2022 JLW Preview

[Ted]

Well, I wanted to finish this before Aaron got his preview up, but I procrastinated, so now at least I get to disagree with him. Being right when Aaron is wrong is one of the reasons I got back into fantasy baseball, so I'm glad we do in fact disagree. On to the preview.

Two things stand out when looking at the JL West. First, aside from my sad sack franchise, this is a division of teams that are one major trade shy of being a contender. Each of the top three teams has a huge flaw that will almost assuredly make them a playoff also ran. Second, this division is bad. You could cherry pick the best players at each position off each team and still not assemble a team that could compete with Calgary, Carolina, Louisville, Brooklyn, etc. You would end up with a monstrous offensive team, but excluding the JL Midwest, this division probably has the worst pitching in baseball.

The top three teams are basically mirror images of each other, one with a bit more offense, one with more pen, one with slightly more defense, one with a bit more starting pitching, etc. A caveat here, if Vancouver doesn't lose Oliver to crazy retirement and have an injury to Alexander, they run away with the division, so lets start our stroll through the blandness that is the JLW in Canada.

VANCOUVER

Last year: 91-71, 1st, lost in Doubleday

Key additions: none (Irishman Patrick "Red" O'Malley is a career underachiever and needs to get off the damn sauce, and defensive additions like CF Hopkins and RF Cheney don't count when you already have a bitchin' defense)

Key loses: Oliver (see above), Alexander (half season)

That's a TON of pitching to miss. The rotation is now a shambles with a bunch of poor man's 3rd and everyone else's 4th starters taking turns hoping the defense continues it's Hoover impersonation, and prays to not give up the long ball. Manuel Guerra, Ace Young, Jose Linares and Juan Lucero don't scare anyone. While they are all quality big leaguers, they all belong safely nestled behind a dominant ace or two. These are all 4+ FIP guys, and a solid defense can help that a little bit, but won't protect you from the likes of the hitters in Hawaii's and Valencia's lineups day in and day out. Vancouver's bullpen is gonna take a lot of abuse. Skip Stewart, Tomas Guerrero, and Mike Clifton are all solid, but not spectacular, and after them there isn't much. (Except Patrick O'Malley and screw that guy)

The position players are going to have to carry this team, and it's possible they can. Elroy Hinson is a legitimate up and coming stud. Vancouver looks to currently have him DHing. I think this at this point he is a marginally better defensive C than Peter McClure, who is aging but is still a very good power bat. I'd DH McClure and start Hinson, unless this is a move to preserve Hinson's knees. Johnny Snyder is a terrific glove at 2B with a better than adequate bat. Pierce is still a tremendous two way CF and A'amakualenalena (if I spelled that right I should get an extra 5 PP) is a nice power bat to squirrel away in the bottom of the order for some unexpected pop. Vancouver has a diamond glove caliber player at every position as well.

The rotation should get a boost when Alexander returns, but I just don't see this team making it until then. I could see a strong start out of the gate, then a slow slide back in the standings as the pitching staff gets worn thin. I think by the time Alexander gets back it will be too little to late. Oliver's retirement is really a shame. That one move transforms this from a club with a solid staff front-lined by a true stud, with veteran position players all hitting their stride to a team whose window will soon be closing that has a big hole. Can anyone else see that bit of randomness being the beginning of a downward spiral that leads to a bunch of ill advised moves as Vancouver tries to hold on, gets into debt, and eventually has to relocate to get away from the disenfranchised fan base? It's like a 30 for 30 waiting to happen.

Prediction: 81-81
Silver Lining: This is a team one trade for an ace away from being the division favorite. I don't see the move out there, but if Brett pulls it off, he should get GM of the year.

HAWAII
Let's take a little swim out into the Pacific to find our next club. Hawaii has looked like a team making slow progress every season, but was that jump over .500 the real deal or an aberration for what has essentially been a .450-.480 ball club for years? The Pythagorean record would have you believe that it is, but I'm not sure I buy it.

