Cubs Season Flow Thread

[Derwood]

Doesn’t fWAR have “better” defensive metrics as well?

[mul21]

These things may all be true but the pop ups and other garbage on the Fangraphs website makes it a real pain in the ass to use. Therefore, i declare bWAR to be better. 🤣

[Jason Ross]

5 minutes ago, Derwood said:

Doesn’t fWAR have “better” defensive metrics as well?

fWAR uses OAA and other statcast data where as bWAR uses DRS. Generally speaking OAA and statcast is superior to DRS, yes. I do think the gap between the two is less then that of OPS+ and wRC+ simply due to the wonky nature of defensive metrics and how far along we are on offensive data points. But its another factor.

[We Got The Whole 9]

6 minutes ago, mul21 said:

These things may all be true but the pop ups and other garbage on the Fangraphs website makes it a real pain in the ass to use. Therefore, i declare bWAR to be better. 🤣

I feel like a need a map and compass to navigate BR

[Wilson A2000]

3 hours ago, Geographyhater8888 said:

Slash $40 million off of the payroll, sign Matt Boyd instead of Max Fried, Tom vetos signing Bregman over deferrals…… If Tucker isn’t a Cub next year it’ll confirm  The obvious.

There’s no way they will outbid everyone else for Tucker. They’ve never even spent $200 million on a player before, let alone $500-600 million. Ricketts has said he just wants to spend enough to get into the playoffs like some of the medium budget teams over the last few years. 

[Jason Ross]

8 minutes ago, Wilson A2000 said:

There’s no way they will outbid everyone else for Tucker. They’ve never even spent $200 million on a player before, let alone $500-600 million. Ricketts has said he just wants to spend enough to get into the playoffs like some of the medium budget teams over the last few years. 

I understand any skepticism on the end of Ricketts. He deserves no defense for his spending patterns. Fair worry. Fair critique.

What I think we need to move away from is the "they've never spent X, so they wont spend Y" form of belief as to why the Cubs won't sign Tucker. That's how records work. The Mets had never signed a $400m deal, let alone a $700m deal. The Jays had never signed a $500m deal let alone a $200m deal. This is how record spending works. 

Its pretty widely believed the Cubs offered over $400m to Ohtani. I get that people like to throw around accusations like "that's just a token offer to appease fans that they knew hed never sign" stuff, but thats just conjecture. Reality is they seemed to be willing to go there. We will see how in-the-weeds they get with Kyle. I don't think he will command over $500 and I do think the Cubs will be in the ballpark ultimately. But I am optimistic.

[Geographyhater8888]

1 hour ago, Jason Ross said:

OPS by nature is flawed because of how it weighs OBP (on a scale from 0-1.000) and SLG (on a scale from 0-4.000), By weighing them equally, despite different scales, it creates an issue. Think of it like this, if I have two quarters, and four nickels, saying I have "six" is misleading. Having four quarters and two nickels is also "six" but one of these is more than the other. By saying that "one coin = the same" we create a situation where it could look like both are the same "six" but in one situation I have $.70 and the other I have $1.10. In a similar way, this is how OPS treats OBP and SLG, where two 100 OPS+ hitters are not really the same.

The "+" is helpful, in any data set, as when you see "+" it means that the number is adjusted for things such as hitting environment (we know that the steroid era was a different offensive environment than, say, the dead-ball era) and ballpark. Things like that. It can help us compare data sets year to year, but also player to player. We know hitting 80 games in Coors isn't like hitting 80 games in, say, Seattle. OPS+ does do this because it has the "+" involved, but the scaling issue is the prime culprit of where the issues are.

