clock menu more-arrow no yes mobile

Filed under:

survey says: cards by 10 games

in late january the Replacement Level Yankees Blog simulated the entire 2006 season 100 times, using the Diamond Mind computer game. the cards averaged 96 wins in that exercise.

today RLYB is back with a more thorough set of simulations. he ran the 2006 season 3000 times, using three different sets of player projection data. it was a similar exercise, you'll recall, that two springs ago projected the cards to run away with the nl central --- at a time when most pundits were picking them for 3d place. and last spring this projection had the cards winning handily again. so for 2006?

w l runs runs
won div
stl 94 68 771 649 76 85
chi 84 78 747 715 13 31
mil 82 80 738 727 7 17
pit 78 84 716 743 3 8
hou 77 85 710 749 2 7
cin 72 90 760 846 0.3 1

no other NL team projects to win more than 89 games, which says a lot about the state of the national league.

i'll look at these more closely later and probably post about them in the morning. but the thing that shines through: pitching. look at those runs allowed --- 65 better than any other team in the division. under all three sets of projection data, Diamond Mind has the cards allowing the fewest runs in the majors ---- which is bizarre, because two of those data sets (ZIPS and PECOTA) forecast large era increases for all of the stl pitchers.

take this stuff for what it is worth. the cards still open the year with an 0-0 record and a whole league trying to shoot them down. realize, too, that even as glowing as these forecasts are, they give the cards a one in four chance to finish out of first place --- about as often as yadi molina got a hit last year --- and a one in six chance to miss the playoffs.

and for all we know, the RLYB's simulations assumed 24 starts for anthony reyes and 10 for sid ponson. i guess all in all you'd rather that the computer loved your team than hated it . . . . but either way it don't mean a whole lot in the standings.