MLB 2019: The Good, The Bad, and The Mets

[DMC]

3 minutes ago, Demetri said:

He is also not solid in exit velocity and hard hit%, he is actually below average. For hard hit %, he is the 25th percentile of MLers, his exit velocity is at the 37th percentile. This is the "generic power hitter who doesn't do anything well" thing I was saying. The baseball savant link shows that he is below average in both categories. His hard hit % is actually a major cause for concern, as is his exit velocity.

Sure, but my point was he's rock solid in barrels.  Which coupled with those hard hit/exit velo percentages suggests when he does get a hold of one, it's hit particularly well.  Which explains his inflated ISO.  And conforms to the eye test that he's got bat speed like Sheffield.  

6 minutes ago, Demetri said:

Yes, there are some players who inherently have high BABIP and xwOBA players. They are the exception, not the rule.

Gotta admit, I find it a bit amusing that your argument is so invested in these metrics, but the fact is Frazier still has a very small sample in terms of major league experience.  24 year olds often improve - and even often considerably improve - as they progress.  The empirics bare that out as well.  Projections are inherently conservative, even Szymborski will tell you that.

11 minutes ago, Demetri said:

He has simply gotten lucky. And almost all FOs (We miss you Dave Stewart, we can also probably include the Orioles) are savvy to this. They won't pay for half a year's worth of luck-driven stats that only make him 20% better than average. I concede that his defense will bounce back (it virtually has to) but his offensive stats are gloom and doom.

So, you get to this later in the post, but it's hard to carry on this discussion because it seems to be you're making self-contradictory arguments.  You said you agreed Frazier could be the centerpiece of a trade for Bauer in the last post (and yes, certainly for MadBum, who I agree isn't worth much.  I wouldn't even want them to give up Frazier for him, was just making a point in those prior links I cited).  But you continually argue that front offices view him as a backup outfielder.  It can't be both.  Nobody is trading Trevor Bauer for a package that has an expected 4th outfielder as the key piece.

15 minutes ago, Demetri said:

But industry consensus is that Florial provides almost the same value (the calculator has it off by $500,000 which is consistent with various other prospect valuation figures). These figures ARE hypothetical, but they're grounded in common practice and past trades.

This is all nonsense.  I am very confident that different teams evaluate Florial in very different ways - just as different public scouting sources evaluate and rank him in drastically different ways.  You can't put a number on it and assume every team is going to follow the same valuation.  That's not how the real world works.  And the facts are he's just getting back from an injury, and hasn't even stood out thus far at High A, which he's repeating.  You are not going to convince me that that type of prospect has equal or more value than Frazier at this juncture, full stop.

BTW, this goes for all your estimates on the FV and how that translates to salary worth.  First of all, this is not the only consideration when teams trade players, even prospects.  Second of all, you're acting like the evaluations or "industry consensus" is how all teams value any given player, and that's simply not true.  What was the "industry consensus" on Didi Gregorius when Cashman traded for him?  Or Aaron Hicks?  Or Luke Voit?

31 minutes ago, Demetri said:

Did you see the Newcomb come-backer? (https://www.mlb.com/video/comebacker-hits-newcomb-in-head) That....is scary.

I did not.  That is indeed scary.  And "OUCH" and sorry about the story with your friend's dad.  God damn.

[Demetri]

1 minute ago, DMC said:

Sure, but my point was he's rock solid in barrels.  Which coupled with those hard hit/exit velo percentages suggests when he does get a hold of one, it's hit particularly well.  Which explains his inflated ISO.  And conforms to the eye test that he's got bat speed like Sheffield.  

No, it's simply the profile for a generic power hitter (that's why that K% isn't going anywhere) who doesn't make much contact in general, but sometimes really gets a hold of one. But so what? There are better stats unless you're trying to prove he can hit HRs. I concede he can, that's what generic power hitters do my definition, and not much else. Sure he barrels balls a lot, but he also has a lot of weak contact. It is illogical to suggest his barrel % suggests amazing bat speed but disregard that his bat speed still doesn't save him from overall weak contact. Sure, you can focus on literally 25 balls he's hit (while claiming SSS on other stats?), or we can use the more holistic xwOBA or the more commonly occurring hard hit %, which is below average entirely because he makes so much weak contact. 

