Like most of my most interesting studies, this one started with a comment on Braves Journal, this time from Stampton.
A lot of us Braves fans have felt the team is bad at that sort of situational hitting for years, but I would like to see some stats on it. Not just batting average with runners on third and less than 2 out but success rate at scoring runners from third with less than 2 out. A couple of years ago we were among the worst in the league at scoring the Manfred man. We seemed to have improved since then and it could’ve been a fluke.
A Brief Methodological Digression
If you just want to look at some metrics and don’t really care what they are, well, you’re not my kind of person, but you can skip this section.
It’s easy enough to use the Retrosheet Database and select every instance in which there’s a runner on third and less than two outs in some historical period. I’m using 2000-2023. There is a field in the database that tells you the ultimate fate of that batter (out on the bases, never left third, or scored.) So far, so good.
But now there is potentially a double-counting issue, or not, depending on the question you’re trying to answer. Take a situation: man on third, no one out. The next batter walks. Is that a new situation (first-and-third, no outs) or a continuation of the old situation? The man on first steals second. Is that a new situation with the force play removed? The next batter strikes out. Now it’s one out, not two outs. New situation? The batter after that strikes out and the batter after him hits a home run, finally scoring our man on third.
The most naive way is to treat every event in the Retrosheet database as a new situation. In the case above, you’d have three events with a man on third and no outs, and one event with a man on third with one out and all four would register as successes in eventually getting the runner on third home, even though only one run scored from third. This would seem to overcount the success rate.
But it might also understate the success rate. If that last guy strikes out instead of hitting a homer, you end up with three failures. Overall, you will understate more often than you overstate, since most at-bats are failures.
On the other hand, if you ignore the history aspect of the way the inning played out, treating every situation as different makes sense. Every different situation is treated abstractly and not part of a sequence and gives you a better handle on that particular situation. In the case above, this really was an example of a guy on third with nobody out, a guy on first and third with one out, a guy on second and third with no outs and with one out and in all four situations, the man on third eventually scored.
So if you treat this as one single man-on-third situation, you have to pick one of the events as the background. The natural one to use is the first event as the background event, ignoring all the subsequent play, but you’ll be throwing away a lot of information about other situations. In particular, the only times you can assess the ability to get a runner home from third with one out, you are restricted to those circumstances in which the runner arrived on third with one out. On the other hand, it’s the method you have to use if you really want to figure out aggregate runs gained or lost by proficiency at bringing runners home. You have to count every particular man-on-third event just once. I think answering the question in this way is the spirit of Stamton’s question.
A third possibility is to treat every opportunity to get a man home from third from the batter’s perspective. Thus, every batter either succeeds or fails based on whether or not he brought the runner home. You wouldn’t treat a walk or HBP as a failure, but you’d ignore it. Wild pitches and double steals and the like would be credited to the batter as a success. More importantly, outs which bring home a run are successes. The problem with this metric is that bringing home runners more quickly increases your success rate. A team that hits a sac fly every time they get a runner on third with on outs will get a 100% success rate, while a team that follows two strikeout with a single will get 0% even though both are equally good at getting the man home. This method is fine for assessing players, but not particularly good at assessing teams.
So we have three potential methods: the situation method, the team method and the player method. I use the team method here, but note that I am throwing out about 20% of the data by doing so. In particular, bringing home a runner from third with one out is considerably less precisely estimated. Still we end up with around 200,000 such situations.
