White Sox Rotation vs. Cardinals Rotation

[OmarComing25]

I was looking through Fangraphs today and I randomly noticed that the White Sox are tied for 2nd in WAR for starting pitchers with the Cardinals (0.3 behind the Dodgers).

 

http://www.fangraphs.com/leaders.aspx?pos=...=&players=0

 

Cardinal's starters' ERA: 2.75 (3.31 FIP)

White Sox starters' ERA: 4.22 (3.63 FIP)

 

Heck, even over the last 30 days we're 4th in WAR despite an ugly 4.95 ERA.

 

I know we have an advanced stat thread but I thought this warranted a separate discussion. I'm as big of a proponent of advanced stats as anyone, but it still just seems completely ridiculous to me to claim our rotation has performed as well as theirs (just like you can't convince me Samardzija has been the 38th best starting pitcher this season). Yet, when you really look into it, the gap isn't as big you might think at first glance. Our defense has been mostly s***, and the Cardinals have experienced a boatload of "cluster luck" (they're on pace to shatter the all-time record for preventing runs with RISP, see link below). On the other hand that ERA gap is just so large that it feels hard to argue the rotations are even kind of comparable.

 

I'm interested to see others' thoughts on this. What would a Fangraphs writer say? Can we really just explain that ERA gap away so easily?

 

http://grantland.com/the-triangle/2015-mlb...s-cluster-luck/

Edited August 19, 2015 by OmarComing25

[cabiness42]

In MLB in 2015, Sox starters are:

 

3rd in IP

3rd in K/9

7th in HR/9

9th in BB/9

8th in FIP

11th in HR/FB

 

So how is a starting rotation with those numbers only 13th in ERA?

 

26th in GB%

25th in LOB%

29th in BABIP

 

Translation: While the starting rotation is excellent at striking guys out and not walking them, a very low percentage of the guys they don't strike out are hitting the ball on the ground. Compound that with poor defense and you have an extremely high percentage of ball in play ending up as hits. That is compounded with the fact that the rotation seems to get a lot worse once guys are on base.

 

I think WAR is probably overvaluing the role that poor defense and luck play into the elevated ERA and undervaluing a bit the role of the pitchers themselves, but there is no doubt that the rotation has pitched better than their ERA shows.

[Balta1701]

I took a swipe at these numbers this morning just to see if there was some additional insight that could be gained.

 

I compiled the WAR stats for pitchers from Baseball-Reference to compare to fangraphs. B-R seems to me to underweight things like strikeouts in favor of overweighting what actually happens on the field whereas I've always felt that Fangraphs underestimates some things happening on the field in favor of value for pitching a lot of innings and piling up strikeouts - Fangraphs to me might underestimate the skill of creating weak contact or keeping the ball where your fielders are playing while B-R may overestimate it.

 

It's a little hard to make a direct comparison because I had to use "wins above average" for B-R since that's what's in the table. However, there are some consistencies - the 2 datasets are strongly correlated with an R^2 of 0.69 (statistics, ignore if you don't care). the same teams are on top and on the bottom in both. However, there is a group of teams - the Padres, Tigers, Reds, Marlins, and White Sox - all of whom are significantly worse on the B-R numbers than they are on the FG numbers, enough to constitute a notable trend. The White Sox have the largest separation in baseball between their B-R numbers and their FG numbers - Fangraphs thinks they're very good, B-R thinks they're mediocre.

 

These teams have in common that they're generally not that good, but I have a tough time making a case that this separation is all due to the teams having a crappy defense. In fact, fangraphs's defense stats find 3 of those 5 teams to be in the upper half of the league, only the White Sox and Padres are thought to have generally poor defenses.

 

There also is basically no correlation between being in this group of teams and having a high babip - the White Sox have one of the worst babip's in the game as noted above and the Tigers are weak on that also, but the Padres are about the middle of the league and the other 2 teams are above average.

 

I basically can't find any single metric where these teams notably stand out or group together in the rankings for their pitching staffs except for the fact that they're all pretty lousy ballclubs. The White Sox, for example, don't have a notably bad ground ball rate per the team stats, they're 11th out of 30 teams. They don't have particularly bad HR/FB rates, they don't generally cluster with low rates of leaving people on base.

 

Also to note - HH, I'm not sure where you're seeing those numbers, but under Fangraphs - team stats - pitching - I see the Sox ranking very different from what you just posted.

 

I don't think this is a case of it being clear-cut the White Sox poor defense, I think there's something systematic in the way these 2 different metrics calculate wins that is separating these teams out. Finding out what that is might actually be an interesting exercise because of the fact that the teams standing out are all out of the pennant races in their respective leagues - if you want to be competitive you don't want to be in this group.

 

If anyone has any ideas for other things I could look at to try to see where this group might separate itself out, I'd love suggestions, I'm out of ideas.

[cabiness42]

I took a swipe at these numbers this morning just to see if there was some additional insight that could be gained.

 

I compiled the WAR stats for pitchers from Baseball-Reference to compare to fangraphs. B-R seems to me to underweight things like strikeouts in favor of overweighting what actually happens on the field whereas I've always felt that Fangraphs underestimates some things happening on the field in favor of value for pitching a lot of innings and piling up strikeouts - Fangraphs to me might underestimate the skill of creating weak contact or keeping the ball where your fielders are playing while B-R may overestimate it.