Last year: 84-78, third

Key additions: 2B Bryan Vogel, RF Matthew Stuart?, RP Juan Castillo?

Key losses: none

Hawaii runs a pitch to contact scheme that is essentially cut from the same cloth as Valencia's. The staff features a bunch of junkballers in Floyd Carter, Chris Lee, Jorge Paz, Max Fischer and Gabriel Medina who will rely heavily on Hawaii's stellar defense to pick them up. But like Vancouver, no one is scare of these pitchers. These are a bunch of 4 FIP guys as well. What gives them the slight edge I feel over Vancouver is that Hawaii has consistently gotten them to outperform their raw stats at a better rate than Vancouver, but I still wouldn't trust my playoff hopes to league average pitchers. The one wild card is late bloomer Jose Olivarez, who has struggled with command through out his career. He has made progress in the last year, but he'll need to make significant progress this year to contribute. If he does, the Tropics have their ace.

The pen depends a lot on how good Ramon Barbosa can be and if Jose Hernandez can figure it out. Otherwise there's not much there besides a lot of 6/7/6 failed starter types.

Hawaii's offense clearly will go as far as SS John Galt takes it. I'm sure the prospect of facing him the next day will cause no small number of poor night's sleeps for opposing right handers. This could be the year he breaks out and puts up some ridiculous line like .300/.400/.600. Or it could be thee year he establishes himself as "merely" a perennial all star, but not the MBBA megastar he could be. His game does have some flaws. He struggles against lefties, is quite slow, and is only a slight bit of defensive regression (not uncommon in the mid to late 20's) away from being a liability at SS. (I'm sure he doesn't care. I can think of worse things than being a rich superstar on a club that lives in Hawaii. He even has a twin on the team to act as a diversion when he doesn't want to talk to the press. I sure hope he is taking advantage of this, and were I Jack Galt, I would definitely be lying to the beach hotties about which brother I was.) Galt has a good supporting cast. LF Fernando Ortiz is one of those OBP doubles machines that make coaches drool. 1B Jorge Caballero's value will defy his stats due to his great glove, power, and ability to see a lot of pitches. RF John John Miranda is a prototypical speedy lead off RF. CF Emory Sharp is probably a little out of position at CF, but is a nice player. I'm really not sure how much free agent pickup Brian Vogel improves this team. He is a slightly better defender than his predecessor, and has tremendous pop, but he has a career ops+ of 97 and has been a 1 to 2 WAR player for the last several seasons. Defensively, this club is solid, but not as good as Vancouver. Offensively, I'm not sure they're as good as Vancouver on paper, but this lineup just fits together very well and as mentioned before, if Galt can reach that next level will just absolutely wear down opposing clubs.

This years Tropics club on paper looks like they improved from last year's squad by a few wins, which if you take that pythagorean record into account makes them look like the division winner. I'm a pessimist though, so Mike will have to forgive me if I'm just not sold on this team. On interesting move would be to trade Olivarez for an established ace in a package or at least for a number two type (more likely given his advanced age, but I don't know the Tropics farm system that well)

Prediction: 84-78, 2nd

VALENCIA
For our next stop we'll head back the mainland to talk about the Stars. I'm a big believer in track records, and it's hard to argue with a team that hasn't finished worst than second since 2012, and has finished in first 4 of the last 5 years. This is especially true when the team that knocked them off of their pedestal just got jobbed by circumstance while the Stars bring back their same winning formula.

Last year: 87-75, 2nd, lost in Geoghegan

Key additions: 3B Dexter Morgan, RF Stephen Ireland (sort of? maybe? He's probably a better glove than what's been out there in the past, but he's hardly "impact")

Key losses: none

I'll admit it. I'm a starting pitching guy, and Valencia has what passes for it in this division anyway. Tony Suarez, Chan Master Yen, and Jose Leon are all better than any pitcher on either Hawaii or Vancouver's rosters to start the season. Sure, Brian Middleton is getting long in the tooth, but he stunk last year and it didn't matter. After those four, Vancouver has a fleet of 7/7/7, 8/6/7, 6/7/8 types that are as good as, if unproven, as anything on Hawaii's or Vancouver's rosters. Keep that in mind for later.

