Introducing MARVIN (AL Edition)
For I don’t know how long I’ve wanted to make a season simulator for the MLB. I can’t tell you why, but I’ve wanted to make this for so long, and I finally bear-ed down and did it (kinda).
I will go into detail about my methodology at the end, but as a preface, I’ll provide a quick synopsis of how I did it:
First, I take the projected WAR of each team from my personal favorite WAR predictor, Steamer. I combined both pitching and hitting WAR into a team “score”. Then a home/road advantage, as well as a constant is added and I simulate each game based on these scores and the odds that come from them. I simulate these games 25000 times and take the median wins for each team as well as playoff and divisional berth odds.
And it's named…. MARVIN
It doesn't mean anything, but I like the name, maybe something like
Machine Automated R…. IDK
AL East:
New York Yankees:
Projected Record: 94.5 - 67.5
Win Division: 63.7–64.8%
Make Playoffs: 85.2–88.6%
Now for clarity, the record has a decimal because it is the median of the sims, and I saw projection systems like PECOTA left the tenth place, so for congruency, I left it in too. The range of division and playoff stems from a scaled confidence interval, so teams with close to 0% chance won't have a giant interval relative to their position.
Now, for each team, I’ll give a brief overview of my feelings on each.
The New York Yankees are clearly the best team in the AL. Their offense might be the best in baseball with a lethal back end to the bullpen. Their only true weaknesses are injuries and starting pitching, and those tend to overlap. Gerrit Cole is a bonafide stud, but a mix-match of Corey Kluber, Jameson Taillon, Jordan Montgomery, Deivi Garcia, Domingo German, and half a season of Luis Severino doesn’t offer much security. They also have the problem of a very strong AL East, including four potential playoff teams. But even after that, the Yankees should still both win the division and the AL, rather handily. If they don’t I will clown them.
Toronto Blue Jays:
Projected Record: 89.5–72.5
Win Division: 26.4–27.5%
Make Playoffs: 63.3–66.7%
Another note (the last one I promise): You’ll notice the Win Division % won’t exactly add up to 100%, it’s quite close, but this number includes tieing for the division lead, just without the Game 163 Simulation.
The Blue Jays went out and bought big with the signings of OF George Springer and SS Marcus Semien. They brought in a bunch of low-mid-level arms but will require more bolstering at the deadline. MARVIN has them as a team right at the forefront of the rather tense AL WC race, and I agree.
Boston Red Sox:
Projected Record: 83.5–78.5
Win Division: 6.1–7.2%
Make Playoffs: 25.4–28.8%
The Red Sox, (the best team ever btw), added numerous new pieces to a team that had a disastrous 2020, leaving them with the fourth overall pick in this year's draft (Please let Kumar or Leiter fall). They added pieces to the pen, versatile bats, and finished the eradication of the 2018 World Championship outfield. But the biggest pieces coming back are SPs Chris Sale and Eduardo Rodriguez. This leaves Boston with a rotation that suddenly is flowing with depth. They are still rebuilding per se, but don’t sleep on the 2021 Red Sox.
Tampa Bay Rays:
Projected Record: 82.8–79.2
Win Division: 4.9–6.0%
Make Playoffs: 22.2–25.6%
The Rays are an odd team, they are one of, if not the brightest teams in baseball, and I see expectation from the fans to compete after falling two games short of a title, but let’s face the facts here. They replaced Blake Snell, Charlie Morton, and Yonny Chirinos (injury) with Chris Archer, Colin McHugh, Rich Hill, and Michael Wacha. In a loaded division, that’s not enough to compete for a title. Maybe a wild card spot if they have some breakouts, but I am not confident in this team.
Baltimore Orioles:
Projected Record: 61.8–100.2
Win Division: 0.0–0.0% (oh boy)
Make Playoffs: 0.0–0.0% (oh no)
Oh, Baltimore what are we going to do with you. a 0% chance to make the playoffs is… sad, but don’t worry you’re not alone in this world. There are tanking support groups, maybe contact Houston? Is this rebuild of sadness a result of the albatross of a contract they laid upon Chris Davis? Is there some curse? At least you have Adley? Grayson Rodriguez!
