For nearly three quarters of the season, the Houston Astros were leading the AL West division.
At one point, in early July, the Astros actually had a commanding 5 game lead over the 2nd place Angels.
In addition, the Astros had the second best record in the AL.
Although the Astros fell into a mini-slump just before the All-Star break, they still led the AL West for the majority of the third quarter of the season (Figure 1).
Suddenly, with only a handful of games left in the season, the Astros are likely out of contention for the division and their hold on the final wildcard spot is hanging by a thread.
Over a month ago, this team was the MLB Network's media darling. Now, the ‘Stros are on the verge of one of the more dramatic late-season meltdowns in recent memory.
The reason for the team's current struggles can simply be explained by analytics.
Figure 1: AL West Standings throughout 2015
Although, only a year ago, the Astros were one of the worst teams in baseball with 70 wins, the team's management has used an interesting philosophy in constructing this season's team.
Based on my analysis, it looks like the team’s roster construction is maximized on the positive value of batting power and has brushed off the negative value of strikeouts.
As shown by the model below, a home run can be worth nearly 1/10th of a win. In other words, batting power in large amounts can be instrumental to wins.
On the surface, the production of runs may not seem to be a problem. Currently, the Astros are ranked 11th in runs scored per game.
Although the team is producing a home run every 24.6 at-bats, the team is also striking out every 3.9 at-bats.
Based on this fact, there is a clear opportunity cost between home runs and strikeouts. While, the lineup gets on-base by walking 8% of the time (ranked 10th).
The team, for the most part, lives and dies by hitting home runs and walking.
Thus, their high propensity for striking out absolutely cripples their ability to get on-base at an above average rate (ranked 20th in OBP).
To the Astros' credit, the management has found a niche in the market by taking on undervalued players that fit the big power, high walk rate, and high strikeout rate mold (Figure 2).
However, the team has so many players that fit under this similar philosophy. Essentially, the Astros do not have the lineup diversity necessary to win consistently.
Figure 2: Starting Lineup for the Astros
Equation 1:
(HR*0.1057664)+(BB*0.0543431)-(SO*0.0247072)-(TE*0.1630308)+(DRS*0.0523805)-(CS*0.150058)-(FE^2*0.0016618)+85.07701 = Total Wins
HR: Home Runs BB: Walks SO: Strike Outs TE: Throwing Errors DRS: Defensive Runs Saved CS: Caught Stealing FE: Fielding Errors
With the Astros' overall construction philosophy in mind, the understanding of analytics comes at the forefront of their shakiness and, possible, late season meltdown.
It is hard to know how much "sabermetrics" plays into the Astros' roster.
Still, it seems that management has followed a typical "sabermetric" philosophy of valuing power over the smaller nuances of the game such as timely hitting (simply getting on base) and, even, fielding.
In no question, analytics can help produce results in the win column. The peripheral analytic numbers show the Astros as above average in all categories of hitting, fielding, and pitching.
Yet, the Astros are barely holding onto the second wildcard spot. This is where the error term comes into play.
The error term can be thought of as the factors that affect winning, which are not readily explainable.
The error term can include team chemistry, lineup diversity, communication on the field, and etc.
Though the Astros are fundamentally sound in statistical terms, the team has its issues on the field.
In a better explanation, they are not a pretty team to watch on television.
For example, while Scott Kazmir did not do the team any favors with his poor outing on September 25th against the Rangers, the Astros committed continual gaffes throughout the game.
In one instance, Preston Tucker missed the cut-off man on a throw from left field, which gave Chris Gimenez of the Rangers a triple.
On another play, Jonathan Villar was tagged out running from second to third on a ground ball to the shortstop.
These two plays were only a small microcosm of the many bad plays committed by the Astros during the game.
The Astros' late season meltdown can be summed up, again, in analytics.
In order to succeed, like anything in life, you must prepare for the anomaly.
By lacing a team with power and seemingly forgoing the negative value of strikeouts, the Astros are missing several elements of playing “smart baseball”, which is essential during crunch-time situations.
"Smart baseball" can simply mean the idea of getting on-base (whatever means necessary) or communicating on flyballs.
Although it might seem I am being rough on analytics, my econometric model still paints a harsh picture of the Astros with a projection of 85 wins by the end of the season.
With 81 wins and 7 games left in the season, it will be interesting to see whether their "fly or die" strategy can hold up and give them a spot in the MLB playoffs for the first time in ten years.
References:
1. "Baseball Reference." Baseball-Reference.com. Baseball Reference, n.d. Web. 14 Apr. 2015.
2. "Baseball Statistics and Analysis | FanGraphs Baseball." Baseball Statistics and Analysis | FanGraphs Baseball. N.p., n.d. Web. 24 Apr. 2015.
NOTE: All statistics accurate as of 09/27/15
By Sanjay Pothula
AriBall.com