If you see an "x" this is "expected"; also using league data to determine what we should "expect" - useful to determine luck. 

wRC+ looks to fix the scaling issues within OPS while also still adding in the "+" that adjusts for hitting environments. It scales the two more accurately and thus creates a better silver-bullet one data point. In my coin example, it would determine one of us had 70 cents and the other a dollar-ten. From a personal standpoint, it's the best way to value offense in 2025. Using anything else is purposefully using outdated methods. OPS+ had it's time in the sun a decade or more ago, but we're beyond that. It's like using an iPod Touch instead of an iPhone with spotify on it. You-do-you, but there's more efficient ways of doing it, ya know?

bWAR fo pitching will factor in ERA, or at least, a version of it. But this is inherently where my issues come from bWAR; results based data doesn't really always equate well to prediction models. If the goal of data is not only to determine current value, but also help us predict value in the future, using results based like bWAR is less useful. FIP, which looks to eliminate noise around a pitcher, is better, IMO. Now, it does create an interesting conundrum with someone like Ben Brown. xFIP can tend to hide pitchers who give up contact that doesn't land as a HR; a double isn't really found within xFIP's data set, for example (FIP looks at things only a pitcher controls, such as Ks, BBs and HRs as a double would involve defense and that creates that "noise"). But the idea behind xFIP is that in situations like we see with Brown, eventually he'll give up home runs and we'll find the data. 

Thanks. What’s interesting is wRC+ and OPS+ don’t really defer a whole lot. Usually there’s a 2 point gap at best when contrasting a specific player.

Edited June 24, 2025 by Geographyhater8888

[Jason Ross]

7 minutes ago, Geographyhater8888 said:

Thanks. What’s interesting is wRC+ and OPS+ don’t really defer a whole lot. Usually there’s a 2 point gap at best when contrasting a specific player.

They're fairly close most of the time, but not always. Even a 5 point swing is a bit more than it would seem, as the data is really a percent. So a 110 wRC+ hitter (or a 110 OPS+) is considered 10% better than league average (average is 100). Being a 123 wRC+ hitter vs a 118 wRC+ means a 5% swing above average.

Regardless, even if they're close, wRC+ is the better indicator. OPS+ isn't terrible, but if our goal is to be as accurate as possible, wRC+ is just more accurate. Which helps in both fWAR and in terms of just looking at offensive value. 

And you're welcome! Always glad to help. Baseball data is a vast pit and there's some red herrings within it. Many data points can sometimes seem relevant but used incorrectly paint a wrong picture. I'm a big fan of "it takes a community" and "pay it forward". There's a ton in the data I learn from others, too! Together we can all understand baseball better.

[Geographyhater8888]

40 minutes ago, Jason Ross said:

They're fairly close most of the time, but not always. Even a 5 point swing is a bit more than it would seem, as the data is really a percent. So a 110 wRC+ hitter (or a 110 OPS+) is considered 10% better than league average (average is 100). Being a 123 wRC+ hitter vs a 118 wRC+ means a 5% swing above average.

Regardless, even if they're close, wRC+ is the better indicator. OPS+ isn't terrible, but if our goal is to be as accurate as possible, wRC+ is just more accurate. Which helps in both fWAR and in terms of just looking at offensive value. 

And you're welcome! Always glad to help. Baseball data is a vast pit and there's some red herrings within it. Many data points can sometimes seem relevant but used incorrectly paint a wrong picture. I'm a big fan of "it takes a community" and "pay it forward". There's a ton in the data I learn from others, too! Together we can all understand baseball better.

 

Does this at all apply to opposing pitchers? Do they make a distinction between hitting a homerun off of Paul Skenes vs hitting a homerun vs whoever their number 5 starter is? And of course, say hitter a goes 1/4 with a homerun and 4 bases with a 1.250 OPS be weighed more than hitter b 2/3 with 2 singles and a walk for 3 bases and a 1.416 OPS, same ballpark for example? Just to get the best possible understanding.

Edited June 24, 2025 by Geographyhater8888

[Bull]

2 hours ago, Jason Ross said:

OPS by nature is flawed because of how it weighs OBP (on a scale from 0-1.000) and SLG (on a scale from 0-4.000), By weighing them equally, despite different scales, it creates an issue. Think of it like this, if I have two quarters, and four nickels, saying I have "six" is misleading. Having four quarters and two nickels is also "six" but one of these is more than the other. By saying that "one coin = the same" we create a situation where it could look like both are the same "six" but in one situation I have $.70 and the other I have $1.10. In a similar way, this is how OPS treats OBP and SLG, where two 100 OPS+ hitters are not really the same.

