General Offseason Priorities

[Stratos]

7 hours ago, squally1313 said:

I'm not outright dismissing your point here because it's becoming a bit of a pattern, but another way of looking at this is that individual hitters, by nature, are going to have good and bad stretches over the course of the year and for the last two years we've had all our hitters essentially get hot at once. I don't know how you gameplan in the offseason to fix that problem, there doesn't seem to be any way to correct it/predict it. I don't really buy into like, one player's success rubbing off on another's.....they're all individual players going through individual events. There's maybe some marginal benefit to the rest of your team putting up an 8 run lead and letting you face their center fielder on the mound, but that's minimal at best. So maybe it is a bit of unluckiness that gets reflected in the pythag numbers?

Let's hope it's just randomness and luck.  They hit well in April this year.

[Stratos]

6 hours ago, Transmogrified Tiger said:

Pythagorean record isn't gospel, but 'their pythag record is inflated because they only blew out teams for a certain portion of the year' is kinda missing the point of pythagorean record.

Right but the Cubs underperforming their pythag record was used by someone here as evidence that the team could do better in wins next year.  My argument is that they've underperformed their pythag record 2 years in a row in large part because the last 2 seasons the distribution of their run-scoring throughout the year has been quite inefficient since they've scored in big bunches in the summer and followed virtually the exact same pattern 2 years in a row which might not be just bad luck/randomness.

[jumbo]

On 9/22/2024 at 11:39 PM, Cuzi said:

Thank you for reminding me who I'm talking with. I didn't limit Gilbert to Jackson Holliday, clown. I said "anything short of Jackson Holliday." That means he's the floor. The Cubs aren't competing in that ballpark.

Chiefs game just ended. Time for bed.

No offense, but I have to laugh as I scroll by someone arguing with Tom

[Cuzi]

It's less about arguing and more about teaching someone how to read. It's impossible though, so why try. You can literally quote what you already wrote down and explain what it means and he still doesnt get it.

[mul21]

1 minute ago, Cuzi said:

It's less about arguing and more about teaching someone how to read. It's impossible though, so why try. You can literally quote what you already wrote down and explain what it means and he still doesnt get it.

I'm all for bagging on Tom but he was right this time.

[Cuzi]

2 minutes ago, mul21 said:

I'm all for bagging on Tom but he was right this time.

Literally the only thing he said that was right was Gilbert is out of min salary years.

[squally1313]

16 hours ago, Stratos said:

Right but the Cubs underperforming their pythag record was used by someone here as evidence that the team could do better in wins next year.  My argument is that they've underperformed their pythag record 2 years in a row in large part because the last 2 seasons the distribution of their run-scoring throughout the year has been quite inefficient since they've scored in big bunches in the summer and followed virtually the exact same pattern 2 years in a row which might not be just bad luck/randomness.

Not to sound snarky, but do you have a better explanation for that last sentence besides 'bad luck/randomness'?

And to split hairs a little bit, the (rough) theory behind pythag is that hitters and pitchers can't really control ordering of events, clutch isn't actually a thing, etc. And so if you took the cumulative performances from everyone on the team last year and then had them spray out in a random order 10,000 times, the theory is that our average record would be better than what it currently is. No one is necessarily claiming that this is going to translate into next year, we have age regression, changes in roster, BABIP gods, etc. Just that the overall production of our offense and defense/pitching (measured in terms of runs scored for the offense and runs allowed for the defense/pitching) implies a higher level of talent than the record currently shows. 

[Bertz]

4 minutes ago, squally1313 said:

Not to sound snarky, but do you have a better explanation for that last sentence besides 'bad luck/randomness'?

And to split hairs a little bit, the (rough) theory behind pythag is that hitters and pitchers can't really control ordering of events, clutch isn't actually a thing, etc. And so if you took the cumulative performances from everyone on the team last year and then had them spray out in a random order 10,000 times, the theory is that our average record would be better than what it currently is. No one is necessarily claiming that this is going to translate into next year, we have age regression, changes in roster, BABIP gods, etc. Just that the overall production of our offense and defense/pitching (measured in terms of runs scored for the offense and runs allowed for the defense/pitching) implies a higher level of talent than the record currently shows. 

I think on top of this, even if there is something to it that is not luck, the roster turns over fast enough to generally wash it out after a year.  Obviously the team already swapped Ross for Counsell.  And the bullpen, which I'd put 90% of the blame for this on, will have almost entirely turned over from the start of '23 to the start of '25.  

