Twins make offer for Chris Archer

[diehardtwinsfan]

 

But their biggest need is a front of the rotation starter. (Darvish or Archer.) A number 4, (Cobb/Lynn), or 5 (Tillman/Garcia), doesn't really move the needle much.

OK... sorry.. but Cobb/Lynn aren't 4s... they were both better than average, even if their peripherals weren't ideal. 

[Jham]

 

Is your home run rate adjusting for league-wide tendencies?

 

Because MLB total home runs rose just shy of 50% (!!!!!) from 2014-2017.

 

It's based on comparison of FIP (actual HR rate) to xFIP (expected HR rate by league average).  I think that's how xFIP is calculated anyway... I'd assume league average is based on that season, but don't know for sure.  Could be historical average.

That's the same argument I use against Sano when comparing his whole skill set.  The HR are nice, but 30 or 35 isn't what it used to be, or it is what it used to be, depending on which era it we're comparing to.

[yarnivek1972]

I like Kepler too, but I guess I am not understanding the angst over trading a guy who is the third best overall outfielder on the team. Trading Kepler would likely mean that Grossman starts in RF (or possibly LF with Rosario switching to RF), Vargas is the primary DH and Granite is a late inning defensive sub a lot. I am not at all convinced that Gordon or Romero will pan out. Romero has substantial injury history and seems destined to be in the bullpen IMO. It’s doubtful Gordon will stick at SS and not at all certain he will hit well enough to stick anywhere. If trading three questionable upside players is what it takes to get a pitcher like Archer, I’d do it everyday and twice on Saturday.

[Mr. Brooks]

OK... sorry.. but Cobb/Lynn aren't 4s... they were both better than average, even if their peripherals weren't ideal.

4.16 and 4.82 FIP, respectively.

[Mr. Brooks]

I like Kepler too, but I guess I am not understanding the angst over trading a guy who is the third best overall outfielder on the team. Trading Kepler would likely mean that Grossman starts in RF (or possibly LF with Rosario switching to RF), Vargas is the primary DH and Granite is a late inning defensive sub a lot. I am not at all convinced that Gordon or Romero will pan out. Romero has substantial injury history and seems destined to be in the bullpen IMO. It’s doubtful Gordon will stick at SS and not at all certain he will hit well enough to stick anywhere. If trading three questionable upside players is what it takes to get a pitcher like Archer, I’d do it everyday and twice on Saturday.

If that would be enough for Archer (I don't think it's close), I'd hope the deal would already be done.

I don't think you even get a return phone call for any package that doesn't include Lewis.

[old nurse]

 

ftfy

Disagree if you want, but I did not say he was not a quality pitcher. If you believe fip as the end all, be all you will over value Archer.  There are other measures that do not like Archer.

[AlwaysinModeration]

Part of it is the initial reaction. I think people were good with the contract when it was signed. It worked out poorly but we needed pitching and it was a market contract.

 

Hughes’ extension was a head scratcher from the get go

Hughes made a brilliant move by not asking to go back out for another inning in his last start (which would have triggered a incentive bonus). TR was so impressed by his selflessness he gave him an extra $48 million.

[Brock Beauchamp]

What I find striking is that he consistently underperforms his FIP... Nolasco was the same way, though not to the extreme. I think there's something to be said for something other than luck here (dont' get me wrong, I like Archer, as he's not 'unlucky' enough to avoid acquiring)... but bottom line is that pitchers have more influence in BABIP than stat heads are willing to accept. There are too many anomalies.

I don’t find a 0.18 difference very compelling.

 

But I think you’re right that some pitchers seem to pitch outside the parameters recorded by FIP, I’m just not sure Archer is one of them.

[diehardtwinsfan]

 

Saying he underperforms his FIP assumes ERA is his actual performance.  Basically, it goes back to saying ERA is the best stat to evaluate performance.

