Wednesday, February 25, 2009

FireHaroldReynolds

It doesn't have quite the same ring to it as FireJoeMorgan, but I'm confident that with his current position on MLB Network, Reynolds will do an even more consistent (Morganism!) job of providing us with sheer, unparalleled hilarity and frustration. Here's an excerpt from a video Grif posted in one of his recent posts.

"There aint no way Albert (Pujols) is getting 133 RBI's with that club. It aint happenin!"

Ah, Harold Reynolds. So distinguished. So well-informed. So lucky to still be having a job on television. I'd like us to take a closer look at this specific part about Pujols and the mention of RBI. I must warn you though, if you share the same way of thinking as that of Reynolds and Larkin, you may want to stop reading now. Because what I'm about to do involves statistics, none of which come from a crumbled piece of paper provided to me by means of an old man in a straw hat.

2008 St. Louis Cardinals
.350 OBP (2nd in NL)
#1 Batter: .345 OBP
#2 Batter: .371 OBP


Albert Pujols bats third, which means the amount of RBI he gets is largely dependent on whether or not the first and second batters get on base. The last time Albert Pujols had at least 133 RBI was in 2006. Let's see what that team looked like.

2006 St. Louis Cardinals
.337 OBP (5th in NL)
#1 Batter: .335 OBP
#2 Batter: .338 OBP


So what Harold Reynolds is saying is that there is no way that Pujols can get 133 RBI with the 2009 club, even though their roster is darn near exactly the same as it was last year, and even though he was able to do it with a team that was inferior in 2006. There are two things to be learned from this:

1) RBI are dumb, unpredictable, and depend on too many outside factors.
2) Harold Reynolds is an imbecile who should stick to what he knows best: teaching kids how to throw a ball accurately to first base from 20 yards away.

5 comments:

Grif said...

This is a fantastic, over-looked facet of Reynolds's idiocy in that segment.

I would LOVE this, except, I can't trust you if you didn't get this information from someone with a stogie in his mouth.

Nemo said...

I'm not sure if this criticism is warranted. Commentators have a HISTORY of making inaccurate predictions and the like. I mean, have you EVER listened to Bobby Murcer or Phil Rizzuto. Rest their wonderful souls - and I LOVED listening to them, they were a great team, and great individually, but wow did they say some awfully stupid things.

Sure part of your fire against Reynolds is for his sexual harrassment charges at ESPN, but..c'mon. Let's move on, shall we?

Rook said...

It probably wouldn't even be worth mentioning if Reynolds was simply making predictions about various players just for the sake of making predictions. I'm aware that plenty of people make stupid predictions all the time. That's fine. What's different here is that he is specifically outing an established, reliable projection system (PECOTA) and HE'S the one doing the criticism in the first place, which is completely unwarranted mind you. If you're going to tear down an entire statistical system and the people who created it, you better have your facts straight.

Also, here at ATH we're not big fans of "moving on" or "letting things go." If that were the case, we would have nothing to write about.

Skye said...

I'm the only female contributor to this blog and I hold no ire toward Reynolds for his alleged indiscretions involving female co-workers.

His statements against PECOTA, however, are unforgiveable.

Goose said...

I guess Harold can relax a little. The latest update has Pujols's forecast for RBI at 122. Again, that's speaking about RBI and involves much more volatility than context-neutral metrics.

I mean, with any projection from any system, there will be lines where you go, "Hmm, that might be a bit optimistic/pessimistic," but that's expected. It's another thing, as Rook mentioned, to take one of those and to totally throw out the validity of it as a result, particularly when look at a metric with relatively high variance even given similar context-neutral measurements.