Statistics

Any ideas or comments on baseball statistics

Red Sox again “most .500 team” in 2025

tomisphereLeave a comment

We are at nearly the halfway point in the season, and I’ve been noticing the Red Sox do some similar things to last season. They sure seem to be hitting the .500 mark a lot. They’ve been 18-18, 19-19, 20-20, and more. Now they’re 40-40. So I wondered, how are they stacking up to last season, when I granted them the title of the Most .500 Team in Baseball?

They are excelling, actually. Right now they have the lead.

If you measure by who has the most nearly .500 record, they lead all of baseball there:

If you look at that “Games above .500” column, the Red Sox are the only team at .500. And while teams that have played an odd number of games can be no closer than 1 game from .500, no teams are that close. The Sox are clear “winners” in this category.

We can also look at the number of times a team has been at .500 during the season. This gives an idea of whether the team has consistently played like a .500 team through the year. This time we’re looking at that “Times at .500” column:

Showing just the top 8 teams here. The Red Sox and the Reds are clear leaders in this category.

One more. We can look at which teams have the smallest difference between runs scored by them and runs scored against them. This we have in the Run diff column. Negative numbers mean the team allowed more runs than they scored.

The Red Sox are among the 4 leaders here.

But what I find more interesting is to consider balancing a team’s record against their run differential, because some teams have a positive run differential but a losing record, or vice versa. We have four good examples of this, this year. Now there are those who take a team’s run differential and convert it into a number of expected wins and losses for a team. So I took the difference of those for each team to get an expected number of games above .500 for those teams (see the “Expected GA .500” column). Then I averaged this with their actual games above .500 to see where the expected and actual balance out. These averages are listed in the “Ave GA XGA” column.

The Braves, Rangers, Blue Jays, and Gaurdians all have this contradiction of either a losing record and more runs scored than allowed, or vice versa. And there with them in the top 6 spots are the Red Sox.

There’s one team near the top of all these lists – The Boston Red Sox. If they go on to have a mediocre second half, they could repeat as Most .500 Team two years in a row. But with Alex Bregman coming back before long, and the rookies and second-year players getting better, and the pitching getting better, the Red Sox are looking like a team that will finish pretty well above .500 this time, so a repeat may not be in the cards. But they sure are off to a “good” start.

Garrett Crochet pitched the best start of the year but didn’t get the win. We can fix that.

tomisphereLeave a comment

The merit method of awarding wins fixes all the injustices that the current method of awarding wins punishes starting pitchers with. The latest of these injustices happened to Garrett Crochet, ace of the Boston Red Sox, last night. He threw 8 scoreless, 3-hit innings before giving up a game-tying solo home run to baseball’s best hitter, Aaron Judge, with one out in the top of the ninth. With only one run of support from his teammates, that tie score made Crochet ineligible to earn a win, as he was removed from the game at that point, and the current rules say the win goes to the pitcher who was the active pitcher when the winning team took its final lead. So the win went to Garrett Whitlock, who retired the two batters he faced in the top of the 10th inning. Whitlock performed well. But who did more to earn the win? The guy who faced 2 batters over 1 inning of work, giving up no runs, or the guy who faced 30 batters over 8 and 1/3 innings, giving up 1 run to a top offense and the Red Sox’ arch rivals?

If you agree with me that Crochet did more to earn that win, then you’ll like the merit method for awarding wins.. I explain the method in my post The how and the why of awarding wins to pitchers by the merit method.

Using that method, we take the 2 runs that the Red Sox scored in Friday’s game and divide by 9 and 2/3, which is the number of innings the Red Sox were at bat. This gives us the average number of runs the Red Sox scored in each inning, the fraction 6/29. Then we simply credit each Red Sox pitcher with this number of runs for every inning they pitched. And then we subtract from this the number of runs they gave up. This gives each pitcher a number of “Runs Ahead”. Then we give the win to the pitcher with the greatest number of Runs Ahead.

The cool thing about this method is that adding the Runs Ahead of all the winning team’s pitchers always gives you a positive number, and adding all the Runs Ahead of the losing team’s pitchers always gives you a negative number. This method also assigns the losing pitcher as the one with the most negative Runs Ahead value.

Here’s a table showing the numbers discussed above for the Red Sox pitchers in last night’s game. In the first three columns in the table below we see IP, RCr/IP, and RCr, which stand for Innings Pitched, Runs Credited per inning pitched, and Runs Credited, respectively. You get the third column (Runs Credited) by multiplying together the first two.

PitcherIPRCr/IPRCrRRA
Garrett Crochet8 ⅓6/2950/29121/29
Aroldis Chapman⅔6/294/2904/29
Garrett Whitlock16/296/2906/29
Runs Ahead (RA) calculations for Red Sox pitchers in victory over New York Yankees, June 13, 2025

Then you subtract runs allowed (R) from this to get each pitcher’s number of Runs Ahead (RA) for that game. Because Garrett Crochet had the highest number of Runs Ahead for the winning team, he would be awarded the win by the merit method. But by current rules, the win went to the other Garrett (Whitlock).

