After the Red Sox scored two runs off Mariano Rivera in the top of the ninth -- stealing four bases in the span of 11 pitches! -- and took a 3-2 lead, Jonathan Papelbon got the ball for the bottom of the ninth. And got royally fucked by home plate umpire Phil Cuzzi.
(Charts are from the amazing Brooks Baseball site.)
Derek Jeter fouled off a fastball and flew out to right. One out.
Nick Swisher got ahead 3-0 (ball one should have been called a strike). Papelbon battled back to 3-2 before Swisher lined a single to right field. Marco Scutaro dove to his left, but it was just out of his reach.According to f/x, Papelbon struck out Alex Rodriguez on three pitches. However, Cuzzi called two of those pitches balls -- and Slappy eventually walked on strike 5, loading the bases.
Papelbon threw strike one to Robinson Cano, but Cuzzi called it a ball. After a pitch in the dirt, Cano singled to right to tie the game at 3-3.Cuzzi blew two pitches in Jorge Posada's at-bat, calling a ball on strike 1 and not ringing up Dumbo on pitch #4. Posada ended up swinging and missing on the next pitch for the K.Papelbon's 1-0 pitch to Lance Berkman was well within the zone, but Cuzzi muffed that one, too. Berkman flew out to right field to end the inning.I counted eight pitches that Cuzzi botched. Of Papelbon's 30 pitches in that inning, the Yankees took 19 of them -- so Cuzzi made the wrong call on 42% of the time.Going back to the top of the ninth, I was flat-out shocked that Cuzzi did not call a handful of pitches off the plate strikes, as umpires often do for Rivera. I went back and looked at the f/x data for the Red Sox at-bats.
First of all, Rivera came into the game with two outs in the eighth. Adrian Beltre was up with runners at first and second. It was the 2-2 pitch that I expected to be called a strike, but f/x does have it well outside. Beltre grounded out to second.Jed Lowrie began the ninth by belting a 0-1 pitch to deep right-center that Swisher caught with an ungainly effort on the warning track.
Ryan Kalish took a ball and a strike and singled to center.Cuzzi may have blown three of the four pitches Bill Hall took in his at-bat. Pitches 2, 4, and 5 should have been strikes 2 and 3 (and 4!) -- but they were called balls. Kalish stole second on the first pitch and stole third on the fourth. With a full count, Hall singled to left to tie the game. He should have been the second out.Mike Lowell batted for Lars Anderson. Hall stole second on the second pitch and stole third on the fourth pitch. It looks like Cuzzi got Rivera's 2-0 pitch wrong, calling it a ball when it should have been a strike. Would Hall have tried to steal third on 2-1? Maybe not. Lowell's sac fly to center gave Boston a 3-2.Rivera went right after Scutaro, with a called strike, three fouls, and a fly ball to right-center.It looks like Rivera was robbed on four pitches, three of them during the important Hall at-bat. All four were in the same lower corner of the strike zone.I looked at Okajima's pitches in the 10th, and Cuzzi got every call right.Cuzzi still fucked up, though. After Curtis Granderson singled, Brett Garnder bunted. Gardner ran to first on the inside of the baseline -- that is against the rules -- but Cuzzi failed to call it. Martinez's throw hit Gardner and the MFY had runners at first and third with no one out.
Every pitch that Rivera and Papelbon threw should have been judged according to the actual strike zone, but they was not. It sure looks like Hall and Rodriguez should have been called out on strikes rather than singling and walking, respectively. I do not recall Cuzzi's calls being so horrifically bad in the previous eight innings.
I know I'm a fucking broken record when it comes to this. Give me a process that gets the correct calls as close to 100% of the time as possible. It has been shown time and time again that humans cannot do it. So let's use technology. In track meets, we don't have guys muttering "one one thousand, two one thousand, three one thousand ...", we have sophisticated machines that can tell us, with certainty, the winning time was 28.735 seconds.