moonslav59

I'd love to see stats like that: L-R splits Late & Close Late & Close in the playoffs or key games in Sept. Stuff like that.

I explained it already, and once again, you misunderstood my position. And you wonder why I need 80% more words.

Yeah. That's probably the right answer.

I showed Papi's "Late & Close" numbers and agree, they should be used, but Papi's career Late & Close numbers are 60 points below his career OPS, meaning if you separated the two sample sizes (Late & Close vs non Late & Close) the differential would be more than 60 points. High leverage shows Papi in a better light- slightly better than non high leverage. I don't think any of these numbers prove anything one way or the other, except that maybe as the sample size gets large enough to approach or reach validity, they seem to get closer to the norm (as statistical probabilities would project in a random generator).

The Guardians just signed Boone Logan. I thought there pen was already very strong and deep!

100% correctomundo. Other than Hanley's excellent half season in 2013, his 2016 season was the first time he exceeded his career norm OPS of .861 since 2009! On the otherhand, EE has not been below his career norm OPS of .850 since 2011!

Or, one person could look like he's clutch, when it was really the other guy choking. I have no issues with people saying, "that was a clutch hit." or "Man the Patriots really choked in that game!" (LOL). I do think there are mental and emotional issues involved with high pressure situations, and how players perform under these circumstance can certainly be influenced by these factors. My issue has just been about assigning definitive labels to players based on small sample sizes that are usually associated with playoff numbers. Then, even if the sample size is large enough, it is hard to prove conclusively that the results are based on a human factor or just randomness. Papi seems like the best example to use, since we all are well ware of his heroics. To me, his playoff sample size is way too small to make any definitive judgement, but he was certainly on his way to showing (or maybe "proving" to some posters) he was "clutch". Let's look at his larger sample size of regular season "Late & Close" numbers. One would think that if Papi was really "clutch", he'd have equal or better numbers than his career high in Late & Close situations( Plate Appearances in the 7th or later with the batting team tied, ahead by one, or the tying run at least on deck). Here is the data: Career OPS: .931 Career Late & Close: .870 in 1451 PAs, which to me is at least very close to a valid sample size number. One can argue he was facing better pitching, so we can expect a player not to do better than the norm and maybe still be called "clutch". I get that point, but he's 60 points down. If I wasn't a Sox fan, I might use this as evidence to say his playoff numbers are just random numbers, and that nobody can "prove" he's "clutch". BTW, he's .942 in High Leverage (1665 PAs), which could be used to prove he "is clutch" after all. All this data is enough to make your head spin, as some seem to contradict the others, but all in all, I'd say Papi's numbers in high pressure situations during the regular season look pretty supportive of his playoff numbers (.947 in 369 PAs). If any ML baseball player can ever be definitively called "clutch", it would have to be Papi. I just think it's hard to prove when looking at it from a statistical probability standpoint and sample sizes that are not large enough to be valid. I know one thing, I was always glad he was on our team, especially when it counted. (Same with Manny.)

Just follow the conversation- that's the context. When I try to keep it simple, you misunderstand me. When I go on and on trying to explain it, you say, "blah, blah, blah". If you don't know my position on sample sizes by now, you never will.

HRam has 155 PAs as a DH. That's a small sample size, but so far there is no indication being a DH has a negative impact on his offense: DH: 1.014 SS: .882 1B: .845 3B .762 LF: .690 Note: DH is the smallest sample size on this list. (LF and 3B are between 350 and 375 PAs.)

It's hard to make a case for playing at the ML level when you're batting .664 in AAA with 2 HRs in almost 400 ABs. Maybe he's just given up hope.

I had really high hopes for Rusney. I'm not criticizing Sox management & scouting, but I don't think he got a very long look at the ML level. I'm not arguing he should get a chance now, since his contract would put us right at the luxury limit, if we were to have him on the 40 man roster all season.

Castillo is hitting .392 in the winter league, but he has no HRs in 51 ABs (.882 OPS).

I get caught up in a conversation and sometimes leave out statements that define the context. I can understand someone reading "Clutch does not exist" in isolation and think I mean I do not even think the term is applicable in any situation. I also assume everyone has been following the whole conversation, sometimes spreading over more than one thread, so I assume people have read and remembered that my position is that it is okay to use the terms "clutch" and "choke" to describe events, but not to definitively define a player or team as such, perhaps even with large enough sample sizes. I have been consistently arguing (perhaps more than anyone else) the position that using small sample sizes to define someone is unfair. I assumed everyone knew that about me, since I've been a bit of a broken record on the subject for years. I feel my idea of a significant sample size is probably larger or much larger than just about everyone here.

Winter League Updates: http://news.soxprospects.com/2017/02/fallwinter-roundup-castillo-leads.html Christian Vazquez went 2 for 4 with a double and a walk before being replaced in the ninth inning. Rusney Castillo helped Caguas grab the victory by going 2 for 6. Castillo will be the lone Red Sox player to partake in the Caribbean Series by virtue of Caguas' victory. Fernando Abad pitched in four of the five games last week for Aguilas, facing one batter over the minimum in 3 2/3 innings and striking out five batters, with the lone batter to reach doing so on an error. Hanley Ramirez and Marco Hernandez [are not] listed on the club's roster for the [upcoming] Caribbean Series.

