Reds interested in Castellanos

[NeilHamburger]

Plus, in all likelihood this means Winker is getting dealt somewhere. And I think Senzel (especially if he is in RF), Shogo and Aquino all have the chance to be above average defenders.

And the potential lineups look good even before a SS:

Shogo CF
Votto 1B
Suarez 3B
Castellanos LF
Moustakas 2nd
Senzel RF
Galvis SS
Barnhart C

And if you somehow improve either SS or C in the above then you're freaking stacked.

You could also flip Senzel and Shogo if one is better suited for CF than the other. And you still have Aquino as depth or a potential defensive replacement.

[kaldaniels]

Soooooo...if the Reds do land him are we thinking 1 yr or multiple years?

[kaldaniels]

He OPS'd just over 1 (as in 1.000) after his move to Chicago and that was over a total of 51 games.

I scrolled thru all his career game logs and don't see a 50 game stretch of such a good performance. (could be wrong)

So to those smarter than me, what gives? Is it statistical noise that his sample size of 51 games in Chicago were better than any other 51 game stretch in his career? Awful coincidental.

Cause my simple brain just defaults to "he just had to get out of Detroit!". And my smart brain says "nah, it isn't just that."

[RedsManRick]

I’ve never been a believer in the “stack up the WAR” philosophy of team building. Castellanos can hit, and they desperately need hitting, they can likely hide his defense in left field and be fine with it.

When it comes to scoring more runs than your opponents, there's nothing magical about scoring a few more vs. preventing them. They count in equal proportion.

If player A generates 20 more runs than player B with his bat but allows 10 more than with his glove, your team is going to come out about 10 runs ahead with player A compared to player B. That's not philosophy. It's basic math.

Meanwhile, there's no such thing as hiding defense --certainly not in the NL. There are only 8 defensive positions on the field and the balls find each of them at predictable rates. Playing LF doesn't make your bad defense stop counting. "I know that ball scooted by the OF and you're on 2B and drove in a run. But the defender was in LF. So we're not going count that one, OK?" Is that how it works?

The fact that it is the position on the field where teams tend to play their less talented defenders only means that the pool of players you compare him to is a bit different. We still see a difference of about 20 runs between the best and worst LFs each year. If he's worse defensively than the alternatives, that difference still dings his ability to help your team. The extra runs that score as a result will still count, will take the shine off the extra offense he's providing.

Now, will his glove give back more runs than his bat creates relative to the other guys the Reds may run out there? Probably not. But burying your head in the sand when it comes to defensive production because of some platitude about behind able to "hide" a defender doesn't help you build a better team. You should at least attempt to account for it.

Ultimately, the balance of run scoring and run prevention is almost irrelevant. A team that scores 700 runs and allows 650 generally wins the same number of games as the team that scores 800 and allows 750. So if you go out there and start throwing money around while only looking at one side of the equation, that's a recipe for spending your resources quite inefficiently.

The argument for Castellanos is not his defense won't matter. It's that he's actually going to be much better than a 2 WAR player and/or that the Reds will trade from their OF depth to shore up another genuine weakness.

[InsaneinthBrame]

I am not sold on him being a fit especially for the money. His career high is 27 hrs. Also lot of his numbers are based on high babip and high line drive rate. Last year his line drive rate was the lowest of his career and gabp will suppress babip.

With his sub par defense and lower expected babip his overall value may not be significantly better than what we have.

I guess if it allows us to make trade to improve elsewhere it would make sense.

[RedsManRick]

He OPS'd just over 1 (as in 1.000) after his move to Chicago and that was over a total of 51 games.

I scrolled thru all his career game logs and don't see a 50 game stretch of such a good performance. (could be wrong)

So to those smarter than me, what gives? Is it statistical noise that his sample size of 51 games in Chicago were better than any other 51 game stretch in his career? Awful coincidental.

Cause my simple brain just defaults to "he just had to get out of Detroit!". And my smart brain says "nah, it isn't just that."

This is one of those cases where the idea of regression to the mean is helpful. Whenever we see an extreme swing like that, it's almost always a combination of both something "real" happening combined with "lucky" statistical noise. The question is what proportion of each.

Recent performance is the generally more predictive than older performance. But more performance is more predictive than less. Combine those notions and you get something like the the Marcel projection system -- a quite simple system that basically just looks at the last 3 years of data, weights each year back a bit less, and takes an average. The backbone of the projection is basically giving a weight of 5 to most recent year, 4 to the prior, and 3 to the one before that.

