2018 Playoff Odds

[Rockies GM (Dan)]

These are early projections. I'm planning on releasing updated odds shortly before the season.

Methodology

My program reads the current Fantrax rosters and FanGraphs' Depth Charts Projections to project expected stats, with depth taken into account. Teams not projected to average enough ABs or IPs per week are heavily penalized in those areas. Then I run the season 100,000 times using a Monte Carlo algorithm with the actual schedule (which does heavily affect odds due to strength of schedule). Thanks to Eric for sharing the schedule early.


Results

NL West	Division%	Wild Card %	Total Playoff %	Average Wins
Arizona	0%	0%	0%	4.78
Colorado	67.47%	32.03%	99.5%	17.74
Los Angeles	0%	0.01%	0.01%	5.23
San Diego	3.37%	63.91%	67.27%	13.64
San Francisco	29.16%	67.05%	96.21%	16.22

NL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	11.61%	20.29%	31.9%	11.97
Cincinnati	0.03%	0.08%	0.11%	6.46
Milwaukee	9.85%	16.72%	26.57%	11.53
Pittsburgh	78.5%	9.95%	88.45%	15.03
St. Louis	0.01%	0.03%	0.04%	5.83

NL East	Division%	Wild Card %	Total Playoff %	Average Wins
Atlanta	0%	0%	0%	4.78
Miami	0%	0%	0%	2.27
New York	2.65%	22.23%	24.89%	11.68
Philadelphia	13.46%	53.78%	67.24%	13.85
Washington	83.89%	13.93%	97.81%	16.85

AL West	Division%	Wild Card %	Total Playoff %	Average Wins
Houston	19.06%	30.82%	49.88%	12.17
Los Angeles	0.72%	2.72%	3.44%	8.58
Oakland	80.15%	11.88%	92.02%	14.84
Seattle	0%	0%	0%	1.03
Texas	0.08%	0.35%	0.43%	6.92

AL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	26.82%	49.97%	76.79%	13.74
Cleveland	46.48%	40.98%	87.46%	14.55
Detroit	2.22%	17.25%	19.47%	10.53
Kansas City	22.37%	50.34%	72.71%	13.55
Minnesota	2.12%	16.51%	18.62%	10.54

AL East	Division%	Wild Card %	Total Playoff %	Average Wins
Baltimore	0.18%	13.83%	14.02%	10.13
Boston	0%	0%	0%	3.55
New York	0%	0.08%	0.08%	5.87
Tampa Bay	3.68%	61.56%	65.24%	13.19
Toronto	96.14%	3.71%	99.85%	17.95

Top 5 World Series Contenders

Toronto Blue Jays (won World Series 26.73% of simulated seasons)
Colorado Rockies (25.96%)
Washington Nationals (13.12%)
San Francisco Giants (10.03%)
Cleveland Indians (5.41%)

[Giants GM (Griffin)]

Looks like the BrewCrew gonna be out this year!!

[Diamondbacks GM (Ethan)]

I demand a recount

[Rockies GM (Dan)]

So, at some point in the past week, FanGraphs must have updated their projections. Automating the power rankings spat back some odds that were different from the ones above but consistent. I double checked to make sure that the automated download was working correctly and as expected. And they are, so I'll be updating these now, even a few days after I first posted.

In the meantime, check out the posted Power Rankings

[Rockies GM (Dan)]

NL West	Division%	Wild Card %	Total Playoff %	Average Wins
Arizona	0%	0%	0%	3.86
Colorado	75.69%	24.1%	99.8%	18.34
Los Angeles	0%	0%	0%	1.24
San Diego	1.63%	56.87%	58.5%	13.51
San Francisco	22.68%	73.37%	96.05%	16.34

NL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	7.67%	11.63%	19.3%	11.42
Cincinnati	0.01%	0.02%	0.03%	5.95
Milwaukee	26.33%	23.73%	50.06%	12.93
Pittsburgh	65.99%	14.99%	80.97%	14.61
St. Louis	0.01%	0.02%	0.03%	5.87

NL East	Division%	Wild Card %	Total Playoff %	Average Wins
Atlanta	0%	0%	0%	5.08
Miami	0%	0%	0%	4.98
New York	5.75%	34.75%	40.51%	12.73
Philadelphia	9.93%	47.87%	57.8%	13.67
Washington	84.32%	12.65%	96.97%	16.72

AL West	Division%	Wild Card %	Total Playoff %	Average Wins
Houston	27.17%	24.21%	51.39%	12.3
Los Angeles	2.67%	5.2%	7.87%	9.54
Oakland	69.91%	13.85%	83.76%	14.11
Seattle	0%	0%	0%	0.94
Texas	0.26%	0.52%	0.78%	7.6

AL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	50.29%	40.47%	90.77%	14.99
Cleveland	24.46%	51.18%	75.64%	13.68
Detroit	4.63%	30%	34.62%	11.64
Kansas City	20.3%	53.49%	73.78%	13.72
Minnesota	0.32%	5.04%	5.36%	9.09

AL East	Division%	Wild Card %	Total Playoff %	Average Wins
Baltimore	0.04%	6.1%	6.14%	9.2
Boston	0%	0%	0%	3.46
New York	0%	0.06%	0.06%	5.93
Tampa Bay	4.21%	65.73%	69.95%	13.5
Toronto	95.74%	4.14%	99.88%	18.05

[Deleted]

What? Only a 99.88% chance I make the playoffs??? Maybe i should think about blowing it up???