But let's say he does have legendary bat speed. His plate discipline stats are HORRID. He whiffs at an above average amount of pitches in the zone, and makes well below average contact for pitches out of the zone. This means he isn't extending ABs and isn't getting a lot of quality contact. It further reinforces that his BB% will remain low. 

He still has a below average xwOBA, meaning that even with all his barreling (which is included in the stat) that once luck settles his xwOBA of .323 is below the average of .324 for major league hitters. His barrel ability and power merely make him an average hitter, propping up poor plate discipline and bad contact. The barrel % is incredibly limited (Again...25 data points...) and doesn't indicate a renaissance. In fact, it indicates that he's probably a finished product. Plate discipline doesn't really change too much, nor does poor contact (barring a swing overhaul), so the fact that power and ability to barrel keep his head over water means that he has developed the skill that all generic power hitters use to get by: ISO. 

It would be foolish to not bend to an inclusive and more holistically defined stat like xwOBA because a tiny subset of that data shows promise. I'll spin it and say that he's learned to do the one thing he does well and can actually do it (power and barrel). But it balances the stuff he does poorly, it doesn't outweigh it. 

24 minutes ago, DMC said:

Gotta admit, I find it a bit amusing that your argument is so invested in these metrics, but the fact is Frazier still has a very small sample in terms of major league experience.  24 year olds often improve - and even often considerably improve - as they progress.  The empirics bare that out as well.  Projections are inherently conservative, even Szymborski will tell you that.

So you're citing literally 25 balls hit as barrel rather than hard hit % which has more data points and xwOBA which is more holistic (and simply a better metric) and then citing sample size?

It isn't so much that I'm "invested in the metrics" so much as they explain what Frazier is doing perfectly. Sample sizes for advanced batting metrics stabilize VERY quickly. It might not be the best sample size, but it is no longer insufficient. It also just so happens to reflect past performance. It would be an incredible coincidence if two flawed data sets spit out essentially the same data. 

Szymborski (who is a super nice guy who pops by the Braves blog almost everytime his name is mentioned- we joke there is some sort of Szymborski symbol) would certainly say that projections are conservative, but they are also pretty damn accurate. He'd also tell you that the sample size for Frazier is sufficient as offensive stats stabilize pretty quickly. 200 PAs is generally sufficient to start saying that the information you're looking at has statistical validity offensively, with more data being better. Combining years, the stats are pretty consistent if you aggregate the nearly 400 PAs. If there were wildly different things among the two major data sets (2017 and 2019) then perhaps we should just look at 2019 (which we could, as it is over 200 PAs and thus has some good value) and ignore the other set. Because they describe the same thing, we can aggregate them. We're now at just under 400 PAs, and they tell about the same story, but with ridiculous luck in 2019. But more importantly, Szymborski himself points out that sample size requirements vary by the stat being looked at. Most of what I've looked at is in the lower end of PA requirements: (https://blogs.fangraphs.com/when-samples-become-reliable/) (By Dan himself: https://library.fangraphs.com/principles/sample-size/)

So, yes, you're right that we don't have a great sample size. We have an alright sample size. But more importantly, we have data! It has a value beyond 0, obviously, and is starting to have predictive value. The projections you call inherently conservative actually say his BB% will increase (while everything else regresses).

And where in the world are you getting the "inherently conservative" notion and how does that affect accuracy? As in that they don't generally predict crazy outliers? That strengthens, not diminishes, the accuracy of projections. What you call "inherently conservative" is simply placing projections closer to league average. That's good statistics unless you have a crystal ball that can tell you which ones are conservative and positive for the player and which are conservative and underestimate the player. But look at Acuna. He was projected at around 125 wRC+, which people thought was crazy low given the phenom that is Acuna. He's currently at 129.

