Tuesday, August 27, 2013

MLB Ejection 148: Greg Gibson (3; Joe Maddon)

HP Umpire Greg Gibson ejected Rays Manager Joe Maddon for arguing a strike three call in the top of the 4th inning of the Rays-Royals game. With two out and three on, Rays batter David DeJesus took a 1-2 slider
Maddon expresses disagreement with Gibson.
from Royals pitcher Jeremy Guthrie for a called third strike. Replays indicate the pitch was located letter-high and over the inner half of home plate (sz_top 3.24, pz 3.438), the call was incorrect. At the time of the ejection, the Royals were leading, 5-1. The Royals ultimately won the contest, 11-1.

This is Greg Gibson (53)'s third ejection of 2013.
Greg Gibson now has 6 points in the UEFL (8 + 2 + -4 N = 6).
Crew Chief Jerry Layne now has 10 points in the UEFL's Crew Division (10 + 0 Incorrect Call = 10).

This is the 148th ejection of 2013 MLB season.
This is the 70th Manager ejection of 2013.
This is the Rays' 6th ejection of 2013, 3rd in the AL East (TOR 12; BOS 8; TB 6; BAL 5; NYY 3).
This is Joe Maddon's 4th ejection of 2013 and first since May 8, 2013 (Scott Barry; QOC = Incorrect).
This is Greg Gibson's first ejection since August 23, 2013 (Robin Ventura; QOC = Correct).

Wrap: Tampa Bay Rays vs. Kansas City Royals, 8/26/13
Video: After a called third strike against DeJesus to leave three men on base, Maddon earns a heave-ho (TB)

28 comments :

Lindsay said...

I saw an espn replay that showed their K zone and the ball was in the zone. I say too close to take with 2 strikes. Swing the dn bat

Lindsay said...

Wonder if the magic words were, "I love you".

Lindsay said...

Maddon is such a drama queen.

Lindsay said...

Question for Gil that doesn't have anything really to do with this play: How do you determine a particular hitter's bottom and top of the zone numbers with the Pitch f/x system? For example, if I pull up a hitter's at-bat on Pitch f/x and it shows the rectangle box, how do I know that hitter's actual bottom and top "limits" of their zone so I don't have to go off of the box alone? I know you guys have said before that the box doesn't really tell the true story of the zone vertically.

Lindsay said...

Yeah, that wasn't "letter high." It's only "letter high" when he squats down for a moment to watch the pitch go by, which is clearly not his his position "as he is prepared to swing." You don't get to duck under a pitch as it passes and change the strike zone, you guys running this thing should know better than that. Go adjust your plot to when he is ready to swing at the pitch and then run your little pitch f/x thing again, which technically is also irrelevant considering that's not what MLB uses anyways and these guys could care less about your pitch f/x. Although congrats to Fagan for having a steller game, not because your plot said he did, but because actual umpires know he did. But what do I know, I'm just a lowly "guest." Someone who's actually in this league needs to challenge based on those parameters.

Lindsay said...

I had it as a strike, quit looking at strikes then blaming the umpire.

Lindsay said...

http://oi44.tinypic.com/2czdq1y.jpg

Lindsay said...

While I have not taken the time to watch any other of DeJesus' at-bats, I have to agree he does appear to duck the pitch and without the ducking the pitch looks like a strike.



I would be interested to know if the pitch f/x numbers used were normalized for all of DeJesus' at-bats or if they were normalized for this specific instance. If it's true that he ducked the pitch and this was not his normal pre-swing stance, we should be sure to use the normalized numbers for all of the pitches to him.

Lindsay said...

The batter's sz_bot and sz_top are determined during the first pitch of the at bat in which the batter enters his natural hitting stance, when he is prepared to swing.


There are two zones, normalized and non-normalized.
Normalized = Relative accuracy; distances are inaccurate but K Zone is right.
Non-Normalized = Absolute accuracy; distances are accurate but K Zone is wrong.

Lindsay said...

How do you find each batter's sz_bot and sz_top for a particular at bat?

Lindsay said...

MLB Gameday, gd2.mlb.com/components/game/mlb/


Pitch 1) DeJesus's sz_top was 3.35.
#2) sz_top was 3.19
#3) sz_top was 3.24
#4) sz_top was 3.24 (carried from #3 since #4's stance was thrown out).


#4's pz was 3.438, which is higher than DeJesus' sz_top for any of the pitches during this AB, including pitch #1, a foul ball.

Lindsay said...

Pitch 1) DeJesus's sz_top was 3.35.
#2) sz_top was 3.19
#3) sz_top was 3.24
#4) sz_top was 3.24 (carried from #3 since #4's stance was thrown out).
^Emphasis on the fact that #4's stance was thrown out and replaced with #3...

