There generally tends to be a lot of mystery surrounding an MLB front office. Fans know that they are responsible for signing free agents, making trades, and gathering information, but they do not know how it was done.
I was recently able to sit down with Cardinals’ Assistant General Manger Michael Girsch, Baseball Development Analyst Chris Correa, and team Arbitration Lawyer Hal Welford after they gave a presentation on statistical modeling in baseball.
The casual fan does not realize how much the front offices do to try to model any decision that they will be making. Some of these models are very accurate while others are just used to give the front office a good sense of what is going on.
One of the first things that a front office does is look to project the net annual cash flow of any player they sign. To calculate this they look the player’s production minus their salary. They use a discounted cash flow model to try to determine the player’s production value. For the Cardinals, this value is determined by the player’s Runs above Replacement value. They then discount it by using calculations of how much a win today is worth versus a win in the future.
For a team like the Royals, with their strong farm system, a win in two years is likely worth more to them then a win now, while with a team like the Yankees, a win today is worth a lot.
Each team has a different dollar amount that it plans to spend for each run above replacement level. This is generally impacted by market size.
In addition to these calculations, the teams are also looking toward calculations that they will be able to do in the future. One thing in development that will provide much more data is Field F/X. This is a system that uses hundreds of cameras to take hundreds of measurements a second. This system will be able to give front offices information about location, speed, and direction of every player and the ball at all times. In turn, this will allow teams to see things such as whether or not an outfielder consistently gets a good jump on the ball.
Even with all this data, the best analytics may only provide a slight edge. Part of this is related to the fact that there have become more teams who are taking a deeper look at the numbers. According to Correa teams with a large analytical department can still have an advantage. He stated this is because, “most teams have few guys working with data”.
However, the competitive advantage of using statistics has changed since it was revolutionized by Billy Beane and the Moneyball strategy. Now, according to Girsch, there are “fewer mispricings in the market. The quality of data available now versus the ‘90’s is so much different. On-base percentage is no longer undervalued”.
Some teams have even chosen not to have an analytical group in their front office. The Twins don’t have one and yet they are able to compete because they have very strong scouting and player development.
Another interesting topic that came up during the presentation was the arbitration process. It was noted that for relievers, their salary is driven by saves, holds, and ERA and that for starters, wins played a large role in addition to awards they have won.
When it was asked why sabermetrics are not used in this process, Welford noted “many arbiters are either economics professors or lawyers, not baseball fans. The information needs to be toned down”.
In addition, Welford said that fan appeal is a factor in the arbitration process, but more so for pitchers. He mentioned the example of Fernando Valenzuela. He said “More than 10,000 fans showed up when he pitched regardless of whether it was at home or on the road. He had a significant impact on the gates”.
From there, the conversation branched off into a conversation about how Clayton Kershaw and David Price could redefine the arbitration market. Ben Nicholson-Smith brought up this concept earlier in the week over at MLB Trade Rumors.
Welford said that the market for these two would be incredibly difficult to establish. He said that “both will be compared to Tim Lincecum and even then, it is still hard to use that as a comparison. If they go year to year, that is very different than the two your deal Lincecum got”.
The conversation then moved toward scouting. When I asked about the impact of luck and stats such as BABIP, Correa said that they are looked at very differently depending on the level. He said, “we don’t really used BABIP (for hitters and pitchers) much in the MLB because we have a lot of batted ball data. However, it is very useful when it comes to evaluating players in the minors, college, or even high school”.
This is not the only time where there is a difference between the majors and other levels when it comes to statistics. It was asked if sabermetrics, Pitch F/X, and scouting reports were used in combination to make determinations about players. Girsch noted that this is done during the draft because it is “a discreet event and we can combine data”. However, he also stated that it is much harder to do at the pro level “because your looking at that data and the information can change between April and June”.
Finally, the questions moved onto information relating to the Cardinals. A question was asked about how Adam Wainwright would be evaluated since he has already established that he has superstar potential but his future is now uncertain. Correa said that the team uses a model to project a player’s ability when they return from specific injuries. According to Girsch, the team has also developed models that “try to pick out injuries to a pitcher based on injury history, fastball velocity, and mechanics”. However, Girsch also stated that these models are not that great but they are still useful. He also said that the team even attempted to model Matt Holliday’s return from his appendectomy but there was not even data available.
The last set of questions was related to the topic that is on everyone’s mind in St. Louis: Albert Pujols.
Pujols means so much to the city and has so much intangible value. However, it could not be modeled by the front office because according to Girsch, the team “doesn’t have analytics for that. We try to talk to ticket sales and marketing, but there are not enough examples like Pujols to actually do any analysis”.
Then, I asked the big question about Pujols’ value and future production. Girsch responded by saying, “It’s really hard. We try to develop an aging curve using all players in the past fifteen years and use that. We can try to limit it to just All-Stars. However, every guy is unique. He could be like Frank Thomas and just drop off, or like Barry Bonds and miraculously get better with age, or he could be like Hank Aaron and stay consistent as he gets older”.
The team has also stated that it will not talk about the Pujols contract situation.
It is my belief that it is more about if the owners decide to open up their checkbook to pay Albert what he wants than anything else. The number that they guys present really doesn’t matter because Albert will be able to get what he wants out on the open market.