Barry at Miami tonight 6 PM

[UMAlum21]

This board is easily the most miserable board of all the miserable boards.

It doesn't matter one iota what the Canes do on the diamond, the same three to four posters will be in here *****ing about it.

The Canes offense has a .420 OBP and .438 slugging %.

Duke- .338 and .318.
Louisville- .378 and .397
Georgia Tech- .349 and .373
Florida State- .394 and .410
Pittsburgh- .374 and .385
Virginia- .360 and .420
Virginia Tech- .323 and .368
North Carolina- .381 and .416
Clemson- .380 and .412

The Canes lead the ACC in on base percentage and slugging on offense. Let's all just chill out a little bit on all of the criticisms and fatalism with our program.

You reliaze all of our stats are inflated by a 25, i repeat 25, run performance!

So they didn't happen? Are you removing other teams' best games? Do you account for the fact that we pulled our most talented hitters once that barrage of runs occurred?

I've never understood the wanting to remove data from a sample size when it's relevant and valid data.

Do you understand the definition of an outlier?

 

[HurricaneVision]

Do you understand the definition of an outlier?

That depends on what you're looking for. Generally speaking, an outlier would be considered- in a normal distribution- to be the tail of the bell shaped curve on either end. More than two standard deviations from the mean. Normally, one standard deviation to either side would encompass 68% of the data. Two SD's would encompass another 32% of the data for a total of 94% of all outcomes. Some simplify it and say 5%, or 2.5% in each tail.

Now, when deciding which data to exclude, are you wanting to exclude all data from that game? Just the innings that are more than two standard deviations from the mean? Or, have you decided that this data is not relevant to the data set because it's more than two SD's from the mean? What about just innings where the starters are playing? How about just against pitchers who throw 90? Maybe limit it to just those pitchers first out of the bullpen?

You decide and I'll run the numbers again. I believe that all teams face this same criteria, therefore all the data is valid, but I'm willing to be flexible.

Additionally, that was a crude subset to show our offense at a glance is not this terrible slug it's made out to be on this board. OPS has limitations- such as the fact it underweights On Base Percentage and over weights Slugging Pervrntage. This is because a double is not worth two times as much as a single to scoring runs, a triple isn't worth three times as much, and so on.

Using linear weights to establish how much each outcome is worth to scoring runs is a better way to do it. You can do this in a quick and dirty way by multiplying OBP by 1.7 and adding it to slugging if you don't want to do the math.

I think you'll be surprised at the outcome- especially if you exclude each teams best game from the data subset.