#4's pz was 3.438, which is higher than DeJesus' sz_top for any of the pitches during this AB, including pitch #1, a foul ball.

Lindsay said...

Pitch FX on Brooks Baseballs had it as a strike. Not sure why it is a ball here?

Lindsay said...

Thanks Gil...the number's don't lie. I have to say that Gibson's job was made very difficult by the ducking of the pitch. I most likely would have called the pitch a strike, but I would have been wrong.

Lindsay said...

Brooks uses a non-normalized plot for individual ABs, which is absolute in that distances are accurate (e.g., the pitch in question was 3.438 feet above home plate), but the strike zone boundaries as drawn are static—the same for all batters—and therefore inaccurate. The non-normalized AB assumes every batter's strike zone runs from 1.500 feet to 3.500 feet, which fails to take into account the individual batter's stances.

A normalized plot runs a normalized "stretched" y-scale, such that the strike zone runs from -1.000 normalized units to +1.000 normalized units. Therefore, while the distances are inaccurate, the strike zone itself is accurate because it accounts for batter individuation.

sz_bot (hollow of the knee) and sz_top (midpoint) are variables that are not included on the non-normalized plot but are present and vital to the normalized plot. pz, which represents the pitched ball's height as it crosses the front edge of home plate, is used in both graphs (though pz becomes norm_ht in the normalized plot dependent on the amount of "stretch").

Meanwhile, because such graphs are misleading if used improperly (as in Brooks' non-normalized plot for strike zone analysis), the UEFL membership in last year's rules summit adopted a change to the Miller Rule 6-2-b-2; we now go straight to the horse's mouth on issues of balls/strikes and use raw numbers. With regard to the PFX margin of error, the general rule employed for QOC is that a called ball (for instance) with a pz value greater than sz_bot is likely incorrect whereas a called strike with a pz value greater than sz_bot is correct. (pz > sz_bot = likely strike; pz < sz_bot = likely ball).

To illustrate the difference for the DeJesus at-bat, in ascii art form...

Non-normalized (the Brooks chart):
------ top of strike zone
o pitch

------ bottom of strike zone

Normalized (the PFX/Gameday chart):
o pitch
----- top of strike zone

----- bottom of strike zone

Lindsay said...

So.. The only accurate and impartial method we have of determining where the pitch is located in the strike zone is irrelevant? Good to know! What shall we use instead?

Lindsay said...

And your opinion of the pitch?

Lindsay said...

I do not argue the numbers.

Lindsay said...

I use this from brooks baseball to get my numbers. You have to click on the individual pitcher and know the pitch sequence to the batter to find the pitches you are looking for but I seem to get the same numbers Gil does. This is an example of one below.




http://www.brooksbaseball.net/pfxVB/tabdel_expanded.php?pitchSel=425386&game=gid_2013_08_26_tbamlb_kcamlb_1/&s_type=&h_size=700&v_size=500

Lindsay said...

they ALL are - but Maddon is a prime example! Terry Collins is getting there also...

Lindsay said...

Ahhh thanks. That's what I was looking for. I wasn't sure how he was finding the raw numbers for the bottom/top of the zone on each batter.

Lindsay said...

I don't know, man. If the guy is not crouching like that (which does not alter the zone anyway), it is difficult to tell. But I imagine if he were standing upright that would be right at his waist. A strike.


Regardless, arguing that shit is doltish. It's like trying to talk your way out of a speeding ticket.

Lindsay said...

The numbers don't lie, but they are prone to lead to misinterpretation. In this particular case the fact that the pitch is off speed means that it likely would have a lower plot number if plotted at the back of the 3 dimensional strike zone.
Additionally, since he swung at pitch 1 I would suggest that is his "real" sz_top and not any of the subsequent "take" pitches. If that were used as his "proper sz_top then the difference between it and the actual pz of the pitch (3.428) would be only .088 which would make the pitch fall within the radius of the baseball (meaning the bottom edge of the ball passed through the top edge of the zone at the front edge of the plate).
I'm not playing the league but if I were I would argue that sz_top for a pitch should be measured only on pitches at which the batter offered, and those values be used for pitches taken since a batter is least likely to "duck" on a pitch at which he swings.
In practice, on the diamond, I have no problem with this being a strike.

Lindsay said...

Dale Sveum was just tossed by Barksdale arguing a check swing. Lance called a no swing during Puig's at-bat, but he clearly went. You never here much about Barksdale but he has not had the best season.

Lindsay said...

I agree with you. He's had somewhat of a down year and the picture shows it was clearly a swing by puig

http://gyazo.com/cee0f63436231e1fc45930dd8478a91f

Lindsay said...

Which zone do you use when calculating the umpire's accuracy behind the plate, normalized or non-normaized?

Lindsay said...

I have said this for some time now. Barksdale has not impressed this season.

Lindsay said...

Normalized

Post a Comment