 

 

Except that your example demonstrates why OBP and SLG should be weighted equally on different scales. The HR is the quarter and the walk is the nickel. 

Edited June 24, 2025 by Bull

[Jason Ross]

14 minutes ago, Geographyhater8888 said:

Does this at all apply to opposing pitchers? Do they make a distinction between hitting a homerun off of Paul Skenes vs hitting a homerun vs some dime of dozen Rockies scrub in Coors field? And of course, say 1/4 with a homerun and 4 bases with a 1.250 OPS be weighed more than 2/3 with 2 singles and a walk for 3 bases and a 1.416 OPS, same ballpark for example? Just to get the best possible understanding.

No. That all evens out in sample size. Those are all anecdotes, and can simply be a luck based outcome. Remember, how I said outcome based predictions can be an issue previously? Well, this is a great example. The reality is mediocre and poor hitters in the MLB are still professional hitters and are capable of running into a pitch. Just like Paul Skenes is human and capable of hanging a ball, or leaving something with bad placement. That means eventually, there's going to be a mediocre talent over their career of 15-20 PA's against Skenes who looks really good, but the rest of his career isn't. He got lucky to run into some pitches and Skenes was unlucky to hang a few balls to that same guy; it's a convergence of the reality of luck. We don't want anecdotes to win the day. 

A good example of this is former Cub Matt Sczur. Over his career, he had an 81 wRC+; so he was roughly 20% worse than league average. That's bad. Against LHP, he had a 90 wRC+, so, better, yet, still bad. However, in his 8 plate appearances against (likely) HoF pitcher Clayton Kershaw he had a home run and two doubles! Looks like a star! But we know even against lefties he's 10% worse than league average. We wouldn't want those PA's to make him out to be something he's not. Chances are with a larger sample size we'd realize he got lucky to get those outcomes in such a small sample. That'd be a bad data set. We need more than anecdotes. 

In a 600 plate appearance sample size, for example, anecdotes get washed away with a deluge of data. Large sample sets, like, playing half of your games in Coors field will be fixed with the "+" but over the course of a full MLB season, a few PA's against player X won't change anything (as it shouldn't). 

[Jason Ross]

3 minutes ago, Bull said:

Except that your example demonstrates why OBP and SLG should be weighted equally on different scales. The HR is the quarter and the walk is the nickel. 

You're looking far too deeply into that analogy, my friend. That was to demonstrate on the very surface level how we simply can't weigh everything equally, not demonstrate every depth of how OBP and SLG are not to be weighed the same. It's not a full 1:1. 

I would caution to not throw the baby out with the bathwater here or make the analogy more than it needs to be. To be fair to myself and my post, that would be a fairly strong misrepresentation of the data I am trying to help people understand.

[Geographyhater8888]

14 minutes ago, Jason Ross said:

No. That all evens out in sample size. Those are all anecdotes, and can simply be a luck based outcome. Remember, how I said outcome based predictions can be an issue previously? Well, this is a great example. The reality is mediocre and poor hitters in the MLB are still professional hitters. Just like Paul Skenes is human. There's going to be a mediocre talent over their career of 15-20 PA's against Skenes who looks really good, but the rest of his career isn't. We don't want anecdotes to win the day. 

A good example of this is former Cub Matt Sczur. Over his career, he had an 81 wRC+; so he was roughly 20% worse than league average. Against LHP, he had a 90 wRC+, so, better. However, in his 8 plate appearances against (likely) HoF pitcher Clayton Kershaw he had a home run and two doubles! Looks like a star! But we know even against lefties he's 10% worse than league average. We wouldn't want those PA's to make him out to be something he's not. Chances are with a larger sample size we'd realize he got lucky to get those outcomes in such a small sample. That'd be a bad data set. We need more than anecdotes. 

In a 600 plate appearance sample size, for example, anecdotes get washed away with a deluge of data. Large sample sets, like, playing half of your games in Coors field will be fixed with the "+" but over the course of a full MLB season, a few PA's against player X won't change anything (as it shouldn't). 

Makes sense. As far as my player a vs player b hypothetical, would they place more value on player a with 4 bases in as many at bats but only one run scored max or player b with only 3 bases in 4 PA’s with the higher OPS and more changes to score but dependent on his teammates in a luck based outcome? Which I assume is part of the calculation. 

