So here we stand at the All Star Break, the Astros a level 44-44. Eighty-eight games the Astros find themselves tied for third in the NL Central with the Chicago Cubs and a scant 3.5 games behind the St. Louis Cardinals. I doubt that there are many out there who would have seen this coming in March, but nonetheless, the Astros find themselves in the middle of a division race in which there is no favorite. All these teams have battled underperformance and injury, and none seems to be the complete package.
The question, obviously, is what does this portend for the Astros chances to find themselves in the top of the NL Central or the NL Wildcard (where they are 5.5 games back with four teams between them and the Giants).
The last time we did this, June 16th, the Astros were 4.5 Games out of the Central and 4 out of the wild card, and things have gotten better:
| Model | Odds |
| BPro Unadjusted | 5.25729% |
| BPro Schedule Adjusted | 10.43277% |
| BPro PECOTA ADJUSTED | 6.07269% |
| Cool Standings | 11% |
|
SCS Weighted |
4.4% |
| SCS 50/50 | 17.7% |
The odds kind of bounce around from model to model, but in every instance, there is improvement:
| Model | Delta |
| BPro Unadjusted | o.531% |
| BPro Schedule Adjusted | 2.17459% |
| BPro PECOTA ADJUSTED | 1.97474% |
| Cool Standings | 5.5% |
|
SCS Weighted |
1.9% |
| SCS 50/50 | 5.4% |
No matter how we decided to look at it, the Astros have managed to keep themselves in a position to make things interesting for us in the second half. A lot of emphasis has been placed on the first month after the break, and perhaps it will be make or break, but for now the Astros sit on very manageable odds at providing us with some second half excitement. I, for one, can't ask for anything more than that from this team.
After the jump there are explanations for what the various models are doing for anyone who is new to my postseason odds posts.
The BPro Unadjusted model uses a standard monte carlo simulation and runs it a million times. When the Astros have enough wins to make the playoffs, they are credited with with a 1, and when they're not, they get a zero. While I have placed a percentage in front of it the BPro numbers, they could be read as the Astros made the playoffs x (remove the decimal point and percentage) times over 1,000,000.
The BPro Schedule Adjusted model utilizes Nate Silver's (now primarily of 538.com) interpretation of the popular soccer odds making model, ELO. It takes into account the strength of teams schedule by looking at their performances between teams to weight the monte carlo. PECOTA adjusted just means it uses the PECOTA prediction of a team's talent to weight its probability for success later. Since we all vociferously disagreed with PECOTA's assessment of the team's true talent level, I say screw'em.
Both Cool Standings and Sports Clubs Stats have excellent explanations at varying degrees of technical proficiencies, and I'll leave the truly curious reader to delve into them. Just know that the SCS 50/50 model isn't very sophisticated.