So far this year, the Braves have played in 28 1 run games. After going 9-9 in the first 18 of them, they proceeded to lose the next 10 in a row, and are now 9-19. In baseball, everything to do with winning happens at somewhere in the ballpark of a 50 percent probability, so losing ten in a row of anything is pretty unlikely, as long as you’re at least mediocre. But unlikely certainly doesn’t mean impossible: unlikely stuff happens all the time, given how many teams are playing and given how every team’s win is another team’s loss. Still, though, it’s hard to put things like this in context unless you understand the context. It’s the context that I intend to lay out here. Application to the Braves’ situation can then be done in a contextually appropriate fashion.
How Common Are 1 Run Games?
How many one run games does a team expect to have in a season? This is a pretty easy question. From 2000-2024 (excluding 2020) teams average 46 one-run games per year. This value has been coming down over time, since more powerful offenses create a greater variance in run differentials: from 1961-1975, it was 52 games a year. Overall, it looks like this:

There is, however, one factor that shifts the expected number of one-run games per year. Mediocre teams play more one-run games than either great team or horrible teams, both of whom play in more blowouts. The effect is pretty small, though. Even with this amount of data, the number of one-run games a team plays in a year doesn’t have a great explanation beyond the overall run environment.

The effect really shows up when you look at the level of individual games. If you look at a game and the absolute value of the difference between the two team’s winning percentage (at the end of the year, not at the particular time of the game) you do see that bigger mismatches have somewhat lower probabilities of a one run game, but the effect isn’t huge:

Pay no attention to the big divergences on the right hand side of this graph (which, for those of you care, is a logit regression on about 135,000 games from 1961-2024.) There are just very few games in which the mismatches are this large. In any case, the effect is fairly small.
So, through 65 games, 28 is certainly well above average, but not amazingly so. In 2021 the Amazin’ Mets played 66 one run games. The 162 game record (set in a low run era in a low run ballpark) is 75 for the 1971 Houston Astros. The Braves are currently on pace for 70, though. I suspect this is just one of those early season anomalies, though, and I would predict no more than 60 or so. Above average, but not historically so.
How Often Do Teams Win One Run Games?
It should come as no surprise that the probability of winning a game with a one run difference is exactly 50 percent. Every single time.
But which of the two teams will win? It should come as no surprise that the better teams win these games. What may be a little surprising is how much it explains. From 1961-2024 (dropping 2020) there were 40,110 one run games. I selected the road team and used the difference between their winning percentage (for the year) to predict the winner. The results were quite good:

Note that winning percentage when the power difference is 0 is somewhat below 50 percent — that’s just the home field advantage. As before, ignore the seemingly bad fits on both the right and left of this diagram — there are just very few games at those levels of power divergence and the actual has a large random component.
Takeaways
You can’t really apply this model to the Braves’ current situation directly because you don’t know their final winning percentage nor that of their opponents. But applying their current winning percentage and that of their opponents as a proxy, you can get some idea of the likelihoods here.
Here are the stylized results (calculated as of games through 6/10):
The Braves have a 28-38 record. That’s a 0.424 winning percentage. Using this and every other team’s current winning percentage, the Braves should have played around 17 one run games instead of the 28 they have played. This pretty unlikely. Using a simple simulation, I find that the chances the Braves would have played 28 or more one run game at this point is only .00406, 406 out of 100,000 .. a little under 1/2 of one percent. But since 1966, the Braves have played lots of 66 game sequences… they play almost 100 every year so they’ve played about 60,000 such sequences, so they should have had quite a few by this point.
I also calculate that in the 28 1 run games they’ve played, they should have won about 14, instead of the 9 they have won. This isn’t nearly that unusual. Again, simulating 100,000 such sequences, I find that the Braves would be expected to win 9 or less 3.7% of the time. That said, the probability that they would have lost the last ten in a row is really unlikely. Repeating the simulation just for the last ten 1 run games, the probability of winning none of them is well under 0.2%, 165 out of 10000. Given that there are far fewer sequences of 10 consecutive 1 run games, this really does show that Raisel Iglesias has sucked, if you didn’t already know.

