View This Article on Crashburn Alley
Back to Crashburn Alley
YOU DID ASK FOR IT, KENI'm usually reluctant to criticize anything Ken Rosenthal writes
because it's usually well-researched and well-defended, unlike a lot of
what's published in newspapers and magazines to be read by millions
countrywide. Mr. Rosenthal, however, has written an article
defending his selection of the Braves as 2008's World Series selection
waving the red flag at the bull that is the Sabermetric community (not
to imply that said community thinks in lockstep).
He starts off his article waving a raw steak just outside the cage where it can't be reached:
Bloggers, it's your lucky day.
Not that you ever need prompting to rip apart the latest
ill-informed splattering from the mainstream media, but here's an
invitation on a gold-engraved, all-but-autographed platter:
I feel like I really want to punch him*, but he's begging for it so much that I don't want to give him the satisfaction.
* I'm actually a pacifist and likely don't have that great of a punch.
Embarrassing as it is to admit, my annual column predicting
which team will win the World Series often defies sabermetric
orthodoxy, not to mention conventional logic. Sort of like baseball
itself.
The way Rosenthal writes this, it's like he's proud of writing stuff
that defies logic. "I know conventional logic says that if you throw
something up in the air, gravity will bring it back down, but I think
that's balderdash."
Statistical analysis is an invaluable tool; that discussion
is over. But we've gotten to the point where everyone from the casual
fantasy player to the shrewdest GM wants to know the end of the story
before Chapter One is written.
Mercifully, that's not how the game works.
Well, Ken, I don't think anyone with a working knowledge of
Sabermetrics is using them like a crystal ball. Humans, sadly, have
this limitation where they can't see into the future and put all their
money into Bear Stearns.
I often liken traditional statistics and Sabermetrics to different
prescriptions of your eye-enhancement of choice (well, are you a
glasses person or a contacts person?). Traditional statistics like
batting average, RBI, runs scored, won-lost records, saves, etc. all
provide a portion of the picture, but not a clear one. To make an
analogy to the analogy, traditional statistics are a television circa
1985 with the bunny-eared antenna. Sabermetrics provide a clearer
picture, like an HDTV circa 2008. It's not perfect by any means, but
it's the best we have right now and incredibly useful - they provide an
amazingly lifelike picture.
It seems almost as if Kenny is discrediting Sabermetrics for not
predicting the future correctly 100% of the time. That's impossible,
for obvious reasons. But they come close relative to the other options
we have (guessing, rolling dice). After the 2007 regular season ended,
I made an Excel file comparing the results with PECOTA's pre-season
projections and I found that the number of games between PECOTA and
reality was...
- 0 games: 2 teams, 6.7% (both Chicago teams, oddly enough)
- 1-5 games: 17 teams, 56.7% (ARI, ATL, BAL, BOS, CIN, DET, KC, LAD, MIL, NYM, NYY, PHI, SD, STL, TEX, TOR, WAS)
- 6-10 games: 8 teams, 26.7% (CLE, COL, FLA, HOU, LAA, OAK, PIT, SF)
- 11+ games: 3 teams, 10% (MIN, SEA, TB)
You can download the spreadsheet here, if you'd like.
I don't know how well PECOTA fared in previous years, but its
performance in '07 is impressive: it foresaw the dreadful decline of
the White Sox, and the return to Earth of the Tigers, for instance.
So, Ken's point that you can't predict the future is valid, but it's
not valid without crediting how much more accurate the predictions can
be with the use of Sabermetrics.
The 2005 White Sox, '06 Cardinals and '07 Rockies were among
the recent World Series clubs that defied the supposed experts, myself
included. Some other team will do the same this season, reminding us
again that baseball's unpredictability is part of what makes the game
so much fun.
The paradoxy of saying that baseball is unpredictable and then
predicting that a team will defy predictions aside... saying that a team
will defy predictions to discredit those predictions doesn't mean much.
It's like using a fortune teller to place all your bets for a week of
NFL games, and you get the first 14 games right, and rake in a ton of
money. As you sit and watch the Monday Night game, your fortune teller
errs and the 49ers somehow beat the Patriots. Despite the fact that the
teller has selected 93% of the games correctly* you decide to dwell on
the one mistake and throw the baby out with the bath water.
* Obviously, that scenario is entirely facetious. Do not use fortune tellers to help you in your NFL get rich quick scheme.
Bloggers, man your keyboards!
My Spidey Sense is tingling, and I sense derisiveness from Mr. Rosenthal.
My choice to win it all is the Braves.
That's absolutely fine. I await to see how you back it up with facts.
As the accompanying sidebar suggests, I've been largely unsuccessful with my pre-season selections over the years.
An ad hominem on yourself? Unprecedented!
But then, who hasn't?
PECOTA and other Sabermetric-aided predictions.
The proper time to write a predictions column is actually
Aug. 1 or even Sept. 1, after teams adjust their rosters through trades.