The projections are generally accurate. The onus is on you to explain why you think the projections are wrong, because the projections are finely crafted statistical models that are balanced and fed new information (remember, until two posts ago, you were uncertain about what data these projections included). They are in the business of being accurate, and are quite good at it. 

More importantly, how wrong do you think these projections are exactly? And why do you think he's going to retain this level or improve it when stats, projections and luck-defining metrics all suggest they're going down? I need a more substantial critique either on the systems as a whole, or for you to tell me why Frazier's is just wrong. Very rarely are they just completely wrong. I'm not saying projections are infallible religious texts, but they ALL think that Frazier is playing over his head. All of them, independently, saw the same things I saw. Mostly, it is the xwOBA and the BABIP which suggests that he's merely average.

That's costing him in the projections and that is frankly how BABIP works. We need like 300 more PAs before it stabilizes, but I'd bet everything I own that it does not stay above .340 (it doesn't for anyone, really, that's top 20 in all of baseball and Frazier has no skills suggesting he is an exception.) When that BABIP falls, so will Frazier

1 hour ago, DMC said:

So, you get to this later in the post, but it's hard to carry on this discussion because it seems to be you're making self-contradictory arguments.  You said you agreed Frazier could be the centerpiece of a trade for Bauer in the last post (and yes, certainly for MadBum, who I agree isn't worth much.  I wouldn't even want them to give up Frazier for him, was just making a point in those prior links I cited).  But you continually argue that front offices view him as a backup outfielder.  It can't be both.  Nobody is trading Trevor Bauer for a package that has an expected 4th outfielder as the key piece.

So Frazier is currently a starting OFer for the Yankees? They didn't expressly trade for a player and demote a guy who is having a good offensive year for an aging player to fill a starting OF role? The Yankees JUST TREATED HIM like a 4th OFer. You're speculating, but we KNOW the Yankees went out of their way to replace him. Is that how starters are treated? The Yankees demoted the guy after a purposeful trade, I'd say it is contradictory for you to suggest that he isn't a 4th OFer.

Hell, if you simply extrapolate his current numbers, he is, by definition, a sub player (War generally considers 2+ as starter and Frazier would have to keep up his offensive pace and improve his defense dramatically to even get to 2.) His WAR and his treatment by the Yankees PROVE he is not viewed as a starter.

They also really aren't contradictory if you understand the valuation stuff, which I've provided several links for. His 4.5 years of team control give him an estimated surplus value of $20.4 million. That is somewhere between 2-3 wins above what they're paying him. He is still pre-arbitration, that isn't going to be hard to do. 

1 hour ago, DMC said:

This is all nonsense.  I am very confident that different teams evaluate Florial in very different ways - just as different public scouting sources evaluate and rank him in drastically different ways.  You can't put a number on it and assume every team is going to follow the same valuation.  That's not how the real world works.  And the facts are he's just getting back from an injury, and hasn't even stood out thus far at High A, which he's repeating.  You are not going to convince me that that type of prospect has equal or more value than Frazier at this juncture, full stop.

BTW, this goes for all your estimates on the FV and how that translates to salary worth.  First of all, this is not the only consideration when teams trade players, even prospects.  Second of all, you're acting like the evaluations or "industry consensus" is how all teams value any given player, and that's simply not true.  What was the "industry consensus" on Didi Gregorius when Cashman traded for him?  Or Aaron Hicks?  Or Luke Voit?

It certainly isn't nonsense. First of all, the calculator is ba

sed on past trades (including gregorious). Am I saying that value is properly captured in individual trades? Of course not. But does that mean we don't have any data to get an idea how the industry works and that there aren't general rules? Absolutely not. There is a ton of scholarship in this area. Beyond the calculator I already provided.

