If Replay Existed: Retro Recap, COL @ SF, 5/25/2013


Credit: Bob Stanton-USA TODAY Sports

As Colorado sports fans collectively get over their figurative and/or literal Super Bowl hangovers, it’s a good time to think about something that baseball now has in common with football: an instant replay system based on challenges. Major League Baseball introduced the system two weeks ago. Others have identified how instant replay might have swung World Series games (and even a World Series itself) and critical playoff games. And, of course, we’ve been reminded of Armando Gallarraga’s perfect game that wasn’t. Still others have pointed out high profile instances where instant replay couldn’t change anything. Some see it as a good thing, while others would rather have the old system stick around. I’m in the former camp, and I’ll explain why below. But what I really want to explore is how instant replay might have played a role, or not, in a semi-randomly chosen game from last year: the Colorado Rockies at the San Francisco Giants on May 25, 2013.

The game was not chosen totally at random. I had a few conditions while scouring through Baseball Reference’s 2013 game results. I wanted to identify a close game separated by, at most, two runs. Second, I searched for a game with not too many strikeouts and a lot of hits, as more balls in play would provide more opportunities for a missed call on the base paths. Finally, I wanted a game that was relevant at the time. Because I limited myself to the Rockies, that meant games in April or May. I settled on the ten inning May 25 game, in which the Rockies had five runs on 11 hits, and the Giants six runs on 13. At the time, the Rockies were tied with the Giants and Diamondbacks for first place, so it was a high stakes game in May. As a bonus, I remembered this game because it was the one where Angel Pagan hit an inside-the-park walk-off home run and caused Duane Kuiper to have a conniption on air. It was fun, despite the Rockies loss.

In re-watching the game, I kept an eye out for any reviewable play that might have been challenged from innings one through six, and whether or not it would have been worth it to challenge the play. Per the rules, from the seventh inning on all reviews are initiated by the crew chief. I watched for anything reviewable in those innings as well. I was most interested in identifying anything that could have altered the course of the game. If you haven’t internalized it yet, you can read the full explanation of the replay system here.

The game’s umpires were Alfonso Marquez behind home plate, Chris Conroy at first, Mike DiMuro at second, and crew chief Ted Barrett at third. Barry Zito and Juan Nicasio started the game—hence all of the balls in play. To be honest, when I decided to do this exercise I thought that there would be one, maybe two inconsequential calls that could have been reviewed but without materially affecting the game. But that was not the case at all with this game. There were three plays in the first six innings that could have been challenged with good reason, and then there were two more from the seventh inning on.

None of the calls took place in a vacuum, so we have to know what’s going on, as the score and situation influence whether or not a play will be challenged. So it is important and satisfying to note that in the top of the first inning

Carlos Gonzalez

hit a glorious two-run home run to right field. It was, in fact, the first time someone from the Rockies hit one into the Cove. The evidence:

Heading into the bottom of the first, it was 2-0 Colorado, and it was here that the first close play, one that might be challenged in 2014, occurred. Angel Pagan led off with a double. Marco Scutaro then generously laid down a sacrifice bunt. But Nicasio was quick to the ball and threw Pagan out at third. Probably. He was called out at least, and the home team broadcasters, pre-hysterical Kuiper and Mike Krukow, didn’t object. “It could have gone either way,” Krukow said, “but I think they had him.” Baseball Reference’s win probability added suggests that this out reduced the Giants chances of winning by six percent.

If this were 2014, would Bruce Bochy challenge? Dan Brooks and Russell Carleton argue that mathematically, managers should challenge every close play regardless of when it occurs because fewer chances will present themselves as a game progresses. Then there is the matter of the means to win a game: runs. The difference, for example, between a single and an out is 0.70 runs. As they say, that’s not nothing. In this specific case, if the call were reversed the Giants would have had runners on first and second with nobody out in a game where they were already down two runs. The next four batters were Pablo Sandoval, Buster Posey, Hunter Pence, and Brandon Belt. Baseball Prospectus’s run expectancy matrix for 2013 indicates that a situation such as that, without even considering the specific context, was worth 1.6 runs. So yes, if this game were to take place in 2014, Bochy should most certainly challenge. The risk would have been worth taking, even though he probably would have lost the challenge.

A 2010 study by ESPN’s Outside the Lines concluded that any given game has an average of 1.3 calls close enough to warrant review. On May 25 at AT&T Park, there was two in the bottom of the first inning. After the force at third, Sandoval came up and knocked a single to left, moving Scutaro to second. Runners on first and second, one out, and Posey up. This is where the second reviewable play happened. Posey hit a hard groundball to Nolan Arenado at third; he scooped it up, stepped on third, jump threw to Jordan Pacheco at first for a close play. Posey was called safe, as it seemed that Arenado’s throw drew Pacheco off of the bag just enough. But replays showed that he was on the bag, and Kuiper and Krukow agreed that they had Posey by a quarter step. This is the type of bang-bang play that is likely to be reviewed most often. Nobody can legitimately fault Chris Conroy for missing it, but replay showed that Posey was out. 2014 Walt Weiss most certainly challenges this play (I hope at least), and it is almost certainly overturned.

The thing about reviewed plays is that they have a cascading effect, dramatic or subtle. If Arenado’s play were overturned it might have changed the course of the game. In 2014, we might not notice exactly how, but we can guess what might have happened in a game from last season.