There's no "proper time" to make predictions. A prediction is
saying, "Based on the information available, I think that [insert
premonition]."
I think what Kenny was trying to get at is that your predictions can
be more accurate if you wait a long time to see how things unfold.
Thanks.
But such a late analysis would be a copout, and even then, there would be a decent chance of looking like an idiot.
Amateur psychoanalysis here, but it seems like Rosenthal is preoccupied with "looking like an idiot."
In my NCAA bracket, I had Duke getting to the Elite Eight. I'm such
an idiot for thinking that. But other than that, all four of my Final
Four teams were alive up until Wisconsin lost to Davidson a few minutes
ago. If you're making a lot of predictions, you're going to end up
getting some of them wrong, and you're going to end up looking like an
idiot on some picks. Bob Knight picked Pittsburgh to win it all, and
they lost in the second round. He looks like an idiot but it doesn't
discredit him from ever coaching again or making more predictions.
Grow a pair, Ken, make some predictions and tell us your reasoning
behind it. At least if you get it wrong, you can feel good about
getting it wrong. Why do I feel like my guidance counselor?
Few imagined last Sept. 1 that the Rockies would make the playoffs and the Mets would not.
Because people lack access to a time portal.
Anyway, here are my general rules for a preseason forecast,
knowing that Eliot Spitzer stands a greater chance of being president
in 2012 than I do of nailing one of these suckers outright:
An Eliot Spitzer joke. Ken is topical!
And he's self-deprecating. Me likey.
Never pick the Red Sox.
Never pick the Yankees.
Why? Because they're good teams? Because they have high payrolls?
Why would you not pick these teams to succeed? I mean, if you are
scared about looking like an idiot, it seems like you'd want to go with
the obvious picks.
Never pick a National League team unless under the influence of imagination-enhancing drugs.
Why? This isn't the NBA - the National League isn't the Eastern
Conference and the American League isn't the Western Conference. The
best in the NL can compete with the best in the AL.
The Red Sox, winners of two of the last four World Series,
probably are the best team on paper. But picking them is like picking
the smartest kid in class to finish with the highest SAT.
It's highly likely that your pick will end up correct, making you look like a genius instead of an idiot?
Besides, the only way for a team to win back-to-back Series
is to keep its pitching intact through three postseason rounds for two
straight years. Hard to do.
It's not the only way; it's a way, albeit a highly good way.
According to this logic, last year's Red Sox could swap Beckett,
Schilling, Matsuzaka, et. al. with Johan Santana, Jake Peavy, Brandon
Webb, etc. and not increase its chances of winning it all, since their
pitching staff is noticeably different.
The Yankees, who have not won the Series since 2000, almost could qualify as a surprise team at this point - almost.
It's just my own subjective observations, but it seems like a lot in
the media are picking the Yankees to make the playoffs. Personally, I
have them missing out because I think they are depending too heavily on
unproven arms, but I would not be surprised if they won the AL East.
They have an offense that will match the heavily-lauded Tigers.
They are a surprise team in that they're not currently better than
the Red Sox or either of the Indians and Tigers, leaving them second in
the Wild Card chase at best.
But now that they're trying to incorporate younger, less
expensive players, the Yankees are like the rich kid in the
neighborhood who tries to act cool by dressing down. Sorry, the rich
kid is still a rich kid - and with dubious pitching, I might add.
These are two sentences that are ripe for amateur psychoanalysis, but I'll restrain myself for now.
The analogy falls apart because the Yankees aren't using young
pitchers to fit in with the crowd; they're doing so out of necessity. A
better analogy would be the rich kid having all this stuff because his
parents own Bear Stearns and then having to find clothing at Goodwill
because of, well, you know.
Actually, the NL has produced three of the past seven World
Series champions - the '01 Diamondbacks, the '03 Marlins and the '06
Cardinals.Frankly, I'm sensing another NL breakthrough [...]
That's it! Write it down! Ken's feelin' it, and he's feelin' an NL
team winning it all! Dump your Bear Stearns stock and put it in KenRo
Inc.
[...] and not simply because two of the best pitchers in the
AL, Johan Santana and Dan Haren, were traded to NL clubs. None of the
AL contenders looks as dominant as the '07 Red Sox; I can't quantify
it, but the disparity between the top teams in each league might not be
as great in years past.
The '08 Red Sox don't look as dominant? I guess if you think losing
Curt Schilling for a half-season (potentially more) is damning. It's a
loss, no doubt, but he's 41 and not anywhere near as dominant as he
used to be. Call me crazy, but I think this year's rotation of
Beckett/Matsuzaka/Lester/Buchholz/Wakefield will be nearly as good as
last year's Beckett/Schilling/Matsuzaka/Wakefield/Tavarez-Lester.
Some (not I) would argue that this year's Tigers look dominant with
the addition of Miguel Cabrera. Some (not I) would also argue that this
year's Mariners look dominant with the acquisition of Erik Bedard.