Not much else to talk about, I guess. We've beaten every other Sox issue to death ten times over.

My position is clear. The term "clutch" is generally used to describe someone or a team coming through in a key moment to do what it takes to win a close game or come from behind. The term "choke" is generally used to describe someone or a team that blows a big lead or comes up short "when it counts" and loses. I'm fine with using those words in that context. This debate has been about labeling a player a "choke" or "clutch". Here is where I disagree. I'm not sure why you have such a hard time understanding a simple position. I realize when I said "there is no clutch" in the context of the conversation on labeling a player one or the other, I probably should have added that to my statement, so those that have a hard time understanding context could follow along. I have always been very consistent on not going along with making definitive judgments based on small sample sizes, and my idea of "small" is much bigger than many posters. To put it simply: I'm okay with people saying an act was "clutch" or "a choke". I'm not okay with labeling a player or a team "clutch" or "choke" based on sample sizes that are too small to be valid. My extended point is this: If "clutch" or "choke" sample sizes were large enough, let's say 700 or 800 PAs or more, it might still be hard to determine if a player is a choke or a clutch based on the fact that if you set a random generator based on common baseball outputs and set the PAs at 800 to create a 100,000 sample sizes, you would find some sample sizes seemed overly high or low from the norm. Let's say 1,000 sample sizes were 25% or more above the norm, and 1,000 sample sizes were 25% or more below the norm. Then, you look at the 100,000 "real" MLB sample sizes and find that the amount of players 25% above or below the normal is about the same as the random generator, then a mathematician could say that those group of players performances fell within the statistical frame and could be just random vs any sort of influence from a mental ability to rise above the pressure. Even if you found that the number of baseball players significantly better and/.or worse than a random generator says should be the norm, you'd still have a hard time determining if a single player truly had something special in his make-up that caused the variance, or whether he might just be that statistical subgroup that randomly did better or worse. All baseball players go through hard times and good times. It's entirely possible that a player just happened to have that good or bad time at the precise moment the playoffs started with no emotional or psychological factors influencing that result. It could happen again 3 years later, when that player reaches the playoffs again. It could happen a third time- just as one could flip a coin and get heads 5 times in a row. In my opinion, There's just no way of knowing for sure if a player is "clutch" or "a choke".

Yes, I did not realize I had two windows open, and I responded to the wrong thread.

I was referring to labeling a player "clutch" or "choke". Using the terms "choke" or "clutch" to describe an event is another argument. Has Papi had "clutch hits"? I'm fine with using that terminology. Based on many "clutch" hits and performances scattered among fewer down performances, I would not "play the game" of trying to label him a clutch player or someone else a "choke". Hell, I was arguing with posters claiming JBJ was a bum based on his first 500+ PAs, so I'm not about to make determinitive proclamations based on significantly less sample sizes than that.

No definitive determinations (or "conclusions") can be made on samples of this size, and even if the sample sizes were large enough, a random generation would produce extremely low and high numbers in a few cases, which would make a claim that someone is a "choke" unprovable. You are right. The reverse cannot be "proved" either. It's like the movie War Games, where the best choice is "not to play the game". If you want to call not playing the game a "conclusion" on my part, so be it.

The real problem is that nobody really has a large enough post season sample size to definitively determine a label like "choke" or "clutch". The other problem is that even if you had a million large enough randomly generated sample sizes, there would be a few that looked far away from the norm on both ends of the spectrum. If a random generator produces a few, then finding a few MLB players at the extreme proves nothing about their mental make-up. It could just be they ended up on the extreme end of randomness. Look, I'm not arguing a player's mental make-up has nothing to do with their numbers. I never have. I'm just saying nobody can prove the reason a player's numbers are at an extreme is due to mental make-up and not randomness. The fact that the sample sizes are way too small and scattered over several seasons to even be considered mathematically determinitive (is that a word?), I see it as an exercise in futility. It's fine to point out that certain players came up short or large in their small and important sample sizes, but I cannot see that judging them so definitively as a choke or clutch goes too far.

1) Sample sizes don't "work"; they just are. 2) Anybody with half a sense of statistics knows that sample size matters, and even a large group of rather large random sample sizes will produce a few very low and very high results. 3) My argument has never been about denying there are players that have done very badly "in the clutch". That's not "senseless". The fact that you can find a few horrible samples proves nothing.

Nobody said there were not players with bad numbers in pressure situations, but that proves nothing. A random generator would produce extreme numbers in many of the many small sample sizes one can dig up.

I think DH'ing should help HanRam stay fresher and hopefully healthier. I doubt being DH changes anything to do with his power or OB skills. I know the mentality is different, but HanRam has DH'd before and said he liked it.

Other Sox RP'ers vs LHBs: Sox Pen vs LHBs in 2016 (OPS against): .413 Thornburg (.531 career) with Milwaukee .461 Abad .500 R Scott .514 Kimbrel .545 R Ross .741 Barnes .890 Hembree .899 J Kelly

I'll go one step further and say it is not a science at all. 1) It's nonexistent. 2) Even if it existed, the sample sizes can enver be large enought to determine anything definitively.