Larry Mahnken and SG's

Replacement Level Yankees Weblog

"Hey, it's free!"


The Replacement Level Yankees Weblog has moved!  Our new home is:
http://www.replacementlevel.com

Featuring:
Larry Mahnken
SG
sjohnny
TVerik
Sean McNally
Fabian McNally
John Brattain


This is an awesome FREE site, where you can win money and gift certificates with no skill involved! If you're bored, I HIGHLY recommend checking it out!



Web
yankeefan.blogspot.com

Disclaimer: If you think this is the official website of the New York Yankees, you're an idiot. Go away.


March 20, 2007


The 2007 Diamond Mind Projection Blowout
by SG

It's that time of the year again, where I run some different projection systems through a multitude of Diamond Mind Baseball simulations to see how the MLB 2007 season projects.

I've done this the last few years, and here are the past seasons, and it's been hit and miss.

2005
2006

This year, I've added a new projection system to the mix. In addition to Dan Szymborski's ZiPS, Baseball Prospectus's PECOTA, and Diamond Mind's own projections, I'm also adding Sean Smith's new system CHONE. So I ran 1000 iterations of Diamond Mind simulations for each of the four projection systems listed above.

Before we get to the results, there are a few things you should keep in mind.

1)Projection systems are inherently limited in their accuracy, particularly for pitchers. We can get a rough idea of how most players perform by looking at their past histories and how similar players have performed, and factoring in aging and regression, but abilities/talent can change in ways that can't be forecasted.

2) Playing time distribution in these simulations will not match the actual playing time of the players involved. I used the rosters and depth charts available at Baseball Prospectus as my guide to set these up as realistically as possible, but it's a possible source of error. Rosters were set up to have 31-35 or so active players per team, and to get a reasonable amount of playing time from the bench and extra pitchers, to more closely model reality.

3) We cannot predict injuries and/or roster changes. These simulations do include projected playing time based on past health issues, and also random injuries which leads to some of the outlying results you see, but there's no way to account for all the fluctuations that will happen with rosters in 2007.

4) These are the averages of 1000 seasons, so the results will tend to regress towards the mean. The final standings will not look like this, because they only play the season once.

5) These are NOT my predictions. These are projections based on running a computer simulation thousands of times with projection data that is inherently limited. If your favorite team doesn't project well, don't blame me, blame the computers and spreadsheets that projected them. You can take heart in the 2005 White Sox projecting to win 79 games, or the 2006 Tigers projecting to win 80.

6) Since this is all automated, I don't break ties. I simply award all ties a share of either the division title or wild card when it happens.

So with all those caveats out of the way, here are the results.

First up, 1000 iterations of CHONE.



Next up, 1000 iterations of Diamond Mind's projection disk.



Third up, PECOTA.



Lastly, ZiPS.



And here's how it looks if you roll up the results of all four into one giant dataset.



Here's what the column headings mean:


One other new thing I've included based on requests from some readers is the average record vs opponents. This is too big to display in full on the blog, so use the scroll bars or read a little further for the link to the full spreadsheet with all the data.



To read this, look at a team on the very left column, and read across. The first number under a team on the top row is the average number of wins the team on the left column had against that team, the second number is the average number of losses. So if you read across from Baltimore, you see they averaged 7.5 wins, and 10.5 losses vs. Boston across the 4000 sets of runs.

You can look at a full Google spreadsheet with all the results here.

Here's how the divisional races shape up graphically (click on the images to enlarge them).




Here are my quick thoughts by team.

American League
Yankees Made playoffs 80.8% of the time. 95 wins seems about right to me, although I have more faith in the offense scoring 900+ runs than the pitching/defense being the best in the AL East.

Red Sox Made playoffs 50.5% of the time.The Red Sox seem lower than I'd have thought, although a lot of that has to do with some pretty bad bullpen projections. They probably have more upside in their rotation than any team in baseball, as far as the projections for Josh Beckett and Jon Papelbon and what they may end up actually doing.

Blue Jays Made playoffs 20.6% of the time.Diamond Mind really likes the Blue Jays, predicting them to finish second in the East. The other systems are less bullish.

Orioles Made playoffs 2.4% of the time.I'm not sure why, but I think they'll be better than that. Not a ton better, but better. Bedard, Cabrera, and Loewen could all exceed their projections.

Devil Rays Made playoffs 1.1% of the time. It still sucks to be them, although they're getting better.

Twins Made playoffs 50.3% of the time. Everyone thought it was nuts that the Twins were picked to win last year too, but look what happened. That being said, I'd be surprised if their pitching performs that well.