Success Rates
So first, a little level-setting: I looked at every situation with a runner on third and less than two outs since 2000. The runner from third scored 93% of the time with no outs and 65% of the time with one out. So what about the Braves over the same period? It turns out that those values are… 92% and 64%. The Braves are pretty much an average team in this regard. Here are the team-by-team results:
| 0 | 1 | |
|---|---|---|
| SEA | 0.9454 | 0.6440 |
| COL | 0.9406 | 0.6512 |
| NYA | 0.9391 | 0.6697 |
| ANA | 0.9372 | 0.6719 |
| TEX | 0.9325 | 0.6713 |
| BOS | 0.9321 | 0.6643 |
| KCA | 0.9303 | 0.6681 |
| HOU | 0.9301 | 0.6529 |
| MIN | 0.9295 | 0.6619 |
| CLE | 0.9294 | 0.6592 |
| OAK | 0.9273 | 0.6513 |
| FLO/MIA | 0.9266 | 0.6485 |
| CIN | 0.9266 | 0.6337 |
| PIT | 0.9258 | 0.6479 |
| Average | 0.9255 | 0.6527 |
| NYN | 0.9252 | 0.6475 |
| TBA | 0.9239 | 0.6524 |
| ATL | 0.9235 | 0.6432 |
| TOR | 0.9226 | 0.6625 |
| CHN | 0.9224 | 0.6316 |
| PHI | 0.9211 | 0.6549 |
| DET | 0.9211 | 0.6457 |
| SLN | 0.9207 | 0.6559 |
| SFN | 0.9204 | 0.6548 |
| BAL | 0.9195 | 0.6624 |
| LAN | 0.9187 | 0.6554 |
| SDN | 0.9186 | 0.6349 |
| CHA | 0.9172 | 0.6612 |
| MIL | 0.9121 | 0.6243 |
| MON/WAS | 0.9112 | 0.6519 |
| ARI | 0.9061 | 0.6421 |
So the Braves are almost exactly the average team over this period. But the Braves have had 24 different teams from 2000-2023. What if we look over time? Here’s the chart:
Ignoring 2020 (which was only 60 games) the Braves only had one really bad year getting guys home: 2011. That was of course Fredi Gonzalez’ first season and the year of the epic September collapse that kept the Braves out of the Wild Card. I think you’d be hard-pressed to think of things that went right that year.
Stamton has hit on a curious thing. Commenters often talk about batting average with runners in scoring position as if it’s a goal in itself. It obviously isn’t. The point is to score runs, so a single with a guy on second who is thrown out at the plate is a bad thing. A ground out that scores a guy from third isn’t as good as a hit, but it scores a run, which is a reasonably precious thing, far more precious than a hit. You can single with a man on second and see him progress only to 3rd. That’s good, but not great. If a team gets a lot of hits but can’t score runs, they’re a bad team. (Trivia question: what’s the record for hits in a game in which that team failed to score a run? Answer below.[i])
What about the base situation in general? Here’s the table:
| Man On Third Scored | ||
|---|---|---|
| Outs | Bases | |
| 0 | 123 | 0.8829 |
| 1_3 | 0.9258 | |
| _23 | 0.9528 | |
| __3 | 0.9782 | |
| 1 | 123 | 0.6567 |
| 1_3 | 0.6341 | |
| _23 | 0.6718 | |
| __3 | 0.6554 |
Manfred Men
We can also use these data to get the percentages of placed men on second who score. We now have data running from 2020-2023.
As it happens, 57 percent of Manfred Men score. When they score, the probability that their team wins the game is 70 percent. In 11 percent of cases, the Manfred Man is put out on the bases. (Note that in this calculation, the question is whether the team eventually wins the game, not whether they win it in the inning in question.) This causes the winning percentage to drop to 39%. The worst case, however, is when the Manfred Man is stranded on second or third, since in those cases there is no chance for subsequent runs to score in the inning. This accounts for 31 percent of innings, and the winning percentage drops to 16 percent.
The result for the Braves in these games is similar — it’s difficult to jusge how different the Braves are given the fact that there have been only 63 Manfred Men. That said, the Braves are slightly less likely to score the Manfred Man: 33/63 have scored, for a rate of 52 percent. They have won a very healthy 79 percent of those games when they do score the Manfred Man.
