re: LSU Baseball 9 Spring Scrimmages Hitting and Fielding Errors Stats Including Team Totals

BABIP is almost useless with a 20 AB sample size

Can you either enlarge the stats or convert sheet to a table? Certainly going to be cross-eyed soon

quote:

---

Batting averages, overall are not very good.

---

If the pitchers aren’t ahead of the hitters at this point in the year then your pitchers aren’t very good.

quote:

---

I think more accurately it illuminates the fact that BA itself is useless with a 20 AB sample size.

---

I think they both probably provide about the same amount of usefulness. This is only two weeks worth of data throughout a regular season.

You should alternate the shading fill of the rows so it's easier to stay on the correct row, like

Not really 2.4 per game since the tigers are batting the entire time. So realistically it’s 1.2 hr per game. Pitchers doing well facing these hitters. Once real games start Tigers will have at least 3.50 hrs per game.

According to FanGraphs, at the Major League level BABIP normalizes (reaches relevant sample size) at 820 balls in play for a hitter and 2000 balls in play for a pitcher.

A college player may never reach those numbers in their career, unless you take into account summer leagues.

quote:

---

You don’t think the additional detail he’s providing is useful? The detail on XBHs and slugging is particularly valuable.

---

It’s not though.

ETA: and it would be easy to ask Bay to add anything we thought useful. Or maybe post it in a thread that already exists. It’s pretty redundant. At first, I thought he just didn’t see those threads. But this is the second time now. Maybe Russian doesn’t like Bay.

quote:

---

ETA: and it would be easy to ask Bay to add anything we thought useful. Or maybe post it in a thread that already exists. It’s pretty redundant. At first, I thought he just didn’t see those threads. But this is the second time now. Maybe Russian doesn’t like Bay.

---

I am pretty sure Bay has been asked to add in slugging and he indicated it would be too much work for him to change up his spreadsheet. I see nothing wrong with someone else supplementing that information given that.

Thanks, Project, for both of your comments. 820 balls in play (to reach a relevant statistical sample size) will likely require (conservatively) 1,000 plate appearances; and so, conservatively it will require 200 games minimum. So for most college players, the statistical sample will never reach a relevant size. For those that it does it will likely be in their fourth year of play.
You note it is essentially trying to quantify "luck". Well, I wish them luck with that. And luck is the ability to defy statistical probability. So again, I don't see its usefulness in analyzing a player's physical ability. I am truly confused about its supposed usefulness in the long run. I don't need BABIP to know Clayton Kershaw and Mike Trout are exceptional ballplayers.