What's new
Fantasy Football - Footballguys Forums

Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.

Advanced stats for old guys - or, how to relate to the next generation of baseball fans (1 Viewer)

Tom Servo

Nittany Beavers
I guess you would call me old school; I've been a fan since the late 60s and grew up on the traditional stats. Now, I have nothing against these new fangled stats; I'm really at a loss to understand them.  I do understand WAR, BABIP and OPS+, but FiP, wRC, etc. leave me completely shuked. Can someone explain to me how I can understand these stats so I can read a dang baseball blog with blindly nodding like a drooling fool?

 
FIP is pretty basic.  It is meant to determine how good a pitcher is independent of the defenders behind him.  It's scaled to look like ERA so that it's easy to follow since we are familiar with what a good pitcher's ERA is.

As far as wRC go here.  Fangraphs actually has good explanations of the majority of advanced statistics.

ETA:  One you should become familiar with if you haven't already is wOBA.

 
Last edited by a moderator:
FIP is pretty basic.  It is meant to determine how good a pitcher is independent of the defenders behind him.  It's scaled to look like ERA so that it's easy to follow since we are familiar with what a good pitcher's ERA is.

As far as wRC go here.  Fangraphs actually has good explanations of the majority of advanced statistics.

ETA:  One you should become familiar with if you haven't already is wOBA.
Nice.

 
rodg12 said:
FIP is pretty basic.  It is meant to determine how good a pitcher is independent of the defenders behind him.  It's scaled to look like ERA so that it's easy to follow since we are familiar with what a good pitcher's ERA is.
To expand, FIP looks at the three true outcomes and weights each of them (HR very very bad, BB bad, K good).  It is meant to be measure of the things that a pitcher can control, because balls in play have a higher variability of outcomes that can have nothing to do with the pitcher.

xFIP takes this stat and replaces actual home runs with more of a league-average expected home run.

Any time you see a stat with x in front of it, it is presumably trying to reduce the impact of variability in events that typically regress to the mean for most players.  Or in other words, trying to measure the players true skill by removing luck from the equation (at least as much as they can).

 
I'd love to subscribe to advanced stats more than I do now, but I guess my biggest issue is which one is most accurate? Why is there a difference between WAR (bWar?) and fWAR, and which one should be trusted more?

I think back to Northern Voice's post on the AL Cy Young race and how the advanced stats were all over the place on who were the leaders. So which should we believe?

On top of that, I love a good defensive player and it seems the advanced stats put way too much emphasis on the offensive (or pitching) side of the game. 

 
I'd love to subscribe to advanced stats more than I do now, but I guess my biggest issue is which one is most accurate? Why is there a difference between WAR (bWar?) and fWAR, and which one should be trusted more?

I think back to Northern Voice's post on the AL Cy Young race and how the advanced stats were all over the place on who were the leaders. So which should we believe?

On top of that, I love a good defensive player and it seems the advanced stats put way too much emphasis on the offensive (or pitching) side of the game. 
it's been extremely difficult to quantify defense, so it's not that they have not tried, but there seem to be problems with it.  Now with the MLBAM video-tracking equipment, there is much more data to be processed in order to try to refine the defensive metrics.

Also not sure why there has to be one stat that is supposed to quantify a player's full worth.  The purpose of advanced stats isn't really to find some unified theory of everything.  It is to try to find some insights beyond simply looking at outcomes, which are inherently variable.  It is helpful to see whether something is more likely a fluke or if it will sustain.  It's just to try to understand the game better.

 
I'd be interested in an analysis of the margin of error inherent in various advanced stats.  A single season is still a relatively small sample size and I've seen some sources report a potential margin of error of one win above replacement per player season (I assume it's +/- 0.5 rather than +/- 1.0).  The aggregate WAR for a team in a single season has a correlation coefficient of 0.83.  I guess this is OK but it's still a pretty large potential variance.

WAR is a fun stat but I agree with Longball Larry that the search for an all encompassing single stat involves some leap of faith in assessing the relative value of very different skills.  I haven't completely bought in to advanced fielding, particularly over a single season.  It's probably more accurate valuation in baseball than other sports because the actions of preventing runs are directly observable and generally attributable to an individual defender.  But I still don't believe a corner OF with an OPS of under .650 can be above replacement value, no matter how good a defender Jason Heyward is 

 
Good topic.

I would be interested to hear what advanced stats people use for fantasy. The last couple of seasons I have paid close attention to contact-rate and line-drive percentage when seeking players for batting average.

 
Good topic.

I would be interested to hear what advanced stats people use for fantasy. The last couple of seasons I have paid close attention to contact-rate and line-drive percentage when seeking players for batting average.
exit velocity for hitters.  Also look at some of the xHR stuff Mike Podhorzer has done.

SwStrk % and Soft contact for pitchers, as well as SIERA.

