Is MIN-MAXING Optimal?
Why Stars and Scrubs is not the One True Way

For many years denizens of this blog and others, including myself, have insisted that the optimal way to build a MLB team is to acquire a bunch of high-salary elite players and then fill in the rest of the roster with cheap veterans and young team-controlled prospects ("Stars and Scrubs"). This is supposed to provide a high degree of flexibility and upside compared to loading up on solid mid-range veterans ("25 Honda Civics"). It has also been assumed that the savvier teams in the league aggressively pursue this strategy. These beliefs have fueled a lot of frustration towards Jerry Dipoto for his numerous signings and trades of "Civics", with commentators arguing that A) he is taking an extreme approach away from acquiring "stars" and B) this is necessarily a poor approach that smart teams would avoid. But are any of these beliefs true? I have never seen any research that takes a look at this issue, and no one has provided any context with which to judge Dipoto's roster makeup with other teams in the league. This should give us pause. We should be very careful about being dogmatic about things when we don't actually have any information to suggest that our theory is correct. And in this specific case, if we think critically about roster building strategies it will become clear that there can NOT be a single, optimal strategy.

Let's think this through. If Stars and Scrubs is inherently optimal, then every team should pursue it, right? But what would happen if that were the case? The result would be that the cost of stars would go way UP, the cost of scrubs would go way UP, and the cost of Civics would go way DOWN. The problem should be abundantly clear. Acquiring stars or scrubs would be horribly inefficient, while Civics would be a bargain. Therefore, going after Civics would be optimal. That means the optimal strategy is going to depend on the specific market conditions at the moment. And in a rational market, an equibilrium will be reached which will make either strategy about equally valid. Therefore, "Stars and Scrubs" can not possibly be inherently optimal.

Now let's look at some actual data to better understand what strategies teams are actually pursuing. What I did was look at this year's opening day rosters for every team and divide the players into three categories: Stars (those making over $15 million dollars), Civics (between $15 million and $5 million) and Scrubs (less than $5 million). Note, I was concerned what a team was actually paying a player, so a guy with a huge salary that is mostly paid by a former team would be counted as a Civic or Scrub depending what his current team was on the hook for. I also am counting Denard Span against both the Rays and M's because he started with the Rays but he was traded early enough in the season that the M's may be paying him enough money that he still counts as a Civic for them.

TABLE IS SORTED BY FEWEST CIVICS/MOST STARS AND SCRUBS

THE TOP 6 TEAMS BY WIN% ARE IN BOLD

Team Stars Civics Scrubs
San Diego 1 1 23
Milwaukee 1 2 22
San Francisco 7 3 15
Texas 4 3 18
Detroit 3 3 19
Cincinnatti 2 3 20
Philadelphia 2 3 20
Atlanta 1 3 21
Arizona 1 4 20
ChiSox 1 4 20
Miami 0 4 21
Kansas City 2 5 18
Oakland 0 5 20
Tampa Bay 0 6 19
NY Yankees 4 6 15
LA Angels 3 6 16
ChiCubs 4 7 14
Colorado 3 7 15
Toronto 3 7 15
Pittsburgh 0 7 18
Seattle 4 7 14
Minnesota 1 8 16
Boston 5 8 12
Washington 4 8 13
St. Louis 3 8 14
Baltimore 3 8 14
LA Dodgers 3 9 13
Cleveland 1 10 14
NY Mets 3 11 11
Houston 2 11 12

Comments:

1) The Astros are possibly the best team yet they are the most dependent on Civics

2) Seattle (7 Civics) is right near the average of 6 civics

3) The other top teams in the AL (Yankees and Indians) are either average or above average in their reliance on Civics

4) In the National League, the division leaders (Braves, Brewers, Diamondbacks) are Stars and Scrubs heavy, but the big spenders (Cubs, Dodgers, Nationals) are average or above average in Civics

Conclusion:

The empirical evidence makes clear that Stars and Scrubs is not necessarily the strategy pursued by the best or wisest teams, nor is Dipoto's strategy in anyway extreme.

Comments

1

Also interesting that you put up some figures to support your thesis.  Good info.

Here's where I think Stars & Scrubs shines, and why it's so popular with many: you don't necessarily know when a Civic is going to pumpkin, but a Star is a lot more predictable in terms of generating enough value to warrant the roster slot (if not the salary).  The flipside of Stars & Scrubs is that the volatility of Scrubs is through the roof; you never know if a Ben Gamel is a real deal, Major League average player, or a flash in the pan ala Willie Bloomquist's first call-up.

But if you cultivate a system where you've got a few key 'anchor' pieces, it becomes easier to arrange the rest of the pieces in order to compete.  Most Stars & Scrubs teams employ a handful of civics when they're near the PEAK of their competitive arc, simply to fill out the glaring holes (as this team should have done with the rotation coming into the season) and turn them from abject weaknesses into non-issues.

It's all about being dynamic and responsive, but I agree wholeheartedly with your thesis that Civics (especially bullpen civics?!?) are currently being grossly undervalued.  There are players who are rock solid average plus contributors who couldn't get perceived market value this last offseason.  That's not a problem, it's an opportunity to go against the market and earn a tidy profit (both in the W/L column and on the payroll) for a savvy team.

Great post :-)  Always a good idea to re-examine our thinking on core issues.

2

Good Shtick CPB,

I think the main detriment to a Civic roster is when a Civic player tanks he gets way too long of a look before he gets DFA'd or traded.  Scrabble was a recent example of such a player.  Its worth remembering that stars, such as Hanley Ramirez, tank as well.  The true art, besides finding good players, is learning how to creatively punt on a toxic contract.

3

Undervalue any commodity and you'll find your Moneyball-esque advantage in exploiting its cheap availability. That said, I think the general trend toward Stars and Scrubs as a preferable starting point is solid... it's just that the pendulum is swinging too far toward that preference.

All in all a great article and a nice eye-opener. I still find myself hoping DiPoto knows how to bring in the right Star or two when the time comes... there are some unique opportunities for that approaching, and his stated reasons for going trade over FA are about preferring an ideal fit for his team rather than a good-enough one. I have hope that the 'right' Star player does exist out there this offseason for him... there are quite a few excellent options available!

Add comment

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

shout_filter

  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.