That's Why They Play The Games

Some guy at a Yankee blog (gasp!) did an interesting thing with all of the projections floating around. Instead of just using runs created predictions and the Pythagorean Wins formula, he actually simulated the games using the Diamond Mind program. I would gladly have done this, but my Baseball Mogul simulator is far less advanced and didn't give very realistic result. The simulator was run 1,000 times each for six different projection systems; Chone, Diamond Mind, Hardball Times, Cairo, PECOTA, and ZiPS. This is about as close as you can get to predicting the standings, I think.

The average over all six, and (my projections)

AL East:
NYY - 95 (95) [Yaye, I got one!]
BOS - 92 (94)
TOR - 85 (83)
TBR - 82 (79) [Those extra O's win have to come from somewhere.]
BAL - 67 (72) [I'll admit that emotion probably added a couple wins here.]

AL Central:
DET - 91 (89)
CLE - 89 (90)
MIN - 75 (74)
CHW - 74 (75)
KCR - 73 (72)

AL West:
LAA - 88 (89)
OAK - 80 (77)
SEA - 77 (77) [HA!]
TEX - 74 (73)

NL East:
NYM - 95 (93) [They aren't the best team in baseball - 95 wins seems a bit too high with the injury risks.]
ATL - 87 (88)
PHI - 86 (85)
WAS - 70 (71)
FLA - 68 (73) [I'm probably too optimistic here, but we'll see.]

NL Central:
CHC - 88 (89)
MIL - 85 (87)
STL - 78 (73) [If they win 80 then Pujols deserves the MVP.]
CIN - 77 (77)
HOU - 75 (73)
PIT - 70 (71)

NL West:
LAD - 85 (86)
SDP - 84 (85)
ARI - 84 (86)
COL - 83 (83)
SFG - 73 (71)

He even breaks down how often each team won their division. The Mets were most dominate, winning the NL East 70% of the team. The NL West looks like the most competitive division (minus the Giants) with each team winning between 20% and 28% of the time. The Orioles were the only team that never won its division, and only managed 6 Wild Cards (1 in 1,000 is about that odds I'd put on the O's playing in October). The most positive projection had Seattle winning 83 games (well, 82.6). No system had Tampa Bay winning less than 80 games - that is a pretty amazing turn-around.

Will multiple computers/analysts do better then me (with a sprinkling of PECOTA - hey, I'll still take credit for picking a system to augment my information)? I'll review at the end of the year.