The Rockies’ Blockbuster Night

In last night’s fifth inning, the Rockies threw punches and punches until the Giants were frontless. They scored 13 runs, which, as was noted at Purple Row, was a team record for runs scored in an inning. Oh, did I mention that this game was in San Francisco, and not in Denver? Because it was, which makes it all the more surprising. Let’s walk back through their blockbuster night, and use it to show what the Rockies are doing right this season.

First, let’s put this game into some context. Here are all the teams who have scored 15 or more runs in a game at AT&T Park, which as you probably know has been open since 2000.

15+ Runs Scored by Single Team, AT&T Park History
Date Tm Runs Opp Runs Barry Bonds?
5/6/2016 COL 17 SFG 7 No
7/10/2015 SF 15 PHI 2 No
9/13/2014 LAD 17 SF 0 No
8/31/2014 SF 15 MIL 5 No
8/24/2010 SF 16 CIN 5 No
9/24/2008 COL 15 SF 6 No
7/23/2005 FLO 16 SF 4 No
9/3/2004 SF 18 ARI 7 Yes
4/9/2003 SF 15 SD 11 Yes
5/24/2000 SF 18 MON 0 Yes
SOURCE: Baseball-Reference Play Index

As you can see, this doesn’t happen very often — happens even less when Barry Bonds hasn’t been involved. For reference, over the same time span, a team has scored 15-plus runs at Fenway Park 37 times. Across the bay at whatever Oakland’s ballpark is called now, it’s happened 16 times. At Camden Yards, it’s happened 27 times. Runs are simply harder to come by in games affected by the marine layer.

It wasn’t just the game that was special, though. The inning itself was pretty special, too, as Rockies PR noted:

Let’s take a look at the game graph, too, while we’re at it:

Source: FanGraphs

When the inning started, the Giants had a win expectancy of 33.6%. By the end, that number had declined to just 0.2%. Trevor Story’s homer to lead off the inning was the biggest WE blow of the inning, dropping the Giants’ chances by 12.2 points. Five batters later, the Giants’ WE was down to 2.2%.

To me, though, the most interesting WE tidbit came in the bottom half of the inning. Chris Rusin walked Kelby Tomlinson to load the bases with one out, and Gregor Blanco promptly hit a two-run single. Two-run singles are good, right? Well, usually. Here, the single lowered the Giants’ WE from 0.6% to 0.4%. The Giants were behind by so much that they needed more in that situation, which is a little bit nutty.

Going back to the Rockies, something one notices from the past week is that the Rockies have finally started putting DJ LeMahieu where he belongs in the batting order — at the bottom. With that tweak, the Rockies are officially stacking their lineup in the most optimal way. Here is the projected wRC+ of each batter in their batting order (via the Depth Charts ROS), as it has been the past week:

  1. Charlie Blackmon: 93
  2. Trevor Story: 91
  3. Carlos Gonzalez: 116
  4. Nolan Arenado: 118
  5. Gerardo Parra: 92
  6. Mark Reynolds/Ryan Raburn: 105/109
  7. Nick Hundley/Ben Paulsen: 92/85
  8. DJ LeMahieu: 82
  9. Starting Pitcher

As you can see, the Rockies are avoiding getting cute with the lineup. This, to me, is important. I know that over the course of the season, a lineup simulator will only show a difference of 5-10 runs no matter how you run the scenarios, but the pattern here shows a willingness to operate sabermetrically. The five constant starters who can hit are at the top of the lineup, complemented by a strong platoon in the six-hole. Then comes the lesser hitters (despite Hundley’s hot start, I’ll take the under on that 92 wRC+). Blackmon hasn’t come around yet, and if he doesn’t, we might see changes, but that’s a problem for later. The point is that the Rockies are doing this right.

Another thing that the Rockies did last night — and of which they’ve made a habit lately — is that they let their relievers work for more than one inning. Now, last night the decision was a little easier, given the cushion, but even a 10-run lead wouldn’t stop some managers from mixing and matching. Furthermore, for the season, the Rockies have allowed their relievers to get four or more outs more frequently than they have two or fewer.

Rockies Relief-Pitcher Breakdown, 2016
Filter # RP # 4+ outs # 2 or less outs
Total 88 26 24
In Losses 45 19 11
In Wins 43 7 13
1-3 RP/G 36 12 6
4-6 RP/G 52 14 18

The Rockies are still carrying 13 pitchers, which I don’t care for, but they are at least trying to get the most out of their relievers. While they have been more specialized in wins than losses, the overall trend is clear: the Rockies are giving their relievers the opportunity to put away more hitters. Part of this is by necessity, as they’ve had some starters unable to make it out of the third, but even in games when they haven’t needed a lot of relievers, they’re still giving their relievers some leeway. Again, this might not be a huge deal, but it does show that they aren’t being a slave to a one-inning reliever/specialist pattern. It’s possible that, given some success with longer relief appearances, the coaching staff could be convinced to carry just 12 pitchers, thus opening up roster flexibility on the offensive side.

