Minor League Equivalencies
Brad Stewart asked if I would talk to you guys about Major League Equivalencies this week. I thought it was a great idea, so, here I am. If you’ve never heard of MLEs before, you’re in for a treat. When many people project younger players, there are several approaches I’ve seen taken that really put these people at a disadvantage… approaches that don’t allow the projection to be as accurate as it could be.
The first is taken when a minor league player who doesn’t have any major league experience, like an Evan Longoria or a Jay Bruce this year, is downgraded to such an extreme extent that the projection is calling for him to have, essentially, no fantasy value at all. The projector says that because the player doesn’t have any major league at-bats and because prospects are so risky, it would be irresponsible to give them anything but an extremely conservative projection.
Similarly, the second is when someone projects a player going into his sophomore year using only his rookie numbers, because “those are the only major league numbers we have of him.”
The third is when minor league numbers are used, but they are used at face value or some subjective changes are made in the projector’s mind.
If you have ever taken on one of these three approaches, you’ve done yourself a major disservice. While it’s true that many minor leaguers never make it to the majors, or fail once they do so, and that minor leaguers are a little trickier to project than major leaguers, it can be done. And that’s where Major League Equivalencies enter the picture.
Major League Equivalencies were first introduced by Bill James – often thought of as the godfather of sabermetrics and current Senior Baseball Operations Advisor to the World Champion Boston Red Sox – in his 1985 Baseball Abstract. He outlined his method for hitters in the Abstract, and since then MLEs have really taken off.
Let’s take a step back, though, and explain what MLEs are. MLEs, essentially, approximate the numbers a given minor league hitter would have produced at the major league level. This is not to be confused with a projection. It serves the same purpose as, say, the 2007 stat line of Alex Rodriguez. You wouldn’t take A-Rod’s 2007 line and use it as a projection for 2008. After the minor league line is translated, though, it can be looked at on the same plane as major league stat lines. Projections can be made from there, the same as you would project major league players.
You don’t need to worry about the specifics of creating MLEs. Others have already done that for you. I’m sure you’d like to know a little about the concepts used to produce them, though. To simplify it, I’m going to put them in list form. What MLEs essentially do to get the final adjusted line is this:
· Adjust for the quality of the league, or the run environment
· Adjust for the minor league and major league ballparks
· Adjust for competition
Adjusting for the competition is kind of a tricky one to do. What you need to do is look at the stats of every player who has moved between levels within a season and weight them based on playing time at each level.
Another thing you need to know about MLEs is that a Triple-A MLE will be much more accurate than, say, a Low-A MLE. This is because when you adjust for competition, far more players jump from Triple-A to the majors within a season than from Low-A. You have a far smaller sample size to judge by.
To increase the sample size, then, you need to develop “intermediate relationships,” as Jeff Sackman of MinorLeagueSplits put it. That is, you look at players who jumped from Single-A to Double-A, and then at players who jumped from Double-A to either Triple-A or the majors. While this solves the small sample size issue, it does still take away accuracy.
Double-A and Triple-A MLEs are quite reliable, but once you get into A-ball or short season leagues you need to be very cautious. Of course, in most fantasy leagues, you’re not going to be projecting too many guys who played in the Gulf Coast League the year before, so this shouldn’t be a major concern. It does occasionally come into play, though, so you need to be aware of it. Take Cameron Maybin as an example. He had just 20 at-bats at Double-A and none at Triple-A before being called to the majors, where he had just 49 at-bats. Be very cautious when projecting him for 2008.
Major League Equivalencies can also be applied to players coming over from Japan. Because almost all of these conversions take place in-between seasons, though, you need to add one additional step to the process: adjusting for age.
Okay. At this point, you might be asking: “Where can I find MLEs?” As I said earlier, people have already created them for you. MinorLeagueSplits.com is one of my favorite websites, and MLEs are available on every player’s page. They don’t yet have 2007 MLEs, and they don’t have a comprehensive list of them, so if you’re looking for them in bulk you’re going to need to go elsewhere.
The Ron Shandler Baseball Forecaster has MLEs this year, and I believe if you buy from them (as in, not through Amazon or somewhere like that) you will get an electronic version of the book.
Or, if you’re really, motivated, check out the Baseball Think Factory article in the list below. Get your base minor league stats from somewhere like MinorLeagueBaseball.com, and then follow along as Dan Szymborski walks you through exactly how to create an MLE.
Closing up, this article, really, has only scratched the surface of MLEs. Its purpose was simply to make you aware of them, because they are a truly invaluable tool.
For further reading about MLEs, or if you’re really interested in the specifics of how they can be calculated, check out these sites. The final one isn’t so much about MLEs as I discussed them, but about potential improvements that could be made in the future.
http://seamheads.com/blog/2008/01/19/major-league-equivalencies/
http://www.baseballthinkfactory.org/btf/scholars/czerny/articles/calculatingMLEs.htm
http://minorleaguesplits.com/mle.html
http://www.hardballtimes.com/main/article/rethinking-mle-the-role-of-experience/
Also, if you’re interested in the minor league park factors, here are a couple of resources for them. I know I’ve seen league factors before, but I can’t seem to locate them tonight. If I run across them, I’ll put the link in a future article.
http://www.minorleaguesplits.com/pf.html
http://firstinning.com/pf/
That wraps up this week. One final note, unrelated to this week’s topic. Two weeks ago I discussed K/BB ratio. Today, I posted an article at THT Fantasy Focus that looks into how pitchers achieve their K/BB ratios. Turns out, pitchers who do it in a certain way tend to put up better ERAs. If you're interested, you can find that here. K/BB ratio is a very important topic and one I’m sure I’ll be writing more about in the future.
