THE BOOK--Playing The Percentages In Baseball

There is certainly some luck involved here. His FB BABIP of .086 vs. career .132, his LD BABIP is .611 vs. .763 for his career. The HR/Air Batted ball is 2.3% vs, 5.4% for his career (calculated quickly).

I’d be more interested in after how many batted balls we start strongly considering his tendencies instead of considering the league (for HR rate). Basically the Matt Cain question.

Okay, look.

Jered Weaver, HR/CON, career: 0.037910323

League, HR/CON, weighted by Weaver’s contact allowed in those seasons: 0.037057946

You can do the exact same thing with Matt Cain. They don’t have some mysterious home run skill, people just keep expecting their HR rates to regress to some utter fiction based on batted ball data that we know to be systemically biased. The real Matt Cain question is why do we still use junk stats like HR/FB at all.

Colin,

Are you rejecting the idea that Jered Weaver is a fly ball (air ball, whatever you want to call it) pitcher? That seems, well, dumb. If he is and he allows HR/Con at a league average rate, well that does mean that he allows fewer HR/FB (Air Ball, whatever) than the average pitcher. Now of course we know that this FB pitchers on average allow fewer HR/FB, so maybe he isn’t special in this sense, but that’s not really the point you seem to be trying to make.

I wouldn’t come to the conclusion about “dumb”.  I think it would be more helpful to list the Angels starters over the last 5 years, and show the percentage of their batted ball outs that came from outfielders and infielders.  And just let the evidence speak for itself as to whether Weaver is or is not a flyball pitcher.

What’s a Kershaw gotta do to get some love around here?

When I ran my poll last year for best young pitcher you’d like to have, I had Felix, Josh Johnson, Verlander, Lincecum (NO WEAVER), plus Ubaldo, Kershaw, Greinke, and Strasburg on the ballot.

http://www.insidethebook.com/ee/index.php/site/article/poll_survivor_pitcher_island/

But I agree, I don’t mention Kershaw (or Ubaldo) enough.

Tango/4,

2010:

http://mlb.mlb.com/stats/sortable_player_stats.jsp?c_id=mlb§ion2=1&statSet2=2&sortByStat=GO/AO&statType=2&timeFrame=1&timeSubFrame=2010&baseballScope=mlb&prevPage2=1&readBoxes=true&sitSplit=&venueID=&subScope=teamCode&teamPosCode=LAA

2009:

http://mlb.mlb.com/stats/sortable_player_stats.jsp?c_id=mlb§ion2=1&statSet2=2&sortByStat=GO/AO&statType=2&timeFrame=1&timeSubFrame=2009&baseballScope=mlb&prevPage2=1&readBoxes=true&sitSplit=&venueID=&subScope=teamCode&teamPosCode=LAA

2008:

http://mlb.mlb.com/stats/sortable_player_stats.jsp?c_id=mlb§ion2=1&statSet2=2&sortByStat=GO/AO&statType=2&timeFrame=1&timeSubFrame=2008&baseballScope=mlb&prevPage2=1&readBoxes=true&sitSplit=&venueID=&subScope=teamCode&teamPosCode=LAA

Every year, Weaver is at the bottom of GO/AO.

Dave G: Looking at it again, I think it was pretty good that I had Kershaw on the ballot last year.  But, yeah, I don’t talk about him enough.

And the reason I didn’t have Weaver (not that you asked) was that he was too old.  I didn’t realize how old he is (b. 1982) compared to Felix (1986) and Strasburg (1988).

***

David Gassko: great stuff.  That’s how you make a case!  So, definitely an airball pitcher within the universe of Angels pitchers from 2008-2010.  And unless the Angels pitchers were big time GB pitchers, then Weaver must be considered an airball pitcher among MLB pitchers.

So, he gives up much more airballs than normal, but also gives up much less HR per airball than normal. 

And Colin’s stat showing his HR per contacted ball being exactly league average shows that those two facts cancel out!  Love it.

By the way, that a great pitcher like Weaver who gives up alot of airballs but not alot of HR obviously means that the distance of his airballs are shorter than other pitchers.  That is, it’s not surprising that his airballs don’t go out of the park.

By the way, that a great pitcher like Weaver who gives up alot of airballs but not alot of HR obviously means that the distance of his airballs are shorter than other pitchers.

