Both Strat-o-Matic and Baseball Reference are running simulations of the yet to start 2020 season. These can be fun to follow, but a single simulation can give a skewed view of the real strength of a team. If we combine the two, we’ll get a slightly better idea of the strength of teams. Note that what we’d really need is thousands of simulations. That’s how Baseball Prospectus comes up with their playoff probabilities.

This spreadsheet combines the two simulations through games of May 5th. Some team, like the Yankees, Indians, Braves, and Dodgers are excellent in both simulations. Some teams, like the White Sox, Mariners, Cubs, and Padres are great in one and terrible in the other. The Giants and Tigers are just bad no matter the simulation.

Going by the combined winning percentage, the Yankees would be the best team in the majors at .681, with the Dodgers second at .657. No surprise there. Cleveland is dominating what should be a very good Twins team, as the Indians lead the AL Central with a .605 WPct to Minnesota’s .542. Athletics lead the Astros .592 to .583 in the the tightest 1-2 race in the AL.

The Braves have played four more games that the Nationals so Washington has games in hand to make up the difference as Atlanta leads .605 to .588. The rest of the division is under .500. St. Louis easily controls the NL Central, in what I thought would be a very tight four team race.

Detroit is the only really terrible team, with a combined winning percentage of .347. That would be good for 56 wins over the entire season. Texas and San Francisco are the only other two teams playing below .400 ball.

538 Looks at how a shortened season and an expanded playoff format changes the probabilities of making the playoffs and winning the World Series. I found this part fascinating:

An interesting change from a normal season to a much different one was that the greatest gainers included a number of clubs that have young and promising cores (the Jays, Rays and Padres) and clubs that made significant offseason additions — like the White Sox, who signed Yasmani Grandal, and the Angels, who gave Mike Trout the best teammate he’s ever had in Anthony Rendon. We wondered in December if such signings would make baseball more competitive; an expanded playoff field would perhaps benefit those teams that tried to improve this winter and serve as a disincentive to tanking.

FiveThirtyEight.com

I would expect young core teams to do better in a short season format because other teams have less time to adjust to the weaknesses of those players. In addition, the younger players would not need to adjust to a longer season than they played in the minor leagues. I assume that is not built into the simulation, so these young core teams might do even better than indicated here.

With no baseball, there are a number of active simulations of the season. Strat-O-Matic is offering one such virtual season, and the Red Sox pounded the Blue Jays in the opening series.

At the heart of the simulation lies the batter-pitcher match-up. Strat-O-Matic combines batter and pitcher events in a straight forward way simply adding the two probability distributions together. So if a batter hits a home run every 0.064 PA, and a pitcher allows a home run every 0.032 PA, then when facing each other, the home run rate would be 0.048 per PA.

That is not the only way to model the match-up, however. The Log5 method combines the probability distribution for the batter with the probability distribution for the pitcher in the context of the league average to produce the new distribution. For example, the same batter and pitcher in a league with a 0.034 HR/PA rate would expect a home run rate of 0.060 according to Log5. A really good home run hitter facing a pitcher near league average should still be a really good home run hitter.

So a nice way to simulate the 2020 season might be to start with projections for players, and calculate the Log5 probabilities for events. This spreadsheet contains Musings Marcels rates for batters and pitchers for the 2020 season, and a third sheet that does the Log5 calculation, and displays the results both as a rate and per 600 PA. Feel free to make a copy of the spreadsheet and play with match-ups. Copy a batter line and a pitcher line to the Log5 sheet and see how one would hit against the other.

The current sheet shows Mike Trout against Gerrit Cole. These are two three-true outcome players, and the calculation indicates that 361 of the 600 PA would end in a home run, strikeout, walk, or hit by pitch. Trout doesn’t get many hits against the Cole, posting just a .238 BA. He does walk a ton, however, and over half his hits go for extra bases. That gives him a .353 OBP and a .506 slugging percentage. Trout would in fact be a very effective hitter against Cole, just not the greatest of all time.

This version of Log5 is not very dynamic, however. Strat-O-Matic offers different distributions for every batter and pitcher based on the handedness of the opponent. It builds in situational hitting and park effects. There are special probability charts for various strategies.

