revisionist look at pitching stats mean anything because over a pitcher's career my guess all the blown saves or offensive help probably average out. To me the bottom line is what does the team do when a certain Starter is on the mound. When Cookie step to the rubber the Rox won 19 times accounting for 25.7% of the total team wins.

If the above study looked at the underbelly of starting pitching (who has luck, no luck, or just happens to pitch that day the offensive was asleep) this next look is what does the Starter do for themselves (i.e. keeps the team in the game by keeping the lead or keeping the game tied). Of the starters, Jimenez pitched with the lead 70% of the time and Cook pitched with the lead 67% of the time. The team overall pitched 62% of the time with the lead or tied (versus 67% in 2007). Finally what did the starters throw to during the season.

Well as you guessed it Cook led the team in Groundballs and Jimenez and de la Rosa had lots of strikeouts. Interesting to see that Jimenez and Cook have very few flyballs which probably leads to their success.

Monday, December 15, 2008

Rox Talk - Big Bat Theory

Winter Meetings
Boy Howdy, Las Vegas was real exciting especially for the Roxs who picked up 38 year old relief pitcher, Alan Embree, and called it a success (?), at least that is what the email to season ticket holders said. Realistically, I didn't expect the Roxs to pick up Manny Ramirez but where is that 4th starter to shore up the rear end of the rotation (Tim Redding???...ughhh scrap heap pitchers ain't the answer)? Just a thought, but with the money drying up in the free agent market, don't you think that next year Holliday might wish he stuck around for the $85 million the Roxs offered up?

Big Bat Theory?
The Denver Post and Patrick Saunders led off Sunday's paper with an article asking "whether a team can thrive without (a top-flight slugger)"? Anytime a question is thrown out like that and then not back up by anything but baseball antidotes I have to look into the numbers. The obvious difficulty in answering a question like this is how do you classify a slugger? Is it your home run king, your RBI leader, combination of HR/RBI leader, or some other stat like OPS? In the Rox's 16 years, the HR and RBI leader was the same only 8 times (Bichette, Galarraga, Castilla, Helton (x3), and Holliday).

In my thinking and looking at 16 years of Rox's data, I will classify a slugger that hits greater than 20% of their teams HRs or hits in more than 16% of the team's RBIs. So for 2008, Holliday and Hawpe hit 25 HR which individually accounted for 16% of the team's HR (160). In 2007, Holliday's 36 accounted for 21% of the team's total (171). So Holliday was a slugger in 2007 but not so much in 2008. For RBIs, Atkins had 99 or 13.9% of the team's total (714) and in 2007 Holliday had 16.6% (137) of the team's total (823). The All-Time leader is Helton in 2000 when he accounted for 26% of the team's HR at 42 (team 161) and Castilla in 1998 when he had 18.2% of the team's RBIs (144 of 791). So back to the analysis...if you take the years the Rox had a slugger (i.e. > or equal to 20%) then there were 10 seasons and those teams had a collective record of 767 - 791 (0.492). If you move the definition to 22% then there were 5 seasons with a collective record of 361 - 391 (.480). RBI leader team collective records for greater than equal to 16% is 760 - 798 (0.488). From this analysis it would appear having one player collect all of your big hits doesn't really make for a great Rox's team. With the other 5 seasons of not having a "slugger" the Rox were a collective 422 - 550 (.434) so I guess in some manner of speaking having a big bopper might help the Roxs. Unfortunately, the Rox's history isn't probably the best example as they have had only 5 winning seasons! It is interesting to note though that for four of five of those seasons the HR and RBI leaders were the same...

OK so maybe the Rox's history isn't the best place to look. In 2007 and 2008, the MLB average was 18% and 19% respectively for HR leaders with respect to their team totals and 14.3% and 14.5% for RBIs. It is interesting to note that in 2008, only 8 of the 30 teams had different HR and RBI leaders (Marlins, Reds, Cubs, Mets, Rox, Dodgers, Blue Jays, and Mariners). In summary then for the two years noted if you look at collective team records for those that had one batter have more than 18% of the team's total HR then the record was 2431 - 2429 (.500) and evenly split with 30 teams having one batter > 18% vs 30 with no batter > 18% of total team HR. For RBIs when one batter with more than 14% of the team's RBI total then the record was 2372 - 2325 (.505) with a 29/31 split. If you increase the HR to 22% then you have a record of 987 - 957 (.508) with a 12/48 spilt and increase RBI total to 16% you have a collective win total of 1226 - 1205 (.504) with a 15/45 split. At a 22% HR and a 16% RBI slugger, than these teams would have 10 - 8 more wins above .500 respectively. Note if you modify my statement above and change a slugger definition to > 22% then the Rox had 5 seasons with a 'slugger; and the collective record was 361 - 391 (.480).

