Can I Get a Little Help?
by beaudog on Sep 1, 2010 4:34 PM CDT
Email
Print
I posted this in another thread and a couple guys suggested I post it here too.
okay guys, here’s the deal. i’ve been a huge baseball fan probably longer than a lot of you have been alive, and have died a thousand deaths rooting for these rangers. i like the heck out of the way you all get behind the rangers and how you can talk smack about them but woe be unto the troll who attempts it. but, i grew up on W/L’s, ERA’s, and BA’s so all the metric stuff you folks speak around here makes my head spin. since i can’t spew stats with acronyms i don’t understand, i’m not less of a fan than y’all, just less informed. That being said, i have a question. is there an easy way for an old dog to learn sabermetrics, or enough to not feel like an idiot?
34 comments
|
3 recs |
Do you like this story?
Comments
Yes
but I just come to LSB for the recipes, grilling tips, and the best new microbrews.
by I am Neftali Feliz on Sep 1, 2010 4:36 PM CDT reply actions 1 recs
try Fangraphs glossary
You don’t have to understand the details but a general understanding of stuff like UZR, OPS, wOBA among others will help out.
Well
There are different stages to understanding advanced statistics. You don’t need to be able to spout the formula for wOBA or tRA to get by. Really, all sabermetrics is is creating stats that are less affected by noise and other players.
1) Recognize which common statistics are bad and understand why. Wins and Losses – bad. RBI – bad. R – bad. The reason behind these is that they don’t really tell you much more than other stats and depend a lot on other players. Two guys who are identical hitters can have vastly different RBI totals based on what lineup they are in. So it doesn’t do any good to look at RBIs.
2) There are some ‘okay’ common statistics, but you can do better. Batting average and ERA fit this category. They aren’t awful, but they don’t give you the full story. Basically, a guy with a good ERA is probably a good pitcher, but you may be deceived, or vice versa.
3) There are slightly more advanced, but better easy statistics: SLG and OBP tell you more about a hitter than BA, for instance; because not all hits are equal and walks matter too. If you cite SLG and OBP, you won’t get laughed at, but you can do better
4) Then you have your raw stats, called peripherals. HR, Ks, BBs – these are good, because the player has a lot of control over them and not much noise affects them. You should normalize to playing time though – HR/AB for hitters, per 9 for pitchers.
5) Based on these peripherals, you can get some more advanced statistics, essentially based on the expected value of each result and so forth. wOBA is a weighted statistic for hitters similar to OBP or BA that computes the actual value of each outcome the hitter hits into. FIP for pitchers is a combination of HR/9, K/9, and BB/9, for example. It is defense independent, since a bad fielder cannot affect it. But it is still susceptible to park effects, etc.
These formulas sometimes get tricky since they’re trying to be predictive for future years, or to apparently drive home the point that OBP is better than BA (in the case of wOBA). But the details don’t matter.
6) There are then statistics based on these described in #5 that relate to wins on the field. WAR is the most common, as it essentially tries to quantify the # of wins that player is worth above replacement. It’s pretty intuitive if you ignore the details. Josh Hamilton is worth a lot of WAR. Jeff Francoeur is not.
7) Ignore people trying to quantify a $ value to WAR
Go Rice Owls!
by JBImaknee on Sep 1, 2010 4:52 PM CDT reply actions 1 recs
This is a pretty good summary
For me, at least, a lot of advanced statistics are really meant to be used together simply to tell you more about a player and whether or not he is actually doing as well or as bad as the ‘standard’ measures say. For instance, a player may be batting .350 but if I don’t watch him play everyday I would probably assume he is a good hitter. However if he has a 100% (unrealistic, unless I’m the hitter) GB % then I know he has probably gotten lucky because they have all been seeing eye singles.
So how do we use this information then? We look at stats like BABIP (batting average on balls in play) in conjuction with other stats like fly ball rates, ground ball rates and line drive rates. This helps you determine how often a player is able to square up on the ball and drive it, how often he puts it on the ground, and how often he hits it in the air. This information is much more useful than simply knowing a guy has gotten a base hit 35% of the time. Now I know, in general, how likely he will be able to sustain that 35% or not.
I’m going to stop here because I feel like I’m rambling and I’m not even sure if what I said makes a whole lot of sense. Sorry about the sloppiness.
by TheHuntforRedOctober on Sep 1, 2010 5:15 PM CDT up reply actions
This is a pretty good summary
Yup.
"I support you, Wash; I’ve always supported you," Young said
"Back on the scene, with a gangsta lean" RW
Good stuff.
I’ll disagree with seven and say ignore people who are using it incorrectly and try not to do the same yourself. The dollar values aren’t strictly “this is how much this player is worth.” It’s how much the best estimate of his production goes for on average on the open market. It’s a useful tool, but way too often used wrong or too verbatim. If FG is telling you Michael Young was worth 10 million this year, it’s not necessarily telling you that MY is unquestionably overpaid. If it tells you he’s worth 20 million it’s not necessarily telling you he’s unquestionably underpaid. All it’s telling you is that teams, on average, got that production (by WAR, which has its own separate issues) for X amount of dollars.
