Neyer stands corrected!
Woo hoo!
All has been restored in the world!
Update [2006-8-17 18:0:4 by baseballgirl]: From today's chat:
Ray (San Francisco): Okay Rob, now that I've booked Texas into the playoffs, per your instructions, lets be realistic: The A's are going to win the West. It's pretty clear, run-differentials be damned. I like stats as much as the next guy, but in this instance I don't think RD tells the whole story. Agree?
Rob Neyer: (1:12 PM ET ) Of course it's true that run differentials don't tell the whole story, particularly regarding the final standings. Obviously I must have been in some sort of drug-induced haze Tuesday -- prescription drugs, I swear! -- when I said the Rangers were going to be in the playoffs. Because at this point the A's are close to being a lock in the West, and the Rangers aren't winning the Water Closet.
From me:
That being said[...]during your chat, you picked the Texas Rangers to win the West. Now, I'm obviously an A's fan, so I can't commit to a winner just yet, but it seems to me that Texas is going to finish third in the West, leaving the Angels and the A's to battle it out for the playoff spot (or leaving the Angels to fade away into the night, letting the A's have the West uncontested).
You probably looked at the A's run-differential and said what all of us do every day: How in the HECK do the A's have the record they do with THAT RD? (Pssst...the answer is 'The Seattle Mariners'.)
Believe me; not an hour goes by when someone doesn't pull up the numbers and try to reconcile the A's record. We know how it works; we know that the A's shouldn't be where they are, but the fact remains that they have a 5.5 game lead on the Angels, and a 6.5 game lead on the Rangers, with 43 to play. They have a less-than-good offense, but they have the pitching (especially bullpen), and the defense, and somehow they are winning the games, even when they're not supposed to (see: two huge game-changing homeruns for just about their only hit in the each game this week).
[...]I've seen or heard 119 A's games this year, and probably 50 or so Angels games (I live in LA), and I follow Texas and Seattle whenever possible, so I have a pretty good idea of the West this year, and I will be the first one to say that the numbers don't back this one up, but here we are anyway.
[...]I enjoy you on ESPN...keep up the good work[...]
His response:
Anyway, I corrected myself today; the A's are going to win. -rob
p.s. Thanks for the kind words; much appreciated -r
Yep, I knew I loved Rob Neyer. So there you have it.
<goes back to nervously looking at the standings>
Good grief; the A's just better win every game from her on out, because I think my fragile state does not need a down-to-the-wire playoff race this year.
39 comments
|
0 recs |
Do you like this story?
Comments
I have a crush on Neyer.
It's just a harmless crush!
I love...
Well, whadya know
Good job, bbg!
The [...]s
Missed that
by peanut gallery on Aug 17, 2006 3:41 PM PDT up reply actions
No no no no...
by baseballgirl on Aug 17, 2006 7:11 PM PDT up reply actions
I know you were nothing but class, bbg
by peanut gallery on Aug 17, 2006 9:35 PM PDT up reply actions
3 things
- Neyer is a good guy. I've emailed him four or five times over the course of the past year, and he's always responded thoughtfully (to both criticisms and compliments). I wish ESPN would give him a test run on Baseball Tonight - with Gammons on the DL, it would be nice to see a reasonable counterpoint to Kruk's bluster once in awhile.
- Anyone who offhandedly employs the phrase "inherent iconoclasm" is okay by me.
- 119 games? Which one did you miss?
by MKX on Aug 17, 2006 3:15 PM PDT reply actions
The email's a couple days old
by baseballgirl on Aug 17, 2006 7:11 PM PDT up reply actions
Don't worry
Actually, the A's only need to win 36 to clinch the AL West. I'm sure we can keep up our current 9-1 pace for the final six weeks of the season.
Or as Rob Neyer would say, "Book it."
I was going to post this as a diary...
Is the use of run differential is a good predictor of a team's success based on the RD over a 162 game season? If so, I would assume that it is less accurate when applied over fewer games.
Additionally, I'd assume that it would have a tendency to even out over the course of the year. When Neyer wrote his initial prediction, the A's RD was 4-5 runs. Now it is 18 (LAAAAAAAAAA is now at 11!). With four games against KC this weekend, it could easily be 30 on Monday (keep yer fingers crossed!). Within a week, the RD starts looking more palatable.
I think Neyer made the mistake of taking a stat that is based on performance over the course of a full season and applying it to a situation that involved an incomplete season.
It does have predictive value.
Or to put it a different way, given two teams with different run differentials, the team with the better differential will likely win more of the remaining games.