So perhaps there really isn't any true statistical basis for determining if having a slugger is worth more wins. From my perspective having a big bat accounting for alot of HR and RBIs is probably not statistically relevant but I bet mentally it has a huge factor. How many late inning rallies center around that number 3 or 4 batter? Also it is difficult to measure how a team feels if behind by one or two runs with their slugger coming up 3rd or 4th in the order versus having your 7 - 9 hitters? I tend to think of it has the reason baseball as sided with one dominant closer. I don't think it really matters who pitches the 9th but so much of baseball has become mental that not having your "closer" come in makes your team more vulnerable?

Monday, December 8, 2008

Rox Talk - Solid Pitching Starts

Short post this week as I just got back from a seven day cruise in the Caribbean! Needless to say, I didn't think much about anything except which direction the sun was facing and where the bar was.

From the Stat of the Day blog, I came across this nugget from Steve Lombardi. In short he was looking at the NY Yankees starting pitching and wondering just how bad they've been from 2004 - 2008. At first he was looking for starting pitchers who had started 90% of the time and had at least 28 starts and an ERA+ of greater than or equal to 110. He then dropped this to 100 to compare to the Red Sox over the same time period and Yanks had 7 solid seasons versus the Red Sox's 14 (obviously the Red Sox have 2 world series rings over this time period)!

And therefore the question is how many solid seasons did the Roxs have over this time span? The answer is actually 6 (I was surprised!) with Jason Jennings 2006 season leading the way with an ERA+ 130. And to think we supposedly had an awful pitching year in 2008 and yet we had 2 solid pitching seasons from Cook at an ERA+ 117 and U-Ball at an ERA+ of 116. The other three came from Jeff Francis in 2007 with an ERA+ of 114, Cook in 2007 with an ERA+ of 116, and another Cook in 2006 with and ERA+ of 116 (look at Mr. Consistent....).

Of course while we had six solid seasons and the Yanks had seven over that span, the Yanks won 476 games while the Roxs only scraped by with 375. Although the Roxs were pretty pathetic in '04 and '05 (note Kennedy's ERA+ of 134 in '05 doesn't count because he didn't start 28 games). Yanks scored 4,470 runs over that time span with the Roxs scored only 3,993 (an average of almost 95 more runs a season). Outscoring your opponent can have its advantages especially when their pitching isn't better than yours!

All in all I guess I take comfort that in 2009 we do have a starting rotation of Francis, Cook, and U-Ball which is at least a good we just have to find starters for the other 66 games (assuming the big three can get 32 starts each....)

Monday, December 1, 2008

Rox Talk - It's the Weather

Hot Stove News
Not much from the Rox front. Nice to see Baylor back in the purple pinstripes. Hurdle will have two former managers ready to take over when the Rox struggle out of the gate next year. At least it will keep him honest.

It's the Weather, Stupid!
To further my cause in esoteric baseball research I decided to look at the effects on homeruns and weather at Coors Field. Weather data comes from baseball reference. Data includes 2007 and 2008 numbers (grey boxes and line) and includes Rox and opponents homers. Circles are the 2007 and 2008 data points. The rate is determined by taking the total number of homers for the point and dividing this by the number of times that point occurred (i.e., it has been 80 degrees 5 times during the last two years and a total of 9 homers have occurred so the rate is 1.80). So below is temperature and generally the trend is that has it gets warmer the rate of homers increases. The weighted mean temperature at Coors Field over the last two years has been 74 degrees. The average homers hit at Coors over the last two year has been 2.2 a game. Eighty-three degrees has had the most home runs hit (27)

The next graph shows wind speed and home run rate. Generally as the wind increases homers decrease. This graph is independent of the whether the wind is blowing in or out. The weighted average of wind is about 7.5 mph.
The final graph gives the wind direction. Generally we don't have a Wrigley effect. When the wind is blowing in (147 homers, 67 times) a rate of 2.2 and when blowing out (106 homers, 48 times) for a rate of 2.2. Also it seems when that when the wind blows out the wind speed averages about 6.5 mph while when the wind blows in it is about 8 mph. The most homers occur when the wind is blowing in from centerfield (85) although the highest rate is when the wind blows from left to right (which also has on average the higher wind speed).