Nike’s green glow “Kyrie-Oke” releases today (Pics)
February 17, 2016Gloves are popping in Goodyear
February 17, 2016Projections are always a dicey endeavor when factoring in all of the variables that can shift during a 162-game season, but PECOTA is using all of the information they have at current day to give the Cleveland Indians the league’s highest win total for the 2016 season. And before you tin hat-wearing folks want to try and compare this to a regional cover on a national magazine, this is much, much different.
The Sports Illustrated story made for great fodder—young, inexpensive team on the come up, ready to shock the world—but didn’t take in to account the complete lack of run scoring and the front office’s desire to sit on a young star until the middle of the season. PECOTA, however, is a bit different. While the algorithm is not shared, the foundation is: Who will play, how well they will play, and then converting those individual performances into team wins. Though projected to be in the middle of the road on offense (which is going to be the case given the less-than-ideal additions this offseason), PECOTA (Player Empirical Comparison and Optimization Test Algorithm) has the Tribe winning 92 games largely due to league-best run prevention thanks to stellar pitching and a defense that they believe will be the second-best unit in the game.1
For comparison purposes, the system feels the Indians will allow 100 fewer runs than the Detroit Tigers. And for those of you who listened closely to Jon Steiner back in the day, 100 runs is effectively 10 wins right out of the gate.
Like any good analysts, we can’t put all of our stock in one projection system (albeit a widely respected one). Good news is that ZiPS (despite comparing Carlos Santana to Nick Swisher), also gives the Tribe pitching staff rave reviews. You could do a lot worse that comps of Javier Vazquez, Curt Schilling and Josh Beckett.
If PECOTA’s forecast holds true, the Tribe will not be spending the entire summer vying for that second wild card spot while telling us that being over .500 is a win. No word from either forecast, however, on the potential for Tribe manager Terry Francona to stop bunting so damn much.
- The caveat here, of course is trying to project how much we will see of Michael Brantley. [↩]
34 Comments
wtf…cancel xmas. What did this say about us last year? We need to take back control of the home record, and hit above .500 on the road. Possible. They have KC dropping to the bottom of division?
The “Tribe’s offense sucked” myth needs to go away. The team ranked #6 in offensive WAR, but #18 in runs scored. The team had season long battle with hit sequencing, on both the pitching and hitting side. That’s why I was happy the team stood pat, with the Napoli signing as a low-risk upgrade, because we can’t get historically unlucky like that again.
For the record, here is a good resource on cluster luck, which compares runs scored to runs created (i.e. expected runs scored). It’s worth a gander.
Say what you want, the SI curse is 2-0 vs the Tribe.
Does PECOTA have a 1.000 winning pct?
I think not.
Team Snark reserves the right to wear it’s tin foil hat until convinced otherwise.
https://media.giphy.com/media/1daQkNsFlE9uE/giphy.gif
https://media.giphy.com/media/SxkYad09qgVzy/giphy.gif
Need to turn that clusterluck around!
Top of the clusterluck list…KC, number 2, STL, best team record in baseball
hi NATE … the tigers are back on their way up. should be a good battle in AL central.
I’m in a great mood, but this has me laughing
3-0. Manny on cover in 1996
You’re right about hit sequencing and bad luck, but the yearlong totals are at least a little inflated by the last 2 months of the season, which was a very different team to watch than in April/May. I still expect some bounceback from that bummer of an offensive year and have great hope that the vastly improved defense will make the difference.
They had us at 81 wins.
Player Empirical Comparison and Optimization Test Algorithm
Seriously. It’s come to this. Get Ken Burns on the phone. We have a compelling next installment for his nostalgic documentary. Ah, baseball.
Very interesting. They have the Royals projected to finish last in the AL Central, the Rays tied for first in the East and the Mets just outside the post-season.
http://www.sbnation.com/mlb/2015/1/29/7945803/2015-mlb-projections-pecota
OOPS THOSE ARE LAST YEAR’S pecota PROJECTIONS! MY BAD!
The name is a baseball nerd joke for those who are familiar with the work of journeyman Bill Pecota.
The Tigers were our kyptonite last year. I guess technically the entire central was, I don’t think we had a winning record against anyone in it except the White Sox.
Interesting piece over at http://fivethirtyeight.com/features/is-2015-the-year-baseballs-projections-failed/
“Paradoxically, in an age of unprecedented baseball data, we somehow appear to be getting worse at knowing which teams are — and will be — good.”