You will note, howerver, than almost all the teams in first place are playing better than their differential would suggets. (All but LAN)
http://www.hardballtimes.com/main/article/tht-daily-all-star-pitchers-win-again/
I looked at this a little while ago
I knew I had read something about this before
One of your last paragraphs states that "we have some evidence that leveraging relievers is one way to outperform your Pythagorean projection" - my off the cuff reaction was to think that this almost certainly applies to the A's. I think, if you combine this with the blowout factor mentioned below and a good bit of random chance, it is pretty easy to accept the fact that the A's have their current record despite their differential.
So, it looks like Neyer's reliance on the pythagorean projection based on run differential was bitten in the ass by a confluence of exceptions to the rule. Interesting...
by RickeySteals on Aug 18, 2006 12:27 AM PDT up reply actions
BBG
Great job BBG.
PS Now can you assure me that EI will show the first game of the DH tomorrow? I keep checking my digital TV guide and it isn't there yet.
righter of wrongs...
ditto that
by whiteshoes40 on Aug 17, 2006 8:23 PM PDT up reply actions
seen her on the radio?
Run differential
Inconsistent due to injuries
This is a team that's 23-9 since the ASB, and has a 10-game win streak, but has also gone through huge fallow periods. Just look at Thomas, Swisher, and Chavez as microcosms, and you see guys that have been alternately really good or really bad due to their health. Their numbers don't impress because when they've been bad, they've been so very bad. But if you've watched the whole year, you know how good they are when they're right.
All the "experts" picked us to win the division (and some, even the World Series) for a reason--when this team is healthy, it can be really good. I hope all the pundits keep chocking up the recent success to luck, instead of resurgent health, because when we get to the playoffs (barring more injuries, of course), nobody is going to know what hit them.
is it me
BBG = no heart, not Mexican enough
But...but...we've been grooming Louismg...
wait a second
She better not do what other former A's have...
Next AN Day SGA, BBG talking bobbleheads.
You can't get it to shut up!
by baseballgirl on Aug 18, 2006 9:00 AM PDT up reply actions
I actually sent an email to Caly Davenport
I pulled out all of the games with RDs of 8 and over, and came up with a Pythag W% of around .533 (3-4 games ago), which wasn't far from their actual W% at the time.
While he didn't like eliminating games altogether, he said he would take a look at possibly capping blowouts to see how much more realistic of a W% it gives.
Just from looking at the teams that are playing below their Pythag W%, you see teams that score runs in bunches, like Texas, and the ones tha are below it are pitching/defense teams for the most part.
Re:
by Marc Normandin @ Athletics Nation on Aug 17, 2006 4:48 PM PDT up reply actions
yeah
I'm thinking about doing it for all of the teams, but since BP went to the new-style reports pulling the data for run differentials in an easily usable format has gotten a lot harder.
Instead of pulling them out
Is it really that hard to understand?
When a team has so-so offense, they're not going to score a lot of runs either.
Whenever said offense scores more runs than the opponent, they pick up a win, but rarely by many runs.
It's more common for good pitching and defense to have a terrible day and lose a lot of runs than for so-so offense to have a great day and score a lot of runs.
Seems reasonable to me. :)
I posted this
"I was more about making the point that our run differential as a whole better reflects the team we've been playing as for most of the season, a mediocre, hurt, underperforming team, worthy of a .500ish record. I think with guys like Bradley, Loaiza, and perhaps Chavez healthy, and better performances starting to show from Ellis, Swisher, Scutaro, Blanton, and perhaps even a guy like Harden returning indicate to me, at least, that we should expect to outperform the standard RD we've estbalished for this season, thus rendering any analysis of our current RD in an analysis of our current team and it's future expectation sorta pointless.
Basically my point, without all the convolutedness, is if the Royals had a disgustingly bad run differential and were 30-60 or something, but then all of a sudden were given Albert Pujols, David Ortiz, Miguel Cabrera, Johan Santana, Justin Verlander, and other stars, and then played a few games where they won (let's say 35-60 now) because they're obviously now a much better team, but their run differential still sucked, you wouldn't take that run differential as a basis of your expectation for the team's play the rest of the way, would you?
I'm not advocating removing outliers from RD or anything of that like, just illustrating that it should be evident that our RD is a poor indicator of the team we're actually now fielding out there."
Again, my point is basically that I think it's pointless to use RD as a basis to predict future success with if a team has largely changed, especially within a season where one flukey week can GREATLY skew an RD. And I think the 06 A's certainly apply if this is really the case.

by 