The Indians haven't changed much and C.C. Sabathia is in a contract year.
The Braves have constructed an AL-type offense.
They have a DH? They are refusing to bunt with their pitchers?
Their bullpen will get a boost if lefty Mike Gonzalez returns from elbow-ligament transplant surgery at mid-season.
That's great, but what are they going to do in the meantime?
Their rotation features enough options to absorb ineffectiveness and/or injury [...]
John Smoltz will start the season on the disabled list and is
nearing age 41. Tom Glavine is 42 and his '07 season was about as bad
as his '03 season (his first with the Mets). Mike Hampton hasn't
pitched in two years and is 35.
Really, the only sure thing is Tim Hudson.
I'm not saying the Braves will again trade for this year's
Mark Teixeira, but they should be able to get the piece or pieces they
need.
How do you know what they'll need? So far, you've said that they
won't really need any starting pitching ("enough options") or bullpen
arms ("boost from Mike Gonzalez"), and the Braves are set at catcher,
first base, second base, third base, center field, and right field. So,
barring catastrophic injuries, the Braves would be trading for a
shortstop or left fielder. Otherwise, they're not really trading for
anyone of consequence.
Yet, the Braves aren't the only legitimate NL threat.
Really? Who'da thunk it?
The Cubs could be a World Series team if they add Brian Roberts.
They won't:
Baltimore president of baseball operations Andy
MacPhail told reporters in Ft. Lauderdale Wednesday that a Brian
Roberts deal with the Cubs is off the table.
"We worked at it this long and we don't have deal," MacPhail
said. "There's other sides characterizing it as an impasse. You make
the judgment."
The Phillies and Brewers will be very good if their run prevention reasonably complements their run production.
Translation: The Phillies and Brewers will be very good if they score more runs than their opponents.
These observations are reaching John Madden levels.
The Diamondbacks' young position players should improve offensively, and the Dodgers are just too talented to ignore.
"Guys, who are you all picking to win the NL West?"
"D-backs."
"Rockies."
"D-Backs."
"D-Backs."
"Rockies."
"Agh! I can't take it anymore! The Dodgers! The Dodgers!" (Falls on floor, crying) "They're too talented!"
Also under consideration: The Mets, who must contend with
age and injury concerns, and the defending champion Rockies, whose
rotation is a bit of a wild card.
The Rockies' rotation was a wild card last season and they went to the World Series.
Mostly healthy last season, the Sox already are without Curt
Schilling and could start the season without Josh Beckett. Daisuke
Matsuzaka's '07 load - he averaged more pitches per start than any
major-league pitcher - might be another warning sign.
Beckett is shooting for April 6. Unless you think the post-season hopes of the Red Sox will be made or broken by one game, this isn't really a huge issue.
Will Hideki Okajima be as dominant a reliever this season?
He sure looked dominant last season.
Will Manny Delcarmen emerge as a legitimate late-inning weapon?
44 IP, 1.023 WHIP, 232 ERA+, 41 K, 17 BB in '07. Looks good to me.
All of the projections besides CHONE have him finishing the season with
a sub-4.00 ERA and all of them have him pitching 50+ innings.
It's also difficult to imagine their top three relievers -
Joe Borowski, Rafael Betancourt and Rafael Perez - being as good again.
"Ken, we think Borowski, Betancourt, and Perez are going to be good
in '08, but before we publish it, we wanted to check with you. Can you
see them being good?"
Ken: (Closes his eyes, grits his teeth, and tries to imagine what
'08 will look like for those three) I see... Abraham Nunez hitting 20 HR,
and unicorns, and Adam Eaton winning the Cy Young. But I'm just not
seeing those three being nearly as good as they were in '07. Sorry,
guys."
Borowski wasn't good last season, by the way. It's a great
illustration of why the save statistic is so flawed. He had 45 saves
last season, but he had a 5.07 ERA and a 1.431 WHIP in nearly 66
innings.
On the other hand, Rafael Betancourt has been dominant in each of the past five seasons. He had a 312 ERA+ last season.
Rafael Perez was almost as dominant as Betancourt last season, but
he's only had one full season in the Majors and it is reasonable to
expect a decline from him.
A baseball season amounts to 162 episodes of 30 different reality shows.
Why is this comparison even necessary?
Those who think they can figure out the scripts in advance are kidding themselves.
I have a few friends who are very into Rock of Love 2. They have
predicted with amazing accuracy which girl is going to get the boot.
Why can they do this? They notice how they interact with Bret Michaels,
they pay attention to body language and the intricacies of the
conversations.*
Similarly, if you do your research, you can be accurate in your predictions.
* This will be the one and only time I will ever mention Rock of Love.
The stats reveal trend lines and tendencies, but in the end the game is played by human beings.
Played by human beings who create those trend lines and tendencies.
I like the Braves ... I think.
Bloggers, fire away.
Hope you liked it, Kenny.