Then pen is by far the best in the division anchored by Skip Glendenning, Jesus "who cares why he's effective" Flores, and old but strangely effective Iestyn Pilgrim. The rest of the pen is 7 or 8 more of those 7/7/7 starter I mentioned, and that's what huge about this team . They have pitching depth. Unlike the other teams in their division, they are not one pitching injury practically guaranteed) away from having to give significant innings to a total scrub. Strong starting pitching and pitching depth equals wins.

The Stars have a fairly solid offense as well. C Jason Dunn is obviously a beast, but he does likely cost the Stars some games defensively . Frankly I'd DH him and start Leonardo Rosa purely to get Rosa's glove in the game. This would give a superior DH, rather than the mediocre Elliot Spencer, and Rosa's glove and pitcher handling abilities probably make him a better than league average catcher. Around Dunn you have star SS Kevin Minuso, and a couple decent platoons at 1B and LF with Gary Knowles/Luis Navarro and Arlen O'Rourke/Tim Wilson. Dexter Morgan is an improved defensive option at least at 3B. The 2B, CF are spots are not great, but they aren't complete holes either.

It is no doubt that both Vancouver and Hawaii have better defenses and offenses than Valencia, but I think Valencia has just enough of each, and vastly superior pitching, to make it 5 out of 6 years in first. Valencia's move would be to get halfway through the year, and trade some of that pitching depth for a middle of the order bat.

Prediction: 91-71

CALIFORNIA
Last and definitely least we have my team. Ugh. You can skip this if you like and head to the wrap up.

Last year: 69-94, last

Key additions: Jorge Franco/Cris Vasquez at DH (both promoted from AAA)

Key losses: Gabriel Campos, Jeremy Ginley

2022 begins life without Gabriel Campos. It will suck. However, it might not suck as bad as it could. Ricardo Diaz has disgusting stuff and will anchor the rotation, but he's actually not expected to be any better than last year. What IS expected to be better is the combination of Gow Cho whose lack of movement doesn't matter as much in spacious California, Juan Lopez (ditto, and both were in the pen for the first half of last season and absolute garbage was starting), and Enrico Macias whose presence will give the chance for a win on the occasional days when and if the offense wakes up. Then pen, outside of Jeremy Casamonica, who is a MR at best, is atrocious and will be discussed no further.

Speaking of atrocious, ladies and gentlemen, the California Crusaders Lineup!!! Again, it will be better than last year, but not nearly good enough. LF Mark Stuart is a solid middle of the order guy. LF Juan Fernandez probably can be a leadoff type eventually, but he's being converted to RF this year (due to Stuart) and will likely be a net loss because of that. Bartholomew Bailey is actually interesting, in that while a poor offensive, Diamond glove corner outfielder, he has the range to be an adequate hitting super defensive CF. Rien Vis is okay at some infield position, and the 1B/DH combo of Vasquez/Franco/Larson has to be better than the clusterf%^$ we had last year. The defense would have been better but RF and CF while Fernandez and Bailey are learning will stink up the joint enough to wipe out any advantage over last year.

So yeah, we're gonna lose a lot of games. How many? I don't really know. I picked us to be a net neutral division (worse than the Atlantic, but equally better than the Midwest), so my record below reflects that, but I still feel it's a bit optimistic.

Prediction: 68-94


WRAP UP
Hey, if your still reading, great, because somehow at 830 am (5 red bulls in 4 hours) I'm still typing. The only certainty is that California will be last. Injuries and trades will be huge. As I look back on this I really think I'm selling Vancouver short, but I'll stick with my initial impression just for the sake of argument and hope they prove me wrong. Any team not named the Crusaders could take the division, but I have to pick some kind of order, so...

