First, I needed to determine everyone's playing time. For this, I used the depth charts (may be behind a premium content firewall) over at Baseball Prospectus as a guide and adjusted where I felt necessary. Using those percentages and the average number of plate appearances per team last season I arrived at everyone's projected plate appearances. I used a similar idea for the pitchers' innings, again tweaking where I felt best.
Once I had everyone's playing time set, I turned my attention to the rate stats. For the hitters that meant walk rates, strikeout rates, rates of extra base hits, etc. Again, the process was very similar for the pitchers. These rates were derived from a loose averaging of the projections available at Fangraphs and the players career averages. Minor adjustments were made for those with very small major league sample sizes. From there, simple multiplication spit out their projections.
For the W-L record, I simply plugged the projected runs scored and runs against into the Pythagorean win theorem formula and out popped the projected record. Last year our Pythagorean record was 62-100, meaning in theory we under achieved by six wins.
Obviously any major changes in roster make up (trades, injuries, etc) could change things dramatically.
I feel like it's a little optimistic, but I can't find any major errors in the formulas or calculations.
Screw it, I'm running with it!
So, how does the offense compare to last season? Well, that would give us 91 more runs scored, 32 more home runs, 66 more walks, 54 fewer strikeouts, and an OPS .023 higher than last year. To put it another way, these projections would give us an essentially league average offense. At first glance only Carlos Lee's numbers look especially optimistic to me.
How about the pitchers?
Nothing here really stands out as unlikely to me. Maybe Happ's numbers are too good, but that's actually very close to his career averages. It's very possible one or more relievers is particularly ineffective this year, but who? And if so, they wouldn't get many innings anyway. So again, how does that stack up to last year's squad? We're looking at a team ERA just 0.17 better as they only give up 26 fewer runs that last year. Walks are up by 41 and strikeouts are down by 12. This squad does allow 17 fewer homers, according to my projections. To again frame these projections against last year's competition a 4.34 ERA would have only put us ahead of one other team.
So, assuming my math is correct, 706 runs scored and 770 runs against would give us a projected team record of..............
Yes, I question my own sanity for actually publishing that, but there it is. Led by a potentially much improved offense it is entirely possible that the Astros will finish not last in 2012.