But what if want to account for a mid-season trade? Well, he played 2/3 with the Tigers and 1/3 with the Cubs. But we can give the Cubs time more credit, so let's just split that 2019 performance 50/50. So for Castellanos:

2017 Tigers: .811 OPS (3 weight points)
2018 Tigers: .854 OPS (4 weight points)
2019 Tigers: .790 OPS (2.5 weight points)
2019 Cubs: 1.002 OPS (2.5 weight points)

Now we just multiply, add, and divide: (.811 * 3) + (.854 * 4) + (.790 * 2.5) + (1.002 * 2.5) / (3+4+5) = .876 OPS

That's compared to a straight average of his last 3 years of .842 or a weighted average that doesn't break-out Chicago separate of .846.

So a system like Marcel would say his Chicago performance would have you bump your estimate of his talent by about 30 points of OPS. Much less than the 150-200 the straight comparison suggests. So Marcel sees a regression to the mean of about 80% from his Chicago performance.

If you want to better account for park effects (Detroit is a rough place to hit), you might use wOBA or wRC+ instead. The story is somewhat similar. Straight and weight 3-year average of ~121 wRC+. Breakout Chicago and that goes up to 125. Meanwhile, his wRC+ in Chicago was 154. So it bumps the estimate up ~5 points compared to the ~30 points observed during his Chicago time.

All that is to say that the sample size is small enough that it doesn't move the projection needle all that much. If you want to make the case that we should expect something resembling his Chicago performance, it's a case founded in scouting or scouting-like narrative -- not the math.

[kaldaniels]

This is one of those cases where the idea of regression to the mean is helpful. Whenever we see an extreme swing like that, it's almost always a combination of both something "real" happening combined with "lucky" statistical noise. The question is what proportion of each.

Recent performance is the generally more predictive than older performance. But more performance is more predictive than less. Combine those notions and you get something like the the Marcel projection system -- a quite simple system that basically just looks at the last 3 years of data, weights each year back a bit less, and takes an average. The backbone of the projection is basically giving a weight of 5 to most recent year, 4 to the prior, and 3 to the one before that.

But what if want to account for a mid-season trade? Well, he played 2/3 with the Tigers and 1/3 with the Cubs. But we can give the Cubs time more credit, so let's just split that 2019 performance 50/50. So for Castellanos:

2017 Tigers: .811 OPS (3 weight points)
2018 Tigers: .854 OPS (4 weight points)
2019 Tigers: .790 OPS (2.5 weight points)
2019 Cubs: 1.002 OPS (2.5 weight points)

Now we just multiply, add, and divide: (.811 * 3) + (.854 * 4) + (.790 * 2.5) + (1.002 * 2.5) / (3+4+5) = .876 OPS

That's compared to a straight average of his last 3 years of .842 or a weighted average that doesn't break-out Chicago separate of .846.

So a system like Marcel would say his Chicago performance would have you bump your estimate of his talent by about 30 points of OPS. Much less than the 150-200 the straight comparison suggests. So Marcel sees a regression to the mean of about 80% from his Chicago performance.

If you want to better account for park effects (Detroit is a rough place to hit), you might use wOBA or wRC+ instead. The story is somewhat similar. Straight and weight 3-year average of ~121 wRC+. Breakout Chicago and that goes up to 125. Meanwhile, his wRC+ in Chicago was 154. So it bumps the estimate up ~5 points compared to the ~30 points observed during his Chicago time.

All that is to say that the sample size is small enough that it doesn't move the projection needle all that much. If you want to make the case that we should expect something resembling his Chicago performance, it's a case founded in scouting or scouting-like narrative -- not the math.

Good stuff Rick. To me (and I’ve said this before), the last 3 years of a guy’s career pretty much define a player so far as what he is at that point. (That may not be exactly what you are saying here, but I digress). But dang it’s strange the switch flipped once he left Detroit. It deserves a deep scouting dive.

[redsfandan]

One thing that I like about Castellanos is that his power has been improving. But's not his homers. It's his slg% and, mostly, his doubles. One of the arguments against Larry Walker in the other thread is that he wasn't the same player before he started playing for the Rockies. Except both his slg% and doubles took a big jump in his last year BEFORE switching to Colorado. So, the jump in homers shouldn't have been that surprising. Well, it's the same with Castellanos. His slg% has improved (granted, not by a lot, .496 in 2016 to .490 to .500 to .525) but there has been a big increase in the amount of doubles he's had (25 in 2016 to 36 to 46 to 58). His homers increased in Chicago (along with avg, obp, slg). A better lineup, a better home park, and a winning team probably helped. But his doubles had already been increasing. Maybe that's nothing but it's at least notable and maybe an indication that he could be ready to have his first 30 homer season.

[JaxRed]

Soooooo...if the Reds do land him are we thinking 1 yr or multiple years?

I'm assuming it's 3-4 like Moose.