[CubinNY]

At the risk of turning this into something it shouldn't here are some axioms

1. We live in a cause and effect universe

2. Behavior is a dynamic property that is frequently changing and subject to multiple variables

3. Reality is never wrong

4. All mathematical modes are subject to measurement error. The errors can be large or small. When the model deviates from reality, the model has errors. See #3. What some call luck/variance is an attribute or explanation not based on a model.  It's also called an explanatory fiction. 

----------------

The fact that the Cubs underperformed or some team overperformed their Pythagorean record doesn't mean they got lucky/negative-positive variance. It means the model isn't perfect and other unidentified factors may have contributed to the deviation. It doesn't mean the model is useless either. 

[squally1313]

20 minutes ago, CubinNY said:

At the risk of turning this into something it shouldn't here are some axioms

1. We live in a cause and effect universe

2. Behavior is a dynamic property that is frequently changing and subject to multiple variables

3. Reality is never wrong

4. All mathematical modes are subject to measurement error. The errors can be large or small. When the model deviates from reality, the model has errors. See #3. What some call luck/variance is an attribute or explanation not based on a model.  It's also called an explanatory fiction. 

----------------

The fact that the Cubs underperformed or some team overperformed their Pythagorean record doesn't mean they got lucky/negative-positive variance. It means the model isn't perfect and other unidentified factors may have contributed to the deviation. It doesn't mean the model is useless either. 

Is this just an argument against the concept of luck, or I guess using the term 'luck'?

[CubinNY]

Just now, squally1313 said:

Is this just an argument against the concept of luck, or I guess using the term 'luck'?

luck is a default explanation. it's what we say when we don't know. It's an attribute to a mystical force for which there is no tangle proof. Is an unlikely outcome due to voodoo or witchcraft? Most people would say no. But we talk about luck all the time when it's the same explanation.

When we analyze or explain outcomes in sports we are dealing with complex systems, so we invent things that help us (models, descriptive and inferential statistics, spray charts, etc.). Those inventions are often useful, but never perfect. 

I saw a post saying the Brewers got lucky this year because they didn't have a guy have a huge year. When in fact, the brewers are the organization that the Cubs aspire to be. They know how to pick players that fit their model and plug them in to get the best outcomes. It's not luck that they continue to dominate our division. 

[Andy]

13 minutes ago, squally1313 said:

Is this just an argument against the concept of luck, or I guess using the term 'luck'?

I don't care for the use of 'luck' when discussing Pythag records and BABIP and the like. Most of the time, one-run wins, in the moment, are earned. That said, it's not a sustainable strategy - you wouldn't go out of your way to try to build a team to win the most one-run games or have the highest BABIP, because such a plan almost certainly wouldn't work. You don't see teams consistently lead baseball in those statistics because it's borderline impossible to do repeatedly.

Edited September 24, 2024 by Andy

[Tim]

51 minutes ago, CubinNY said:

At the risk of turning this into something it shouldn't here are some axioms

1. We live in a cause and effect universe

2. Behavior is a dynamic property that is frequently changing and subject to multiple variables

3. Reality is never wrong

4. All mathematical modes are subject to measurement error. The errors can be large or small. When the model deviates from reality, the model has errors. See #3. What some call luck/variance is an attribute or explanation not based on a model.  It's also called an explanatory fiction. 

----------------

The fact that the Cubs underperformed or some team overperformed their Pythagorean record doesn't mean they got lucky/negative-positive variance. It means the model isn't perfect and other unidentified factors may have contributed to the deviation. It doesn't mean the model is useless either. 

Do you agree that variability exists? It seems like you are suggesting that every process is deterministic

[squally1313]

15 minutes ago, Andy said:

I don't care for the use of 'luck' when discussing Pythag records and BABIP and the like. Most of the time, one-run wins, in the moment, are earned. That said, it's not a sustainable strategy - you wouldn't go out of your way to try to build a team to win the most one-run games or have the highest BABIP, because such a plan almost certainly wouldn't work. You don't see teams consistently lead baseball in those statistics because it's borderline impossible to do repeatedly.

Sure, I don't disagree with this. I think we use 'luck' as a high level term to describe non-sustainable events or pieces of events that have been largely proven to be non-sustainable. Wins are wins, however you got them, they are banked and credited to your team, and the Cubs haven't gotten enough of them one way or another. The simplest way I can describe it is if you declared tomorrow that the MLB season would go another 100 games starting Monday, you could make a strong argument that relying on pythag over W/L records YTD would be a better indicator of what your record would be for those 100 games. Obviously we're coming to a significant cut off point and rosters will churn and guys will grow 6 months older and pick up and lose skills and injuries and I think we're all aware of that. 

