Peter Keating: O's Rebound a Mirage

[SteveA]

OK, I just got home and haven't read the entire thread, just the first page, so maybe I'm going over some ground again.

I think it's fairly obvious that our starters' performance under Buck can't be sustained because it is so RIDICULOUSLY good. It's as good or better as the Phillies rotation containing Hamels, Holliday, and Oswalt might be expected to do.

When we had around 30 games under Buck I looked at our quality start percentage and it was comparable to Lincecum in Grienke when they won the Cy Youngs last year. In other words, the COMBINED work of our 5 starters was turning in roughly as many quality starts as two CY winners did in their best seasons ever. I don't think anyone realistically thinks that can be sustainable.

But you don't need 5 guys pitching at a Cy Young type pace to have a good team either. The Yankees are going to win the division with Burnett having a crappy year including a month long stretch where he was by far the worst starting pitcher in the majors.

So there will undoubtedly be some regression to the mean with the performance of our starters. Of course there will also be some improvement to counteract that as Arrietta, Bergeson, Matusz, and Tillman are all on the upward moving side of the age curve. They should be better next year just by virtue of being a year older, more experienced, and still on the upslope where players tend to improve with age. That could negate some of the regression.

But there will be regression. And that's why, if we want to play .600 ball next year like we have since August 2, we have to make some improvements in our hitting. Given that we had a black hole offensively at SS, and one of the weakest 1B positions at the plate in baseball, as well as weakness at 3B, and probably only one hitter who is on the downward side of the age curve that had a career year (Scott), there's no reason to expect that a competent GM can't improve our hitting for next year.

So yeah, a little regression from our starters is to be expected, but if we have a good offseason we should have a better offense next year. Then you add luck with injuries which every team needs (the Red Sox lost this entire season to injuries for example), and I think this team COULD contend next year.

[SteveA]

What the stats crowd usually calls "luck" they really mean is randomness.

As in, while it is entirely possible to flip a coin and get heads 10 times in a row, the true probability is 50% ... so over a large enough sample, the amount of heads will grow closer and closer to 50%, rather than remaining at 100%.

When stat-heads say that a player is "lucky" or "unlucky" because he has a currently unsustainable BABIP it simply means that there is a current streak of abnormal numbers. Just as Drungo always says that it's possible for a a true 85 win team to win 90+ games, it's also possible for a true 4.50 ERA pitcher to have a "lucky" season and end up with a 3.50 ERA. It's just not likely he'll repeat that performance, as the results will tend to regress toward his actual talent.

I don't understand why everyone is so up in arms about this article.

I think one reason some people are viscerally upset with this article is that a guy like this chooses now, when we are having the most fun we have had in a dozen years, to go out of his way to piss in our cornflakes, as it were.

Early this season, we were having some tremendously bad luck. We were underperforming most projections such as Pecota by a ridiculous amount. We were doing things that were nearly unprecedented since the pitching mound was lowered 41 years ago, such as going an entire month without scoring more than 2 runs in an inning, or going 44 games without a 3-run homer, or hitting so much worse with runners in scoring position than even the poor overall numbers we were putting up. All of these things were ridiculously amazing statisitcal outliers, every bit as much as our starter ERA since Buck took over.

But when that happens, were the statheads coming to the Orioles' defense and writing articles about the ridiculosly, nearly impossible, bad luck we were having, and that we were really a much better team? If they were, I don't recall a lot of such articles being linked here.

But now that the pendulum has swung, and we are overperforming in some areas in ways that are unsustainable, this guy decides to come along and write an article that seems to want to "put us back in our place". And while the text of the article (at least as quoted here) seems fair and balanced, the headline "O's Rebound a Mirage" makes a much stronger statement.

It's the old situation where you can insult your own wife like Henny Youngman, or make jokes that apply to your own racial or ethnic background, but when an outsider tries to insult your wife or your ethnicity, you'll be up in arms. I think that reading htis board every day, there is not some groundswell of opinion that our starters will perform all year next year like they have under Buck, or that we don't need to improve the team in the offseason. A few people have started to broach that possibility but it's not the general opinion.

But when an outsider puts up an article that says in the headline that our entire last 50 days of enjoyement we have had is "a mirage", and then cherry picks every stat possible to back up his point of view (pythagorean theorem, fly ball %, FIP, etc)... there's bound to be an emotional reaction of getting defensive and standing up for our guys. Especially since when the "luck" was going in the other direction they weren't falling all over themselves to write articles about how the Orioles were better than the .280 win% they were putting up.

[SrMeowMeow]

Your last two posts are excellent.

[Steve I]

piss in our cornflakes is a great description. Basically, the article attempts to take statistics to disprove reality. I always find that completely fascinating.

Take reality over an extended stretch, in baseball, 6 weeks or so, and try to frame it as a fluke. Buck's O's stopped being a fluke a few weeks back. Now to argue that it is unsustainable is just a logical statistical dance. ie. meaningless. You can make stats say whatever you want. It is all bias and proves nothing about the future. In baseball, your previous history is a pretty good indicator. Thus, the dude is pissing in our cornflakes.

Have the starters kicked raw butt? Big time. Just check the quality starts metric. Has Milwood? Ah, no. Same old Milwood and yet, he is factored into the starter's record. Will he be back next year? Ah, no. Has the youth blossomed a bit? Yep. Bergie has healed, Brian has matured, Jake has matured. Guthrie has regained Guthrie form. Are those things likely to regress? No. His whole argument is that this a statistical fluke from the norm. He refuses to grant that this is growth and luck and good old fashioned fear for your job. Screw this writer. I expect him to write a big mea culpa next summer when the O's are far improved and contending, of course, he will blame it on whoever we acquire in FA. This guy just wants to piss in our cornflakes. Maybe Buck whipped him in poker?