Edited June 24, 2025 by Geographyhater8888

[CubinNY]

Whoever said the Cubs need some guys in the pen who can miss bats hit the proverbial nail on its proverbial head.

 

 

[Jason Ross]

14 minutes ago, Geographyhater8888 said:

Makes sense. As far as my player a vs player b hypothetical, would they place more value on player a with 4 bases in as many at bats but only one run scored max or player b with only 3 bases in 4 PA’s with the higher OPS and more changes to score but dependent on his teammates in a luck based outcome, which I assume is part of the calculation.

Anecdotes. Neither OPS+ nor wRC+ deals in such a small sample size well. These are season-long statistics. Even at a game level, when we're boring down on what is a single game's worth of chances, you'll get funny things. Bad players have good games. Good players have bad games. wRC+ and OPS+ strive to peer through the anecdotal game-by-game outcomes and tell you what they're like on the overall. 

There is a concept in statistics in which is considered "stabilization". Essentially, "at what point is this data set beyond just pure, dumb, random luck?". Each statistic requires a different burden of proof, so there's no magic bullet, but every data point is well above the four-plate-appearance level. Most (beyond just wRC+ and OPS+, think like hard hit rate, barrel rate, pull rate, k-rate...) need at least 50 PA's of outcomes and there are some that require even more to really paint an accurate picture. Prior to that, they're anecdotal and not really important. It's cool to go 3-4, and have a double, but in a one-game sample size, even bad MLB players do that sometimes. So we need more data so that one good game from, like, Leodys Tavares isn't skewing the data. We have more than enough data to know he's a bad hitter, and that he's not magically a .750 guy now.

Scoring runs, however, in general, is not something we want to attribute to a single player and all current analytics ignore this. I had mentioned "noise" before, and these are "noise" data, they create static because they add a second player into the mix. If I hit a ground ball to 2b and a runner scores from 3b, for example, all I did was hit a ground ball; that's bad. That there was someone on third base...I didn't do that, someone else did. A common counter argument to this is "situational hitting", and, yeah, I guess, but if we're being honest with ourselves, "situational hitting" still creates and out, limits more runs scoring later, and the batter is still trying not to make an out in almost all scenarios. It's also anecdotal and not a relevant sample size, so it's something we can usually ignore. The outcome of the PA by the batter (weak ground ball) is more predictive than "but yeah, he got the run home!"

Similarly, if I score from 3b, cool, I should get credit for what I did to get to 3b (say, a double and stolen base) but scoring on that ground ball still required someone else. A different "someone else" in both situations and maybe a run doesn't score. We want to, as accurately as we can, determine only what one, singular player is capable of doing. So, we cut out the static, the noise, as much as possible. These are viewed as "team statistics" and not player specific ones. You can use these concepts to understand why things like ERA (which is affected greatly by team defense for example), RBI, runs scored, etc are largely considered obsolete. They had their moment but we're beyond them as meaningful data points in player evaluation. I think from a team perspective they may share some interest, but for individual players, they do not.

[squally1313]

Brain dumping here because last night's game sucked and then it was hot as hell in our house so I lingered on it more than I/anyone should.

We're 79 games in, and I think, though the particulars have changed a little, my opinion of the true talent level of the team as currently constructed is largely unchanged from what it was going into the season, which is that it's roughly an 88 win team. PCA found another level, but it's important to remember that the FG projections loved him (like 4.5-5 fWAR if I remember), and I think that step up is more or less offset by losing Steele. My analysis of the rest of the offense hasn't changed much based on first half performance. Horner/Suzuki/Swanson are all trending to be worth between 3 and 4 wins, Busch a shade under 3, Happ had a slow start but I still think he's the same player going forward. Carson Kelly is still Just Carson Kelly, but we were able to bank some wins off of his Barry Bonds stint, which is nice.Pitching is, minus Steele, overall the same for me too. Horton is better than I thought, but I'm downgrading Shota a little based on what we saw earlier in the year. Bullpen is 