 

It's a little hard to make a direct comparison because I had to use "wins above average" for B-R since that's what's in the table. However, there are some consistencies - the 2 datasets are strongly correlated with an R^2 of 0.69 (statistics, ignore if you don't care). the same teams are on top and on the bottom in both. However, there is a group of teams - the Padres, Tigers, Reds, Marlins, and White Sox - all of whom are significantly worse on the B-R numbers than they are on the FG numbers, enough to constitute a notable trend. The White Sox have the largest separation in baseball between their B-R numbers and their FG numbers - Fangraphs thinks they're very good, B-R thinks they're mediocre.

 

These teams have in common that they're generally not that good, but I have a tough time making a case that this separation is all due to the teams having a crappy defense. In fact, fangraphs's defense stats find 3 of those 5 teams to be in the upper half of the league, only the White Sox and Padres are thought to have generally poor defenses.

 

There also is basically no correlation between being in this group of teams and having a high babip - the White Sox have one of the worst babip's in the game as noted above and the Tigers are weak on that also, but the Padres are about the middle of the league and the other 2 teams are above average.

 

I basically can't find any single metric where these teams notably stand out or group together in the rankings for their pitching staffs except for the fact that they're all pretty lousy ballclubs. The White Sox, for example, don't have a notably bad ground ball rate per the team stats, they're 11th out of 30 teams. They don't have particularly bad HR/FB rates, they don't generally cluster with low rates of leaving people on base.

 

Also to note - HH, I'm not sure where you're seeing those numbers, but under Fangraphs - team stats - pitching - I see the Sox ranking very different from what you just posted.

 

I don't think this is a case of it being clear-cut the White Sox poor defense, I think there's something systematic in the way these 2 different metrics calculate wins that is separating these teams out. Finding out what that is might actually be an interesting exercise because of the fact that the teams standing out are all out of the pennant races in their respective leagues - if you want to be competitive you don't want to be in this group.

 

If anyone has any ideas for other things I could look at to try to see where this group might separate itself out, I'd love suggestions, I'm out of ideas.

 

You used entire team pitching stats. I was looking at starters stats since the OP was specifically talking about starting rotations.

[Balta1701]

You used entire team pitching stats. I was looking at starters stats since the OP was specifically talking about starting rotations.

Ah, thanks.

[Eminor3rd]

I compiled the WAR stats for pitchers from Baseball-Reference to compare to fangraphs. B-R seems to me to underweight things like strikeouts in favor of overweighting what actually happens on the field whereas I've always felt that Fangraphs underestimates some things happening on the field in favor of value for pitching a lot of innings and piling up strikeouts - Fangraphs to me might underestimate the skill of creating weak contact or keeping the ball where your fielders are playing while B-R may overestimate it.

 

Another way of putting it is that FanGraphs assumes the pitcher gets NONE of the credit for what happens wit the defense, and B-R assumes that the pitcher gets ALL of the credit. Both entities acknowledge that the truth is somewhere in the middle, but they don't agree on how to deal with the error. B-R thinks that your WAR isn't useful unless it's a complete answer, even if that means that there's significant error in the actual measurement. FanGraphs would rather give you an incomplete number and say "hey this doesn't tell the whole story, but what it DOES tell you is accurate" and then have you insert your own caveats for the rest. The "rest" essentially comes down to a pitcher's ability to induce weak contact and/or control where the ball lands.

 

Most people (myself included) tend to favor the FG approach, and just acknowledge that there will be exceptions to the established norms.

[cabiness42]

Another way of putting it is that FanGraphs assumes the pitcher gets NONE of the credit for what happens wit the defense, and B-R assumes that the pitcher gets ALL of the credit. Both entities acknowledge that the truth is somewhere in the middle, but they don't agree on how to deal with the error. B-R thinks that your WAR isn't useful unless it's a complete answer, even if that means that there's significant error in the actual measurement. FanGraphs would rather give you an incomplete number and say "hey this doesn't tell the whole story, but what it DOES tell you is accurate" and then have you insert your own caveats for the rest. The "rest" essentially comes down to a pitcher's ability to induce weak contact and/or control where the ball lands.

 

Most people (myself included) tend to favor the FG approach, and just acknowledge that there will be exceptions to the established norms.

 

You could just give the pitcher half of the credit for what happens with balls in play and you probably get closer to the truth than B-R or FG.

[OmarComing25]

You could just give the pitcher half of the credit for what happens with balls in play and you probably get closer to the truth than B-R or FG.

So just take the average of the two WAR numbers? I think that's probably the best way to give credit to what "should" have happened (FG) and what "actually" happened (BR).

[Balta1701]

So just take the average of the two WAR numbers? I think that's probably the best way to give credit to what "should" have happened (FG) and what "actually" happened (BR).

If that were the case I'd think that the data I looked at yesterday would look different. Instead there are some teams where this is a big deal and some teams where it's not a big deal at all.

[OmarComing25]

After Sale's 14K performance yesterday The Sox starting rotation has now passed the Dodgers and Cardinals for the top spot in the MLB.

 

http://www.fangraphs.com/leaders.aspx?pos=...0&sort=11,d