Sabermetrics, if you've never heard of it, is the use of various statistical methods to evaluate baseball players and teams. It is actually somewhat controversial because many people in MLB like to stick to their traditions, and sabermetrics flies in the face of that. There are countless jokes about sabermetricians all living in their mom's basement and don't actually watch the games. In spite of the resistance, the use of sabermetrics has actually become more common in recent years. Several teams, most notably the Oakland A's, Tampa Bay Rays, and Boston Red Sox have used high level analysis to help build their teams.
I decided to start blogging with the hopes I can add something to the science. If nothing else, I could provide a unique perspective to my hometown team, the Colorado Rockies. Before I go further with my work there are a few thoughts on the science I would like to share.
First of all, I actually believe that sabermetrics backs up a lot of what traditional baseball thinking has always taught. For example, the old saying "Don't make the first or third out at third base" can be supported using sabermetrics. Of course you shouldn't be willing to make any outs, but the 1st and third are especially damaging. I also think that things that happen on the field can be explained using sabermetrics, such as a player who seems to always find the hole may indeed have a high batting average on balls in play (BABIP).
I don't necessarily agree with everything that sabermetrics tends to support. For one thing I don't believe there is a tell all statistic. Every stat tells you something about a different player's ability. Even though, on-base percentage (OBP) is more valuable than batting average (AVG), I don't think it's totally useless. I do think that it can show the likelihood of a batter driving in a runner in "scoring position." Of course a high slugging percentage (SLG) will indicate a batter is more likely to drive in a runner who is NOT in scoring position. There is also the notion that sabermetrics is only about walks and home runs. I would disagree and feel that it is also about singles, doubles, and the occasional triple. Another common belief among stat guys, that I don't really believe, is that pitchers have no control over their BABIP.
I am doing this for the fun of it and to hopefully gain more insight to the game. But if any team's GM reads this and wants to hire me, please send me a message. I'll get back to you right away. I also welcome any constructive comments about my work. However, if you're going to drop a "momma's basement" joke on me, you can get lost.
There are a number of projects that I plan to work on as I write. I plan to look closer at park factors, BABIP from the hitter's and the pitchers perspective, and would even like to do some work with Pitch F/X data. I am really excited about all of this. I only wish I had started doing this sooner. There's a lot of discoveries to be made, so I had better get to work.
Thursday, October 22, 2009
Sunday, October 18, 2009
Franklin Morales, LOOGY
It's funny how a bad outings (or a streak of them) can affect people's perception of a player. Case in point, is Franklin Morales who had a breakout season in a relief role. He pitched well enough that when Huston Street was injured, he took over the closer role. He recorded saves in his first 6 opportunities, then began to struggle. His struggles, highlighted by a 5 run, 7th inning in Los Angeles left many fans feeling that he should have been left off the roster. There was one good reason he made it: he had been very tough against left handed batters. With Philly's lineup of left handed, the move to include Morales, only made sense.
The move paid off, with Franklin pitching 2 and 1/3 perfect innings in the first three games of the series. All of that good work was quickly forgotten, however, thanks to his 3 walk (1 intentional) performance in game 4. Of course he did do one very important thing that inning - get Ryan Howard out. Many fans were again ready to show Morales the door, but I say hold on.
First off, I am never one to judge a player based on one stretch. It's much more telling to look at the player's body of work. One thing that pops out about Morales' career is how good he has been against lefty batters. I feel the Rockies should keep him since he has a great chance to be an outstanding left handed one out guy (a LOOGY).
Over his 3 year career, Morales has given up an AVG/SLG/OBP of .185/.276/.277 vs left handers (.175/.247/.275 in 2009). His numbers vs. righties are .274/.373/.396 in his career and .277/.366./405 in 2009. While facing righties he has clearly been hittable, although they have not hit him for much power. Against lefties he has been flat out dominant. That kind of arm is not easily replaced.
One of the big criticisms of Morales has been his tendency to walk too many batters. This is certainly justified, since has was walked 12.4% of the batters he has faced in his career. Interestingly, his splits again tell a deeper story. While walking 13.3% of right handed batters, Morales has walked only 9.2% of lefties, a more acceptable rate.
Hopefully the Rockies can take a close look at his effectiveness vs left handers and keep him around. More importantly, I hope they know to use him vs lefty bats and limit his work vs righties. After all, a misused reliever can do as much damage as a bad one.