What would be great is a way of building all that into a probability distribution without writing special code to handle strategy or game situations. A calculator that knows to return a zero probability of a sacrifice fly with two out or no one on base. A probability distribution that is much more likely to sacrifice with a poor hitter at the plate, or in close and late situations. A function that accounts for parks, and handedness and men on base.

Some experiments will be on the way soon.

A shorter season could be a boon to some middle of the road teams. Fewer games gives luck more of a chance to shape a pennant race.

The 95% confidence interval provides the range of wins (or losses) expected for a series of binary events that covers the middle 95% of a probability distribution. If a team is considered a .500 team, that is they are expected to win and lose in equal amounts, we can calculate the range of wins that would fit into this interval.

For a .500 team playing a 162 game schedule, we would expect the team win between 69 and 93 games in 95% of their seasons. In terms of winning percentage, that is .426 to .574. In a 100 game season, the 95% confidence interval ranges from 40 wins to 60 wins, or .400 to .600. So it’s easier for a .500 team to post a division winning record in a short season. We’ve seen examples over the years of a team winning the first 100 games easily but winding up in second place.

You can also see this by playing with the Baseball Musings Season Simulator. Play with the number of teams and the number of games. Change it from assigning all teams a .500 winning percentage to assigning them random winning percentages. It’s a nice way of visualizing how evenly matched teams can finish with very different winning percentages, and how teams with lower intrinsic winning percentage sometimes finish on top of the division.

The Book Blog discusses what leagues would look like if every team were evenly matched. A few years ago, I wrote this season simulator to show this very thing. You can set the simulator to use random intrinsic winning percentages, or set all teams to .500.

I received a copy of MLB 2K10 for the Wii in the mail recently. I’ve tried to play it, and it takes some getting used to. I’m not a gamer; the Wii is for my daughter, I use it to work out with the Wii fit, so I’m not used to using all the controls. MLB 2K uses the full capabilities of this game console, including the nunchuck controller.

On offense, you can swing, and control if you want to try to hit a grounder, hit for power, check a swing, or bunt. When in pitching mode, you can select a pitch, throw and defend. Throwing the pitch is easy, although timing is everything. The game provides a target that shrinks, and you want to release when it shrinks to a green circle. Batting is tougher, as you need to time the pitch, just like in real life.

I had problems after the ball was put into play. On defense, you need to move the fielders with the nunchuck joy stick. That took some getting used to. You also need to direct the throws to the right base with the plus control pad. I kept hitting the lower leg of the controller to throw the ball in to the infield, which sent it all the way to home plate. It would be nice if there was a button for default throw, that threw to first on a grounder, or second on a double play, or second on a ball to the outfield.

I also had problems advancing runners. On what was an obvious double, the runner on first stopped at second. When I later figured out to hit the Z key to advance runners, they didn’t stop at the obvious base. You can make them runner faster and slide.

There are also options you can set make things more automatic, but I didn’t explore those.

All thirty teams are available, and you can create players as well. You can also let the game play itself to simulate seasons. Operation Sports did just that, and their top fantasy picks are here.

The graphics are very good. Most of the players look like their real-life counterparts, and the stadiums have the feel of the actual ballparks. The colors are rich and vibrant.

There was a huge learning curve for me to play this game well. Someone more used to this type of play will undoubtedly pick it up faster. My daughter enjoyed it. As a software developer, I think the game should be more obvious. If you are going to play, take the time to work through the tutorials and practice pitching, hitting and fielding. It will be a lot more fun to play that way than just jumping right in.

Jeremy Greenhouse at Baseball Analysts tries to figure out the system Tampa Bay used to decide Willy Aybar could hit Zach Greinke, even though they never faced each other.

A few years ago I helped a team out writing software for a idea that they had to compare batters to pitchers similar to the one they were going to face that day. It was a way to expand the sample size of batter vs. pitcher. This sounds similar, but with better technology.

I’m taking part in a Strat-O-Matic simulation of the 1986 season at SportingNews.com this winter. You can follow the fun here.

