Not the 2017 Standings

What will the 2017 standings look like at season's end? Probably not like the chart below. This is Fangraphs' Zips projections for each team, which has every bit as much value as every other projection system, but less than toilet paper, which has at least one other important use.

Still, it's a starting point for understanding the upcoming season.

Some items to note: 
1. None of the projection systems see much in the way of division races. Zips has the Red Sox winning the AL East by 4 games, but after that there's no drama. It's got Cleveland by 8 in the Central and Houston by 10 in the West.  In the NL, it projects the Nats by 8 and Cubs by 17, and the Dodgers over the Giants by 7 in the West.

2. If all goes according to form, which it literally never does, the Wild Cards are a pair of shootouts at the October corral. This chart rates the Angels Mariners, Rangers and Rays as a toss-up for the second Wild Card behind the Blue Jays. In the NL, the Giants, Mets, Cardinals and Pirates will battle for the play-in game.

3. The most volatile projections belong to the Anaheim Angels, Baltimore Orioles and St. Louis Cardinals. Zips puts the Halos at 83 wins and a Wild Card while Baseball Prospectus's PECOTA rankings have them at 77 wins and 85 losses. PECOTA doesn't believe in the Orioles either, tabbing them at 74 wins and the AL East cellar. And PECOTA is unfathomably dubious about St. Louis, forecasting just 77 wins and a finish behind Milwaukee.

So where do all those wins go in PECOTA? To the basement. Its worst team -- the Royals -- still win 71 games, more than four teams in the ZIPS projections

4. No one is much buying the Rangers, who ran away with their division last season. The advanced metrics suggested that was more a matter of serendipity than talent and regressed the Rangers to their natural ability level. A healthy Yu Darvish could mess with that.

5. The Mets' outlook is dampened severely by health concerns on the mound. They are rated as just the fourth best staff in the NL, barely ahead of Washington in fifth. But if the breaks go their way -- or actually, if there are no breaks, and pulls, tears, strains, tweaks, stiffness, contusions and syndromes are kept to a minimum -- they could be scary good at preventing runs. Of course, if the queen had testicles she'd be king.

6. The White Sox will look to deal whoever isn't nailed down mid-season, so their numbers could suffer in the short term in order to improve quickly down the line. The projections already like the path the Braves are on.

Are you ready? Opening Day's just a couple of days away.

MLB Demonstrates the Uselessness of Projection Systems

Major League Baseball is touting today the Steamer baseball projections for 2016. Steamer is one of several Sabermetric projection systems -- including Fangraphs' ZIPS, Baseball Prospectus's PECOTA and the Hardball Times' Oliver -- that offer a crystal ball of sorts on future player and team performance.

These systems generally work the same way, projecting the accomplishments of players by comparing their careers to similar players in history, averaging the performances of those similar players in their next season and adjusting that to the player in question.

The MLB.com story is not a particularly interesting read. It projects that Miguel Cabrera will win another batting title, Clayton Kershaw will again top the pitching charts and Giancarlo Stanton will once more outslug the field. 

Shocking.

This is what you get with projection systems, which have no ability to predict the unpredictable. The 2015 accomplishments of Jake Arrieta, Dee Gordon, AJ Pollock, Dallas Keuchel, the Kansas City Royals and New York Mets were no less a mystery to these projection systems than to you.

They also have no ability to predict outlying performances. They will never project any player will slam 58 home runs, bat .360 or accrue 21 pitching wins unless he has done that regularly in the past. 

Moreover, Steamer does not attempt to project playing time, so even though Stanton has missed 40 games a season in his career, the projection assumes he plays 150 games. (Some of these systems do take playing time into account, which means they are wildly wrong slightly less often.)

These systems gave us the Angels, Dodgers and Nationals as baseball's best teams in 2015. They got the AL East and West almost exactly wrong. They offer some mild insights that can be almost completely encapsulated in 13 simple rules.

So don't bother reading the article. (Read this one about Statcast instead.) And let's just enjoy the raging unpredictability of the game that served us so well last year.

The Projector Was Broken: Pre-Season Projections Got Everything Wrong

History will little note, nor long remember, what we say here." -- Abraham Lincoln, Gettysburg Address. Voted worst prediction in history.

In February, we examined the PECOTA projections for MLB 2015. Based on everything the computers could crunch prior to the start of the season, this is how Baseball Prospectus stacked up all the teams for the 2015 season. Remember that the projections can't predict trades and injuries. They can't predict much else, as you'll see.

The projections had Washington, St. Louis and the Dodgers winning the NL divisions with the Mets, Giants, Marlins and Padres competing for the Wild Cards. That's two of three division winners and two of four Wild Card contestants.

PECOTA saw the Cubs, Pirates and Braves as .500 teams. To that we say: ha!