I hope someday to convince MLB league officials to change to the merit method for awarding wins. It fixes so many things that are just not right about the current method.

Xander Bogaerts back on pace to reach 200 hits, win AL batting title

tomisphere1 Comment

Back on Wednesday morning, I showed that Xander Bogaerts and Miguel Cabrera were hitting at paces that would cause Bogaerts to (most likely) surpass Cabrera for the AL batting title. Though I didn’t mention it at the time, these projections also showed that he’d reach 200 hits even if he sat out a couple of games, and a few more than that if he played all the remaining games. After a pair of low-hit games knocked Bogaerts off that pace, his 3-for-4 performance last night has put him right back on it.

In trying to project future totals using “the pace at which a player is producing right now”, how many games do you use to determine what that pace is? The last 5? The last 10? 20?

I circumvent that question by using all of them … I calculate his pace of production over his last 5, 6, 7, 8, etc. games, then use that pace applied over the remaining number of games to be played to see what final numbers he’s headed for. This gives a big collection of possible final numbers; you then choose one in the middle.

On Wednesday I did that for Cabrera and Bogaerts using their paces of production as established by their last 8, 9, 10, etc. up to their last 20 games. That gave 13 paces of production for each player. I then applied these to their remaining games assuming they’d not sit out any games, and then again assuming they’d each sit out two games. I got these results:

If playing all remaining games
Bogaerts Cabrera
Low 0.327 0.324
Median 0.329 0.326
High 0.332 0.331
If sitting out two games
Bogaerts Cabrera
Low 0.327 0.326
Median 0.329 0.328
High 0.331 0.332

In all but one of these 26 projections, Bogaerts would end up with at least 200 hits.

I just updated these numbers, and now they look like this:

If playing all remaining games
Bogaerts Cabrera
Low 0.327 0.325
Median 0.329 0.326
High 0.330 0.332
If sitting out two games
Bogaerts Cabrera
Low 0.327 0.327
Median 0.328 0.328
High 0.329 0.332

Here are Bogaerts’ projected numbers of hits:

Bogaerts projected 2015 hits
# of recent games used If playing all games If sitting two games
20 204.0 200.8
19 203.3 200.2
18 203.0 200.0
17 203.3 200.2
16 203.6 200.5
15 204.0 200.8
14 205.1 201.7
13 204.9 201.5
12 203.8 200.7
11 204.4 201.1
10 204.0 200.8
9 204.7 201.3
8 204.3 201.0

Longer term projections (based on his last 40 or more games) almost all have him finishing with 200 hits exactly if he sits out 2 games, 203 hits if he plays all remaining games, and a .327 average.

If they play it out, and stay on pace, Bogaerts probably will win the batting title and will get to 200 hits.

Thanks to Baseball-Reference.com for the gamelog data I used for this article.

My previous mathematically-oriented baseball posts

tomisphereLeave a comment

I’ve been away for a while.  But I’m returning.

Where have I been the past year?  Mostly over here, and sometimes commenting over here.  But during that year, I’ve done a lot of mathematics to study questions I had about the baseball I was reading about and following, and some of that has filtered into some of those posts.  I thought I’d provide a selection here of some of the more interesting ones.  Some of these contain hints to posts I plan on putting up here in the coming weeks, posts that will include discussions of the math of streaks, and just how much a small sample size actually tells us.  There will be other new topics too, not previewed in any of these posts.  Stay tuned.

Here is that selection of my mathematically-oriented posts of the last year or so:

A post from August 2014 explaining why batting 15 points above league average is sometimes actually hitting at league average. This in the context of examining one upcoming player.

A comment titled “Expected frequency of reverse platoon splits exceeds the actual numbers” to the article “Are reverse platoon splits sustainable?” on Beyond The Box Score. In this comment (scroll to the bottom of the comments section) I used binomial theory to come up with what would be the expected number of players, based on random chance alone, having a reverse platoon split in on-base percentage for the years 2012 and 2013. I show that the actual numbers were less than the numbers you’d expect by random chance, seeming to indicate that reverse platoon splits are unsustainable.

A comment titled “No, because starters face more batters” to the article “The Hidden Perfect Games of Relievers” on Beyond The Box Score. In this comment (scroll to the next-to-last comment), after making two points about the right way to compare starters and relievers for the purpose of the article, I discussed my first attempts at producing an expected number of “wrap-around” perfect games that will occur in a given season for starters and relievers, to help clarify any meaning that might be attached to the reported results. I did complete that work, which I plan to publish later on this blog, in a post about the math of streaks.

A post from September 2014 that argues that a certain young player is better than his overall numbers say he is, by analyzing his advancement as a hitter at each new level he played at.

A post from September 2013 explaining why one baseball team’s chances of making the playoffs were ridiculously close to, but not quite exactly, 100%.

A post from later in September 2013 which explains (in more detail than anyone probably cared to read) why that same baseball team’s chances of having home-field advantage were about 7 out of 11 (washing dishes at night gave me a lot of time to listen to baseball and think about this stuff).