Overall, the Rockies ‘pitching is doing enough to keep them in games. Their starting pitchers have the 14th-best FIP- in baseball right now, as well as the 19th-best ERA- and ninth-best xFIP-. That’s not only pretty good, but their best in awhile:

Rockies Starting Pitcher Minus Stats/Ranks, 2010-2016
Year ERA- Rank FIP- Rank xFIP- Rank
2016 112 19 100 14 94 9
2015 115 27 116 29 113 28
2014 114 27 114 29 112 29
2013 104 17 101 14 111 28
2012 126 28 122 29 117 30
2011 108 19 109 23 108 28
2010 92 4 90 2 96 6

If you don’t remember, 2010 was the best season of Ubaldo Jimenez’s career — he started the All-Star Game that season — and the team also got solid contributions from Jason Hammel, Jhoulys Chacin and Jeff Francis.

This season isn’t shaping up quite as nicely, but in Jonathan Gray, Tyler Chatwood and Chad Bettis, the Rockies have three starting pitchers who are performing well (well, Gray sort of is, he’s been unlucky on the surface) and can be expected to remain good all season. The last two spots are still a bit of a quagmire, but the team has plenty of options to cycle through.

This is what the Rockies need. They need their pitching to keep them in the game (the bullpen is also in the middle of the pack in the three minus stats) and for their offense to do the heavy lifting. Last night was an extreme example of that. And while the offense may not be up to that challenge night in and night out, they’ve been better than expected. We’re now at the point where we’re seeing articles about whether Nolan Arenado is the best player in the National League. While we don’t need to debate that here, what we can say is that he’s on the short list.

Entering the season, it was clear that the Rockies would need to see some players really step forward in order to be even a decent team this season. So far, they’re getting those performances, and Jeff Hoffman and David Dahl — who are both off to hot starts — are looming in the farm system. The Rockies may not quite be ready for another Rocktober — not without importing some more pitching at the trade deadline, anyway. But they may finally be ready to shed their doormat status, as they emphatically showed last night when they ran San Francisco’s jewels.

Paul Swydan used to be the managing editor of The Hardball Times, a writer and editor for FanGraphs and a writer for and The Boston Globe. Now, he owns The Silver Unicorn Bookstore, an independent bookstore in Acton, Mass. Follow him on Twitter @Swydan. Follow the store @SilUnicornActon.

Newest Most Voted
Inline Feedbacks
View all comments
Roger McDowell Hot Foot
6 years ago

I don’t understand how a two-run single — or any non-out outcome of a PA — can ever lower a team’s win expectancy.

6 years ago

Yeah, that seems like a bug in the win expectancy algorithm. It’s not like the single ran time off the clock. It also happens earlier in the same inning: with runners at second and third, Crawford singled to drive in one, driving in one and leaving runners on the corners. Win expectancy supposedly dropped by .001. There at least the double play is in order after the single, so I can theoretically see where it drops the win expectancy, but for the Blanco single that’s not in play.

I don’t know if it’s related, but the play log fails to note Crawford advancing to third on a wild pitch two plays before that, although the base-out state is correctly noted.

6 years ago

I can imagine it in a sophisticated version of win expectancy that takes individual hitters and pitchers into account. That’s not an especially strong statement, though, simply arguing that the intentional walk is not necessarily bad for the “any non-out outcome of a PA.” A two-run single when the bases were already loaded does seem pretty difficult, though the batter on-deck was the relief pitcher, Derek Law.

At the very least it’s claiming that the guy who was on first had a better chance of scoring from first with Blanco batting than he did the rest of the inning starting from 2B with the pitcher up. (It’s saying even more than that, because the guys on 2B and 3B wouldn’t be 100% to score before Blanco.) I could believe it more with two outs, but there was one out.

6 years ago
Reply to  JohnThacker

Fangraphs’ win expectancy algorithm does not know anything about the individuals who play in each game. It only knows base-out state, inning, and score.

6 years ago

Isn’t win expectancy just how often teams have historically gone on to win after facing a given game state? We’re talking about some pretty unlikely game states here so the historical data on them is presumably pretty thin, not enough to take these tiny percentages too seriously.

6 years ago

If I understand how Win Expectancy works, it looks at how often previous teams in identical score, inning, and base-out states have gone on to win the game. (WE is explained here:

Perhaps this is due to small sample size. I’d be interested to know how many games have featured a 12 run deficit in the bottom of the fifth with the bases loaded and one out, and how many featured a 10 run deficit in the bottom of the fifth with men on first and second with one out.

Interestingly, the spreadsheet linked to in the WE explanation (which uses a fit rather than raw data) gives a 0.3% WP before the two-run single, and a 0.7% WP after the two-run single.

6 years ago

Yeah it’s definitely a glitch from a failure to smooth SSS events. The model definitely did not expect Blanco to do *better* than a two-run single such that a 2-run single was a disappointment.

Ian R.
6 years ago

The only way I can see this making sense is if the two-run single also sets up a double play – for instance, if the team went from a second-and-third situation to a runner on first base only. For a team behind by a sufficiently huge number of runs, the possibility of two outs on the next play might outweigh the two runs scored on the single.

That said, this particular two-run single came with the bases loaded, meaning the double play possibility was already there – so that explanation doesn’t fly.

6 years ago

Does the win expectancy chart stop at like 10 runs either way? Maybe it’s extrapolating, and that’s making an error?