New Jersey Institute of Technology professor - Marlins 76 wins
by craig on Mar 11, 2010 4:38 AM EST
in News
7 comments
This is about like all of the projections trying to predict the Marlins seasons, it comes up with a losing season. Granted there are some who predict they will break even. However, back to the good prof's predictions.
The Amazin' Mets will come in third in the NL East, according to Bruce Bukiet, an associate professor of mathematical sciences and associate dean of the College of Science and Liberal Arts at the New Jersey Institute of Technology.
---
;">
Bukiet bases his predictions on a mathematical model he developed in 2000, one that computes the probability of a team winning a game against another team with given hitters, bench, starting pitcher, relievers and home field advantage. For this season, Bukiet has refined his algorithm slightly, incorporating a more realistic runner advancement model. Whatever that is.
The professor claims to have beaten the odds in six of the eight years he's been using the model. According to his predictions, the Philadelphia Phillies, St. Louis Cardinals and Los Angeles Dodgers should all repeat as winners in the National League, with the Atlanta Braves taking the wild-card slot.
Don't have the time to research his past predictions where the Marlins are concerned. Maybe he got them right, but I sincerely doubt it. I would be flabbergasted to learn he had the Marlins in the preseason winning the Wild Card in 2003. But maybe he did.
My guess is that his model is about like all models, suffering from omitted variables basis along with structural flaws. Had he had the proper equation, and nobody does, and estimated it with something like the likelihood function, something I have found to be very hard to estimate, he may have picked up more explanatory power where teams like the Marlins are concerned. Be it known, I have no idea what his methodology is, so I could be flat out wrong about it. But whatever the case, assuming the Marlins don't suffer major key injuries, I think his analysis is short changing the Fish.
Anyways, here is what he came up with for the NL East.
NL East: Phillies – 90-72; Braves – 88-74; Mets – 82-80; Marlins – 76-86; Nationals – 72-90
Let's just say, I'm not buying the results.
0 recs |
7 comments
|
Email
Print
Comments
Marlins and projections don't mix good
Over at fagraphs a while ago they did an article on how we almost always beat out projections, and it’s because of all the young talent we have. When you have so many players with limited MLB sample size it’s hard to make accurate predictions.
I think the Marlins are better than 76 wins… and I’m pretty sure the Nats wont win 60 games, let alone 72
The Nats aren't awful
You’d have to have a pretty terrible team to project below 60 wins, and the Nats have too good an offense for that.
That being said, I think this shortchanges the Marlins too, but there’s so much variance in a given season that we wouldn’t be able to tell anyway. The std. deviation of a given team’s win total in one season is 6 or so wins. I personally think the Marlins are around 83 wins (though I haven’t done the checking yet, could be more like 81), but his prediction is close to those bounds. Last year the team played like an 83-win team and won 87, who knows what will happen in season?
A quick note: projections are not predictions. You can’t predict what happens during a season, given injuries and stuff like that. This professor’s model has the Marlins as a true-talent 76-win team. I don’t buy it, the bigger systems have us at around .500, as do I.
Marlin Maniac, a Florida Marlins blog
Writer, Beyond the Box Score
Writer, Baseball Propsectus Fantasy Beat
Writer, Heater Magazine
just a prediction lol...
but I honestly dont see them winning 72 games, especially in this division. The Mets are healthier (well somewhat), the Marlins are good, the Braves are also good, the Phillies are great.. but who knows maybe they’ll end up becoming the next TB Rays and surprise everyone (including myself), but i highly doubt it
The problem with projections
is that they always forget that real people play the game. Projection systems treat the individuals that play as machines who will have predictable outputs from 5, April to the last game of the world series. I think it’s important to remember that there are incredible immeasurables in each player and in how each player responds to those around him. That accounts for success or failure at a much greater rate than most people want to admit.
Not exactly.
Projections are based on averages. Built into the projection — on a team level — is the assumption that some players will exceed those average expectations. Some will fail to reach them. Others will be injured. And it’s a safe bet that the players exceeding expectations will balance evenly against the players falling short.
But it’s a safe bet like flipping coins is a safe bet — if you flip enough coins, you’ll get something like half heads and half tails. But if you flip 10 coins, you’re not always going to get 5 and 5. You may get 3 and 7, in which case you’ll probably fall a little short of the expectations. You may get 8 and 2 and make the playoffs.
And where the Marlins tend to overachieve is, they’ve generally got a few coins in reserve, and a lot of players with limited track records. I mean, anyone can project a season for Hanley, or Albert Pujols… or for that matter Kevin Kouzmanoff or whoever. Any established player with 5+ years in the majors. This stuff is easy. But a guy like Maybin? Maybin could be 2009 Maybin, or he could be a star. You don’t know. You don’t have the numbers to project from.
And consider our rotation after Josh and Nolasco (who, by the way, kind of came up tails last year…) We’ve got Sanchez, you’re projecting Sanchez’s numbers. Well, okay, Sanchez might equal your projection. He might exceed it. But if Sanchez starts falling short of those numbers, we’re just going to put Sean West, or Badenhop, or someone into the rotation. Flip again.
And we can do that, both because we have the extra players, and because we’re not so tied to the players we have. If the Mets pay some guy $10 million a year to play third base, it’s going to take an extreme situation for them to bench him. We’re paying half our starters $400,000 a year. They screw up… well, if we have a better option to try, we’re trying it.
Bass is a kind of fish.
Comments For This Post Are Closed
Secondary Sidebar
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2010 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.