Valencia
Hawaii
Vancouver
California

The end. (Also I'm stealing Aaron's big font idea cuz I'm so freaking tired I can't read this. Damn old age)

[bschr682]

Hinson is DHing to get more AB's. Playing everyday at C requires more off days than playing everyday at DH. So if I have to pick between McClure and Hinson as to who I want to get more AB's its Hinson by a landslide.

This is the reason I have been so open to moving Hinson for a stud at a different position. All other things being equal, C is the least valuable position simply because they need way more time off during a season.

[Ted]

Solid strategy. Didn't think of it that way.

[aaronweiner]

You're obviously not a big believer in Pythagorean records. You've discounted Hawaii's AND Valencia's - last year, Valencia's run differential was underwater, while Hawaii's was way, way behind their final record.

[Ted]

You're obviously not a big believer in Pythagorean records. You've discounted Hawaii's AND Valencia's - last year, Valencia's run differential was underwater, while Hawaii's was way, way behind their final record.

You're right, I'm not. I was actually going to mention this, but I never got around to it. I feel like Pythagorean records are wonderful in retrospect. When you look at a team and say, "They should have more success than that", and the pythagorean record backs it up, great. I don't think they are good predictors of the future. I have no evidence for this and have not researched it at all. I know you like them a lot, so maybe you have and I am wrong. I just remember an awful lot of "the Washington Sea Dogs under performed their Pythagorean record by ten wins last year" followed by another third place finish.

Seriously though, I think that both middling offenses and middling pitching are tremendously unpredictable year to year, and middling defines every aspect of the JLW. The other thing that would be fun to look into is teams that consistently both under perform and outperform their pythagorean records and see if anything shows up. I'd bet teams with big offenses/poor pitching consistently underperform/outperform their Pythagorean records. An 8-1 win is no more valuable than a 2-1, and vice versa for the loss.

[aaronweiner]

If you ever read Grantland's Bill Barnwell, he's huge on expected records. Most statheads are pretty much all over Pythagorean record as a predictor of future success or failure.

Valencia has been a notable exception, but they've benefited more from other teams not being any good than their own greatness. I'd say 90 wins for Valencia is more than a little optimistic.

[Ted]

Aaron, I'll agree that the baseball statheads Grantland uses are by far more knowledgeable than me, but I just spent 40 minutes searching for and read 3 papers that all claim Pythagorean records predict future success and then do nothing to prove it. All they do is prove that runs scored and allowed are better predictors of future winning percentage for the same team during the same season than existing winning percentage. Well duh. Barring massive personnel changes, you now have an established track record for that club, that year, against it's common opponents.

I understand the desire to extrapolate of this to the next season, but from a scientific standpoint that is a HUGE jump. There are so many things that change on a single team with NO personnel changes from one year to the next. For a division, you're talking multiple player changes on multiple teams, effects of aging, development, injuries. There is a FAR greater degree of statistical variance in the performance of a team in terms of runs scored and runs allowed from Oct 1 to say June 1 than there is from June 1 to Oct 1.

Similarly, there is no good explanation for teams deviance from their Pythagorean record. Most of what I'm reading says that the two most common believed sources for deviation from Pythagorean record are bullpen (interesting in the Valencia/Hawaii/Vancouver situation), but even more commonly just "luck" or randomness. This is a frank contradiction of what Pythagorean prediction is about. The idea is that baseball gives such a large sample size so using Pythagorean records to predict future success is reliable because all the testing models give statistically significant p values. But if that's the case, how can you explain deviation, which frankly is very common (I'd say randomly one team for every four each year), by randomness? It contradicts itself. I'm not saying they are worthless. I'm just saying they are of limited use from one season to the next, especially when trying to predict a less than ten win difference between clubs. In fact, what we really need to be looking at is the standard deviation for Pythagorean winning percentage versus actual record.

[Ted]

Aaron, I'll agree that the baseball statheads Grantland uses are by far more knowledgeable than me, but I just spent 40 minutes searching for and read 3 papers that all claim Pythagorean records predict future success and then do nothing to prove it. All they do is prove that runs scored and allowed are better predictors of future winning percentage for the same team during the same season than existing winning percentage. Well duh. Barring massive personnel changes, you know have an established track record for that club, that year, against it's common opponents.