Once again, this data litters the ground. I concede that there FOs differ, obviously, but you're denying that there is any consensus and that is simply false. Here are just SOME of the work that suggests all the same stuff: 

https://blogs.fangraphs.com/2018-trade-value-1-to-10/ (this deals with MLers more, if you read only one, this is a good one)

https://www.baseballtradevalues.com/valuing-minor-leaguers/

https://blogs.fangraphs.com/an-update-to-prospect-valuation/ (this is very good for prospect valuation)

http://www.thepointofpittsburgh.com/mlb-prospect-surplus-values-2018-updated-edition/ (this has methodology explained in more detail on its original platform, I'm happy to provide)

https://www.drivelinebaseball.com/2019/02/prospect-valuation-much-top-prospects-worth-professional-baseball-teams/

(I could go on and on, again, there is a lot of scholarship here. We are no longer in the days where analytics departments are rare or unknown)

The Yankees, like many teams, have created their own models https://www.beyondtheboxscore.com/2019/4/4/18294438/mlb-yankees-data-analytics-brian-cashman-steinbrenner-chapman-betances-voit-ottavino-world-series

So, yea, FOs differ. The argument isn't that every FO views data the same way, but we do know that they endeavor to value players and that there are finite numerical inputs in the world. Logic follows that the Yankees didn't invite something that isn't already represented (at least in part) by any of the others. Beyond prospects, we already have a system that values players: that is what arbitration is aimed to do (though in an adversarial, all-or-nothing process). We know the value of WAR from free agent figures. This information isn't simply made up, it is derived from and compared against historical events.

But we're talking about how the industry as a whole does things, not what you can trick one GM to give you. Again, if you can trick a FO (like the Braves did with Stewart) then fine and dandy. We know overpays happen, we know some folks value players differently, but there is a general consensus. There is a lot of historical data on top 100 prospects being traded. The same data used for these figures (all shockingly similar....) also suggests that a 50 FV position prospect is worth more than 50 FV pitching prospect. We've seen that play out and also is conventional knowledge.

But here is a guy who knows a thing or two who completely disagrees that there aren't common approaches (Scott Boras):

"How do we know teams are using numbers like ours?

We don’t for sure, but we have enough evidence to suggest they’re using a similar model. Even super agent Scott Boras acknowledged it, in an April 2019 story in The Athletic by Ken Rosenthal (by way of criticism of it):

To hear Boras tell it, the problem is not his negotiating style, but the way that clubs use analytics to value players, often landing at similar dollar amounts in their appraisals.

“These markets are very different because we have got a dynamic where the valuation component is common to all teams by design,” Boras says."

Boras is basically saying that the FA market situation comes from a market "where the valuation component is common to all teams by design." In fact, many agents claimed collusion (which isn't impossible) because of the similar approaches.

Look, what I'm doing is providing an objective baseline. You are fundamentally misunderstanding what I mean by "general industry consensus" if your rebuttal is "well, that doesn't apply to every player and every team." I never said that. I'm saying that this is a useful baseline. Whether you like it or disagree with it does not detract from its validity. Furthermore, the world you suggest (where every team is different and mysterious) then we can't say Frazier should/could be the centerpiece for any trade. We'd have to ask each team "Hey, what's a Clint Frazier" as if stats didn't exist. They do. The Yankees, by action, have indicated that they don't think Frazier is a starting OFer. They probably based this on his luck-driven results, just like I am. That isn't even to mention the bargaining position power. And of course things like bargaining position, urgency, window for contention etc are all factors and all very impactful. I never once denied it.

But Boras seems to think that the finite data is being used in similar ways by teams (hence free agent results). Rather than assertions, I've provided independently consistent models (how independently is debatable at point of origin, but they are not directly affiliated), pragmatic evidence in the form of FA consistency, and a quote by an expert in the field of valuing players (Boras) as well as pointing to a MLB mechanism for player valuation (arbitration). That's a lot of evidence for nonsense!

