Jordan Weems

We know it's not a fluke. When was the last time you saw an organization promote a player who was hitting .166 / .283 / .215? That's insane. I've never heard of anything like that.

Now, we add three pieces of data:

-- He was given a late-second-round bonus and has always been very well-liked by scouts.
-- He was hitting .151 / .258 / .187 (160 PA) on June 3, then hit .250 / .406 / .375 in his next 9 games / 33 PA.
-- They then promoted him, and he's hit .340 / .411 / .440 in 57 PA at the tougher level. In other words, consistent with his last 9 games at Greenville.

Well, the Red Sox are not psychic, otherwise they would have done a lot of things differently.

I'm hoping one of the SP scouting crew can report on what he started doing differently. Because if this were a mental adjustment, you don't promote him after 9 games, especially so if it had to do with self-confidence. It had to be some kind of mechanical alteration that worked for him dramatically and was so obvious that he seemed to be a different hitter -- the kid they gave $500K to instead of the one that had struggled to make any kind of hard contact since the day he signed. That kind of change you can be confident about continuing to work, so they promoted him, and they've been right.

I had more to say about this in one of the game threads -- I'll try to dig that up.

Weems, of course, is one of three or four failed ex-prospects who have had breakthroughs -- the others being (as jrusso mentioned) Derrik Gibson, Carson Blair, and possibly Williams Jerez as a pitcher. Toss in Chris Hernandez doubling his K rate and it's been a good season for guys flying under the radar.

Eric, for posterity, is this the "10% chance" thing, or something else?

We haven't been able to get eyes on Weems, but I was always of the assumption that he was promoted as sort of a "if it isn't happening now, it's never happening" thing. They've done it in the past. Jason Place's 2009 promotion to Portland while he was putting up a .248/.326/.384 in Salem with 85 K in 341 PA comes to mind.

Personally, I'm literally ignoring Weems until he does something next season.

I for one am happy for the young man. He played behjnd swihart's enormous shadow a few years back and I think that messed him up a bit.

Now as to why he is hitting at Salem and was not in the south Atlantic league is beyond my ability to grasp at this point.

But as long as he keeps hitting I am happy.

This is precisely backwards thinking. The larger and more consistent the previous baseline, the less likely it is that something dramatically different is just a fluke.

Your logic is that the large and consistent baseline has completely established who he is. It didn't: it completely established who he has been, up until recently. By interpreting the large, consistent baseline as evidence of who he still must be, you are assuming that he couldn't have somehow changed, which is in fact the question that's being asked.

No one is assuming that he couldn't have somehow changed, just that there is no evidence that he has changed.  You're accusing people of begging the question, but what you yourself are doing is shifting the burden of proof, with no justification.  

The real problem is that people don't understand statistics very well. That's why they make dumb statements that 54 ABs must be a fluke because it's a small percentage of the total outcomes. They especially don't understand the difference between random, non-random, and predictive variables.

Very briefly, the percent chance that an event was caused by a randomly distributed variable or set of variables is the sum of squares difference of between the outcomes in the sample and the expected mean or (x - x(bar)) and the number of outcomes in the sample N. If (x - x(bar)) is large enough you might have an event that has a very small chance of being caused mainly by random variables.

Now just because a variable is a non-random variable doesn't mean that it's a predictive variable. An example of this would be facing really really awful pitchers for 18 games or so. So it would be nice to know if the change in performance is associated with something like a change in a batting stance. If it is, that doesn't necessarily mean that Weems is a top prospect, but it would mean that he's more worth watching than before.

Hilarious. You hate when others speak to you with a condescending tone, but do you ever read your own posts?

Okay, I think this has been said numerous times, and I'm not always great with my descriptions, but lets try this again. These types of analyses are terribly flawed. OF COURSE, if you go into a dataset consisting of a long string of at-bats, and you locate and select (POST HOC!!!) the hottest streak and compare it to a preceding sample of lesser plate appearances, then these two samples will look statistically significantly different from each other. Why are we here making much of p-values (greatly influenced by sample size) anyways? Do you understand how ridiculous that is? Find me any player, take a full season of at bats, locate their best 100 ab sample and compare it to their worst. You can't reasonably conduct that type of post hoc analysis. You're driving up the Type I error and increasing your chance of locating a false positive. For someone who is all high and mighty regarding their supposed statistics knowledge, this is such a fundamental and obvious error.

Since his promotion he is hitting .338/.411/.415 with 8 infield singles. 36 GB, 9 FB, 4 LD

Infield singles? Let's trade him for Jake Peavy.

The idea all along could have been to prove to him that he couldn't hit. In fact, I'd bet good money that was the idea - the whole "perhaps he'll hit better if he doesn't need to worry about catching" thing didn't make sense for him given his career numbers. If he's going to make this transition, he needs to buy in completely, so showing him that, yes, catching had nothing to do with why you can't hit, could've been the plan.

By the way, let's remember this the next time we have a guy to on a "statistically meaningful" hot streak.