The move paid off, with Franklin pitching 2 and 1/3 perfect innings in the first three games of the series. All of that good work was quickly forgotten, however, thanks to his 3 walk (1 intentional) performance in game 4. Of course he did do one very important thing that inning - get Ryan Howard out. Many fans were again ready to show Morales the door, but I say hold on.
First off, I am never one to judge a player based on one stretch. It's much more telling to look at the player's body of work. One thing that pops out about Morales' career is how good he has been against lefty batters. I feel the Rockies should keep him since he has a great chance to be an outstanding left handed one out guy (a LOOGY).
Over his 3 year career, Morales has given up an AVG/SLG/OBP of .185/.276/.277 vs left handers (.175/.247/.275 in 2009). His numbers vs. righties are .274/.373/.396 in his career and .277/.366./405 in 2009. While facing righties he has clearly been hittable, although they have not hit him for much power. Against lefties he has been flat out dominant. That kind of arm is not easily replaced.
One of the big criticisms of Morales has been his tendency to walk too many batters. This is certainly justified, since has was walked 12.4% of the batters he has faced in his career. Interestingly, his splits again tell a deeper story. While walking 13.3% of right handed batters, Morales has walked only 9.2% of lefties, a more acceptable rate.
Hopefully the Rockies can take a close look at his effectiveness vs left handers and keep him around. More importantly, I hope they know to use him vs lefty bats and limit his work vs righties. After all, a misused reliever can do as much damage as a bad one.
Wednesday, October 14, 2009
Rocktober is Over
The Rockies season is over. After splitting two games in Philly, the Rocks came home and took two tough losses. I'm not going to write much specifically about the games. I will however tip my hat to Yorvit Torrealba, who got a big clutch double in game 4 and a home run in game 2. Although I still don't think he is as good as he played over the past month and a half, he did come through in those spots.
A few observations about the playoffs in general so far:
Three pretty good closers blew saves, while one lousy one managed to get two and blow zero. It just goes to show how anything can happen in a short series. Fair or not, Joe Nathan, Jonathan Papelbon, and Huston Street are all now rumored to be with new teams next year. Of course there was probably a chance all three would be gone anyway.
The umpiring was horrible around the league. While it's impossible to say that any team would have or wouldn't have won if the calls had been made correctly, it's reasonable to say that we shouldn't have to wonder. It should be up to the players to win or lose games, not umps. Now is the time to expand instant replay. Before that can happen, a few questions about the procedures must be answered:
Can replay happen quick enough not to slow down games?
What types of plays can be reviewed?
How will the plays be reviewed? (I.e. 5th ump in booth)
How will the Replays be initiated? (Personally, I think it should be up to the umps, although managers should have some way to ask for them. However, I do not think there should be any punishment for a failed review, like in football. I don't want to see anyone lose a game because they asked for a review, and was forced to give up an out. I also would not want to see a play not be reviewed that should, because a manager is afraid of being punished. If you do that you are changing the game too much.)
What will be done if continuity of a play is interrupted and how can you prevent that from happening? (My suggestion is that an ump can give a signal on a close play to "play it out" and that it will be reviewed when the play is over. If the review shows a fair ball, for example, then everyone stays where they ended up. If you do not have this, then a foul call that is over turned will require judgement by the umps to figure out where each runner should be.)
Where does the burden of proof lie? (In football, if a play is inconclusive then the ruling on the field stands. That may be logical, but in some cases, I don't think that would be the best way to handle it. For example if an ump says a fielder came off the base, or a runner missed a base, I feel that a replay should have to clearly show that happened, regardless of what the play was called on the field.)
Hopefully the league will spend some time this winter figuring out these questions and implement replay for next season, before something like this or this happens again.
A few observations about the playoffs in general so far:
Three pretty good closers blew saves, while one lousy one managed to get two and blow zero. It just goes to show how anything can happen in a short series. Fair or not, Joe Nathan, Jonathan Papelbon, and Huston Street are all now rumored to be with new teams next year. Of course there was probably a chance all three would be gone anyway.
The umpiring was horrible around the league. While it's impossible to say that any team would have or wouldn't have won if the calls had been made correctly, it's reasonable to say that we shouldn't have to wonder. It should be up to the players to win or lose games, not umps. Now is the time to expand instant replay. Before that can happen, a few questions about the procedures must be answered:
Can replay happen quick enough not to slow down games?
What types of plays can be reviewed?
How will the plays be reviewed? (I.e. 5th ump in booth)
How will the Replays be initiated? (Personally, I think it should be up to the umps, although managers should have some way to ask for them. However, I do not think there should be any punishment for a failed review, like in football. I don't want to see anyone lose a game because they asked for a review, and was forced to give up an out. I also would not want to see a play not be reviewed that should, because a manager is afraid of being punished. If you do that you are changing the game too much.)