[corkedbat]

One thing that I like about Castellanos is that his power has been improving. But's not his homers. It's his slg% and, mostly, his doubles. One of the arguments against Larry Walker in the other thread is that he wasn't the same player before he started playing for the Rockies. Except both his slg% and doubles took a big jump in his last year BEFORE switching to Colorado. So, the jump in homers shouldn't have been that surprising. Well, it's the same with Castellanos. His slg% has improved (granted, not by a lot, .496 in 2016 to .490 to .500 to .525) but there has been a big increase in the amount of doubles he's had (25 in 2016 to 36 to 46 to 58). His homers increased in Chicago (along with avg, obp, slg). A better lineup, a better home park, and a winning team probably helped. But his doubles had already been increasing. Maybe that's nothing but it's at least notable and maybe an indication that he could be ready to have his first 30 homer season.

If they sign Castellanos, I'd be willing to put up a solid package headlined by Senzel for the right SS (I'd rather deal Winker, but don't see any other club loving that trade out). I'd also like to include some combo of Galvis, Iglesias &/or Barnhart to try and limit adding some of the young talent. Wouldn't mind seeing the Reds bring back a young catcher in the deal as well.

[M2]

I’ve never been a believer in the “stack up the WAR” philosophy of team building. Castellanos can hit, and they desperately need hitting, they can likely hide his defense in left field and be fine with it.

I'm not into stacking WAR either, but you do have to hide bad defense. Put it where it can do the least damage, make sure the rest of your defense is solid so that one goofy glove doesn't overly cost you. Like, you have manage around it.

I'm not sold on Castellanos being the sort of guy who pounds the ball so hard who cares what his defense is. The Reds seems to be hot for him, so go for it. At least they're hot for someone. Yet he's kind of blah vs. RHPs and I wouldn't bank on him repeating his Chicago numbers. A star turn is far from guaranteed. Pat Burrell might be a fair comp for him. Pat the Bat had his uses, but he was probably the 6th most important position player on the Phillies team that won the Series.

[redsfandan]

If they sign Castellanos, I'd be willing to put up a solid package headlined by Senzel for the right SS (I'd rather deal Winker, but don't see any other club loving that trade out). I'd also like to include some combo of Galvis, Iglesias &/or Barnhart to try and limit adding some of the young talent. Wouldn't mind seeing the Reds bring back a young catcher in the deal as well.

I think the right team would have interest in Winker (most likely an AL team like the A's). I wonder if they could deal Winker, Aquino, a few prospects, . . . and get a better rightfielder from one team and another piece/possible upgrade from another team. I know trades within the division are rare but Wilson Contreras has been mentioned as a trade candidate.

[MoneyInTheBank]

Put me in the camp of not being especially optimistic about a guy improving his crappy OF defense in his late 20s. I see a 2ish WAR player being added to a mix full of at other 2ish WAR players at reasonably significant cost. Sure's he got upside if the defense comes around, but you can make similar upside cases for Winker, Senzel, Shogo, and Aquino.

I don't think it's quite that simple. Castellanos has only been playing OF full time for 2 years. Other than Senzel, the other guys have spent their careers in the OF. I'm not saying he's certain to improve his defense or that his defensive stats shouldn't count but there's reason to be more optimistic about an improvement over guys who have been playing OF for years. In fact, statistically, he dramatically improved his defense year over year 2019 vs. 2018

[TeamSelig]

Alright stat junkies... has anyone read any theories on the shift making defense less important? Obviously Suarez is good enough to cover SS in a pinch, but not as a long term solution. But if we run good defensive shifts, wouldn’t that take away from the need for a traditional SS? Obviously you’d still want players with good range to help cover the hitters hot zones, but maybe it isn’t as drastic as traditionally perceived? I haven’t read any theories about this or anything, just a couple thoughts.

[JFLegal]

He OPS'd just over 1 (as in 1.000) after his move to Chicago and that was over a total of 51 games.

I scrolled thru all his career game logs and don't see a 50 game stretch of such a good performance. (could be wrong)

So to those smarter than me, what gives? Is it statistical noise that his sample size of 51 games in Chicago were better than any other 51 game stretch in his career? Awful coincidental.

Cause my simple brain just defaults to "he just had to get out of Detroit!". And my smart brain says "nah, it isn't just that."

it's an interesting question you raise. i'm a big believer in the mental side of sports, so here is my theory: castellanos was finally out of detroit where he had no chance to win and every game was basically a meaningless game. suddenly he's in chicago, in the middle of a pennant race, playing in front of packed houses in one of the most historic ballparks. he felt free. he felt good that the cubs went out and traded for him. his confidence was at an all-time high. playing in a more hitter-friendly park helped, but it's not like wrigley is some bandbox. he was just a better hitter for that 51-game stretch.

and maybe it was just a coincidence. but i tend to believe he just felt different/better due to the situation he was in. he was playing meaningful baseball games for the first time in his major league career. it also helped he was batting in the middle of a really good order. there was no pitching around him.