AL Central:
Chicago White Sox:
Projected Record: 90.7–71.3
Win Division: 49.0–50.1%
Make Playoffs: 70.8–74.2%
The White Sox have turned into a contentious team as of late. Their roster is constructed like its 2000, but their player development staff is one of the best in baseball. They hit the ball on the ground, they pitch to contact, and they just play baseball. And from an efficiency and success point of view, it feels ancient, but from a fan’s point of view? I love it. This team has the pieces to compete in 2021 and should if all goes according to plan.
Minnesota Twins:
Projected Record: 90.3–71.7
Win Division: 45.0–46.1%
Make Playoffs: 68.5–71.9%
I like the Twins, I think they have a well-constructed team with a few holes that can be patched. But, it is the Twins. Do I really need to say much more? They’ll make the playoffs, but if they can’t win a single playoff game again… Something needs to change. Maybe if some of their rookies like Larnach, Kiriloff, and Duran can make a significant difference they’ll lose the ALDS 3–1?
Cleveland Indians:
Projected Record: 81.9–80.1
Win Division: 7.3–8.4%
Make Playoffs: 18.6–22.2%
The Indians made the trade of the offseason when they sent star SS Francisco Lindor and SP Carlos Carrasco to the New York Mets for a million shortstops. Gimenez looks good, but you have Tyler Freeman and 2B Aaron Bracho coming up, and clearly, Amed Rosario cannot play center. The pitching will be good because it’s Cleveland, but hopefully, they finally found an OFer who can hit with Eddie Rosario.
Kansas City Royals:
Projected Record: 74.1–87.9
Win Division: 0.1–0.9%
Make Playoffs: 0.1–2.5%
I absolutely love the Royals’ methodology the past couple of years. They’ve stockpiled pitching talent and are taking low-buy risks on some offensive youth like Andrew Benintendi while adding some veteran bridge pieces in the meantime like Carlos Santana. Their pipeline looks solid with the likes of Bobby Witt Jr., Erick Pena, and Asa Lacy. They’ll compete for the first half I’d imagine, but this team is not built to go 162, and they know that. Expect numerous trade deadline deals to build up the farm. During the writing of this, the Royals have signed (god help me) C Salavador Perez to a large deal, why they did is baffles me, and so I rescind my compliments.
Detroit Tigers:
Projected Record: 66.5–95.5
Win Division: 0.0–0.1%
Make Playoffs: 0.0–0.1%
The Tigers have pitching talent and Jeimer Candelario? Is that it? This team is loaded with aging free agents like Robbie Grossman, Wilson Ramos, Nomar Mazara, and Jonathan Schoop. They have the big three arms all coming up in Casey Mize, Matt Manning, and Tarik Skubal. Those three should be the staples of the rotation for the next five years. They also have number one overall pick Spencer Torkelson who could see some time this year. They won’t compete, but many of these guys are names to watch this year.
AL West:
Houston Astros:
Projected Record: 92.3–69.7
Win Division: 64.7–65.8%
Make Playoffs: 70.0–74.2%
I, and evidently MARVIN see the Astros as the clear second-best team in the AL and the biggest potential rivals for the Yankees in this seemingly one-horse league. Their lineup is as lethal as ever, even with the subtraction of George Springer, as a first five of Jose Altuve, Carlos Correa, Michael Brantley, Alex Bregman, and Yordan Alvarez should shake any opposing pitcher. I was, and to a degree still are worried about their rotation, but I expect big things out of LMJ this year, and Odorizzi should prove to be a solid addition for hopefully 120 innings of 3.5 ERA ball. Urquidy and Javier are both relative wild cards, but maybe we see Forrest Whitley finally emerge or a robo-JV down the stretch.
Los Angeles M̶i̶k̶e̶ ̶T̶r̶o̶u̶t̶’̶s̶ Angels:
Projected Record: 85.8–76.2
Win Division: 19.8–20.9%
Make Playoffs: 30.1–34.3%
Somehow, someway the Angels keep on chugging. This lineup feels potent, headlined by Mike Trout, Anthony Rendon, and Shohei Ohtani. But what really feels brand new is their pitching staff. There aren’t any notable new additions besides Jose Quintana and Raisel Iglesias, but Andrew Heaney, Griffin Canning, Dylan Bundy, and dare I say it… Shohei Ohtani could make big impacts this year for LA. Just for Mike Trout’s sake, I really hope this team sneaks into that second wild-card spot.