What will be done if continuity of a play is interrupted and how can you prevent that from happening? (My suggestion is that an ump can give a signal on a close play to "play it out" and that it will be reviewed when the play is over. If the review shows a fair ball, for example, then everyone stays where they ended up. If you do not have this, then a foul call that is over turned will require judgement by the umps to figure out where each runner should be.)
Where does the burden of proof lie? (In football, if a play is inconclusive then the ruling on the field stands. That may be logical, but in some cases, I don't think that would be the best way to handle it. For example if an ump says a fielder came off the base, or a runner missed a base, I feel that a replay should have to clearly show that happened, regardless of what the play was called on the field.)
Hopefully the league will spend some time this winter figuring out these questions and implement replay for next season, before something like this or this happens again.
Thursday, October 8, 2009
Who Should Start?
The Rockies are down 1 game to none vs the Phillies. Cliff Lee was dominant throughout the game, only giving up one run in the ninth. As with any playoff game where you only get 1 run, fans are obviously calling for changes to be made. The Rockies reputation as being a team the hits left handers poorly only heightens the sense of urgency, with Cole Hamels starting game 2. Changes need to be made to the lineup, but which ones?
I decided to look at splits vs left handers for all of the Rockies hitters to see if an optimal lineup could be found. There are several positions which are considered to be up in the air: 2nd base, catcher, third base, and up to all three outfield spots. I had planned on comparing slugging and on base percentages for each positional battle. If one player had a higher slugging but lower OBP, I would use baseball musing's Lineup Analysis Tool to determine which lineup would score the most amount of runs per game; the one with player A (high OBP) or player B (high SLG). It turned out to not be necessary, as in EVERY situation, the player with the higher OBP vs lefties also had the higher SLG. (There was a different order among the three outfielders, however the top three in SLG were also the the top three in OBP.)
The winners are with OBP/SLG:
C. Chris Iannetta .406/.580
1B. Todd Helton .369/.372
2B. Clint Barmes .314/.496
3B. Garrett Atkins .363/.428
SS. Troy Tulowitzki .382/.519
OF. Seth Smith .368/.500
OF. Dexter Fowler .377/.482
OF. Carlos Gonzalez .343/.466
It's unlikely that these are going to be the starting 8 position players. My guess is that Stewart gets the start over Atkins for his defense, and Spilly gets the start over Seth Smith. It should be pointed out that there may be some sample size issues, as these are 2009 numbers and not career numbers. However, I doubt the Rockies are really looking close enough to the numbers to care.
Hopefully whichever lineup they go with does something. I don't want to be sitting out in the cold on Saturday being down 0-2.
I decided to look at splits vs left handers for all of the Rockies hitters to see if an optimal lineup could be found. There are several positions which are considered to be up in the air: 2nd base, catcher, third base, and up to all three outfield spots. I had planned on comparing slugging and on base percentages for each positional battle. If one player had a higher slugging but lower OBP, I would use baseball musing's Lineup Analysis Tool to determine which lineup would score the most amount of runs per game; the one with player A (high OBP) or player B (high SLG). It turned out to not be necessary, as in EVERY situation, the player with the higher OBP vs lefties also had the higher SLG. (There was a different order among the three outfielders, however the top three in SLG were also the the top three in OBP.)
The winners are with OBP/SLG:
C. Chris Iannetta .406/.580
1B. Todd Helton .369/.372
2B. Clint Barmes .314/.496
3B. Garrett Atkins .363/.428
SS. Troy Tulowitzki .382/.519
OF. Seth Smith .368/.500
OF. Dexter Fowler .377/.482
OF. Carlos Gonzalez .343/.466
It's unlikely that these are going to be the starting 8 position players. My guess is that Stewart gets the start over Atkins for his defense, and Spilly gets the start over Seth Smith. It should be pointed out that there may be some sample size issues, as these are 2009 numbers and not career numbers. However, I doubt the Rockies are really looking close enough to the numbers to care.
Hopefully whichever lineup they go with does something. I don't want to be sitting out in the cold on Saturday being down 0-2.
Monday, October 5, 2009
The Case for Chris Iannetta
Even before this happened, I've been saying the same thing to anybody that would listen: Chris Iannetta should be the Rockies' starting catcher. Unfortunately a lot of people out there don't, most importantly Jim Tracy, don't seem to agree. With the playoffs starting soon, I think now is the time to lay out my case.