Here’s what we know did happen. After Posey was called safe, Nicasio walked Hunter Pence, but then Brandon Belt struck out looking. The call cost Nicasio two extra batters faced and ten pitches, finishing up the first inning with 23 rather than 13. The Rockies put up two more runs in the fourth inning, and the Giants went scoreless until the bottom of the sixth, where they entered with a win expectancy of six percent. Nicasio promptly gave up two doubles before being removed, then the momentarily strike-zone allergic Josh Outman and Adam Ottovino finished up in this manner: single, walk, sacrifice fly, walk, pop-up, walk for a run, strikeout, ultimately yielding three runs.

Now let’s start making assumptions. Assuming that Posey was called out in the first inning and that every batter each Rockies pitcher faced faired the same as he actually did, but pushed forward in the game, what would have happened? Let’s also assume that there are no double plays, and that runners on first don’t score on doubles. Here is what it would look like:

Bottom of second: walk, strikeout, strikeout, double, groundout—no runs

Bottom of third: popout, lineout, popout—no runs

Bottom of fourth: single, popout, groundout, popout—no runs

Bottom of fifth: double, groundout, groundout, flyout—no runs

Bottom of sixth: flyout, double, double, single, walk, flyout, walk, flyout—two runs

Two runs instead of three. Of course, our assumptions cannot be taken at face value. At bats would not necessarily take place the exact same way, but it does provide a good idea about how one challengeable call (one that would likely be overturned in 2014) affected this game.

But we’re only in the sixth inning—the bottom of the seventh and eighth innings also had two plays that would have been reviewed and overturned. Posey led off bottom of the seventh with a single before Pence hit a fly ball to Gonzalez in left; Belt then walked before

Andres Torres

doubled Posey in, sending Belt to third and tying the game at four. At this point, the Giants win-expectancy was 74 percent. It was, at least, before Alfonso Marquez blew the next play.

Brandon Crawford

hit a slowish grounder to DJ LaMahieu at second, who threw Crawford out at home. But he was not out. See:

The Giants lost a run and an out, and that play dropped their win-expectancy by 17 percent. But Marquez was not quite done.

Matt Belisle

opened the bottom of the eighth by getting Pagan to flyout; it was followed by a Scutaro single. Sandoval then blooped a single to left field, and Scutaro attempted to stretch all the way to third. Quick to the ball, Carlos Gonzalez threw him out at third. He did—but Nolan Arenado didn’t really tag him. Examine:

So instead of runners on first and third with one out in a tie game, with a run expectancy of 1.1, there were two outs with a runner on first and a run expectancy of 0.2. This was not an egregiously blown call. The tag-or-not happened out of Marquez’s line of vision, and he was left to make his best guess. He chose wrong, and in 2014 this—for the third time in eight innings—would have been overturned. For what it’s worth, the time it took for Bruce Bochy to argue the call and get ejected likely approximates the time it will take to get calls on the field correct.

And the rest of the game? It went to extra innings tied at four; in the top of the frame

Troy Tulowitzki

mashed a home-run to left field. The only thing better than the audible collective groan at AT&T Park is Tulo’s enthusiasm as he rounded the bases:

In the bottom of the tenth, Rafael Betancout walked Brandon Crawford on base.

Guillermo Quiroz

bunted him over, and in the process decreased the Giants win expectancy by four percent, so thanks for that. It didn’t matter though, as Pagan hit a deep fly to right center, outpacing an ambling

Michael Cuddyer

for an inside the park home run. See, with mixed feelings:

Based on previous research, this game had more close calls than the average game. By my count, three would have been overturned, while at least two more (the play prior to the would-be double play and a Gonzalez steal in the fifth) could have been challenged with reason. We can’t know with certainty how the game would have progressed if Posey were called out in the first inning, which would have completed an inning ending double play. I’m convinced that the rhythm of the game would have been different because the call had material consequences that provided the context for the other controversial plays to take place. Maybe the plays still occur in an alternate scenario, maybe they don’t. Maybe the Rockies win in extra innings if Angel Pagan doesn’t bat when he does, but maybe the Giants win in nine innings with a more productive series of at bats in the sixth, seventh, and eighth innings.

So in this game, Bochy would have challenged and lost one play in the first inning without consequence, and Weiss would have challenged and won a play in the first inning, but without another clear opportunity to use the second challenge. When the new rules were first announced, I thought that the transition to booth initiated reviews in after the sixth inning was dumb. It seemed to suggest that close plays and runs later in the game are somehow more valuable than early in it, which is why replay would no longer be limited. The clips available on MLB.com correlate to this, as the missed call from the bottom of the first is nowhere to be found, even though it was part of a spectacular defensive play. But the more I think about it, the more I am sure that the reason manager’s cannot challenge after the sixth inning is because those are generally the most opportune times to exploit a challenge to ready a bullpen arm if necessary. Placing the impetus on the crew chief disposes of that likelihood. But that also means that we’ll probably see anywhere from two to four challenges in the first six innings of most games in 2014.

My central point is that the calls missed in this particular game will be easily correctable in the future. Instant replay is a good thing. Whether or not the current system of challenges survives even one year of managerial over-use, instant replay will stick around. For those of you concerned about the essence of the game, longing for the next Lou Piniella to drop his trousers in front of the first base dugout, there’s always balls and strikes.