Steve Phillips Challenge: The halfway point
by Two Rs and Two Ls on Jul 13, 2009 11:38 PM PDT
33 comments
Before the season, nine intrepid LL'ers went on a mission to call the records of all 30 teams in MLB. The picks can be found in this post. I have put all the predictions into a spreadsheet, converted them to win percentages, and compared them to the standings at the all-star break.
Functionally, what I did was take a predicted win percentage, and multiply it by the number of games that the team in question has played, rounding down all games (So by your percentage, Cleveland won 46.9 games? I'm only counting it as 46). I then took the absolute value of the difference between each commenter's predicted wins and the actual wins at the break. For example, I predicted the Dodgers would only have 46 wins at the break. Instead, they have 56, giving me a score of 10 for that team. I added those scores together for all teams; low score wins.
The Standings:
Benne (116 off)
Two Rs And Two Ls (123 off)
sironz (124 off)
vivaelpujols (126 off)
cwel87 (128 off)
Jack Moore (129 off)
JI / Poochie (129 off)
hcoguy (132 off)
Trenchtown (135 off)
As a group, we were shitty at picking Cleveland (an average of 12 off), but good at picking all the teams in the AL East (no team an average of more than 3 off), and great at picking Minnesota, the White Sox, Milwaukee, Cincinnati, and Houston (all an average of just 1 off).
1 recs |
33 comments
Email
Print
Comments
I have it all in down as well
The original Vegaswatch used squared mean error as the standings determination and I can post that if you want (not that it treats my predictions any better.) Admittedly, I thought this was going to be much easier then it so far has been, fucking TIgers
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 13, 2009 11:48 PM PDT reply actions 0 recs
My spreadsheet looks
I’m hoping that works.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 13, 2009 11:55 PM PDT up reply actions 0 recs
I actually get a blank page when I click on that
I’m using firefox so I don’t know if that matters
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 13, 2009 11:59 PM PDT up reply actions 0 recs
Same here on both counts
Derosa.
by vivaelpujols on Jul 14, 2009 12:04 AM PDT up reply actions 0 recs
Funny you say that
I am searching frantically through espn looking to see if he actually made a real prediction or not to right down, funny I never did that to start the season
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 13, 2009 11:59 PM PDT up reply actions 0 recs
Hmm. I'll try to look into that tomorrow.
Google documents. I don’t know.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 14, 2009 12:03 AM PDT up reply actions 0 recs
Wrong thing to reply to.
Woops.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 14, 2009 12:05 AM PDT up reply actions 0 recs
Ok, bad news and possibly good news
Bad news is I can’t find Steve Phillips’ predictions. Possibly good news is that I found the original vegas watch article, followed their links, and could not find the 2008 Steve Phillips predictions, but saw how they might have been on that page at one point. This means that it is possible that they have the Phillips predictions recorded somewhere and will post them at the end of the year like they did this past time. If not, between articles and prediction we should be able to piece together as pseudo-Phillips prediction. This is not ideal but better then nothing. However, I am confident that Vegaswatch has the information recorded somewhere
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 14, 2009 12:21 AM PDT up reply actions 0 recs
If I did the math right...
…it looks like predicting .500 would make the sum of the absolute errors 139 right now. (This is without rounding anything and saying that a .500 prediction has a half-game error for a 45-44 team.) ~12% of that 139-game total is from Washington. I knew they were bad, but that’s nearly 2003 Tigers bad.
So it looks like everyone is doing slightly better than the null hypothesis that all teams are equally good.
by ubelmann on Jul 13, 2009 11:57 PM PDT reply actions 0 recs
That's better than most (bad) projection systems actually
Derosa.
by vivaelpujols on Jul 14, 2009 12:03 AM PDT up reply actions 0 recs
Yeah
And if you use the other obvious null hypothesis, that teams should be as good as they were in 2008, I get the absolute error (without rounding any of the partial games) to be about 144 games. I would expect that if you made a regression using the ‘07 and ’08 data you could probably come up with a formula to beat the “everyone is a .500 team” hypothesis, but I’m not sure by how much it would win.
by ubelmann on Jul 14, 2009 12:09 AM PDT up reply actions 0 recs
If I do a regression between 2007 and 2008
(I know, not much data to go on, but I am lazy in compiling records by team.) Then I get the following regression equation:
2008WPct = 0.10*(2007WPct) + 0.45
Which is equivalent to this form, which I think is a nicer way to look at the regression:
2008WPct = 0.10*(2007WPct-.500) + .500
The r is pretty low on that fit, unfortunately, just 0.09. If you figure the same coefficients should hold for ’08 and ’09, the absolute error in the prediction to this point is ~133 games, which hardly seems worth the effort for the 6-game improvement over the “every team is .500” hypothesis. Going back to use more pairs of years to get better regression coefficients might help, but probably not more than another 6 games (total) or so, I would guess.
by ubelmann on Jul 14, 2009 12:25 AM PDT up reply actions 0 recs
Despite the fact that you predicted the Mariners only playing 160 games!
