14 March 2018

How the Orioles win the AL East in 2018

The Orioles could win the the AL East this year.  They could.  I am not kidding.  Such an outcome is one I would not expect.  It is an outcome that I would find shocking, but there is a path.  If you pay attention to projection models incorporated into season models (pick one, any one, they all are about as accurate as any other), then you may know the club is generally seen as a mid 70s win team.  Such a club matches or exceeds 90 wins almost 10% of the time.  In other words, it is not uncommon.

However, 90 wins is nice, but not a guaranty of much of anything with an AL East that looks to have two monsters in the Yankees and Red Sox.  What would be a safe number would be around 95 wins.  That leads us to this idea, what would a scenario look like where the Orioles win 95 games.  That may sound unrealistic, but projection models disagree.  There is a chance.  That chance would require an even push by several Orioles in a more positive direction or a couple big breakouts to shoulder the load.  In this post, I will take the slight push approach.  Chance shown for each player is the probability for a player doing at least that well, so they could actually do better.

Catcher
Projection Change
PA WAR WAR WAR diff
C Chance Sisco 256 0.8 0.8 0
Caleb Joseph 320 0.8 1.6 0.8
The Rest 64 0.1 0.1 0

Chance: 50%
Fangraphs sees both Joseph and Sisco as competent catchers.  The Orioles off season long search to displace Sisco from the roster is not coming up as a good idea in the models.  Here, though, I decided not to push anyone's performance in the positive direction.  Everyone sticks to their 50th percentile projection.  The group does improve by 0.8 WAR.  Why?  Well, they do not consider pitch framing in their estimates and Joseph is routinely an excellent pitch framer, so the gain here is simply recognizing Joseph's total value.

First Base
Projection Change
PA WAR WAR WAR diff
1B Chris Davis 476 1.4 1.9 0.5
Trey Mancini 189 0.3 0.5 0.2
Mark Trumbo 35 0.1 0.2 0.1

Chance: 40% (Davis), 40% (Mancini), 50% (Trumbo)
I nudged each player here.  Davis' 0.5 WAR improvement comes from the idea that perhaps now that his core is healthy again that his first step comes back, improving his defensive play.  Mancini's bump of 0.2 is based on the assumption that last year's level of offensive performance is real.  Trumbo's bump comes from him bouncing back and performing slightly better at first base.  If you are counting, we are now at 78.5 wins.

Second base
Projection Change
PA WAR WAR WAR diff
2B Jonathan Schoop 644 3 5 2
Engelb Vielma 56 -0.1 -0.1 0
Chance: 10%
The model seems to think that Schoop is in for a regression to his career line.  We are going to say last year's performance was not only real, but also add another win for expected modest growth in talent.

Shortstop
Projection Change
PA WAR WAR WAR diff
SS Manny Machado 644 5.3 7.3 2
Engelb Vielma 56 -0.1 -0.1 0

Chance: 20%
For Machado, we are throwing in there a 7.3 WAR which would just barely be the highest WAR of his career.  This assumption is the assumption that Machado is actually one of the best players in baseball, which may not well be true.  In this scenario, he is.  And, we are up to 82.5 wins.

Third Base
Projection Change
PA WAR WAR WAR diff
3B Tim Beckham 644 1.9 3.4 1.5
Danny Valencia 35 0.1 0.1 0
Engelb Vielma 21 -0.1 -0.1 0

Chance: 20%
The projection for Beckham is that his offensive talent level will regress.  Instead, we decided that the average of his entire 2017 is his true talent level.  So, that bumps him up 1.5 wins.

Left Field
Projection Change
PA WAR WAR WAR diff
LF Trey Mancini 385 1 1.6 0.6
Joey Rickard 140 -0.1 0 0.1
Austin Hays 105 0.1 0.2 0.1
Anthony Santander 70 0.1 0.1 0.0

Chance: 40% (Mancini), 70% (Hays) 
In left field, the benefit comes from the aforementioned Trey Mancini replicating his 2017 season. In addition, the assumption here is that Rickard will reduce his playing time and Hays will see more with a plus arm.  The club is now at 84.8 wins.

Center Field
Projection Change
PA WAR WAR WAR diff
CF Adam Jones 630 2 2.6 0.6
Joey Rickard 35 0 0 0
Austin Hays 35 0.1 0.1 0

Chance: 30%
Jones' isn't dead, yet. He feels happy.  Assuming his play in centerfield remains adequate and his bat leans more toward his second half high without actually duplicating it, we could potentially see something like his 2015 season (without the defense).  That would add a little bit to the pot.