Of course, the real world isn't simply plugging in a figure. Who in the world said we plug in an exact figure? First, the valuations aren't simply saying "Prospect of X FV is worth Y". They build in bust rates, position v. pitcher, bands of performance that we see from similarly ranked/rated prospects, bust rate, star rate, median WAR etc etc. This data isn't simply made up, it is based in empirical data and leads to the resulting valuation.

One thing that helps a prospect's value is how much younger he is than competition. You cite Florial as repeating A+ (most recently, he spent 75 games at A+ in 2018 and 11 thus far in 2019) and yet he is still 1.5 years YOUNGER than the average competition. Saying he's "repeating" A+ is not totally true as 75 games in 2018, but furthermore, it misses that he remains younger than the competition which is far more important than playing the level again (being -2.4 years from average in 2018 and -1.5 in 2019). Injury certainly creates uncertainty, but does it move his value to 0 or does it simply shift it around what history suggests? Does playing older competition compensate?  The injury question is a fair unknown, but again, I was providing a baseline stat not a definitive value. But the general valuation I provided SOMEWHAT built in injury questions, as it accounts for his current place on lists that also accounts for injury.

Furthermore, why would they promote a player coming back from injury? Many medical experts suggest that an entire year of playing is necessary for power to fully recover from hamate injury. Given that injury and the second wrist injury that was uncovered after the hamate injury (pardons if the timeline there is a bit off) what's the concern? He is young, has carrying tools, and is coming back from injury. There is a lot of promise there. As said previously, you can value ability over tools. Teams do often divide on such lines, but he has plenty of loud potential tools and I still think arm and speed (and, to a lesser extent, defense) are all carrying tools for him. Those kind of tools even out the wide gaps in expected WAR. 

In terms of FV. It is remarkable how many prospect evaluations come out, and how generally similar they are. I won't argue that scouts might differ. But to suggest that FV ratings have no value is to say that FV ratings are entirely meaningless. And yet, we hear constantly that teams are seeking "a top 100 prospect" for a certain guy or "need a top 50 prospect and a top 100." I could produce countless articles on that. When negotiating a trade, we KNOW FOs reference top prospect lists. 

Do you think that they all compare their independent top 100 lists? What happens when a GM asks for a top 100 guy, only to find out that the other team thinks that the guy is really a top 300 guy? That just doesn't really happy (though, it does happen a bit more in pitching prospects.) We know that such industry lists are relevant because we hear such lists cited EVERY offseason and EVERY trade deadline. It is also remarkable how similar lists are. A guy can probably be fairly considered top 100 if a well thought off prospect service has a guy on a top 100 list.

You're right, that the individual tiers of lists (representing jump from 50 to 55 FV within top 100, for instance) might and do change. But that is not the same thing as suggesting that there is no usefulness to that data at all. We also know that there is movement between scout evaluating professionals and front offices (Kiley McDaniel and Jeff Sullivan are notable examples.) This all suggests that FOs, to an extent, use this data. I also tend to think they are more likely to differ on players in their system versus players from another system. 

But no one is saying it is a perfectly crafted and absolutely predictive number, but history suggests that it represents the approximate value of a guy with a general consensus FV of a certain number. It isn't perfect or directly translates, but it is present and it is increasing. The fact that the data is derived from historical data further gives it application to MLB (whether by informing FO personnel or by expressing why and what happened in a generally correct manner.)

The idea that each team has to take every trade conversation by first having to prove that at top 50 guy isn't actually a guy more worthy of a top 200 spot is ludicrous. The evaluators and the FOs are rational actors and are using the same, limited, published data. FOs are even taking it further by starting to install measurement systems in JUCO and minors. So, yes, they have proprietary data and more access, but the data they're gathering is bringing them in step with what the industry was already, at least in part, doing. 

I think the Florial issue is agree to disagree. I think that you'd really enjoy some of the resources out there, but prospects are a personal pursuit. I personally love it, but arguing about prospects doesn't take us far. Now nearly 25 year old MLers with major question marks defensively and concerns offensively, I think we can meet openly and have a legitimate discussion on that.

[DMC]

Wow dude, that might be the longest post I've ever seen in the general forums.  Um, I'll get back to you.