Hold on, let’s see the home/away split before we go there.  He plays half his games in one park that is a bit deeper than average in most directions…

Excellent point, Greg.

Weaver’s got pretty decent home/away splits.  His K’s slightly more at home, but gives up alot fewer walks and alot fewer HR.

Could be he’s pitching better, or he’s leveraging his park better.  Or it’s luck.

Let’s wait until Kershaw produces another season similar to the last before we put him into the top tier of starting pitchers.

Why do we have to wait?  That was the whole point of that poll with Strasburg included that you don’t have to wait, that you can reasonably decide where he can rank with limited performance numbers, if you focus on his tools as well.

This is the way it works: how much would the guy get paid if he were to sign a free agent contract for the next 4 years?(*) You go through each player, and you give him a dollar value.  And that’s how much he’s worth.  And that’s where you rank him.  It’s as easy as that.

(*) Maybe you do 5 or 7 years instead.  It won’t change the answer much.

@Colin/2 (or anyone)

Do you still believe the hypothesis laid out in this piece?

http://www.hardballtimes.com/main/article/when-is-a-fly-ball-a-line-drive/

After reading through this thread, I recalled this article you wrote and revisited it. Is this the systematic bias you claim exists in batted airball data?

I haven’t been around here in awhile, but how would percentage of batted balls hit in the air be biased?  The distinction between a ground ball and a ball in the air is incredibly easy to make, and i highly doubt it’s suseptible to bias.

I haven’t been around here in awhile, but how would percentage of batted balls hit in the air be biased?  The distinction between a ground ball and a ball in the air is incredibly easy to make, and i highly doubt it’s suseptible to bias.

Hi, Nick!

I’m not sure it’s correct to boil the issue here down only to the question of distinguishing between air balls and ground balls.  (Nor am I necessarily saying you’re doing that, Nick.)

At a minimum, there’s also the question at the other end of the spectrum--distinguishing between fly balls and popups.

But I believe Colin’s original point was at least partly about distinguishing between line drives and fly balls, because when most people talk about HR/FB rates, they are excluding line drives from that discussion.  That’s when Tango suggested backing off to including all air balls. (BTW, it’s not clear to me which types of air balls MLB.com includes in their “AO” category.  Does anyone know?)

But even beyond the LD/FB distinction and the FB/PU distinction, I think this discussion has to include ballpark (which has been mentioned here) and directionality of the batted balls (pull field, opposite field, etc.).

Now, having said all that, I don’t buy the claim that distinctions between ground balls and line drives are “incredibly easy to make”.  I’ve found some evidence in the HITf/x data of catch/no-catch bias in scoring line drives versus ground balls, for instance.  The dividing line between them for the MLB scorers seems to be around a vertical launch angle of 6-7 degrees, but I don’t see any reason why that wouldn’t be susceptible to bias just as other batted ball classifications are.

I agree that in most cases, it is easy to make.  But not always.

Take for example if a SS dives for a ball, or traps it.  You can call them both liners, but the second one will be called a groundball because it hit the ground in the infield.

My definition of a “groundball” would be one where the ball would have landed in the infield if not for the fielder being involved, and the ball never going above a certain height (say 6 feet or 8 feet or something).  The designation of the word “ground” is irrelevant, and only used as a generally useful word.

So, in this illustration, these two balls would be called “groundballs”.  If that’s too hard to accept as a redefinition, then call them “low infield balls”.

There are some borderline balls where line drive/groundball is difficult to determine.  Data do not have to be perfect to come to a general conclusion that Weaver allows more flyballs than most, however.

If it’s misclassified line drives and popups, then does Weaver allow fewer of these than average, thereby inflating his flyball count?

Looking at the wording Colin chose to use in #2 makes it sound like batted ball data are so useless that even saying that Weaver allows more flyballs than Derek Lowe cannot be concluded.

The other source of error is one of transcription.  You can intend to write GB, but mark it something else instead.  Or, not even pay attention, and then have to write something.

At least with outs, there’s a strong incentive to get those correct.

Some good stuff from Jeremy:

http://www.baseballprospectus.com/article.php?articleid=13730

Just bumping, in relation to the other thread.