By The Numbers, Dec 7 (December 8, 2003)
The new issue is out...
(I'll make my comments after I get a chance to read it.)
I would also like to direct readers to the following links, as they relate somewhat to some of the articles in the above issue:
==========================
1 - For catcher game-calling article (but not about game-calling):
http://baseballstuff.com/tangotiger/catchers.html
==========================
2 - For DIPS, the "solution" has been published here:
http://baseballstuff.com/tangotiger/solvingdips.pdf
==========================
3 - For the Mills Brothers stuff, I've introduced my own here called Win Advancements:
http://baseballprimer.com/studies/archives/00000194.shtml
==========================
4 - Attendance
Voros published an excellent piece here:
http://baseballprimer.com/articles/mccracken_2002-08-19_0.shtml
I ran a very quick study here:
http://groups.yahoo.com/group/RetroList/message/1959
I just ran a very simple study that took the top 100 attendance figures (where available) for each team from 1974-1990, and looked at how often each team won. The team win% was almost 55%.
When you realize that the normal HFA is 54%, that doesn't look too impressive.
Re-running the study so that this time I took the top 5 attendance figures per team/year, the home team won 50% of the games.
I'm sure that there are other biases that I did not account for quality of opposition for one), and that there are better tatistical methods.
A regression analysis between wins and attendance (no separation of team or year) was virtually zero.
My bet is that maximizing revenue is probably the best goal, but that you have a secondary goal to maximize your fans so that you develop a loyal following. But if ownership churns every 5-10 years, that secondary goal becomes an easily ignored goal.
--posted by TangoTiger at 09:25 AM EDT
Posted 8:10 a.m.,
December 10, 2003
(#1) -
Alan Jordan
Check out the "Accuracy of Preseason Forecast" article. Out of 5 experts, nobody beats the monkey overall. They do well with the American League, but horrible with the National league. It would be interesting to do this with more seasons. Doesn't Diamond Mind post it's preseason predictions?
Posted 8:53 a.m.,
December 10, 2003
(#2) -
Alan Jordan
Florida's a good baseball city? I would argue that there are two dimensions of being a good baseball city. One is average attendence controlling for win/standing, past/present, exapansion, p4-95 strike. The other is how sensitive attendence is to winning. In a regression you model these two parts as:
attendence=Team + Team*win + win + expans + strike
The first two terms are what we care about and the rest are just there to be controlled for. We want the first term to be positive and the second term to be near 0.
Posted 10:42 a.m.,
December 10, 2003
(#3) -
PhillyBooster
I think a third dimension of being a good baseball city would be actually being a city. Florida is a state.
Posted 11:52 a.m.,
December 10, 2003
(#4) -
tangotiger
The attendance research is one that I was interested in doing, so I'm happy to see something like this.
The best-fit equation is listed as:
ATT/AVE = 2.7525 * (WIN) -.3769
where, WIN = win% and ATT/AVE is the attendance relative to league average.
So, setting the league attendance to 2.4 million, we see that the marginal change of 1 win in a season (.006 wins/game), will increase attendance by 1.7%. If I remember correctly, Pete Palmer, in The Hidden Game of Baseball noted that 1 marginal win changed attendance by 2%.
Assuming that attendance and TV revenue and all those revenues also increase by 1.7%, and assuming 120 million$ of team revenue, 1 marginal win generates 2 million$ in revenue. The 2% rule would specify 2.4 million$, and Voros' research comes out with 2.65 million$.
My own research on replacement level, coupled with the total amount of salaries paid out, shows a marginal win costs, in salary, somewhere around 1.4 to 2.0 million$ / win.
Posted 12:00 p.m.,
December 10, 2003
(#5) -
tangotiger
(homepage)
Baltimore was one of those strange teams that I noticed as well. (See homepage link)
Baltimore was a very good team in the 70s, but they just didn't draw. Even when they weren't playing so well in the 80s, their attendance was poor. Only when they went into Camden Yards did things change drastically.
I think Baltimore should be split into two cities. The Baltimore pre-Camden and the Baltimore/Camden era. I think there was a whole redevelopment of the area, and not just the park. I understand it's the same people, but there's a completely different mindset here. Other areas with new parks don't also do a massive redevelopment of the entire area.
At the least, I would introduce an extra variable called "massive neighborhood redevelopment", and give a "1" to the Baltimore/Camden years.