 

I'm not really saying that at all. I'm saying that there's more skill involved in a pitchers results than stats like FIP seem to recognize. If ERA was as random as is often implied here, there would be no such things as pitchers that could consistently under or over perform their FIP. I'm not saying ERA is the be-all/end-all stat for evaluating pitchers. I'm simply saying that I don't buy into the idea that it is largely luck driven. Usually it's an adjustment that a pitcher makes which gets their results in line with their peripherals, not just more reps.

[Oxtung]

 

I'm not really saying that at all. I'm saying that there's more skill involved in a pitchers results than stats like FIP seem to recognize. If ERA was as random as is often implied here, there would be no such things as pitchers that could consistently under or over perform their FIP. 

 

Yes there would be. Statistics actually demands there will be. Ever flipped a coin 100 times? At some point it is going to land same side up a bunch of times in a row.

 

 

I'm not saying ERA is the be-all/end-all stat for evaluating pitchers. I'm simply saying that I don't buy into the idea that it is largely luck driven. Usually it's an adjustment that a pitcher makes which gets their results in line with their peripherals, not just more reps.

 

Nobody is saying ERA is primarily luck driven but we are saying it takes into account things a pitcher can control and things they can't control. Among the list of things they can't control is the park they play in, is a 4.00 ERA from Coors field the same as a 4.00 ERA at Petco Park? When a flyball is hit to Buxton is that the same as Denard Span? Defense matters. In addition to the onfield differences there will also be statistical variation; sometimes a ball drops and sometimes it doesn't. If Sano gets a pitch over the heart of the plate he can tattoo a ball but if it's hit right at the 3rd baseman does that mean the pitcher did something amazing or did he get lucky? Some years that's going to happen more than others and that will affect a pitchers ERA even though there has been no discernible skill change.

 

So, is ERA entirely luck driven? No. Are there components a pitcher can't control? Yes. Is one of those things statistical variation? Yes. Will there be pitchers whose statistical variation consistently overperform their true talent? Yes. Statistics demands it.

[FormerMinnasotan]

I don't think so. A team receiving him would be getting half a year less of control, and deadline deals don't always fetch a premium, David Price being another Tampa example. I'm sure there are more, but I'm in public transportation and don't feel equipped to do the research.

It may not cost us potentially as much for major league talent, but we may have to give up more higher up prospects if we trade for Acher at the trade deadline. Consider this: right now we may have to give up Kepler, Romero (who may never pan out as a starter), and Nick Gordon. If we trade for Archer at the trade deadline we may have to give up the likes of Lewis Thorpe, Blayne Enlow, etc. That’s what I mean when I say we’d have to give up more at the trade deadline. We’d have to give up higher ceiling guys than we do now.

[Brock Beauchamp]

 

Yes there would be. Statistics actually demands there will be. Ever flipped a coin 100 times? At some point it is going to land same side up a bunch of times in a row.

 

 

Nobody is saying ERA is primarily luck driven but we are saying it takes into account things a pitcher can control and things they can't control. Among the list of things they can't control is the park they play in, is a 4.00 ERA from Coors field the same as a 4.00 ERA at Petco Park? When a flyball is hit to Buxton is that the same as Denard Span? Defense matters. In addition to the onfield differences there will also be statistical variation; sometimes a ball drops and sometimes it doesn't. If Sano gets a pitch over the heart of the plate he can tattoo a ball but if it's hit right at the 3rd baseman does that mean the pitcher did something amazing or did he get lucky? Some years that's going to happen more than others and that will affect a pitchers ERA even though there has been no discernible skill change.

 

So, is ERA entirely luck driven? No. Are there components a pitcher can't control? Yes. Is one of those things statistical variation? Yes. Will there be pitchers whose statistical variation consistently overperform their true talent? Yes. Statistics demands it.

The problem here is that diehard is talking career FIP/ERA (or at least several consecutive seasons) while you're talking about FIP on a yearly basis.

 

There's no explanation for Ricky Nolasco's career FIP discrepancy that doesn't involve the statement "FIP is missing something here". Nolasco has played for three teams in two leagues in front of a variety of defenses, both good and bad.

 

Yet he has underperformed his FIP in ten seasons and overperformed it just twice and never by more than .18 runs (well within a reasonable margin of error).