[…]
“While PECOTA’s absolute prediction errors are getting smaller across the entire population of MLB players, its squared errors — a gauge more sensitive to outliers — have increased over the last five seasons. For that kind of discrepancy to exist, there can be only one explanation: The big misses are getting bigger, at least relative to the normal, everyday misses. And, notably, more of those extreme errors come when predicting the performance of young players.”
I’m optimistic about this season. Winning 92 games seems high but I can see them winning 85 and grabbing a wildcard spot.
The article also admits that it’s not unlikely that we’re just choosing an endpoint that exacerbates the situation.
“Then again, maybe it’s all just luck — we mean literally. By definition, the compression of team records across MLB means that random variance is playing a larger role in the standings than it used to. How much larger? Computing the spread of true talent in a season using the standard deviation of team winning percentages, it turns out that a whopping 64 percent of the observed variation among teams so far this season can be explained by binomial luck — by far the highest single-season proportion of the past two decades.”
Okay, that’s funny, and clever.
I wouldn’t claim it to be anything close to a gut-buster, but the name is intentionally ridiculous.
I’d agree but the trend is significantly degrading over time. The models appear to be missing something.
https://espnfivethirtyeight.files.wordpress.com/2015/08/arthur-paine-feature-royals-21.png?w=575
But how will they do in April? hashtag early games don’t matter
Laugh it up fuzzball, but the algorithms state that based on our Theoretical Batted Balls in Play and the Clutterluck Striking in Situational Matchups (CSiSM) coupled with the jWAR and Offensive Balls Hit into Positional Gaps, we are going to score so many theoretical runs.
And don’t you wise crack about things like actual runs and wins and boring and outdating terms like that.
#aprilgamesmatter
Our past Aprils have been a total clusterluck. Or something like that.
I’m not exactly getting what’s going on with that chart. What exactly does the red line represent? Why does it drop in 2007 and 2013 when it seems that the coefficient of determination is higher? Why does it not drop, and even increase a bit, in 2000, 2001, and 2004?
If you take out that 2015 mark, I’m not seeing anything noteworthy. So that leaves us with that 2015 mark. Did something fundamental happen to the game that requires us to alter our prediction models, or is it a funny fluke? I have no idea, I think we don’t have enough evidence on either side.
The argument could be made that the AL Central is the toughest division in baseball. The NL Central would certainly beg to differ with its three super teams–Cubs, Cards, and Pirates. But the drop off from there is extreme with the Brewers and Reds both capable of losing 100 games this season. Alas, there are no such tomato cans in the Tribe’s division. Every team except the Twins, IMHO, has a plausible chance at winning this division. The Royals are back-to-back American League champions, the Tigers are pushing $200M with their payroll, and the White Sox have a pretty fearsome middle-of-the-order and an ace in Chris Sale. And if the Tribe occasionally hits in the clutch this season, they could be the best team of all.
https://en.wikipedia.org/wiki/Coefficient_of_determination
1.00 means it is perfect, 0 means it has no predictive value.
So it is showing that over the past 20 years the predictive value of the model is getting worse, peaking in 2005 and trailing off sharply from around a .65 to the current about .41…
I understand what a coefficient of determination is. The problem is that the chart is unmarked. I was under the impression that the grey dots were measuring the coefficient for each year individually, and that the red line is some kind of trend. If the red line is the coefficient, then what are they grey dots for? If the red line is a trend, why does it go move in the opposite direction of the grey dots kind of often? I’m assuming it’s some kind of rolling measure, but without labels, its darn near useless.
That’s what I was wondering. Maybe there’s something going on that I just don’t understand, but how is it that the grey dots for ’06 and ’07 are at or above the red line yet result in that dipping lower? That doesn’t seem logical to me.
Considering they just missed the playoffs by a few games, hopefully they at least have a better April.
My fault for not using the sarcasm font. This was directed towards those who were like “just chill, man. It’s only April,” when I was annoyed with the Cleveland Baseball’s weak start.
you were reading the zips projections rather than the baseball prospectus ones which the article is referring to which has us winning 92 games
http://www.letsgotribe.com/2015/1/29/7940135/2015-pecota-projections-cleveland-indians
Not a link to the BPro site itself, but a write-up on the 2015 projection. They projected us to win 81 games last year.