Indians Made playoffs 44.6% of the time. They're looking pretty strong, but they looked strong last year too.

Tigers Made playoffs 40.2% of the time. A lot depends on how Gary Sheffield fights off age and his injury last season. His projections are fairly conservative. They should definitely be in the mix again this year though.

White Sox Made playoffs 3.0% of the time. This one is a real head-scratcher to me. I think they're going to have a tough time with the division they're in, but not this tough. I'd expect Don Cooper to harness some of the talented arms they have into a better pitching performance. They should pick up a few wins there, at the expense of the teams projected ahead of them.

Royals Made playoffs 0.0% of the time. At least they've got Alex Gordon, right? The Royals are the only team who failed to make the postseason at least once.

Angels Made playoffs 52.3% of the time. Their projections seem about right to me, they will be one of the better teams in the league as far as preventing runs.

Athletics Made playoffs 34.1% of the time. Oakland's going to miss the Big Hurt. That being said, if Rich Harden can make more than his 22 projected starts, they should be in the AL West mix.

Rangers Made playoffs 13.2% of the time. Maybe they'll get the 60 HR version of Sosa...

Mariners Made playoffs 7.3% of the time. Felix Hernandez may have to single-handedly carry them to the playoffs if it's going to happen.

National League
Phillies Made playoffs 50.4% of the time. Their projection surprised me too. They've got some of the best front-line talent in the league in Ryan Howard and Chase Utley, but they've got some fairly significant holes in the lineup and bullpen. Luckily for them, their division isn't particularly strong.

Mets Made playoffs 42.8% of the time. I thought the Mets would project better, but given the uncertainty of their rotation maybe I shouldn't have. A lot depends on improvement by Oliver Perez, and Jose Reyes maintaining or exceeding his performance from last year. Both are certainly reasonable possibilities. The sooner they swap out Shawn Green for Lastings Milledge, the better. I suppose Mets fans can keep deluding themselves that their team is as good as the Yankees for now though.

Braves Made playoffs 37.5% of the time. It's all about a full season of Bob Wickman apparently. More seriously, it looks like the rebuilt bullpen and a solid middle of the lineup has them back in the mix after struggling last season.

Marlins Made playoffs 4.5% of the time. Losing Josh Johnson hurts, as does the lack of a real CF. There's enough young talent here to do better than this though.

Nationals Made playoffs 2.0% of the time. I'd be shocked if they won 71 games. This projection assumed a half season of Nick Johnson, which may be overly optimistic.

Cardinals Made playoffs 41.6% of the time. Last year's champs are still projecting as the class of the NL Central overall, although the different systems vary in this. ZiPS loves them, PECOTA doesn't.

Cubs Made playoffs 38.7% of the time. All that spending has improved them by a projected 19 wins, which is a lot. I'd probably give them a slight edge over St. Louis right now.

Brewers Made playoffs 22.2% of the time. They were a trendy pick last year, and it looks like they may be one again this year.

Astros Made playoffs 19.3% of the time. Their projection doesn't include any time for Roger Clemens. If he pitches half a season for them and replaces someone like Wandy Rodriguez, it's probably a 3 win upgrade.

Reds Made playoffs 3.5% of the time. I ran these before they announced Griffey moving to RF. Whether that makes a ton of difference remains to be seen.

Pirates Made playoffs 2.9% of the time. Will this be the year Littlefield gets fired?

Padres Made playoffs 49.3% of the time. They were my gut pick to win the division this year, so I was pleasantly surprised to see them win the simulated NL West.

Diamondbacks Made playoffs 32.2% of the time. They're another trendy pick to win their division, but a lot of it depends on how soon the Big Useless can come back, and if he's no longer useless when he does.

Dodgers Made playoffs 29.1% of the time. I thought Juan Pierre was a ten win upgrade. Apparently not.

Rockies Made playoffs 12.4% of the time. For some reason I thought they were worse than an 80 win team.

Giants Made playoffs 12.0% of the time. When I ran an initial set of 100 simulations in late December, they projected to win the division. However, based on the revised rosters and depth charts, which cut Barry Bonds's playing time severely, they're now bringing up the rear. If Bonds can play in more than the 50% of the games he's predicted to, expect them to win more frequently than this.

Update: Added a divisional placings break-down based on a suggestion by gotowarmissagnes. Click on the picture to enlarge it. If you want to see any other divisions let me know.


And there you have it. The 2007 Diamond Mind projection blowout. Results are not guaranteed.