 

His career FIP/ERA difference is nearly six-tenths of a run.

 

FIP is a very good stat. It's not a perfect stat and doesn't seem to accurately judge the talent level of a few pitchers. I've never seen a good explanation why that is the case.

[Oxtung]

 

The problem here is that diehard is talking career FIP/ERA (or at least several consecutive seasons) while you're talking about FIP on a yearly basis.

 

There's no explanation for Ricky Nolasco's career FIP discrepancy that doesn't involve the statement "FIP is missing something here". Nolasco has played for three teams in two leagues in front of a variety of defenses, both good and bad.

 

Yet he has underperformed his FIP in ten seasons and overperformed it just twice and never by more than .18 runs (well within a reasonable margin of error).

 

His career FIP/ERA difference is nearly six-tenths of a run.

 

FIP is a very good stat. It's not a perfect stat and doesn't seem to accurately judge the talent level of a few pitchers. I've never seen a good explanation why that is the case.

 

Absolutely I was talking about career stats. Statistics says there will be people who under/over perform consistently. That's why tails will come up 9 times in a row when you flip a coin 100 times. That is the definition of a bell curve. If there weren't outliers then I would be worried. 

 

Is FIP perfect? No. Does a player's ERA outperforming their FIP for their career mean FIP is broken? No. Could there be something FIP isn't catching? Sure, but there better be more proof for it than "Random player X is out performing his FIP, it must be broken".

[USAFChief]

Absolutely I was talking about career stats. Statistics says there will be people who under/over perform consistently. That's why tails will come up 9 times in a row when you flip a coin 100 times. That is the definition of a bell curve. If there weren't outliers then I would be worried. 

 

Is FIP perfect? No. Does a player's ERA outperforming their FIP for their career mean FIP is broken? No. Could there be something FIP isn't catching? Sure, but there better be more proof for it than "Random player X is out performing his FIP, it must be broken".

I think the onus is on the FIP proponents to supply more proof than “ignore the counter evidence, FIP isn’t broken.”

[ashbury]

I hope we don't trade for Archer. We'd have to revisit ERA-versus-FIP in every thread about him.

[bcs4]

Questions:

 

1: Is the FIP shown in the comparison above for Archer and another pitcher (sorry, I forget who) a calculation for the following year, or based on current year numbers for that year?  Same question for ERA.

 

2: Is the differential between the 2 calculated by adding the total of the ERA and FIP and then dividing by the number of years shown?

 

Any help would sure be appreciated.  Thanks

 

 

[ashbury]

I think the onus is on the FIP proponents to supply more proof than “ignore the counter evidence, FIP isn’t broken.”

I can accept that FIP correlates better to next-season ERA than ERA itself does.

 

It still leaves me cold. It's not "fielding-independent" as named, it removes fielding by taking out much of what makes baseball baseball. It's kind of the "Garfield Minus Garfield" of baseball stats.

 

[Jham]

I can accept that FIP correlates better to next-season ERA than ERA itself does.

 

It still leaves me cold. It's not "fielding-independent" as named, it removes fielding plus a lot of what makes baseball baseball. It's kind of the "Garfield Minus Garfield" of baseball stats.

 

It also has a failed premise of trying to predict era. The whole point of fip Siera etc is to get away from era which has too many variable to be a great measure of performance. Why not try to predict itself? The wide variation in fip and xFIP from year to year tells us it's not perfect. But who really cares? Pitchers also change from year to year. There is no perfect measure or perfect prediction. Archer is a top quality pitcher by many measures but not results. We hypothesize why. I believe it's mostly competition. But whatever. He's the exact type of player a team like Tampa is trying to acquire, not trade. They'd have to get a prospect projecting to be better than archer and several pieces that are ready now. We don't even have the first part.

[Brock Beauchamp]

 

Absolutely I was talking about career stats. Statistics says there will be people who under/over perform consistently. That's why tails will come up 9 times in a row when you flip a coin 100 times. That is the definition of a bell curve. If there weren't outliers then I would be worried. 