Finally, just for fun, here is a list of every Manfred Man by an Atlanta Brave and their won/loss record by result:
| Number of Innings | Win %age | ||
|---|---|---|---|
| Manfred Man | Result | ||
| Adam Duvall | Scored | 4 | 1.0000 |
| Alex Jackson | Stranded | 1 | 0.0000 |
| Austin Riley | Out On Bases | 2 | 1.0000 |
| Scored | 2 | 1.0000 | |
| Stranded | 3 | 0.3333 | |
| Charlie Culberson | Stranded | 1 | 0.0000 |
| Cristian Pache | Scored | 1 | 1.0000 |
| Dansby Swanson | Scored | 4 | 0.5000 |
| Stranded | 2 | 0.0000 | |
| Eddie Rosario | Scored | 1 | 1.0000 |
| Ender Inciarte | Scored | 1 | 1.0000 |
| Stranded | 1 | 0.0000 | |
| Forrest Wall | Scored | 2 | 1.0000 |
| Freddie Freeman | Stranded | 2 | 0.0000 |
| Guillermo Heredia | Scored | 2 | 1.0000 |
| Stranded | 1 | 0.0000 | |
| Joc Pederson | Stranded | 1 | 0.0000 |
| Kevin Pillar | Scored | 1 | 1.0000 |
| Marcell Ozuna | Out On Bases | 1 | 1.0000 |
| Stranded | 2 | 0.0000 | |
| Michael Harris | Scored | 1 | 1.0000 |
| Nick Markakis | Stranded | 2 | 0.0000 |
| Nicky Lopez | Scored | 1 | 1.0000 |
| Orlando Arcia | Scored | 1 | 0.0000 |
| Ozzie Albies | Out On Bases | 1 | 0.0000 |
| Scored | 4 | 0.7500 | |
| Stranded | 2 | 0.5000 | |
| Ronald Acuña | Scored | 3 | 0.3333 |
| Stranded | 1 | 0.0000 | |
| Sam Hilliard | Out On Bases | 2 | 0.0000 |
| Scored | 3 | 1.0000 | |
| Sean Murphy | Stranded | 1 | 1.0000 |
| Travis d’Arnaud | Scored | 2 | 0.5000 |
| Stranded | 1 | 0.0000 | |
| William Contreras | Out On Bases | 1 | 0.0000 |
| Stranded | 2 | 0.5000 |
[i] The record for hits in a game without scoring is 15, in a game which echoed this season. On August 1, 1918, the Pirates took on the Braves in Boston. Your Braves had 15 hits in a 21 inning game without scoring. The Braves pitcher, Art Nehf threw a 21 inning complete game and lost 2-0. The Pirates used only two pitchers. The 21 inning game took just over 4 hours to play.
Of course, 15 hits in a 21 inning game isn’t even a hit an inning. The record for hits in a 9 inning game is 14, accomplished twice: once by the New York Giants on September 14, 1913 in a game in which Jim Thorpe pinch hit, and again on July 10, 1928 by the Indians, or as I call them now, the Proto-Guardians, which makes them sound like invaders from an alien planet.
If you want something a little more recent, there are more than a dozen 13 hit shutouts, including a game on July 18, 1993 in which Steve Avery, Greg McMichael and Mike Stanton combined to hold the Pirates scoreless on 13 hits. Avery, who obviously didn’t have his best stuff that day, gave up 12 of the 13 hits in 7 2/3 innings pitched. Stanton gave up the other hit in the 9th. Former Brave Zane Smith took the loss.
So… it seems like team propensity to bring men home, the most important aspect of “situational hitting,” is likely not a skill. Just as we expect a pitcher’s strand rate to regress to the mean, we should probably expect a team’s likelihood of bringing in a man on third with fewer than three outs to regress to the mean, and any such failures are frustrating but evidence of little other than bad luck.
Is that fair, or is that overstating the case? Is it possible to be bad at this?
I think that’s a slightly premature conclusion. I started to look not just at did you get the man home, but how did you do it. (It quickly got too complicated so I stopped) There may well be strategies (like suicide squeezes) that work sometimes but fail on average. Sac flies, bunts, making sure you get a bat on the ball, Big strikeout teams ought to be bad at this. Whether there are enough chances to prove it is another thing.
I tend to think that this should be more of a counting stat than a ratio. Good hitting teams will have more opportunities than bad hitting teams. It’s kind of the flip side of announcers complaining about “we’ve left 12 men on base tonight…”. That’s usually a good thing, right? Bad teams don’t get as many runners on base to begin with.
Possibly, but separating these particular “skills” (if it’s a skill) from a general ability to score runs may obscure things. Take an example. The decision to go first to third on a single ought to depend on the difference between scoring from second and scoring from third, among other things. Teams that are more conservative going from first to third because they already have a lot of power will have fewer counting stats, but no fewer runs.