I play in the Postwar Baseball League. The league is getting ready to expand and is looking for new managers, as well as a replacement manager:

The Postwar Baseball League (PWBL) is looking for a general manager
to take over the Seattle Pilots starting in 1966.
Founded before the start of the 1946 season, the PWBL is a play-by-email league that has been in existence for six calendar years. Seven of its original twelve owners are still with the league, testifying to its stability. (One of the original owners describes the PWBL as a “terrific league–best non-FTF league I’ve ever been involved with by a long margin. . . .”) Now consisting of sixteen teams, the PWBL is populated by an exceptional group of GM’s, who have a wide and varied knowledge of baseball and its history (and who get along exceptionally well).
Because the PWBL has played twenty complete seasons, it has a rich history of its own, especially in the accumulated statistics. Last year (1965) Robin Roberts of the New York Mets became the second PWBL player to reach 300 wins when he defeated the San Diego Padres in a game in which teammate Willie Mays hit three homers, giving him 67 for the season and 498 for his career. (Warren Spahn, who pitched his last season for St. Louis Cardinals in ’65, was the league’s first 300-game winner. He retired with 328 victories.) Eddie Mathews is the career leader in homeruns with 546.
The league is remarkably competitive. So competitive, in fact, that no team has ever won back-to-back Mulcahy Cups–named after Hugh Mulcahy, the first major league baseball player to enlist in the armed forces after Pearl Harbor. Twelve different teams have captured the championship in the league’s history.
League games are played in round-robin fashion by those GM’s who volunteer. Other GM’s prefer to stick to the front office, drafting and trading and keeping track of their players. Because of its remarkable stability and quick pace–we finish at least three seasons per calendar year–the league “runs itself,” and GM’s can fluctuate between active involvement and a more detached interest as their schedules permit.
The Seattle Pilots enter the 1964 season with a strong nucleus, including its Gold Glove doubleplay combination of Bill Mazeroski and Luis Aparicio. Third baseman Tony Perez and slugging first sacker Boog Powell round out its infield. Tony Oliva is the team’s leading hitter, while Ken Holtzman and Fritz Peterson are Seattle’s top mound prospects. After winning their division in 1962, the Pilots have fallen under .500 and need a new winning attitude.
Looking ahead to the 1969 season, about one calendar year from now, the league hopes to expand by two or even four teams, if we can find responsible, committed GM’s willing to build teams from scratch. Our previous expansion, from 12 to 16 teams at the turn of the decade, was a successful one, with new teams quickly becoming competitive. By keeping the ratio of PWBL teams to real-life teams high (currently 80%), we have avoided “All-Star-itis,” maintaining player usage at realistic levels by permitting 110% usage for most players, with unlimited usage allowed for mediocre “replacement-level” players.
Ideally, we would like expansion team GM’s to join the league far enough in advance that they can become fully immersed in the PWBL even before they begin building their teams, participating in league decision-making, voting on current-season awards, perhaps even lending a hand with game play.
If you are interested in Sixties baseball and think you might like to join the PWBL, please drop a line to David Myers (dgmyers@tamu.edu), the league’s commissioner.

I really enjoy seeing the boxscores cross every day. I’m not one of the more active members, but the drafts are challenging, as you balance your needs for the season vs. your needs for the future. Check it out.

Larry Mahnken simulated the 2006 season using three different sets of projections. The Twins do surprisingly well. I guess good pitching can take you a long way.
Baseball Musings is holding a pledge drive during March. Click here for details.

Congratulations to the SOMBILLA on their 25th anniversary. Sorry I could not make the party today. 🙁
You can read the history of the Pinto Hatchbacks’ best SOMBILLA season here.

Steve Treder offers up his memories of Strat-o-Matic baseball at the Hardball Times. I started playing the game in college, where I was a proud member of the Thirteenth Floor Strat-o-Matic League. Later, I joined the Sombilla, and participated in two seven-game World Series (I lost both, you can read the summaries here and here). I also played in a play-by-mail league for a couple of seasons. I love the game, and many of my best stories are a result of Strat-o.