[Guardians GM (Matt)]

This makes more sense. I like my team, but I am also running out only 1 CI right now

[Royals GM (Mac - Retired)]

Mar 4, 2018 16:29:22 GMT -8 Guardians GM (Matt) said:

This makes more sense. I like my team, but I am also running out only 1 CI right now

You have a damn good team. Whether it’s with 1 CI or 2

[Padres GM (Amy)]

I prefer the first one

[Former Twins GM (Robin)]

Hmmm. I question either projections or your method, Dan.

.32% for the division and only 5% for the wild card? Projections must hate my team.

Unless you're just trolling me, in which case... Well played!

[Rangers GM (Victor)]

I didn’t see these were out. No wonder in a recent PM someone mentioned me in rebuilding mode. I see why now. 😭 Maybe I should start early and just blow it all up now. 🤔

[Former Twins GM (Robin)]

Don't trust the numbers! Unless Dan proves to be right at the end of the season. AL west podcast coming out on Monday.

[Rockies GM (Dan)]

The numbers didn't correctly predict the A's breakout 2017. Perhaps you're next, Victor?

[Pirates GM (Hollar)]

Wait... KC projects for more wins than the Indians, but the Indians have a higher playoff percentage? Presuming this is a high-variance roster sort of thing, but felt it was worth checking.

[Rockies GM (Dan)]

Mar 19, 2018 20:04:08 GMT -8 Pirates GM (Hollar) said:

Wait... KC projects for more wins than the Indians, but the Indians have a higher playoff percentage? Presuming this is a high-variance roster sort of thing, but felt it was worth checking.

In a future iteration, I'll save results as I sim them. But the problem with basing everything on current projections and rosters is that the program reflects changes instantly with trades/FA and injuries/changes in projection. Now the Indians are at 93.13% with 15.26 average wins and the Royals are 59.15% and 12.92.

I imagine it had to do with tiebreakers when teams had the same records, if I had to guess. If the Indians consistently beat the Royals head-to-head, and there were a significant number of simulations when the two teams finished tied, won by Cleveland, that could explain why that minor anomaly occurred

[Rockies GM (Dan)]

Week 2 Odds

NL West	Division%	Wild Card %	Total Playoff %	Average Wins
Arizona	0%	0%	0%	4.81
Colorado	79.34%	20.19%	99.54%	18.22
Los Angeles	0%	0%	0%	1.19
San Diego	1.96%	48.54%	50.5%	13.63
San Francisco	18.7%	72.68%	91.38%	15.88

NL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	2.83%	9.68%	12.51%	11.48
Cincinnati	0%	0%	0%	5.95
Milwaukee	29.42%	29.18%	58.6%	13.85
Pittsburgh	67.75%	17.98%	85.74%	15.35
St. Louis	0%	0%	0%	4.85

NL East	Division%	Wild Card %	Total Playoff %	Average Wins
Atlanta	0%	0%	0%	4.85
Miami	0%	0%	0%	4.94
New York	5.55%	33.16%	38.71%	13.14
Philadelphia	12.09%	53.91%	66%	14.41
Washington	82.35%	14.67%	97.03%	17.04

AL West	Division%	Wild Card %	Total Playoff %	Average Wins
Houston	24.46%	38.34%	62.8%	12.63
Los Angeles	2.46%	9.46%	11.92%	9.81
Oakland	71.88%	17.44%	89.33%	14.38
Seattle	0%	0%	0%	2.27
Texas	1.19%	5.14%	6.33%	9.26

AL Central	Division%	Wild Card %	Total Playoff %	Average Wins
Chicago	57.05%	35.05%	92.1%	14.89
Cleveland	31.07%	50.18%	81.25%	13.72
Detroit	3.86%	26.34%	30.2%	11.17
Kansas City	7.05%	38.58%	45.63%	12.01
Minnesota	0.96%	11.02%	11.98%	9.77

AL East	Division%	Wild Card %	Total Playoff %	Average Wins
Baltimore	0.41%	9.76%	10.17%	9.55
Boston	0%	0%	0%	2.63
New York	0%	0.14%	0.14%	6.17
Tampa Bay	7.32%	51.66%	58.98%	12.64
Toronto	92.26%	6.9%	99.16%	16.54

[Rangers GM (Victor)]

Wow.. I beat the league favorite in week 1 and the computer now doubles my wins?!? LOL. Thats why I dont put a ton of stock in these estimates. I got really lucky is all.

[Nationals GM (Preston - Old)]

Apr 11, 2018 8:21:08 GMT -8 Rangers GM (Victor) said:

Wow.. I beat the league favorite in week 1 and the computer now doubles my wins?!? LOL. Thats why I dont put a ton of stock in these estimates. I got really lucky is all.

Or your players may just be a bit better than thought and others are worse (and injuries).

[Pirates GM (Hollar)]

The computer upped your wins by 1.6, which is a whole lot less than doubling. I suspect it was predicting you for an L in week one, so there's one of them, which leaves .6 for us to find. David Price proving his arm is still attached to his body could be all the change in itself.

[Rockies GM (Dan)]

Also, projections change. If I used the exact same power rankings/projections each time, it wouldn't change as much. So quit complaining - it just means the projections like your players better than they did pre-season. That's a good thing


Even if I did use the same projections, using the algorithm, Toronto had a 91.3% chance at winning. Simulated out to 100,000 games, that means you were only gaining .087 average wins from that matchup. As Hollar said, with week 1 results added, those 1.6 extra wins include .913 more than you previously had. That's significant.