Things we already knew

So, I took the data that MaxPower posted in another thread and decided to first focus in on which ball parks are the most successful. For my first pass I was focused just on $80M no DH. Once I ran the analysis one thing quickly became obvious and that is don't play in neutral parks. For this purpose I defined a neutral park as any park with all 4 single and HR park chances between 8 and 12.

With those parameters as a filter, there were 39 teams that played in a neutral park and they averaged a -58 run differential a season. Only 8 of the 39 teams had positive run differentials with a max of 101 and a min of -238. Now its likely some of those bad results were from new managers and likely the favorable outcomes were from some of our savvy vets, but it is a good reminder to pick a park that you can tailor your team around, and neutral parks are not the place to do that.

going back to the original purpose of finding the best environments in $80M no DH, there were 8 parks that were used at least 10 times that averaged a 50 run differential or higher. Those parks are:

1. Forbes '54
2. St. Louis '20
3. South End Grounds '11
4. League Park '34
5. Forbes '09
6. Sportsman's Park 1938
7. Penmar Park
8. Polo Grounds '41

Of that parks that were most popular in this group, Braves Field 1938 and Petco Park '05 may not be delivering as expected. In 39 uses (second most behind Forbes '57), Braves Field averaged a negative -4 run differential, while Petco was a break even park for the 36 that used it (on average).

I just found an error in the original data parse for Busch Stadium '67, that was excluding it from my neutral park analysis above. This park was used 21 times and meets those neutral parameters above and came in with a negative 43 run differential. It is the only "popular park" based on usage that isn't at least close to break even or positive.

I was going to make a connection between the players used in some of those successful parks, but unfortunately the individual player data does not have the team ID only the league ID, so I can't easily parse out which players were used in each park environment.

Nice work! I can contribute to this one too. I analyzed parks in terms of their Barnstormers grouping and how they performed at each Barnstormers cap (combining caps $200 and above into one group). The colored cells are how many games above or below .500 parks in each grouping perform at each cap. And the "N" cells on the right are just the number of teams in the sample.

https://i.imgur.com/LRbsCCu.png

Keep in mind these are Pythagorean wins, not real wins.

Lefty parks are the best at every cap except $140. And neutral parks are indeed the worst overall. Moderate hitting and extreme hitting also fare rather poorly.

Max,

The data parse for ball parks under rsi is incorrect for any singles environment above 9. For some reason it dropped the 10's digit in the parse. I reparsed the ballpark data from the park factor's field, but just and fyi.

I was about to do something similar and group like type ball parks to provide a bit better analysis, but the barnstormer's grouping are not refined enough for this analysis, so I am working on a way to isolate similar parks, but haven't quite decided where to draw my cutoffs.

Ahhhh wow that explains why the average rSI has always been so bizarrely low, thanks! Sharing the database paying dividends already...

Ok that rSI data should be correct now, thanks again for flagging that.

And yeah...why is Busch 67 so popular?

This is great stuff. Of course, if everybody in the league uses a lefty park……