I remember that the last time this team had an overabundance of 1 run games in 2011, Fredi Gonzalez ran O’Ventbrel into the ground and the team completely ran out of gas in the September collapse.
Maybe that was the nostalgia of bringing him back. Though at least he is not making decisions with the pitching staff this time.
Excellent work. A couple of statistical notes:
Counting trials based on subsets such as 66 game sequences out of 162 has a problem of partially recounting the same trials. The next 66 game sequence is 65/66 dependent on the prior sequence.
I have run into this problem when trying to figure out the chances of hitting streaks to determine if streakiness is real. I’m not good enough at discrete math to solve it.
Another issue is that there is a little bit of circularity with using final record as an input for chances of winning a 1-run game. A team that gets unlucky in 1-run games has a worse record and a worse chance of winning 1-run games. Using run differential solves this problem.
Both points are well-taken. The estimates of the likelihood of a particular pattern are correct (assuming everything else about them is correct!), but the sample space of trials is clearly overstated. I didn’t want to go into that issue, but it’s why I didn’t make an explicit calculation of the number of such sequences. There are somewhere between about 2 and 96 66-game sections of a season, and the fully-overlapping maximal estimate of 96 can’t be used directly to estimate the probability that you’d see a reasonably rare event…. which is why I explicitly hand-waved there. One could do it “correctly” by simulating seasons, but that maintains a different fiction: that “teams” aren’t continually moving entities without fixed parameters. Better to hand-wave.
As to the second point, using team’s overall records clearly corrupts estimates of their prowess in one-run games by including their prowess in those games. The fairly tight distribution of results, however, makes that less of an issue, and the regression analyses which removed the one-run success games from the calculation made no substantive difference, so I left it in. Leaving it out would leave out what may or may not be a skill: winning close games. If that were the thing I was investigating, I’d definitely leave out those games at the cost of more imprecisely measuring power differentials, but it wasn’t. (By the way, I was slightly surprised to see that Winning Percentages as an independent variable in addition to either ABS(Win differential) of (Win Differential) contributed very little. What little it did contribute was probably mostly an estimate of this effect, but it was very, very small.
In general, I am a fan of using runs to predict wins rather than using wins to predict wins, as you agree should be done. I talked about this in some earlier pieces that I’m too lazy to look up right now because because I’m going out to play golf. That said, the run differential in one-run games should be subtracted for the same reason that you’d subtract wins, and doing so will make the powerful seem more powerful and the weak more weak (since all of these games have one run differentials) and I suspect would have little impact on the results. I didn’t do that either, because the only point of the piece was to provide a context, not to make precise estimates.
But thanks for the critique.
Stampton, thanks for explaining Schwellenbach’s path last post. I hadn’t realized how inexperienced he was before being drafted, being old for A, and therefore not being included on top prospects lists. Makes sense. I would like to see him make 30 starts in consecutive seasons before we crown him, and especially see if he can hold up now that he’s increased velocity (when does this ever work?), but it’s really encouraging.
Excellent conclusion.
As we all predicted in 2023, Orlando Arcia had a walk off single yesterday as a member of the Colorado Rockies.
I think I was one of the earliest people last year to call for the Braves to sell high on Fried. (I also said they should trade Iglesias.)
I’m going to stake out this ground early and say that unless things change dramatically over the next four weeks, I think they should trade Sale. It’s too early right now, as the market won’t really develop for another month. But he’ll almost certainly be the best pitcher on the trade market – and possibly the best player, period – and it should be possible to get a serious return for him.
Right now, we simply do not have enough depth in the organization to sustain a serious playoff run. If Jarred Kelenic is able to figure things out enough to become a league-average regular, that would make a difference. (Same for Profar, vey iz mir.)
But the table scrap approach has left us with an outfield and bullpen just as rickety as we feared in the winter, and the farm badly needs to be restocked. As unlikely as it seemed in the offseason, this team is threatening to win fewer games than we did in our injury-decimated campaign a year ago, and our front office has given every indication of being too cash-strapped to spend to win.
If that’s still the case, standing pat is the worst choice they could make.
Alex, what sucks about the expanded playoff format is that the Braves pretty much can’t sell. If they’re remotely close to contending, they can’t pocket the money they didn’t spend in the offseason and then sell at the deadline because the team was cut off at the kneecaps before the season ever started.
I’d like to see us sell and buy the way my local Rays do almost every year. They frequently will trade an established regular as if they were “selling” and buy prospects and even established regulars at the deadline as if they’re “buyers” They make their own rules.
Sell:
-Sale
-Ozuna or Murphy
-Sell “money” by taking back bad contracts in exchange for prospects
-Albies
Buy:
-Young, non-elite relievers with control
-Pitching prospects
-Stopgap 2B with no upside (this year’s Whit Merrifield)
-Stopgap SPs with control but little upside
I just don’t know if they can sell that to the fanbase the way the Rays can to its. But if they can’t, then do everything above but don’t trade Sale. But definitely trade Albies and Murphy/Ozuna and take back bad contracts.
Rob, I think the approach you’re outlining is fundamentally “retooling” rather than “rebuilding,” and it’s pretty much what Schuerholz did all the time. If they could convince fans to swallow the David Justice trade, they can explain why they traded Sale, Ozuna, and Murphy.
I have to agree that if the Braves are sellers they have to listen to offers for Sale. He is 36 and just has 1 year left on his contract after this year, and in this pitching market should fetch an excellent return…multiple good young players under team control.
lol, Verdugo is batting 2nd today because he’s 4 for 8 against Marquez
Snitker is f***ing high. That’s the only explanation.
Also, today would be a good day to try Baldwin as the DH in the No. 2 spot. But, Snitker.
Speaking of the Rays they just made a shrewd move buying Forrest Whitley from the Astros. He has been absolutely can’t-find-the-zone terrible in several seasons at AAA but he put up a dominant 2024 season in relief at AAA in a hitters’ league. Hasn’t translated this year but I like that kind of move for a former top 10 prospect who finally showed something last year.
Over the last several years I loved the way AA built the bullpen. He seemed to have a plan and built the bullpen as a top priority. For some crazy reason he didn’t do that this year. Our pen is not a dumpster fire but I think it will likely be the biggest reason we are not able to make up the needed games to make the playoffs. Not terrible but not elite like we need. I think we could have done that and stayed within budget.
Figures that Verdugo actually does well in his first two PAs. I’ll take the L on that. You win tonight, Snit.
But those two Riley and Olson ABs, my word. I would call them “trash” but I don’t want to do my garbage a disservice.
This team is dog doo-doo.
Hey… Money Mike… just what the doctor ordered.
EDIT: Yup, 3-run HRs will win you ballgames.
Earl Weaver was right about most things, but especially 3 run homers. That’s sure what this team has been needing.
Recapped