On the other hand, it correctly tabbed Philly, Cincinnati and Colorado as bottom feeders.

In other words, the fancy computers with their gigagoogles knew about as much as you did with your biases and half-baked opinions. Not exactly an endorsement for SABR membership. About the only thing we can say about the projection is that it recognized improvement in Flushing.

It gets worse. In the junior circuit, PECOTA struck out looking on a fastball down the middle. It identified last place Boston and Detroit as division kings and named the Of Anaheims as the best team in the league. Its Wild Card competitors were the woeful Mariners and A's, along with Tampa. In other words, oh-for-six.

PECOTA projected the top teams in the Central -- Kansas City and Minnesota, who finished a combined 32 games over .500 -- as the AL's two worst teams, with a combined 141-183 record. It did the same for Texas and Houston in the West.

That's as stinkin' wrong as you can get.

In the East, PECOTA projected the Blue Jays a couple of games over .500 and the Yankees a couple of games under. About the only thing the projection system nailed was that all the machinations in Chicago still left the White Sox as poseurs.

In other words, all the fancy algorithms don't know squat. A blind squirrel could have picked one of the division winners. This is the very point I made in this post  eight months ago.

It's not that the seamheads are stupid, or have nothing to add to the conversation, or to our understanding of the game, quite the opposite. It's that baseball is a crazy game that no one can predict, project, prognosticate, portend, prophesize, augur, foretell, forecast or in any way anticipate. That's what we love about it.

So let's all, all of us including the great scientific analysts of the horsehide, just take a chill pill and dial down the...um...what's the word...?

An Un-Evan Projection for Longo

What do the projection systems have against Evan Longoria?

The All-Star third-sacker has manufactured a .298 lifetime True Average (TAv) in a seven-year career, dropped below .300 only by last season's struggles following off-season surgery. That not withstanding, Longo has been automatic 30 homers and a .351 lifetime on base percentage.

The projections systems say that at 29 all of Longoria's best days are behind him. Three prominent systems -- PECOTA, ZIPS and Steamer -- say he is more like a .255-.260 hitter, with OBP in the .330s and low-20s home runs. A seven win player his first six seasons, Longoria is grading out as just a three-win player in 2015, almost all of it because of defense.

This suggests that players with Longoria's profile -- medium-speed power hitters with fancy leather -- don't age well starting in their late 20s. Or it suggests that a new $100 million contract will emotionally short-circuit the ordinarily Evan-keeled Longoria and dampen his skills in 2015.

Neither of these premises rings particularly true. Indeed, the most likely event, barring further injury, is a bounceback from last year's scuffles that sustains itself for a couple of seasons before real decline sets in. The former Rookie of the Year has too much talent, demonstrated too consistently, to suddenly fall off the cliff in his prime.

You Can Project Corey Kluber's 2015 Season

Do you aspire to be PECOTA? 

Not Bill Pecota, the late 80s-early 90s Royals infielder who smacked 22 lifetime home runs.

PECOTA, Baseball Prospectus's "Player Empirical Comparison and Optimization Test Algorithm."

It's one of several super-heated statistical projection systems that help filter out some of the noise to determine how today's Major Leaguers might perform next year based on their past performance, age, injury history, and how similar ballers have performed at the same point in their careers.

As I mentioned last post, PECOTA and its cousins have a WAR of about two. They're contributing team players but you're not winning the pennant with them. They excel at pointing out the ravages of age on players. They can't predict big surprise performances. 

In fact, you can be your own PECOTA if you keep these rules in mind:

1. Hitters tend to peak around 26-29 and thereafter decline.
2. Pitchers' peaks are more varied and can occur later.
3. Hitters' batting averages can fluctuate because of varied BABIP, but walk rates tend to remain fairly constant.
4. As they age, everyday players lose speed, walk and strike out more, steal fewer bases, play worse defense and hit more home runs.
5. Lumbering first basemen and DHs tend to lose their skills and their value quickly. Speedsters tend to age more gradually.
6. After about age 36, even the healthiest everyday players begin breaking down and losing time to injury.
7. Players who have one anomalous season after several years of significantly better or worse performance tend to regress to the mean the following year. That is, a third baseman with a .250 lifetime batting average and 15 homers a year who bats .300 and goes yard 35 times this season is likely to bat around .275 with 25 taters the following season.
8. Players who have anomalous first half seasons tend to play to their projection in the second half, not withstanding injuries and other externalities.
9. Players who enter the league late tend to leave it early. Think Evan Gattis.
10. Hitters who enjoy unusually high BABIP, and pitchers who allow unusually low BABIP, are generally hoodoos and fakes who will pay the piper the subsequent season. (The reverse is also true.) "Unusual" is relative to a player's own experience. Ichiro had ridiculously high BABIP every year because of the way he went about hitting.
11. Pitchers can seem to have great or poor ERAs because of the defense behind them. Examining "Fielding Independent Pitching" statistics offers a more accurate window into the future.
12. Pitcher ERAs rise when they move from the NL to the AL (and vice versa) mostly because of the DH.
13. Pitcher ERAs fall when they leave the rotation and enter the bullpen. It's easier to subdue batters for one inning than for seven. Moreover, relievers tend to face same-handed batters; starters tend to face lineups constructed to defeat them.
14. Baseball players are humans. They are unpredictable. That's what makes the game fun.