Iannetta's critics like to point out a number of things why he shouldn't be in there. His .228 batting average and high strikeouts have doomed Iannetta in the eyes of many. Of course the biggest road block for Iannetta has been that his replacement has been on fire.
Since taking over the starting job on Aug 29, Yorvit Torrealba has had an AVG of .319 with an OBP of .373. Both numbers are very good, but his SLG over that span is .404. If we take a closer look we can see that Torrealba's BABIP has been an even .400. The typical BABIP is around .300. Although I do feel there are certain reasons that a player can have a higher than normal BABIP (such as speed, or gap power), YT doesn't have any of those. Even if he did, a BABIP that high is way beyond what any hitter can produce over a long span. It is pretty clear that he has been very fortunate over that span. Even with a .325 BABIP over that stretch, YT would have posted a very light .255/.313/.340 line. So unless the balls keep finding holes at that incredible rate, Torrealba's productivity should be expected to drop.
Meanwhile Chris Ianetta's BABIP has been a very low .253. If he has a more normal BABIP of .300, he add 9 hits to his total, bringing his AVG up to a more respectable .261. What about all those strikeouts? CI finished the year with 75 k's in only 289 at-bats. If his strikeout rate was reduced to the league average of 20 percent, he'd cut his total to 58 strikeouts. That is significant, but at his current BABIP would only give him 5 additional base hits, and a .244 AVG. So while a reduction in strikeouts would add a small increase in productivity, having a few more balls fall in would be even better.
Another advantage that Iannetta has, is the ability to draw walks. In 332 plate appearances, Chris has drawn 43 walks, a rate of 13%. Although his AVG is sitting at .228, his OBP is a respectable .344.
The most important thing to consider in this issue is Iannetta's platoon splits. He has struggled versus right handers this year, batting .202/.320/.413. Versus left-handers he has hit .296/.406/.580 with 4 homers and 9 doubles in 92 plate appearances. Of course his career numbers are only .265/.379/.530. With a team that has struggled to hit left handed pitching, you can't afford to leave that type of production on the bench.
Hopefully the Rockies' management will soon figure out the catching situation before it's too late. In the worst case scenario, they need to set this up as platoon with CI starting against lefties.
Iannetta's critics like to point out a number of things why he shouldn't be in there. His .228 batting average and high strikeouts have doomed Iannetta in the eyes of many. Of course the biggest road block for Iannetta has been that his replacement has been on fire.
Since taking over the starting job on Aug 29, Yorvit Torrealba has had an AVG of .319 with an OBP of .373. Both numbers are very good, but his SLG over that span is .404. If we take a closer look we can see that Torrealba's BABIP has been an even .400. The typical BABIP is around .300. Although I do feel there are certain reasons that a player can have a higher than normal BABIP (such as speed, or gap power), YT doesn't have any of those. Even if he did, a BABIP that high is way beyond what any hitter can produce over a long span. It is pretty clear that he has been very fortunate over that span. Even with a .325 BABIP over that stretch, YT would have posted a very light .255/.313/.340 line. So unless the balls keep finding holes at that incredible rate, Torrealba's productivity should be expected to drop.
Meanwhile Chris Ianetta's BABIP has been a very low .253. If he has a more normal BABIP of .300, he add 9 hits to his total, bringing his AVG up to a more respectable .261. What about all those strikeouts? CI finished the year with 75 k's in only 289 at-bats. If his strikeout rate was reduced to the league average of 20 percent, he'd cut his total to 58 strikeouts. That is significant, but at his current BABIP would only give him 5 additional base hits, and a .244 AVG. So while a reduction in strikeouts would add a small increase in productivity, having a few more balls fall in would be even better.
Another advantage that Iannetta has, is the ability to draw walks. In 332 plate appearances, Chris has drawn 43 walks, a rate of 13%. Although his AVG is sitting at .228, his OBP is a respectable .344.
The most important thing to consider in this issue is Iannetta's platoon splits. He has struggled versus right handers this year, batting .202/.320/.413. Versus left-handers he has hit .296/.406/.580 with 4 homers and 9 doubles in 92 plate appearances. Of course his career numbers are only .265/.379/.530. With a team that has struggled to hit left handed pitching, you can't afford to leave that type of production on the bench.
Hopefully the Rockies' management will soon figure out the catching situation before it's too late. In the worst case scenario, they need to set this up as platoon with CI starting against lefties.
Subscribe to:
Posts (Atom)