Showing posts with label infographics. Show all posts
Showing posts with label infographics. Show all posts

13 April 2015

Does Major League Baseball Need a Stats Lesson?

I saw this data in the St. Louis Post-Dispatch last week claiming that the Cardinals are the 6th most expensive team  to see at their home park. (the "FCI" averages all of the columns from this chart together to get an estimate of what a family of 4 would expect to spend at the ballpark.)

I have no problem with the FCI, but I wonder if this chart's reporting of "MLB LEAGUE AVERAGE" is a little off. That row looked suspiciously in the middle to me, and when I counted rows, it was indeed exactly in the middle.


So what's going on here? Is it a misrepresentation of "average," do MLB teams attempt to group themselves symmetrically around this figure, or is it pure coincidence?

Here's the MEAN of those FCI listings by team:

Could the difference between 211.68 (reported in the table as "average") and my calculation of 211.89 be the result of rounding error in the data they used that I don't have access to in this report? Are there 21 rogue cents floating around in their numbers? 



What's this mean for my students?
I think this graphic and table is a good conversation starter for both mean vs. median AND the role of rounding in getting "different" answers. What's the clue that this CAN'T POSSIBLY be the median? Its not listed in the data of course. 

17 April 2014

Are The Chicago Cubs About to Be Historically Bad?

I have a student in one of my AP Stats sections that is a Cubs fan. Its usually a great opportunity to build relationship with him (and the others) by harassing him about how bad they are.

It's in good fun. And educational.

He was telling the class today about the Cubs 4-10 record through the first 14 games of the season. It sounds bad, and it is, but since we know significance tests now, I thought it'd be fun to through their preseason projected win percentage (.413) against their current (.286) to find the chances of a win percentage that low, assuming that the original .413 was close to correct.



Methods
1. Refer to preseason final standings projections from mlb.com to find the Cub's projection along with the other 30 MLB teams.

2. Find the standard deviation of all the team's projections using a spreadsheet.





















3. Use the standard deviation of win percentage, projected win percentage, and current win percentage values to find the z-score and corresponding p-value on the normal curve. Our's calculated to .0015%. That's the likelihood of the Cubs having a win percentage that low, assuming the original projection was correct.

So what's happening here?
We probably shouldn't rule out that the Cubs could be worse than their projected win percentage, but with the team only 9% into the season, its certainly too small a sample size to declare that the team is headed to a historically poor season.