 

Is FIP perfect? No. Does a player's ERA outperforming their FIP for their career mean FIP is broken? No. Could there be something FIP isn't catching? Sure, but there better be more proof for it than "Random player X is out performing his FIP, it must be broken".

You're confusing your analogies. The coin coming up heads nine times in a row is the equivalent of a single season. It's a statistical aberration but not unreasonably so (1 in 512 chance).

 

The coin coming up heads 70 times out of 100 is the equivalent of a career (1 in 43,000-something chance). There's a huge statistical difference there. The more data points, the more the odds of unusual performance plummet.

 

When you apply that to a real world person who plays for multiple teams, in multiple leagues, in multiple divisions, over the course of a 12 year career, and it happens almost every year, the odds are so low they're not worth mentioning.

[Brock Beauchamp]

 

I hope we don't trade for Archer. We'd have to revisit ERA-versus-FIP in every thread about him.

We shouldn't because FIP actually works for Archer. He had one kinda weird FIP/ERA year but his overall career numbers are close enough to make it irrelevant.

[howieramone2]

 

Hughes is a split grade... A- for the original signing and F- for the extension.
Nolasco was probably a C

I agree with the original signing grade, but would give a B for the extension. Most extensions come after a good year, which it did, and 3 years for 39M was reasonable for both sides. Injuries are part of the game, and I refuse to downgrade a player, for what is part of the game. Also, I'm not big on hindsight.

[Brock Beauchamp]

 

It also has a failed premise of trying to predict era. The whole point of fip Siera etc is to get away from era which has too many variable to be a great measure of performance. Why not try to predict itself?

I view it differently. FIP is a good indication - not perfect, but good - of yearly performance and a fairly good predictor of short-term future performance.

 

What convinces me that FIP is a pretty solid stat is that it does match up with ERA the vast majority of the time given a long enough sample (say, 3-4 years or longer). ERA is the one that fluctuates but, after a time, generally settles somewhere close to actual performance level if you look at a veteran's entire career. What FIP does is tell us more exactly what to expect in the near future and what happened in a single season.

 

But any examination of FIP needs to include the fact that some pitchers, however the hell they do it, "break" FIP. Those pitchers are rare and don't break the statistic, but it means that FIP is missing something that a very select numbers of pitchers can or cannot do.

[howieramone2]

 

Sorry for the nitpick, but Nolasco a C?! We ended up paying him something around $42M for 3 seasons (they sent $6M to LAA, I think?) and got 321 innings of 5.44 ERA with 1.474 WHIP and 11 hits per 9. Baseball Reference gave him a -0.4 WAR for those 3 years. 

 

If that's a C, I want to know what an F is.

Nolasco was marginal his first year. He was 5-1 in the second after a month, then missed the rest of the year, until a late start. He was marginal in the 3rd, before Rob Antony earned his place in Twins history, by dumping him on the Angels at the trade deadline. I don't downgrade players, due to injuries, I agree with the C.

[USAFChief]

Nolasco was marginal his first year. He was 5-1 in the second after a month, then missed the rest of the year, until a late start. He was marginal in the 3rd, before Rob Antony earned his place in Twins history, by dumping him on the Angels at the trade deadline. I don't downgrade players, due to injuries, I agree with the C.

Seems odd to give a passing grade for a signing that also earns high praise for a dumping. In the same post.

[howieramone2]

 

I like Kepler too, but I guess I am not understanding the angst over trading a guy who is the third best overall outfielder on the team. Trading Kepler would likely mean that Grossman starts in RF (or possibly LF with Rosario switching to RF), Vargas is the primary DH and Granite is a late inning defensive sub a lot. I am not at all convinced that Gordon or Romero will pan out. Romero has substantial injury history and seems destined to be in the bullpen IMO. It’s doubtful Gordon will stick at SS and not at all certain he will hit well enough to stick anywhere. If trading three questionable upside players is what it takes to get a pitcher like Archer, I’d do it everyday and twice on Saturday.