I understand the extrapolation of this to the next season from a scientific standpoint, but that is a HUGE jump. There are so many things that change on a single team with no personnel changes from one year to the next. You're talking multiple player changes on multiple teams, effects of aging, development, injuries. There is a FAR greater degree of statistical variance in the performance of a team in terms of runs scored and runs allowed from Oct 1 to say June 1 than there is from June 1 to Oct 1.

Similarly, there is no good explanation for teams deviance from their Pythagorean record. Most of what I'm reading says that the two most common believed sources for deviation from Pythagorean record are bullpen (interesting in the Valencia/Hawaii/Vancouver situation), but even more commonly just "luck" or randomness. This is a frank contradiction of what Pythagorean prediction is about. The idea is that baseball gives such a large sample size so using Pythagorean records to predict future success is reliable because all the testing models give statistically significant p values. But if that's the case, how can you explain deviation, which frankly is very common (I'd say one team out of 4 each year), by randomness? It contradicts itself. I'm not saying they are worthless. I'm just saying they are of limited use from one season to the next

4 papers. Need a stats refresher. Now my head hurts. I sure hope every else is ignoring this boring "conversation". I should sleep but instead I think I'll write up some stuff about what happened during spring training for California.

[agrudez]

Ted,

There have been multiple studies that confirm the correlation between WAR and record. WAR is basically just a simple conversion of runs scored/prevented by a player. PYTH is how many wins should have occurred based on run differential. The logic circle thus loops back around to there being confirmation of correlation between PYTH and predicted success.

There are a few factors that I believe can mislead PYTH, though.
1) Offensive/defensive BABIP - Obviously, any positive swing in either of these will result in a disingenuous run differential and, thus, needs to be accounted for when presuming next year's record based on last year's PYTH.
2) Schedule - It is less so than most sports (due to the larger sample size) in baseball; however, who you play will obviously dictate how well you succeed to a large extent. If there is a large swing in the talent of your division or even conference then last year's success predictors will be less accurate.
3) Bullpen - This is a more of a "feel" observation than one grounded in math; however, I believe a weak bullpen can lead to a -PYTH and a strong bullpen can lead to a +PYTH. The rationale for this is that bullpens have by far the least amount of impact on your team's run differential because they just don't impact a majority of the innings your team plays, so their impact on record isn't properly modeled in the statistic. This can best be seen in a team's 1-run W-L record. A strong bullpen will generally lead to a better record in that category while a weak one will do the opposite. Since PYTH is based on W-L record, an overwhelming swing either above or below .500 in that category will naturally lead to a swing in your PYTH.

[Ted]

Sounds good I guess. I get the fundamentals. But look how much WAR varies year to year. I just don't think it's a simple taking last years runs scored and against, adding or subtracting the previous WAR of added or removed or whatever, and using that as a prediction. One of the fundamental steps in approving a test, or treatment, or predictor, or whatever, should be a non-superiority test. Essentially, compare two standards, and see if the new one provides any useful further information. You can make the greatest predictor in the world, but if it tells you with a hundred percent certainty that if you flip a coin 1 million times, you will get 50 % heads, then that predictor is useless.

Frankly, this is how I feel about Pythagorean records, except they aren't even that good. They tell me which teams are really good, which teams are really bad, which teams are kind good, kinda bad, etc. Beyond that, I don't see much. I looked at last year. 8 teams or one out of three were 5 decision or more off from their Pythagorean records. The deviation decided two divisions.

In 2020 TEN teams were off by more than 5 decisions. Three teams with worse Pythagorean records went to the playoffs.

It seems to me that they give you a good ballpark figure, but that's about it. It's a standard deviation question, and again, in my limited research I have yet to see the standard deviation associated with prediction by Pythagorean record, but in the admittedly small sample size I have seen, it doesn't look too good.

[leejay56]

Great write up. I feel it's going to be close. Down to the wire year again.