You are armed. Go out and predict the future. Be PECOTA.

(Next post, we'll examine some of the interesting projections. With a  grain of salt, of course.)

A Smoaking Gun and Other Stories

At the risk of putting too fine a point on a previous post -- aw hell, there's no risk: we're going to shave the nub to within a millimeter of its life -- Baseball Prospectus released their projections for MLB players today and here's what the Tarot cards suggest for Justin Smoak:

.224/.313/.371 with 14 HR in 470 plate appearances for the Mariners this season. With weak baserunning and slightly below average defense, Smoak grades out essentially to replacement-level at first base, even after factoring in the park he plays in.

In other words, you can pick up any old Tom, Dick or Justin Smoak off the waiver wire and expect about the same production.

Add in the missing 150 plate appearances, which suggest either an injury or a demotion; in either case his contribution during those 30 games is zero.

By comparison, to which the previous post alludes, Matt Wieters's forecast comes in at .259/.330/.422 with 19 HR in 588 plate appearances while dinging him slightly on defense behind the plate. Wieters is worth a pair of wins against replacement in this forecast, which makes him a solid regular but not a star.  

These are projections, not predictions, so even if Smoak breaks out and Wieters breaks down, the system can't be said to be "wrong." (In fact, the system includes the odds of breakout and collapse as well, but the math is obtuse.) It just means that players who have performed the way these two have at their ages in the past have tended to perform like the projection going forward.

According to BP's projection system, the top 10 players offensively in 2013 will be: 

• Albert Pujols
• Joey Votto
• Ryan Braun
• Miguel Cabrera
• Mike Trout
• Troy Tulowitzki
• Adrian Gonzalez
• Andrew McCutcheon
• Robinson Cano

It sees a big bounceback for Pujols, as does Bill James's system, and a massive dropoff for Mike Trout, who has nowhere to go but down after an historic first season. Of course, projection systems don't work well with guys like Trout because he doesn't really have an comparables to project him against.

It's also interesting to see the system projecting a return to stardom for Hanley Ramirez, whom it ranks as the 21st best offensive player. That is based on the Dodgers leaving him at short and giving Dee Gordon the seasoning he needs in Albuquerque.

The system also sees a return to relevance for newly shorn Tim Lincecum with a 3.10 ERA in 174 innings and a return to normalcy for Gio Gonzalez with a 3.67 ERA and the same value to his team as Milwaukee's Mike Fiers.

As for the unprojectable R.A. Dickey? Canada won't want to hear this. The system sees a 38-year-old moving to a tough division in the tougher league and projects 224 innings of a 4.41 ERA with a below .500 record. But then, R.A.'s beaten the odds before.

This is the kind of conjecture we have at our disposal as pitchers and catchers loosen up in Florida and Arizona.

The Final Standings As of March

For what it's worth, below are Baseball Prospectus's projections for the 2012 season.

Keep in mind that projections are not predictions. They are derived by aggregating the expected performances of each individual player and his projected playing time. Each player projection is a meld of his past performance and the experiences of players like them in the next year of their career.

BP runs a million season simulations and averages them for the numbers you see here.

NL East

Atlanta Braves            88  74  .543

Philadelphia Phillies    88  74  .543

Miami Marlins              87  75  .537

Washington Nationals 83  79  .512

New York Mets            78  84  .481

NL Central

St. Louis Cardinals      88  74  .543

Milwaukee Brewers    86  76  .531

Cincinnati Reds          85  77  .525

Chicago Cubs              74  88  .457

Pittsburgh Pirates      72  90  .444

Houston Astros          62  100 .383

NL West

San Francisco Giants     86  76  .531

Arizona Diamondbacks  85  77  .525

Colorado Rockies           80  82  .494

San Diego Padres           79  83  .488

Los Angeles Dodgers      78  84  .481

AL East

New York Yankees           93  69  .543

Boston Red Sox              88  74  .543

Tampa Bay Rays              86  76  .525

Toronto Blue Jays           77  85  .474

Baltimore Orioles            73  89  .451

AL Central

Detroit Tigers          85  77  .519

Cleveland Indians     80  82  .494

Chicago White Sox   78  84  .481

Minnesota Twins      73  89  .451

Kansas City Royals    67  95  .414

AL West

Texas Rangers          99  63  .611

Los Angeles Angels   88  74  .543

Oakland A's               73  89  .451

Seattle Mariners       70  92  .432

The Texas projection shocks me, particularly losing C.J. Wilson and expecting a decline in performance from Josh Hamilton and perhaps some of the rest of the rotation. BP foresees big things from Japanese import Yu Darvish and the Texas offense, which it projects to outscore the Angels by 207 runs.