I've been all for opening up the farm system for trades for the past 2 years, until I heard Kepler mentioned. I view him as a borderline All-Star sometime in the future. I'm all for dealing fat, but it's too early to be dealing muscle.

[laloesch]

 

But their biggest need is a front of the rotation starter. (Darvish or Archer.) A number 4, (Cobb/Lynn), or 5 (Tillman/Garcia), doesn't really move the needle much.

 

I have to disagree on this.  Signing Cobb or Lynn makes this team quite a bit better when you consider what we currently have for starters.  No question about it in my mind and both those starters are better than #4s.

[GP830]

I say sign both on prove it deals and try and acquire a more true number 1 through trade. If Lynn and Cobb are good midseason, you can either extend them or trade them for value, and you'll at least have an idea as to where Gonsalves, Romero, and Thorpe are.

 

Solved it.

[Oxtung]

 

You're confusing your analogies. The coin coming up heads nine times in a row is the equivalent of a single season. It's a statistical aberration but not unreasonably so (1 in 512 chance).

 

The coin coming up heads 70 times out of 100 is the equivalent of a career (1 in 43,000-something chance). There's a huge statistical difference there. The more data points, the more the odds of unusual performance plummet.

 

When you apply that to a real world person who plays for multiple teams, in multiple leagues, in multiple divisions, over the course of a 12 year career, and it happens almost every year, the odds are so low they're not worth mentioning.

 

First let me say I'm happy to agree that FIP isn't perfect.  I actually prefer SIERA, personally (not that it's perfect either). I'd be incredibly interested to see how teams are blending the new fielder tracking data and statcast launch data into their new models. I think that could really help nail some things down. Unfortunately that isn't available to us.

 

That said, what is your basis for the 70/100 comment? You seem to be basing your argument on some mathematical idea but it isn't coming through. 

 

Every large data sample will have outliers. Sometimes that is random variation. Sometimes that is a difference in "skill". If you want me to believe there is a "skill" involved then you need more proof than "some people break FIP" because that is expected in every data sample. What people break it? What is their skill? How are you going to show that with math?

[Brock Beauchamp]

 

First let me say I'm happy to agree that FIP isn't perfect.  I actually prefer SIERA, personally (not that it's perfect either). I'd be incredibly interested to see how teams are blending the new fielder tracking data and statcast launch data into their new models. I think that could really help nail some things down. Unfortunately that isn't available to us.

 

That said, what is your basis for the 70/100 comment? You seem to be basing your argument on some mathematical idea but it isn't coming through. 

 

Every large data sample will have outliers. Sometimes that is random variation. Sometimes that is a difference in "skill". If you want me to believe there is a "skill" involved then you need more proof than "some people break FIP" because that is expected in every data sample. What people break it? What is their skill? How are you going to show that with math?

Where did your nine heads in a row comparison come from? Fine, change 70 out of 100 to 60 out of 100 and the chances of it happening is still much lower than flipping nine out of nine.

 

Random variance that led to a career FIP/ERA discrepancy would likely be a couple of years that influenced the overall number, much like your "flipping nine heads in a row" comparison. A brief period of extreme luck in either direction that influences the overall number. Because if you extend past a single season, you change some assortment of defense, catching, quality of opponent, and perhaps team/league, not to mention the luck/unluck component needs to stay identical to what it was the previous season. That's asking a lot.

 

But we don't see that with Nolasco. What we see is a consistent underperformance of FIP in nine out of 11 seasons (and the overperformance seasons are marginal in comparison, never more than 0.18 runs), with the underperformance gap being 0.30 runs or more in a whopping eight out of those nine seasons (again, across three teams, three divisions, and two leagues).

 

At that point, you need to either admit FIP is missing something (mostly likely quality of contact, IMO) or that Ricky Nolasco is one of the unluckiest men on planet earth.

 

Which one makes more sense?

[Otto von Ballpark]

 

I hope we don't trade for Archer. We'd have to revisit ERA-versus-FIP in every thread about him.

They could switch to FIP on the Target Field scoreboard too!

 

Or, they could switch between FIP and ERA, whichever is most flattering to the Twins pitcher.