I think CubinNY would disagree with the statement that the job of an offensive player is to produce runs. He would say it's to win games, I think. Apologies if not. And, honestly, it's not an incorrect argument. But all of our individual statistic tracking (that drives 'player value') is based on a players ability to produce runs (or avoid outs). And vice versa for pitching and defense. And so a statistic that looks, at a macro level, of how many runs an offense were produced vs how many runs were allowed, with no regard to how those lined up in 9 inning chunks, is beneficial in predicting future performance. 

[CubinNY]

Just now, Tim said:

Do you agree that variability exists? It seems like you are suggesting that every process is deterministic

Variability exists, but we live in a deterministic universe, it's explainable in theory and extrinsic to behavior.  So a guy goes out to a restaurant the night before a day game and has a slight case of intestinal discomfort. The next day his performance is adversely affected. Several factors may or may not account for his performance and how much it deviates from typical performance, none of which depend on luck. We may not be able to quantify the contributions of all the variables with any certainty because of how complex the interplay is between all of them. That's what makes sports fun. 

Variability of behavior and luck are often used interchangeably by people.  

[Tim]

1 minute ago, CubinNY said:

Variability exists, but we live in a deterministic universe, it's explainable in theory and extrinsic to behavior.  So a guy goes out to a restaurant the night before a day game and has a slight case of intestinal discomfort. The next day his performance is adversely affected. Several factors may or may not account for his performance and how much it deviates from typical performance, none of which depend on luck. We may not be able to quantify the contributions of all the variables with any certainty because of how complex the interplay is between all of them. That's what makes sports fun. 

Variability of behavior and luck are often used interchangeably by people.  

When you have a variable process (such as hitting) and the outcomes of that variable process land in your favor...you got "lucky". It isn't repeatable or predictable (other than it will happen a certain percentage of the time). You try to build a team such that the expected results are as good as they possibly can. But variability does exist and it can work to your favor or against it. Hence lucky and unlucky.

Yes, many people use the word wrong. But that doesn't mean that it doesn't exist.

I shall stop my pedantic side conversation. 🙂

[CubinNY]

4 minutes ago, Tim said:

When you have a variable process (such as hitting) and the outcomes of that variable process land in your favor...you got "lucky". It isn't repeatable or predictable (other than it will happen a certain percentage of the time). You try to build a team such that the expected results are as good as they possibly can. But variability does exist and it can work to your favor or against it. Hence lucky and unlucky.

Yes, many people use the word wrong. But that doesn't mean that it doesn't exist.

I shall stop my pedantic side conversation. 🙂

An outcome that is hard to repeat means that's is hard to repeat outcome. The outcome is whatever happened. You might as well say Lucifer made it happen.

[CubinNY]

Getting back on track - The Cubs strategy for the bullpen banks on hard-to-repeat outcomes. Grabbing guys off the DFA/waiver wire and mixing them in through attrition is based on a failure model. Failure in the pen usually means losing winnable games. We've seen it play out now for a couple of years. They are recycling the same strategy this year. We should expect similar results, but hey maybe they'll get "lucky".

[Jason Ross]

I don't think their bullpen strategy isn't working though. This year, they were banking on Julien Merriweather, Adbert Alzolay Mark Leiter, and Hector Neris as the back end. Two of the four of them essentially missed 90% of the season hurt. And while Neris was bad as a "closer", the stabilized most of the year as a "fine" arm most of the time. They also lost Yency Almonte right away after he was looking like a useful weapon . Now, if you want to argue the Cubs should have done a little more, I'm not sure I'd fully disagree there (the rest of the BP opening day was a mish mash). but I think the Cubs were also hoping Ben Brown (thrust into the rotation due to injury) and Cade Horton (who got hurt and never had a chance to help) were going to add to the BP throughout the year. The BP took what feels like a heavy amount of damage early on and the Cubs suffered for it. June 1st on the Cubs had a top-5 BP ERA and the 13th best FIP. They over performed a bit via ERA, but it was a "fine" bullpen once they were able to get it to settle in.