On the other end of the spectrum, it's hard to overstate how little the projection thinks of Houston. To lose 100 games, teams usually experience multiple breakdowns, injuries and unexpected bad play. This projection says the Astros will lose 100 if all goes according to plan. That means a bad season in Houston could mean 110 losses.


The projection sees the Yankees and Red Sox both crushing the ball, but New York doing a much better job of run prevention. (Andy Pettite's return makes little difference in that equation.) It also projects that the Blue Jays' pitching is even more odious than the Orioles', but less putrid than both the Twins' and Royals'. The system has little reverence for the AL Central in general.


With the departure of Roy Oswalt and serious regression from Vance Worley, BP sees the Phillies' rotation as a three-man All-Star Team. It rates the Giants' pitching staff as 19 runs superior, though that may all be ballpark.


These standings suggest rock-'em, sock-'em, top-to-bottom slugfests in the NL East and West, and a three-way dogfight in the NL Central. Not so much in the AL, where the projections, like the Watergate reporters, follow the money.
b

For What It's Worth

As I was saying yesterday, it's fun and futile to make pre-season predictions. Here we go again.

This is the work of Baseball Prospectus, which uses a slightly different method than Fangraphs to make its projections. After projecting individual player performances and aggregating them by team, BP runs a million simulated seasons to give us a range of possibilities. So while the averages say the Pirates will win just 65 games, a sufficient number of things break right for them in about two percent of cases that would get them into the playoffs.

                       Div.     Wild    Total
NL East      W      L   Title %  Card %  Postseason %
Mets       92.2   69.8   39.1    11.8    50.9
Phillies   87.1   74.9   23.9    10.7    34.7
Braves     87.0   75.0   23.7    10.7    34.4
Nationals  78.5   83.5    9.5     5.5    15.0
Marlins    71.6   90.4    3.8     2.5     6.3

                        Div.     Wild    Total
NL Central   W      L   Title %  Card %  Postseason %
Cubs       95.3   66.7   52.0    10.0    62.0
Brewers    86.4   75.6   22.9    10.4    33.3
Cardinals  80.1   81.9   11.6     6.5    18.1
Reds       79.4   82.6   10.4     6.1    16.4
Astros     66.6   95.4    1.9     1.2     3.1
Pirates    64.5   97.5    1.3     0.8     2.1

                        Div.     Wild    Total
NL West      W      L   Title %  Card %  Postseason %
D'backs    91.6   70.4   46.7     6.9    53.6
Dodgers    83.7   78.3   22.5     6.4    28.9
Giants     77.9   84.1   12.1     4.1    16.3
Rockies    77.1   84.9   11.0     3.7    14.7
Padres     74.0   88.0    7.6     2.6    10.3

                        Div.     Wild    Total
AL East      W      L   Title %  Card %  Postseason %
Red Sox    98.0   64.0   38.9    24.0    62.0
Yankees    95.9   66.1   32.1    24.5    56.5
Rays       91.3   70.7   20.3    21.1    41.4
Blue Jays  81.2   80.8    6.4    10.1    16.5
Orioles    73.9   88.1    2.3     4.5     6.8

                        Div.     Wild    Total
AL Central   W      L   Title %  Card %  Postseason %
Indians    83.0   79.0   32.3     2.6    34.8
Twins      79.4   82.6   22.6     2.2    24.8
Tigers     78.3   83.7   20.2     2.1    22.2
Royals     74.7   87.3   13.8     1.5    15.3
White Sox  73.0   89.0   11.2     1.2    12.4

                        Div.     Wild    Total
AL West      W      L   Title %  Card %  Postseason %
Angels     84.3   77.7   42.1     2.2    44.3
Athletics  82.3   79.7   35.4     2.3    37.6
Rangers    71.9   90.1   12.5     1.0    13.5
Mariners   70.0   92.0   10.1     0.8    10.9

Again, this is as of Feb.19 when Manny and others remained unsigned. Should the Dodgers ink him, that would close the gap some with the Diamondbacks, who seem awfully well-regarded for a team lacking a power bat. BP has little faith in the White Sox, who last year vastly outperformed their projections.

This whole thing will be settled with two words: Play ball!