So anyways, 88 win team. Split the next two games, as an 88 win team should, and you're looking at 47-34 at the halfway point. Another 44 wins in the second half gets us to 91. Do I think that's enough to win the division? Pretty definitively yes. It requires the Cardinals and/or Brewers playing at a 93 win pace the rest of the year to overtake us, and I just don't see the talent level there for that to be a reasonable expectation. On the flip side, I think we're still a clear step behind the Dodgers, and probably not quite at the level of the Phillies. There are moves to make to get us to the Phillies level or even better, and I think we should make them, if just to solidify the division chances (the fact that both teams are right there is a little more concerning than simply a two team race). We have the money, we have the prospect capital, we have essentially our whole lineup returning next year (besides, obviously, Tucker). We should lean into this window. The Dodgers will be better, but playoff baseball is stupid and coin flippy and we should maximize our chances at appearances and ideally byes. 

As for the Jed conversation...as someone who has generally been on his side, I think if he doesn't make it he probably has to go. We usually like to appeal to process over results here and I think the org is in a healthy spot. But I also think we, or at least I, tend to appeal to authority sometimes when moves are made/not made and I don't quite understand the reasoning. "Well this guy looks like he sucks, but Jed's found talent in these spots before, etc" and I think after a hypothetical three years of missing the playoffs, you almost have to do the opposite. I look at the overall decisions/process and see a (flawed) playoff team, but....you have to actually make them at some point. Three seasons of some level of 'going for it' is a large enough sample size that the process over results thing starts to lose a little authority. 

[Transmogrified Tiger]

Yeah I think that if anyone told us on Opening Day that at the midway point of the season that  Steele and Shota would have missed a combined 20 starts due to injury, even if we also told you that PCA went supernova we would not have predicted the team would be playing at a 94 win pace.  Shota coming back will help, the schedule easing off will help(33 of the 64 games after the break are against >.500 teams, Cardinals have 43 and the Brewers 45), the deadline addition(s) will help, and there's still some farm contributions they may yet mine.

[CubinNY]

They purposefully construct a team each year to be mediocre. We can argue about the reasons, but that doesn't change the outcome. I don't know exactly how much blame to assign to whom, but I'd be willing to bet that whether Jed stays or goes, the team will be mediocre moving forward unless they hit on some draft picks and/or international players. To have an argument or whatever, I consider 88 wins mediocre. 

Bad = > 82

Mediocre = 83-90

Contender = above 90

That doesn't preclude them from hitting 90th percentiles and making the playoffs or even winning a WS due to the nature of the format, but it's not what they should be doing given their resources. 

[Transmogrified Tiger]

ZiPS projected one (1) team in MLB to win more than 90 games: https://blogs.fangraphs.com/the-official-and-hopefully-not-too-regretful-2025-zips-projections/

[muntjack]

To be fair, that's not really realistic. I don't believe there has been a season in which only one team has won over 90 games in almost 100 years.  

I understand that they use weighted averages and running a trillion sims, but there isn't going to be a season like that anytime soon, so while I might argue that 88 wins is a reasonable (30th percentile?) outcome for a pretty good team,  I understand what he's trying to say.  

[Transmogrified Tiger]

1 minute ago, muntjack said:

To be fair, that's not really realistic. I don't believe there has been a season in which only one team has won over 90 games in almost 100 years.  

I understand that they use weighted averages and running a trillion sims, but there isn't going to be a season like that anytime soon, so while I might argue that 88 wins is a reasonable (30th percentile?) outcome for a pretty good team,  I understand what he's trying to say.  

I can get that, but it circles back to either you're coming up with your own more vibesy expectations for all the teams to get to that distribution, or you're setting a bar insanely high to not have a roster be framed as a failure(setting aside the Cubs, that also excludes the Mets, Yankees, Phillies, Padres, etc).  And the Cubs are playing at a 94 win pace so it's not as if we're just extrapolating the season so far.

[Geographyhater8888]

On 6/24/2025 at 10:50 AM, Jason Ross said:

Anecdotes. Neither OPS+ nor wRC+ deals in such a small sample size well. These are season-long statistics. Even at a game level, when we're boring down on what is a single game's worth of chances, you'll get funny things. Bad players have good games. Good players have bad games. wRC+ and OPS+ strive to peer through the anecdotal game-by-game outcomes and tell you what they're like on the overall. 