Right Field
Projection Change
PA WAR WAR WAR diff
RF Colby Rasmus 410 1.2 2.2 1
Danny Valencia 210 0.1 0.1 0
Joey Rickard 35 0 0 0
Austin Hays 21 0 1 1
Anthony Santander 14 0 0 0

Chance: 30% (Rasmus), 50% (Hays)
Projection models still like Rasmus' potential to reach back and deliver something similar to his 2015 season with the Astros.  His offensive demolition last year with the Rays is a bit more unlikely.  The rest of the pickup here is due to Hays spelling Valencia.  If Hays meets his mark and has a plus arm, he could deliver a win over 200 plus plate appearances.

Designated Hitter
Projection Change
PA WAR WAR WAR diff
DH Mark Trumbo 539 0.6 1.6 1
Chris Davis 112 0.2 0.2 0
Trey Mancini 28 0 0 0
Danny Valencia 21 0 0 0

Chance: 10% (Trumbo)
The model thinks there is about a 10% chance that Trumbo can perform a smidge below his 2016 career year.  This does not consider that he appears to not do well when playing as designated hitter.  Anyway, this gets us to 86.5 wins.

Starting Pitchers
Projection Change
IP WAR WAR WAR diff
SP Kevin Gausman 177 2.6 3.1 0.5
Dylan Bundy 177 2.1 3.1 1
Andrew Cashner 138 0.7 2.2 1.5
Chris Tillman 122 0.4 1.9 1.5
Gabriel Ynoa 111 0.4 2.4 2
The Rest 215 1 1 0

Chance: 15% (Gausman), 20% (Bundy), 10% (Cashner), 5% (Tillman)
A number of assumptions are being made here.  One, Gausman is assumed that he will match his career high performance.  Two, Bundy is assumed to meet last year's performance and add a half a win.  Three, Cashner is pushed to producing a normal year for himself.  Four, Tillman slides back to his 2012-2016 performances.  Five, well, the Orioles sign Alex Cobb.

Relief Pitchers
Projection Change
IP WAR WAR WAR diff
RP Brad Brach 65 1.2 1.7 0.5
Darren O'Day 65 1 1 0
Mychal Givens 65 0.9 1.4 0.5
Richard Bleier 45 0.2 0.2 0
Zach Britton 40 0.5 1.5 1
The Rest 240 0.1 0.1 0

Chance: 10% (Brach), 20% (Givens), 15% (Britton)
The bullpen hopes lie in the Orioles hitting their stride on their fearsome three: Brach, Givens, and Britton.  The likelihood of this seems about as good (or bad) as the starting pitching predicament.

Conclusion
And there you have it. If about 15-20 players hit their upper quartile projection, the Orioles will be expected to win about 95 of their games. That is the optimistic push of the finger. Of course, we have been talking about the chances of all of this happening. That would be less than a 0.000000001 chance of 95 games being won.

6 comments:

Anonymous said...

I was afraid, for a second, that you were saying Ynoa could be a 2 WAR pitcher, LOL. As you know from my last comment, I am all over this. I mentioned improvements you state from Mancini, Rasmus, Trumbo, and signing Cobb. I like to think Manny will outperform his 1st half from last year significantly. I also like the idea that Caleb as the primary catcher might help the pitching staff some. The rest, well, there's a lot of pie in the sky. I'm more with Matt that the pitching likely to top out at "decent". Your projection makes them "great". I can't imagine the bullpen getting more than 4.5 WAR, which would be really good in the scheme of things. And the SP, I hope Cahsner can be that good and that Bundy/Gausman can continue to improve, but I think a more realistic assessment is that one of Bundy/Gausman improves and either Cashner or Tillman but not both can be better than expected. The big add would be Cobb. Overall, I think there is considerable hope for WC-competitive team with 85-88 wins if things go well. Maybe some of Hays/Sisco/Santander can really do something. Maybe the O's acquire Cave and trade Trumbo.....

Unknown said...

How I win the Powerball tonight...

Unknown said...
This comment has been removed by the author.
Unknown said...

I think your odds in the final paragraph are way off. Comes down to something I think you mentioned at the beginning of the article - the odds involved are the odds of each guy performing at least as well as listed. There are a few guys for whom many of the instances of reaching pojections may actually also involve passing projections - Davis would be one who, if healthy enough to reach this projection, could also easily cruise past by another win or 2. And that eases the load on somebody else. So yeah, I think your odds are probably an order of magnitude or 2 low. Still pretty bad odds, but not quite as depressing as you're making it.

Jon Shepherd said...

Distribution matters as well as context issues. For instance the odds of a player exceeding that mark in value compared to another significantly undercutting the 50th percentile.

One could ignore distribution and simply use a generic sigma solution at the team, which would put it around 0.00003%. I think that ignores phenomena related to high and low end win distribution.

Pip said...

Is there a reasonable possibility that Tillman doesn't make the team regardless of performance?
If so, Wright would be 4 and a season of TBAs would ensue.