[Ramsay B.]

What a wild convo. 

So Reds sweep Astros, and Yankees sweep Rays. This has me worried for the upcoming 4 game set somehow. I’ll be at the game Sunday. I was really hoping it would be against anyone but stupid Verlander. 

Scherzer is an absolute stud, and Trout is ok too.

 

[Arch-MaesterPhilip]

1 hour ago, Ramsay B. said:

What a wild convo. 

So Reds sweep Astros, and Yankees sweep Rays. This has me worried for the upcoming 4 game set somehow. I’ll be at the game Sunday. I was really hoping it would be against anyone but stupid Verlander. 

Scherzer is an absolute stud, and Trout is ok too.

 

I hope the Astros have a hangover from their series with the Reds but I'm not counting on it. My job is raffling off tickets for Sunday, I hope I get them.

[DMC]

6 hours ago, Ramsay B. said:

So Reds sweep Astros, and Yankees sweep Rays. This has me worried for the upcoming 4 game set somehow.

Agreed.  Seems set up to be a trap series.

[Jaime L]

Culturally, Montreal and Tampa Bay have always felt like sister cities so I'm glad they can finally share a team that neither team will care all that much about. 

[Arch-MaesterPhilip]

This is the end of the Rays. We've seen this movie before.

[Myshkin]

Eat it Bumgarner!

[Myshkin]

13 minutes ago, Triskele said:

What is this Garlick?

Yup, a big ol hunk of raw Garlick. 

[Myshkin]

31 minutes ago, Triskele said:

These Gigantes aren't done yet despite even more Garlick, Goblin King.  What say you?  

ETA:  And lo, the Goblin King sent forth his minion whom he called Joc, and halted the Giant's charge.  

Nothing can resist the power of JocJamz 2019!

[Myshkin]

Well that got a little hairy.

[DMC]

Gary Sanchez has 23 HRs in 56 games.  That's a 66.5 HR pace in 162 games.

[Teng Ai Hui]

I feel honored to be a Cardinals fan again today after seeing the heart-fell ovation given to Albert Pujols upon his return to Busch Stadium.  They even jeered their own team at one point in favor of this baseball legend.

[DMC]

44 minutes ago, Triskele said:

And Yellich has like 28.   

Yeah 29 in 69 games now, which is good for a 68 HR pace for 162 games.  But I'm not a Brewers fan.

[DMC]

Start spreading the news...

8 in a row.  I know the Stros are without Correa and Springer, and they missed Cole, but that was really nice.  They'll probably lose with Happ v Verlander tomorrow, but I think they can get at him.  They did a pretty solid job of it in April.  Might not matter in terms of the pitching, but hopefully Happ isn't one of their four starters come October. 

This was a very encouraging week in terms of how they can handle the playoffs.  Only thing left to worry about it the Twins, I guess, and I'm pretty sure the Yanks beating the Twins in the playoffs is one of the Ten Commandments.

[dbunting]

Well since the Tigers suck ass, at least those of us in Michigan can root on, well, Michigan in CWS forst time in over 50 years. Have to have some baseball to root for this summer since the Tigers are just depressing.

[DireWolfSpirit]

2 hours ago, dbunting said:

Well since the Tigers suck ass, at least those of us in Michigan can root on, well, Michigan in CWS forst time in over 50 years. Have to have some baseball to root for this summer since the Tigers are just depressing.

I know it's just awful for me. I am reduced to enjoying Verlander outings since he's at least an ex Tiger.

[dbunting]

1 hour ago, DireWolfSpirit said:

I know it's just awful for me. I am reduced to enjoying Verlander outings since he's at least an ex Tiger.

Sad what we have been reduced to in such a short time span. At one time we had 3 cy young winners on staff, triple crown winner and to MVP's. Couple bad signings here and there along with letting the wrong players go to FA and it's all over.

[DireWolfSpirit]

When Mr Illitch passed, we all lost. Best owner ever RIP.