Also like looking at changes in pull/oppo for hitters and really any changes in general.  I think significant shifts like that are good for determining sustainability in a new season.

 
FIP (and the like) won't survive for more than another few years. It will be replaced by something that takes into account exit velocity.

 
Dickie Dunn said:
On top of that, I love a good defensive player and it seems the advanced stats put way too much emphasis on the offensive (or pitching) side of the game. 
Well, see Jason Heyward, as mentioned by Eephus.

 
Well, see Jason Heyward, as mentioned by Eephus.
Good example. And I have to think SS is a more valuable defensive position than RF (even though I played RF almost exclusively all the way through my 20s), but I have to wonder if a guy like Mark Belanger ever would have a chance in MLB these days with so much emphasis on offense.

 
Belanger played in a time where runs were scarcer than today.  His career OPS+ normalized for era and ballpark was 71 (I'm excluding the years at the end of his career where he wasn't a regular and was terrible even by his standards).  That's obviously not great but it's not appreciably worse than some current regular SS.  You could also probably make the argument that run prevention increases in importance when there are fewer runs being scored.

Much of it was probably instinctive but Earl Weaver managed more in line with modern analytics-based tactics than his contemporaries.  Weaver valued OBP and ISO, eschewed low percentage SB attempts and the sacrifice bunt, platooned a lot and was able to effectively use guys like Ken Singleton and Terry Crowley.

 
Good example. And I have to think SS is a more valuable defensive position than RF (even though I played RF almost exclusively all the way through my 20s), but I have to wonder if a guy like Mark Belanger ever would have a chance in MLB these days with so much emphasis on offense.
Adeiny Hechavarria has started for 4 seasons now trotting out a career OPS+ of 74 and he's not an elite defender like Belanger was. :shrug:

 
I'd be interested in an analysis of the margin of error inherent in various advanced stats.  A single season is still a relatively small sample size and I've seen some sources report a potential margin of error of one win above replacement per player season (I assume it's +/- 0.5 rather than +/- 1.0).  The aggregate WAR for a team in a single season has a correlation coefficient of 0.83.  I guess this is OK but it's still a pretty large potential variance.

WAR is a fun stat but I agree with Longball Larry that the search for an all encompassing single stat involves some leap of faith in assessing the relative value of very different skills.  I haven't completely bought in to advanced fielding, particularly over a single season.  It's probably more accurate valuation in baseball than other sports because the actions of preventing runs are directly observable and generally attributable to an individual defender.  But I still don't believe a corner OF with an OPS of under .650 can be above replacement value, no matter how good a defender Jason Heyward is 
They Heyward thing makes me question WAR's ultimate viability every time.  No effing way a RF could possibly be that valuable defensively.

 
Belanger played in a time where runs were scarcer than today.  His career OPS+ normalized for era and ballpark was 71 (I'm excluding the years at the end of his career where he wasn't a regular and was terrible even by his standards).  That's obviously not great but it's not appreciably worse than some current regular SS.  You could also probably make the argument that run prevention increases in importance when there are fewer runs being scored.

Much of it was probably instinctive but Earl Weaver managed more in line with modern analytics-based tactics than his contemporaries.  Weaver valued OBP and ISO, eschewed low percentage SB attempts and the sacrifice bunt, platooned a lot and was able to effectively use guys like Ken Singleton and Terry Crowley.
wouldn't dWAR take that into account?

 
Kevin Kiermaier had 5.0 defensive WAR in 2015, which is staggering.  As I understand it, a common replacement level is set for all three OF positions which I think would inflate the impact of all CF, especially a superior defensive one like Kiermaier.  

 
wouldn't dWAR take that into account?
I believe dWAR and oWAR are adjusted for runs per game so a given offensive output today would result in more wins than during the steroid era.

dWAR for pre-2003 players like Belanger should be taken with a shaker of salt.  They're estimations based on very limited data like total chances, double plays and errors so they have value for comparing players of the era but probably aren't useful to rank vs. modern ballplayers.

 
The FA market for poor defenders suggests there's some merit to the publicly available measurements of defensive value. 

I wonder if both the AL & NL HR leaders have ever been available as FA before.  If this happened 20 years ago, Trumbo and Carter would have attracted more interest.

 
Eephus said:
The FA market for poor defenders suggests there's some merit to the publicly available measurements of defensive value. 

I wonder if both the AL & NL HR leaders have ever been available as FA before.  If this happened 20 years ago, Trumbo and Carter would have attracted more interest.
Do you think Dave Kingman would have a job in today's game? 

 
On top of that, I love a good defensive player and it seems the advanced stats put way too much emphasis on the offensive (or pitching) side of the game. 
I agree.  I think defense is under valued.   Not knowledgeable in these advance stats, but let's take an overly simplistic look at two outfielders.  Now most fly balls are pretty routine and every ML quality OF is going to catch them. But once every other game or so, there will be that one fly ball that the elite defender, player A, is going to track down that the subpar OF, player B, is going to let fall for a base hit.  Maybe that is a bit of an exaggeration, but I don't think the frequency is that far off. 