My Mariners blog - SodoMojo Twitter Feed
by gregrabble on Jul 14, 2009 12:24 AM PDT up reply actions 0 recs
It could happen.
I like using semi-colons; they make me feel smart.
by Llewdor on Jul 15, 2009 12:48 PM PDT up reply actions 0 recs
A couple rainouts against insignificant teams and you've got it.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 15, 2009 1:18 PM PDT up reply actions 0 recs
Ugh Cleveland.
I recall rating the Brewers high but can’t remember.
Thanks for following through on this. My friends and I do something similar every year with the division predictions but don’t go all out like this.
by hcoguy on Jul 14, 2009 12:12 AM PDT reply actions 0 recs
Holy crap, I'm in first place? Sweet.
Is that the light at the end of the tunnel, or the headlights of an oncoming train?
by Benne on Jul 14, 2009 9:29 AM PDT reply actions 0 recs
Heh, we pretty much all whiffed badly on the Indians and D-Backs.
Is that the light at the end of the tunnel, or the headlights of an oncoming train?
by Benne on Jul 14, 2009 9:30 AM PDT reply actions 0 recs
Strangely, we were terrible at the Nationals.
They have 26 wins so far. The lowest guess anyone gave so far was 30 wins (Sironz, who guessed the record as 55-107). Nobody else picked them to lose 100 games this season, though a couple of us picked them to lose 99.
The next two teams to get blown so often were Arizona and the Dodgers; we were an average of 9 games off of each of them.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 14, 2009 12:08 PM PDT up reply actions 0 recs
I guess I was wrong to assume that the Nats' "rock bottom" was last year.
Is that the light at the end of the tunnel, or the headlights of an oncoming train?
by Benne on Jul 14, 2009 2:42 PM PDT up reply actions 0 recs
I doubt you would of found many people who could have predicted Cleveland and Arizona to suck this badly.
The Rise of a Superstar:Justin Upton-.425 wOBA, 21 years old.
by Goose on Jul 14, 2009 12:17 PM PDT up reply actions 0 recs
Which one of us correctly predicted the 10 game range that the Mariners would reside in?
by Robert on Jul 14, 2009 9:36 AM PDT reply actions 0 recs
I went conservative with the M's and predicted 77-85.
Is that the light at the end of the tunnel, or the headlights of an oncoming train?
by Benne on Jul 14, 2009 10:07 AM PDT up reply actions 0 recs
I went and got the PECOTA projections.
PECOTA scores 74 so far. Interestingly, that system was off the most on Washington and Cleveland, just like we were as a group.
On average, we BEAT PECOTA on:
Boston, White Sox, Kansas City, Seattle, Philadelphia, Atlanta, and Houston.
PECOTA was ON THE MONEY (off less than 1/2 of a game) on:
Toronto, Minnesota, and Milwaukee.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 15, 2009 2:16 PM PDT reply actions 0 recs
You mean the Pecota Projections updated July 12th?
yeah thats real fair. Just looking at the original projections of Pecota posted February 19th, they only have two division posted (I let my subscription lapse this winter) for free and were off 60 alone for those two divisions and I doubt they were off only 14 games in the four other divisions combined (not a chance) is the only way they would be at 74. When were those projections you used posted?
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 15, 2009 10:35 PM PDT up reply actions 0 recs
I don't know if the projections themselves have been updated
but the page that they were on had been. Linky.
From looking around other sites, it looks like you’re right, they did update the projections. Well that makes this worthwhile.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains. -- P&P&Z
by Two Rs and Two Ls on Jul 16, 2009 10:06 AM PDT up reply actions 0 recs
It would be fun to check all of this accuracy using the BaseRuns/Pythag records.
...and now I'm here
by Librocrat on Jul 15, 2009 4:37 PM PDT reply actions 0 recs
The Chone/Decatur projection
using the Two R’s method is 136 off, or in other words, dead last
Don't believe the lies Bill!!!! look at the sparkly ERA!!! Sparkly, Sparkly!!! - McCovey Chronicles
by Trenchtown on Jul 15, 2009 10:17 PM PDT reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2009 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