The first is taken when a minor league player who doesn’t have any major league experience, like an Evan Longoria or a Jay Bruce this year, is downgraded to such an extreme extent that the projection is calling for him to have, essentially, no fantasy value at all. The projector says that because the player doesn’t have any major league at-bats and because prospects are so risky, it would be irresponsible to give them anything but an extremely conservative projection.
Similarly, the second is when someone projects a player going into his sophomore year using only his rookie numbers, because “those are the only major league numbers we have of him.”
The third is when minor league numbers are used, but they are used at face value or some subjective changes are made in the projector’s mind.
If you have ever taken on one of these three approaches, you’ve done yourself a major disservice. While it’s true that many minor leaguers never make it to the majors, or fail once they do so, and that minor leaguers are a little trickier to project than major leaguers, it can be done. And that’s where Major League Equivalencies enter the picture.
Major League Equivalencies were first introduced by Bill James – often thought of as the godfather of sabermetrics and current Senior Baseball Operations Advisor to the World Champion Boston Red Sox – in his 1985 Baseball Abstract. He outlined his method for hitters in the Abstract, and since then MLEs have really taken off.
Let’s take a step back, though, and explain what MLEs are. MLEs, essentially, approximate the numbers a given minor league hitter would have produced at the major league level. This is not to be confused with a projection. It serves the same purpose as, say, the 2007 stat line of Alex Rodriguez. You wouldn’t take A-Rod’s 2007 line and use it as a projection for 2008. After the minor league line is translated, though, it can be looked at on the same plane as major league stat lines. Projections can be made from there, the same as you would project major league players.
You don’t need to worry about the specifics of creating MLEs. Others have already done that for you. I’m sure you’d like to know a little about the concepts used to produce them, though. To simplify it, I’m going to put them in list form. What MLEs essentially do to get the final adjusted line is this:
· Adjust for the quality of the league, or the run environment
· Adjust for the minor league and major league ballparks
· Adjust for competition
Adjusting for the competition is kind of a tricky one to do. What you need to do is look at the stats of every player who has moved between levels within a season and weight them based on playing time at each level.
Another thing you need to know about MLEs is that a Triple-A MLE will be much more accurate than, say, a Low-A MLE. This is because when you adjust for competition, far more players jump from Triple-A to the majors within a season than from Low-A. You have a far smaller sample size to judge by.
To increase the sample size, then, you need to develop “intermediate relationships,” as Jeff Sackman of MinorLeagueSplits put it. That is, you look at players who jumped from Single-A to Double-A, and then at players who jumped from Double-A to either Triple-A or the majors. While this solves the small sample size issue, it does still take away accuracy.
Double-A and Triple-A MLEs are quite reliable, but once you get into A-ball or short season leagues you need to be very cautious. Of course, in most fantasy leagues, you’re not going to be projecting too many guys who played in the Gulf Coast League the year before, so this shouldn’t be a major concern. It does occasionally come into play, though, so you need to be aware of it. Take Cameron Maybin as an example. He had just 20 at-bats at Double-A and none at Triple-A before being called to the majors, where he had just 49 at-bats. Be very cautious when projecting him for 2008.
Major League Equivalencies can also be applied to players coming over from Japan. Because almost all of these conversions take place in-between seasons, though, you need to add one additional step to the process: adjusting for age.
Okay. At this point, you might be asking: “Where can I find MLEs?” As I said earlier, people have already created them for you. MinorLeagueSplits.com is one of my favorite websites, and MLEs are available on every player’s page. They don’t yet have 2007 MLEs, and they don’t have a comprehensive list of them, so if you’re looking for them in bulk you’re going to need to go elsewhere.
The Ron Shandler Baseball Forecaster has MLEs this year, and I believe if you buy from them (as in, not through Amazon or somewhere like that) you will get an electronic version of the book.
Or, if you’re really, motivated, check out the Baseball Think Factory article in the list below. Get your base minor league stats from somewhere like MinorLeagueBaseball.com, and then follow along as Dan Szymborski walks you through exactly how to create an MLE.
Closing up, this article, really, has only scratched the surface of MLEs. Its purpose was simply to make you aware of them, because they are a truly invaluable tool.
For further reading about MLEs, or if you’re really interested in the specifics of how they can be calculated, check out these sites. The final one isn’t so much about MLEs as I discussed them, but about potential improvements that could be made in the future.
http://seamheads.com/blog/2008/01/19/major-league-equivalencies/
http://www.baseballthinkfactory.org/btf/scholars/czerny/articles/calculatingMLEs.htm
http://minorleaguesplits.com/mle.html
http://www.hardballtimes.com/main/article/rethinking-mle-the-role-of-experience/
Also, if you’re interested in the minor league park factors, here are a couple of resources for them. I know I’ve seen league factors before, but I can’t seem to locate them tonight. If I run across them, I’ll put the link in a future article.
http://www.minorleaguesplits.com/pf.html
http://firstinning.com/pf/
That wraps up this week. One final note, unrelated to this week’s topic. Two weeks ago I discussed K/BB ratio. Today, I posted an article at THT Fantasy Focus that looks into how pitchers achieve their K/BB ratios. Turns out, pitchers who do it in a certain way tend to put up better ERAs. If you're interested, you can find that here. K/BB ratio is a very important topic and one I’m sure I’ll be writing more about in the future.
0 Comments:
Post a Comment
<< Home