[biggrthngibbons]

Am I the only one who got the literary reference in the author's name?

[SrMeowMeow]

Am I the only one who got the literary reference in the author's name?

You are not.

[biggrthngibbons]

You are not.

So the question is... Did his parent's read the book? If so, why name him after the most worthless character? Maybe this is a pen name?:scratchchinhmm:

[SrMeowMeow]

As long as you don't mind rationalizing away why Buck can't keep them playing at that level going forward, you're good.

[JayThomas]

What I'm saying is, over the course of 162 games and certainly over the course of a career, his HR/FB ratio will tend to remain static. More importantly, the vast majority of pitchers will find that static point to be somewhere between 9 and 12 percent (and most of them would fall in the 10-11 percent range). That's pretty specific.

RVAbird, where can I find this info? How far back are there statistics on the HR/FB ratio for a given pitcher for a given season? I think the relevant distribution is the distribution of (HR/FB ratios for individual pitching seasons). I suspect that some pitchers will stand out in having consistently better HR/FB ratios than the norm and vice-versa.

[VTmook213]

piss in our cornflakes is a great description. Basically, the article attempts to take statistics to disprove reality. I always find that completely fascinating.

Take reality over an extended stretch, in baseball, 6 weeks or so, and try to frame it as a fluke. Buck's O's stopped being a fluke a few weeks back. Now to argue that it is unsustainable is just a logical statistical dance. ie. meaningless. You can make stats say whatever you want. It is all bias and proves nothing about the future. In baseball, your previous history is a pretty good indicator. Thus, the dude is pissing in our cornflakes.

Have the starters kicked raw butt? Big time. Just check the quality starts metric. Has Milwood? Ah, no. Same old Milwood and yet, he is factored into the starter's record. Will he be back next year? Ah, no. Has the youth blossomed a bit? Yep. Bergie has healed, Brian has matured, Jake has matured. Guthrie has regained Guthrie form. Are those things likely to regress? No. His whole argument is that this a statistical fluke from the norm. He refuses to grant that this is growth and luck and good old fashioned fear for your job. Screw this writer. I expect him to write a big mea culpa next summer when the O's are far improved and contending, of course, he will blame it on whoever we acquire in FA. This guy just wants to piss in our cornflakes. Maybe Buck whipped him in poker?

While I agree with the majority of what you're saying, Millwood has also been excellent since Buck took over. His record might not show it, but hey ... same thing as the beginning of the season.

GS: 9

IP: 58.2

ERA: 3.68

Considering his ERA for the year is 5.29, I think 3.68 represents him kicking butt

And once again, I would like to point out that the author did say that the Orioles are playing like a team reborn.

[Sports Guy]

For all the people crying over this article, I find it funny that in a poll of about 120 OHers, that only 2 think this is a 600 team...That tells me that people do believe we have been lucky on some levels and that this team can't sustain this over 162 games.

So, what exactly is the issue again?

[sakata_catching]

Am I the only one who got the literary reference in the author's name?

Calling Ayn Rand "literary" is a stretch.

[Mizerooskie]

I like statistics, sabermetrics, etc., when they're used in the right context. Team wins and losses is never, IMO, the right context. That's where this article (and the subsequent analysis) falls short.

A team never wins or loses more games than it "should" based on what the statistics tell us. A team wins exactly the number of games it should. There are too many factors external to the events that can be analyzed by the statistics. Team morale, work ethic, confidence, and heart can't be quantified. Those are things that leadership and character bring to a team. A manager can have a significant impact on them.

Much of the statistical analysis we see is based on mean regression. I'm curious, how often do teams actually perform to their expected level based on this analysis?

[Slappy]

I like statistics, sabermetrics, etc., when they're used in the right context. Team wins and losses is never, IMO, the right context. That's where this article (and the subsequent analysis) falls short.

A team never wins or loses more games than it "should" based on what the statistics tell us. A team wins exactly the number of games it should. There are too many factors external to the events that can be analyzed by the statistics. Team morale, work ethic, confidence, and heart can't be quantified. Those are things that leadership and character bring to a team. A manager can have a significant impact on them.

Much of the statistical analysis we see is based on mean regression. I'm curious, how often do teams actually perform to their expected level based on this analysis?

Go back and look at the pre-season projections from Bill James or PECOTA. They are usually fairly close for players with significant big league experience and usually pretty far off for rookies and sophomores. All the science, and you really get no value over what any fairly engaged baseball fan could have come up with.

The scientists never predicted Jose Bautista's year. They didn't predict Ben Zobrist and Jason Bartlett would crater. How many stat crunchers went out on a limb and said the Reds and Padres would compete for their division titles? On the other hand, how many predicted Seattle would be a major threat in the AL West?

I think the deep-analysis types definitely changed the game by recognizing the truth behind once-neglected stats like OBP, and have made some progress with BABIP and BB/K numbers. But I haven't seen a lot of evidence that they are any better at projecting future performance than anyone who sticks to slash lines could do.

[wildcard]

For all the people crying over this article, I find it funny that in a poll of about 120 OHers, that only 2 think this is a 600 team...That tells me that people do believe we have been lucky on some levels and that this team can't sustain this over 162 games.

So, what exactly is the issue again?

That was not the poll. The poll is - can the current team be a .600 win team next year? There has been no poll on whether this is a .600 win team now.

Even I had to vote no, the current team can't sustain a .600 preformance level because the team will have more injuries then they currently have. If injuries are what you mean by luck then I agree with you.