So player B gives the other team one additional hit every other game.   What does he have to do offensively to offset that?   Simplistically, he would have to have one more hit than Player A ever other game, which would translate into a BA of about 100 points more.  So I can see how an elite defender batting .230 is more valuable than a subpar defender batting .310.  Now if he is also hitting 40 HRs, that changes things.  But it takes a huge bat to offset defensive deficiencies, and I don't see that in how most teams value players. 

 
Last edited by a moderator:
I used Belanger mainly because his stats are just mind-numbing given the state of today's game. He had consecutive seasons in his prime in which he batted .225 and .226 and earned MVP votes in both. That came one season after he batted .186 over 113 games -- and that wasn't even in the bottom three season averages of his career. His 10-year stretch from ages 28-37 he batted over .228 just once.

I guess I just can't see any team these days willing to live with that in their lineup for more than a couple of seasons, never mind the 13-15 years that Belanger played regularly for the Orioles, no matter how good he was defensively.

 
Belanger played in an era where a purely defensive SS was the norm.   I looked at 1973 when Belanger finished 21st in MVP voting in spite of a .564 OPS.  The other starting AL SS were:

TEX - Toby Harrah .693
CLE - Frank Duffy .703
BAL - Luis Aparicio .633
KC - Freddie Patek .632
OAK - Bert Campaneris .626
DET - Ed Brinkman .609
CHI - Eddie Leon .582
CAL - Rudy Meoli .579
NY - Gene Michael .547
MIN - Danny Thompson .542
MKE - Tim Johnson .502

Harrah's OPS+ was exactly 100. 

 
I agree.  I think defense is under valued.   Not knowledgeable in these advance stats, but let's take an overly simplistic look at two outfielders.  Now most fly balls are pretty routine and every ML quality OF is going to catch them. But once every other game or so, there will be that one fly ball that the elite defender, player A, is going to track down that the subpar OF, player B, is going to let fall for a base hit.  Maybe that is a bit of an exaggeration, but I don't think the frequency is that far off. 

So player B gives the other team one additional hit every other game.   What does he have to do offensively to offset that?   Simplistically, he would have to have one more hit than Player A ever other game, which would translate into a BA of about 100 points more.  So I can see how an elite defender batting .230 is more valuable than a subpar defender batting .310.  Now if he is also hitting 40 HRs, that changes things.  But it takes a huge bat to offset defensive deficiencies, and I don't see that in how most teams value players. 
Personally I think it's a bit of a nod to the fact that we know so much more about offensive outcomes and value.  If we had similarly reliable defensive version, I'd be willing to count it more heavily relative to the overall value of a player.  I think we're getting closer with some of the stuff Mlb did last year. As of like 2014, I think you had to cheat WAR more weighted towards offensive component or even xFIP because stats like UZR are just a guess in comparison. 

 
One thing I find interesting about defensive stats is that advancements on the team level led to much more frequent shifting, which makes individual defensive stats harder to calculate. UZR isn't real helpful when the third baseman is catching fly balls to right.

 
K-bb is my favorite in season pitching stat for fantasy purposes. 
Top 10 in K-BB last year.  Which one of these things doesn't belong?