If I’d add an 8) to this really good list, it would be that the concepts behind the stats are far more important than actually understanding the nitty gritty. If you’re going to cite wOBA it’s great if you understand linear weights, but all that’s really important from an intelligent baseball approach is understanding that avoiding outs is very, very, very important and all hits are not created equally. Similarly, it’s great if you look at all the details of WAR, but what’s important is understanding that different positions on the field are more difficult to fill or have more defensive impact and defense and playing time matter.
thanks
the #7 was really just a silly comment designed to elicit a response from you in particular. I actually think there is probably a good way to get $ values for players, but I don’t think the way they do it now is very good. Too many assumptions of linearity and lack of a valuation of risk…
Definitely agree about the last part – the details don’t really matter, it is the principal of it. As you’ve pointed out before, most of these stats take half a season to a season to stabilize anyway – that means that their error rates are such that they’re only going to be 90%-95% accurate as comparative measures anyway (if two pitchers have FIPs within 5% of one another, I’d be wary of drawing strong conclusions that one is better than another). Understanding exactly what things are weighted to isn’t that important, since the weights are approximations anyway. These advanced stats are fuzzy numbers, but they’re still capturing meaningful info fuzzy.
Go Rice Owls!
I would agree the linearity isn't the best.
I think it’s better than just blind guessing, but I agree a whole lot can be done to improve it and it’s used as an absolute way too often. As are many sabermetrics (and I don’t pretend I’m not guilty of it).
love that commercial.
"Biggest mistake in DFW history?" - Bigger mistake in LSB history.
"Back in Irish's day you had to kill a man before you were taken seriously in polite society." - Aquaman56 06/25/10
One of the only commercials I've found myself rewinding to watch again.
Nothing like a Bieber-haired twentysomething white boy channeling Eddie Murphy as James Brown.
My wife loves the companion commercial targeted to the ladies.
I like the one targeted for the ladies too.
The chick who wants the dark side guy is smokin’.
"I don't really like pitchers." - Nelson Cruz
The saber library
really helped me out. http://saberlibrary.com/
That's why they call them business sox
Good links so far.
I also learn a lot from some of the posters here and by asking questions at Beyond the Boxscore. It’s not hard to pick things up just from experience, as long as you’re not resistant to someone telling you why the opinion you just cited isn’t necessarily correct and you’re interested in reading their support.
"someone telling you why the opinion you just cited isn’t necessarily correct and you’re interested in reading their support."
or telling you that you are full of shit. :)
good
stuff all the way around. i never really wanted to know how to do all the mental gymastics required to arrive at the stats, just understand what they are and their intended use. that, and to be able to hit the post button without getting the snot beat out of me everytime.
"If Nolan Ryan hadn't been born, Larry McMurtry would have made him up."
----Kevin Sherrington of the Dallas Morning News to Bud Selig.
OPS is all you need to know.
'Waiting for a girl and she gets me into fights
Waiting for a girl we get drunk on Friday night'
I forgot WHIP. :)
'Waiting for a girl and she gets me into fights
Waiting for a girl we get drunk on Friday night'
on a lighter note
i fully support the thought of firing joe morgan
"If Nolan Ryan hadn't been born, Larry McMurtry would have made him up."
----Kevin Sherrington of the Dallas Morning News to Bud Selig.
All of this is great stuff.
A funny thing though reading through all the different stats, The SaberLibrary seems to hate Yuniesky Betancourt.
thanks guys
i appreciate all the links. i’ll try to use them wisely.
"If Nolan Ryan hadn't been born, Larry McMurtry would have made him up."
----Kevin Sherrington of the Dallas Morning News to Bud Selig.
Another great link from the Saber Library
http://www.socrated.com/user_courses/48
It kidn of makes it seem like work but seems like a great place to start and very organized way to approach such a big task.
And this post should be rec’d and maybe re-titled to “A Beginner’s Guide to Advance Stats” or something
I tried to go through that "course"
last night and had a tough time with it. Most of the articles reference things that haven’t yet been covered in the “course” yet. I learn better by someone showing specific examples and able to talk back and forth on it, so I’ve never been able to really grasp a lot of the metrics unfortunately.
I wish our stat gurus could do a webinar series or something.
Meh.
My problem is that to understand all the stats
I would have to stay sober more often.
What are we at the park for except to win? I'd trip my mother. I'd help her up, brush her off, tell her I'm sorry. But mother don't make it to third. ~Leo Durocher
Did you ever hear the theory that if you study for a test drunk or stoned,
you would do better on the test if you took it while drunk or stoned?
by lost in space on Sep 2, 2010 6:22 PM CDT up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