Oakland Athletics:
Projected Record: 85.0–77.0
Win Division: 16.2–17.3%
Make Playoffs: 25.4–29.6%
The A’s are an odd team, they always are. They have a plethora of talent with the two Matts, Ramon Laureano, and Sean Murphy. They also will probably lead the Majors in team defense, but is that enough to aid this troubled pitching staff? Are we really going to rely on Mike Fiers and Sean Manaea (VAA Legend) for 300 innings? This is a team that, like the Rays could easily outperform and contend for the division title, but I certainly have my doubts.
Seattle Mariners:
Projected Record: 75.8–86.2
Win Division: 0.4–1.5%
Make Playoffs: 0.1–3.8%
The Seattle Mariners are always a team that I’m gonna root for, its a fun organization usually loaded with star talent, but lacking pieces around them. Think Ichiro, Ken Griffey Jr, Felix Hernandez. The 2021 Seattle Mariners don't have that. They have stockpiled average players which has lead to little to no success. 2021 will be more of the same with a pretty good pitching corps, and solid pieces in Kyle Seager and Mitch Haniger, but nothing to be excited about. But then, you look into their farm and my goodness is it LOADED. Headlined by names like Jarred Kelenic, Julio Rodriguez, George Kirby, Emerson Hancock, Noevli Marte, Logan Gilbert this Mariners pipeline can make your mouth drool. So forget about this season M’s fans and look to the future.
Texas Rangers:
Projected Record: 57.3–104.7
Win Division: 0.0–0.0%
Make Playoffs: 0.0–0.0% (yikes)
Oh Texas, oh no Texas. This team does not have much to be excited about. They have some pretty solid VERY young pieces in Maximo Acosta but in the Majors, it's the Joey Gallo show. He is their franchise currently, and rumblings say he could be moved this year too. In a pretty well-rounded division, this team is going to have a rough, rough year, potentially landing them the number one pick to take a guy like Connor Prielipp. Maybe grab a guy like Lawlar or Mayer in this year's draft to continue their long-term rebuild.
So that completes the AL, here’s what would be the final AL Standings according to MARVIN:
- New York Yankees — AL East Winner
- Houston Astros — AL West Winner
- Chicago White Sox — AL Central Winner
- Minnesota Twins — Wild Card #1
- Toronto Blue Jays — Wild Card #2
- Los Angeles Angels
- Oakland Athletics
- Boston Red Sox
- Tampa Bay Rays
- Cleveland Indians
- Seattle Mariners
- Kansas City Royals
- Detroit Tigers
- Baltimore Orioles
- Texas Rangers
So that's the AL, I’ll make another blog post for the NL in the coming days, but for those that are curious, I’ll describe my metholody:
Step one was to find a schedule for 2021, I forget what site I found it on, but it each teams schedule as a column in a csv that looks like:
Note: I did all of this work in R, besides changing names like “White Sox” to “White.Sox” in Excel because it was just easier.
I then subsetted each team’s column and created a home and away team column to have a clear schedule, and merged them all together, while removing duplicates. This gives me a whole 2021 schedule that encapsulates each Home and Away team on each date.
Then, as described as above I created a team score using combined projected WAR from Steamer, with a Home/Road factor and a constant to even out the playing field, as without it, teams like Texas would win 53–55 games.
I then left_join (basically a VLookup) the home/road scores into the schedule and perform a Monte Carlo Simulation using odds based on the scores. I perform one sim, and then use the replicate function in R to re-sim it 25,000 times.
This leaves me with a VERY large dataset of ~60,000,000 observations. Then with a little dplyr magic I subset the data into indivudual sims. This has led to some fun results, so here’s the AL high and low in wins for the 25,000 sims:
AL High: Yankees — 124 Wins
AL Low: Orioles and Rangers— 35 Wins
35 Wins, dear God is that low, but I guess 25,000 times, you’re bound for something crazy.
I then add whether they win the division and/or make the playoffs* and sum it all into what you’ve seen above
*Ok so, there’s a bit of a problem with this, as I cannot predict G163, and when division winners are tied it adds an extra Wild Card Spot. This is something I am working on, but for now, I added CIs with the bigger value being the estimated playoff percentage I calculated. So I strongly encourage you to use the range there.
Thanks for reading! This is my first article on here, and hopefully more to come. Just want to publish some of the work I’ve done lately. This should be a 3-Part Series, with NL being the next, and conculsions in the third.
Follow me on Instagram:
@jacksondels
@bsbl.stats
Or on Twitter:
@jacksondels2
And Make sure to listen to my podcast:
Deep Drive Pod Available on most major streaming Platforms!