There is a concept in statistics in which is considered "stabilization". Essentially, "at what point is this data set beyond just pure, dumb, random luck?". Each statistic requires a different burden of proof, so there's no magic bullet, but every data point is well above the four-plate-appearance level. Most (beyond just wRC+ and OPS+, think like hard hit rate, barrel rate, pull rate, k-rate...) need at least 50 PA's of outcomes and there are some that require even more to really paint an accurate picture. Prior to that, they're anecdotal and not really important. It's cool to go 3-4, and have a double, but in a one-game sample size, even bad MLB players do that sometimes. So we need more data so that one good game from, like, Leodys Tavares isn't skewing the data. We have more than enough data to know he's a bad hitter, and that he's not magically a .750 guy now.

Scoring runs, however, in general, is not something we want to attribute to a single player and all current analytics ignore this. I had mentioned "noise" before, and these are "noise" data, they create static because they add a second player into the mix. If I hit a ground ball to 2b and a runner scores from 3b, for example, all I did was hit a ground ball; that's bad. That there was someone on third base...I didn't do that, someone else did. A common counter argument to this is "situational hitting", and, yeah, I guess, but if we're being honest with ourselves, "situational hitting" still creates and out, limits more runs scoring later, and the batter is still trying not to make an out in almost all scenarios. It's also anecdotal and not a relevant sample size, so it's something we can usually ignore. The outcome of the PA by the batter (weak ground ball) is more predictive than "but yeah, he got the run home!"

Similarly, if I score from 3b, cool, I should get credit for what I did to get to 3b (say, a double and stolen base) but scoring on that ground ball still required someone else. A different "someone else" in both situations and maybe a run doesn't score. We want to, as accurately as we can, determine only what one, singular player is capable of doing. So, we cut out the static, the noise, as much as possible. These are viewed as "team statistics" and not player specific ones. You can use these concepts to understand why things like ERA (which is affected greatly by team defense for example), RBI, runs scored, etc are largely considered obsolete. They had their moment but we're beyond them as meaningful data points in player evaluation. I think from a team perspective they may share some interest, but for individual players, they do not.

I understand that these 4 at bat sample sizes even out over the course of a 162 game season, but as you’ve stated, OPS is outdated. So even if it evens out with a larger sample size we’ve established that a 2-5 point deviation in wRC+:OPS+ can happen, and the former is preferred. So if you have say 2 players on the same team with identical park factors and an equal OPS+, while one of the players has a 2 point edge in wRC+ for example, what quantifies it? 
 

is it a mathematical formula that has a higher correlation with with player success? like in the NFL for example where passer rating is the most popular quarterback efficiency statistic while Adjusted net yards per attempt correlates stronger with wins and losses. Because a qb who’s 6/9 for 100 yards and 1 td pass and 1 sack for a 6 yard loss has a higher passer rating than a guy who goes 6/10 for 100 yards 1 td pass and no sacks with a low passer rating but a higher ANY/A, if you get my gist.

Edited June 26, 2025 by Geographyhater8888

[Geographyhater8888]

What sticks out a lot between fan graphs and baseball reference is the difference in how they ranks teams. BR uses SOS and RD as a tool to gauge a teams performance.

[squally1313]

Offense YTD: 3rd in wRC, 2nd in baserunning, 6th in defense, 3rd in fWAR

Offense in June (with all these black holes in the lineup): 11th in wRC, 5th in baserunning, 6th in defense, 10th in fWAR

Pitching YTD: 16th in ERA, 17th in FIP, 21st in xFIP, 18th in fWAR

Pitching in June: 1th in ERA, 22nd in FIP, 24th in xFIP, 24th in fWAR

All resources should be going to improving the pitching staff. 

[muntjack]

Agree.  With the dearth of clear 3rd base upgrades,  I'd rather just give Shaw the entire season.  Could pay dividends this year and moving forward and it seems now that the worst case scenario is gold glove caliber defense at 3rd.  I'll take that + upside over Candelario or McMahon.