Yeah it’s tough to try to isolate a “skill” here because it’s such a team thing. With no outs I always felt a bit more pressure on the next couple of guys up. If you’re batting with two outs then you pretty much have to get a hit, while the guys who went before you can score the runner a ton of different ways (or at least not make an out). It’s just good in general to have a deep lineup with guys that aren’t easy outs. The 2024 Braves have a bunch of dudes that make outs, period. I haven’t looked it up but this feels like our worst team OBP-wise in a while.
Where I actually come out, in my heart of hearts, is that batters really should modify their approach depending on the situation: sit on a fastball in a hitter’s count, defend the plate and choke up / take a shorter swing with two strikes, try to get under the ball with a man on third and fewer than two outs. If the pitcher has walked multiple men in the inning, sit on a first-pitch middle-middle fastball, and spit on anything else.
That all seems fairly obvious, but the counterargument is pretty well known by now, too: hitters should play to their strengths, and swing from their shoetops on every pitch: it’s worth trading extra strikeouts for a few more homers. I don’t think any team goes entirely to either extreme, unless they employ Luis Arraez or Aaron Judge. But our Braves hitters have failed to meet the moment much too often this year.
So obviously I agree with Krussel: it seems like this team, more than other recent teams we’ve had, has both a high propensity for swing-and-miss generally, and an inability to change their situational approach. As Matt Olson, Orlando Arcia, and Sean Murphy in particular have had their seasonlong slumps, it has seemed like they not only couldn’t square up a pitch with the bases empty, they couldn’t “do the little things” in any other situation, either.
Just perusing the stats on baseball-reference: last year, the Braves had a swinging strike percentage of 19.8%, tenth-highest in baseball: about one-fifth of all strikes called against the Braves came on a swing and a miss. And 30.3% were foul balls, highest in baseballl. This year, it’s 21.7% swinging strikes, second-highest in baseball, and 29.8% fouls, second-highest in baseball. So we’re swinging and missing more, relative to everything else.
In all, we have the fourth-lowest contact percentage in baseball this year, while last year we were exactly 15th. So our incredible across-the-board improvement in not striking out, which drove our remarkable offensive surge last year, seems to have regressed.
Concurrently, this year, we have the fourth-worst productive outs percentage: that’s advancing any runner with none out, or driving in a baserunner with the second out of the inning. Last year we were 12th in baseball.
In all, I think the cardinal sin this year appears to be swing-and-miss. I’m not sure how much of our regression in that department is luck; much of it is likely explained by the injuries to our best hitters and the woeful performance of their substitutes, but that doesn’t explain why the regulars have all been worse, too.
There’s also the issue of the persistent and sadly unavoidable xwOBA underperformance that has plagued this team all year. The Braves are tied for the 7th-best xwOBA in MLB (.325) but only the 15th-best wOBA (.308). That’s brutal.
They still grade out well in many of the areas that made them such an offensive powerhouse last year (2nd in hard hit rate, 2nd in barrel rate, 1st in xwOBACON, etc.). Sure, they’re not doing those things to the same extent as last year, but that’s due in part to this year’s less friendly run environment and poorer inputs from some key players (Olson in particular). All of this doesn’t fully explain why many of the regulars have been so much worse this year, but it at least partially explains why the offense has seemingly dried up.
Great work, JF. At the risk of stating the obvious, the biggest problem with the Braves’ offense this year is not so much that they aren’t getting the runners who reach base to cross the plate. The big problem is that they simply aren’t getting on base enough in the first place. Their OBP last year was first in the league; this year it’s 12th out of 15.
Losing Ronald was huge in that regard, and Olson’s sixty point drop in BA has led to a nearly 80 point drop in OBP. Those two were far and away the best at getting on base a year ago. But there are plenty of other offenders–indeed pretty much everyone except Marcell is making outs at far too high a clip.
The object is to score runs. The best way to do that is to minimize the plate appearances in which you make outs.
Thanks, and I agree. This piece was really not intended to address any particular problem with this year’s team as it is a general response to “why can’t we get people in from third with less than two outs” and “do we ever score the Manfred man?”