1


Jose Fernandez


Marlins


12.49


2.71


4.60


0.64


34.3 %


7.5 %


26.9 %


.220


1.12


.332


76.6 %


70


58


62


2.86


2.30


0.56


2.56


2.81


2


Max Scherzer


Nationals


11.19


2.21


5.07


1.22


31.5 %


6.2 %


25.3 %


.196


0.97


.255


81.7 %


71


79


82


2.96


3.24


-0.28


3.37


3.05


3


Noah Syndergaard


Mets


10.68


2.11


5.07


0.54


29.3 %


5.8 %


23.5 %


.240


1.15


.334


76.9 %


65


56


65


2.60


2.29


0.31


2.67


2.95


4


Justin Verlander


Tigers


10.04


2.25


4.46


1.19


28.1 %


6.3 %


21.8 %


.204


1.00


.255


79.9 %


72


81


89


3.04


3.48


-0.44


3.78


3.42


5


Madison Bumgarner


Giants


9.97


2.14


4.65


1.03


27.5 %


5.9 %


21.6 %


.209


1.02


.265


79.1 %


69


83


86


2.74


3.24


-0.50


3.54


3.36


6


Chris Sale


White Sox


9.25


1.79


5.18


1.07


25.7 %


5.0 %


20.7 %


.225


1.04


.279


76.6 %


78


79


84


3.34


3.46


-0.12


3.58


3.43


7


Michael Pineda


Yankees


10.61


2.72


3.91


1.38


27.4 %


7.0 %


20.4 %


.264


1.35


.339


70.7 %


114


86


78


4.82


3.80


1.02


3.30


3.40


8


Danny Duffy


Royals


9.42


2.10


4.48


1.35


25.7 %


5.8 %


20.0 %


.239


1.14


.291


80.9 %


81


91


89


3.51


3.83


-0.32


3.79


3.53


9


Corey Kluber


Indians


9.50


2.39


3.98


0.92


26.4 %


6.6 %


19.8 %


.214


1.06


.271


74.8 %


73


76


82


3.14


3.26


-0.12


3.50


3.50


10


Chris Archer


Rays


10.42


3.00


3.48


1.34


27.4 %


7.9 %


19.5 %


.235


1.24


.296


72.5 %


98


92


80


4.02


3.81


0.21


3.41


3.50

 
Top 10 in K-BB last year.  Which one of these things doesn't belong?


1


Jose Fernandez


Marlins


12.49


2.71


4.60


0.64


34.3 %


7.5 %


26.9 %


.220


1.12


.332


76.6 %


70


58


62


2.86


2.30


0.56


2.56


2.81


2


Max Scherzer


Nationals


11.19


2.21


5.07


1.22


31.5 %


6.2 %


25.3 %


.196


0.97


.255


81.7 %


71


79


82


2.96


3.24


-0.28


3.37


3.05


3


Noah Syndergaard


Mets


10.68


2.11


5.07


0.54


29.3 %


5.8 %


23.5 %


.240


1.15


.334


76.9 %


65


56


65


2.60


2.29


0.31


2.67


2.95


4


Justin Verlander


Tigers


10.04


2.25


4.46


1.19


28.1 %


6.3 %


21.8 %


.204


1.00


.255


79.9 %


72


81


89


3.04


3.48


-0.44


3.78


3.42


5


Madison Bumgarner


Giants


9.97


2.14


4.65


1.03


27.5 %


5.9 %


21.6 %


.209


1.02


.265


79.1 %


69


83


86


2.74


3.24


-0.50


3.54


3.36


6


Chris Sale


White Sox


9.25


1.79


5.18


1.07


25.7 %


5.0 %


20.7 %


.225


1.04


.279


76.6 %


78


79


84


3.34


3.46


-0.12


3.58


3.43


7


Michael Pineda


Yankees


10.61


2.72


3.91


1.38


27.4 %


7.0 %


20.4 %


.264


1.35


.339


70.7 %


114


86


78


4.82


3.80


1.02


3.30


3.40


8


Danny Duffy


Royals


9.42


2.10


4.48


1.35


25.7 %


5.8 %


20.0 %


.239


1.14


.291


80.9 %


81


91


89


3.51


3.83


-0.32


3.79


3.53


9


Corey Kluber


Indians


9.50


2.39


3.98


0.92


26.4 %


6.6 %


19.8 %


.214


1.06


.271


74.8 %


73


76


82


3.14


3.26


-0.12


3.50


3.50


10


Chris Archer


Rays


10.42


3.00


3.48


1.34


27.4 %


7.9 %


19.5 %


.235


1.24


.296


72.5 %


98


92


80


4.02


3.81


0.21


3.41


3.50
Jose Fernandez. 

 
I guess you would call me old school; I've been a fan since the late 60s and grew up on the traditional stats. Now, I have nothing against these new fangled stats; I'm really at a loss to understand them.  I do understand WAR, BABIP and OPS+, but FiP, wRC, etc. leave me completely shuked. Can someone explain to me how I can understand these stats so I can read a dang baseball blog with blindly nodding like a drooling fool?
Do what I do and don't care about them :)

 
The game can entertain on multiple levels
Yep.  Was never a deep stat person except for the basics.  I do enjoy more scouting stuff as holes in swings and stuff.

All my years of playing I really cared about avg, obp and steals :)

 
It still comes down to players making plays.  I doubt Brett Lawrie can spell most of these advanced stats.

 
My ideal defensive stat would measure the amount the defensive player had to move to catch a ball and the amount of time he had to get there.  It would be nice to know whether that ball that just dropped in for a hit could have been caught by an above average defender.  A super fast player who does not get a good jump and routinely misjudged balls off the bat, could be over-rated just based on speed.  It seems like we have the technology now, we should be able to come up with metrics which would tell us for a given ball in play, which players we would expect to make the play and which players it would be impossible for. 

 
Good topic.

I would be interested to hear what advanced stats people use for fantasy. The last couple of seasons I have paid close attention to contact-rate and line-drive percentage when seeking players for batting average.
I input a dozen or so into a spread then look for outliers. They all have value, but they all have flaws too. Need the whole picture to figure out what doesn't look right given the rest of that players profile. 

 

Users who are viewing this thread

Top