Part of the issue with BP building is that it's incredibly variable to begin with. With smaller samples come larger fluctuations in performance and it's super regular to see a good reliever have a bad year and vice-versa. Pinpointing who that is going to be can sometimes be a guessing game. You'll likely find a little more of that variance going the DFA'd guy route, but you can find bummers like Matt Moore. Chris Stratton and Joe Kelly in that mid-range as well. The top-top of the market is a little more stable (Stephenson, Hader, Lopez and Jimenez,. for example) so I think you can say "the Cubs should have probably shot a bit higher than Neris all things considered" but I'm not even sure one other BP arm was going to solve 2 months of players dropping like flies.

[Tim]

28 minutes ago, CubinNY said:

An outcome that is hard to repeat means that's is hard to repeat outcome. The outcome is whatever happened. You might as well say Lucifer made it happen.

lol

[squally1313]

15 minutes ago, 1908_Cubs said:

Part of the issue with BP building is that it's incredibly variable to begin with. With smaller samples come larger fluctuations in performance and it's super regular to see a good reliever have a bad year and vice-versa. Pinpointing who that is going to be can sometimes be a guessing game. You'll likely find a little more of that variance going the DFA'd guy route, but you can find bummers like Matt Moore. Chris Stratton and Joe Kelly in that mid-range as well. The top-top of the market is a little more stable (Stephenson, Hader, Lopez and Jimenez,. for example) so I think you can say "the Cubs should have probably shot a bit higher than Neris all things considered" but I'm not even sure one other BP arm was going to solve 2 months of players dropping like flies.

Yeah, you can either 'failure model' your way through league minimum dudes and AAAA guys or you can pay guys $8-$12m a year and also 'failure model' through them but more expensively or you could spend $75m/year on your bullpen. 

[Andy]

28 minutes ago, 1908_Cubs said:

Part of the issue with BP building is that it's incredibly variable to begin with. With smaller samples come larger fluctuations in performance and it's super regular to see a good reliever have a bad year and vice-versa.

I have no quarrels with your post, really, but I quoted this part just to note that this passage illustrates why it was a bad idea to sink a bunch of resources into a guy like Hector Neris, whose underlying numbers last season were screaming that a large regression was coming.

I also just looked up his Astros numbers since he signed there and funnily enough he has a worse ERA than he did in Chicago (in only 13 1/3 innings, granted) despite a far better WHIP. Go figure.

Edited September 24, 2024 by Andy

[Bertz]

Hector Neris has a 3.92 ERA this year.  Josh Hader had a 3.67.

Emmanuel Clase last year blew 12 (!!!) saves despite a 3.22 ERA.

There's just not as much rhyme or reason to bullpen performance as there should be. 

It's not total anarchy, but generally the best approach is to build up as reliable a top 3 or 4 as you can manage and then flank them with as many live arms (ideally optional) as you can.  And with that top group there's a delicate balancing act between buying track record without locking yourself into long term money.

[Jason Ross]

54 minutes ago, Andy said:

I have no quarrels with your post, really, but I quoted this part just to note that this passage illustrates why it was a bad idea to sink a bunch of resources into a guy like Hector Neris, whose underlying numbers last season were screaming that a large regression was coming.

I also just looked up his Astros numbers since he signed there and funnily enough he has a worse ERA than he did in Chicago (in only 13 1/3 innings, granted) despite a far better WHIP. Go figure.

I think it's fair to point out Neris was going to regress. But nothing suggested he was going to faceplant off a cliff. He had that crazy low-ERA, but I think the whole league knew that wouldn't keep up and why he was around for what he was. I mean, if we all thought the <2.00 ERA was going to continue he'd have gotten more and quicker. 

I don't think Neris was horrible, either. He had some really notably bad outings that were magnified because he probably shouldn't have been closing and was (due to a combination of necessity and no one else really separating themselves). Probably was worse than his ERA suggested (his xFIP is 4.576 and his FIP was 4.10), but a lot of that was a terrible April. Past April he had a 4.10 xFIP with the Cubs. Closer worthy? No. Not even what we paid him worthy. But like...okay. 

Regardless I don't think it was a terrible idea to sign Neris. I just wish they would have also added another arm to really lengthen the BP out a bit. 

[CubinNY]

34 minutes ago, Bertz said:

Hector Neris has a 3.92 ERA this year.  Josh Hader had a 3.67.

Emmanuel Clase last year blew 12 (!!!) saves despite a 3.22 ERA.

There's just not as much rhyme or reason to bullpen performance as there should be. 

It's not total anarchy, but generally the best approach is to build up as reliable a top 3 or 4 as you can manage and then flank them with as many live arms (ideally optional) as you can.  And with that top group there's a delicate balancing act between buying track record without locking yourself into long term money.

ERA is not a good metric for a bullpen pitcher.