Advanced baseball stats

[The Croom Diaries]

Something to chew on this week...

http://maroonandwhitenation.com/2014/03/24/baseball-advanced-stats-week-6/

[Saltydog]

nt

[Will James]

FIP don't lie

[RAYn_Man]

FIP is a worthless stat.

[WeWonItAll(Most)]

Whatever-stat-fits-your-agenda don't lie

[Lefthandersrule]

FIP is a worthless stat.

Really? So please let all of us in on what you might think is a viable stat.

This stat shows what only the pitcher (and catcher ) control.

Examples being ground balls, walks, Home runs, etc.

I'm not saying FIP is the tell all stat but it does show what the pitcher is doing regardless of the other position players.

[messageboardsuperhero]

Really? So please let all of us in on what you might think is a viable stat.

This stat shows what only the pitcher (and catcher ) control.

Examples being ground balls, walks, Home runs, etc.

I'm not saying FIP is the tell all stat but it does show what the pitcher is doing regardless of the other position players.

You are right except for the part about ground balls- FIP doesn't directly account for that (although technically if you get a lot of ground balls you theoretically should give up fewer home runs, which FIP does account for).

Anyone who really believe FIP is worthless is dead wrong though.

[preachermatt83]

all of them have strengths and weaknesses but the one pitching stat that , to me, best tells how good the pitcher is , is WHIP.

[lefty96]

The real answer is no one stat, baseball or otherwise, can be used to quantify anything. You always need more than one measurement to keep the others in context. Go back to the "Ross Mitchell is lucky" argument, based solely on babip, that would be the acceptable conclusion - that there's no way he could expect to continue to have that many of his struck balls fielded properly. But if you take his babip in context with his % of grounders and v/s his % of line drives/hard hit balls. How Mitchell has success is more apparent. You can never make a fair judgement with one number, people will always try to find one, but I don't think there ever will be. That being said, I'd like to see our team avg's compared to the SEC on the north half of 50% more often than not.

[Lefthandersrule]

Good answer lefty! I agree that no one stat is tell all but the ole ERA just isn't the only stat that should be looked at. Articles are attempting to show the more recently constructed stats to give a more complete view of a pitchers worth.

[lefty96]

I read this yesterday and thought about Mitchell, the piece about Burke Badenhop. . .
http://www.fangraphs.com/blogs/sunday-notes-cubs-as-sabr-analytics-red-sox-rays/

I will add on Mitchell that he has the lowest babip listed and has a defense behind him that is just above average compared to the SEC - that gives good context for the traditional view of babip.

[esplanade91]

We just got beat in the national championship by a team with a really, really bad batting average. Not going to put much on stats. Nothing more than a talking point one way or another.

[lefty96]

Good answer lefty! I agree that no one stat is tell all but the ole ERA just isn't the only stat that should be looked at. Articles are attempting to show the more recently constructed stats to give a more complete view of a pitchers worth.

Don't get me wrong, I like the articles and I enjoy the stats, but any of them, old ERA or new stuff - you can't just pick one and run with it. It's short sighted. I do stats/analytics for a living and alot of people fall in love with one number then get their ass handed to them for ignoring the rest.

[Johnson85]

You are right except for the part about ground balls- FIP doesn't directly account for that (although technically if you get a lot of ground balls you theoretically should give up fewer home runs, which FIP does account for).

Anyone who really believe FIP is worthless is dead wrong though.

So what is a good FIP? Our numbers are all over the place. I assume FIP is an outlier for Mitchell just like BABIP?

[lefty96]

I would suggest adding innings pitched and plate appearances to this, it will give people a better idea of the sample size.

[The Croom Diaries]

So what is a good FIP? Our numbers are all over the place. I assume FIP is an outlier for Mitchell just like BABIP?

It's adjusted so that it looks like what a good ERA should look like. It only weights what a pitcher can control, i.e. walks, K's, HBP, HR. So a pitcher like Lindgren who gets a lot of strikeouts will almost always have a better FIP than a guy who pitches to contact like Mitchell. With our defensive infield Mitchell has a higher value to a team like MSU despite his relatively high FIP, but to a team with mediocre fielders they would rather have the guy with lower FIP.

[dickiedawg]

Nice try, but .984 is not "just above" average... It's first.

[The Croom Diaries]

I would suggest adding innings pitched and plate appearances to this, it will give people a better idea of the sample size.

The purpose of this weekly post are to give stats that aren't available via HailState's stat book. I assume that anyone interested in this level of detail would also want to look at the official stat listings to gain their own perspective.

[The Croom Diaries]

Nice try, but .984 is not "just above" average... It's first.

We have a nice defense no doubt, but in all honestly if our official scorer knew the difference between a hit and an error we'd have 8-10 more errors this year.

[lefty96]

Nice try, but .984 is not "just above" average... It's first.

So, the distribution is pretty flat then. . . so if first place BA in the league is .250 and the avg is .245, is 250 still "good"?

[lefty96]

The purpose of this weekly post are to give stats that aren't available via HailState's stat book. I assume that anyone interested in this level of detail would also want to look at the official stat listings to gain their own perspective.

That's cool, I was just looking at it and thinking - and being your typical lazy person, had no motivation to look anywhere else.

[The Croom Diaries]

That's cool, I was just looking at it and thinking - and being your typical lazy person, had no motivation to look anywhere else.

Ha, well I'm lazy too. And it takes about an hour each week to look through all the game notes to find RISP and inherited runners, then enter the data and publish the post...that's more time than I spend on any post and this one is by far our least read each week...but I know that's because it's not a mainstream topic. It would probably take about 2 minutes to add that data but I really just don't want to clutter the post too much.

[dickiedawg]

You think a .005 difference in Batting Average is the same as .009 difference in Fielding Percentage?

We're making errors around 2/3 as often as league average. If we were getting out 2/3 as often as league average, we'd be hitting .522.

(Yes, Croom Diaries is right that our official scorer is pretty generous with Hits. But I'm not convinced other official scorers are any stricter)

[lefty96]

You think a .005 difference in Batting Average is the same as .009 difference in Fielding Percentage?

We're making errors around 2/3 as often as league average. If we were getting out 2/3 as often as league average, we'd be hitting .522.

(Yes, Croom Diaries is right that our official scorer is pretty generous with Hits. But I'm not convinced other official scorers are any stricter)

No, I was just pointing out that first does not always equal clearly better than average. Pretty much my entire point in every post in here is that everything needs to be kept in context. That's all I'm getting at.

[Maroons]

all of them have strengths and weaknesses but the one pitching stat that , to me, best tells how good the pitcher is , is WHIP.

WHIP is a nice stat, but even that one is somewhat flawed due to different players and teams' defensive abilities.

[lefty96]

Ha, well I'm lazy too. And it takes about an hour each week to look through all the game notes to find RISP and inherited runners, then enter the data and publish the post...that's more time than I spend on any post and this one is by far our least read each week...but I know that's because it's not a mainstream topic. It would probably take about 2 minutes to add that data but I really just don't want to clutter the post too much.

I certainly wouldn't want to attempt to gather that myself. I'm auditing a class in sabermetrics in the summer, I can't decide if I should be excited or just worried how dull it could end up being. More sql and R work - hurray?

[dawgs]

FIP is a worthless stat.

no it's not. it's harder to quantify in the moment unless you are sticking numbers as they happen into the formula, but it's pretty indicative of which players are likely to see their more easily measurable performance (era, whip, etc) improve or regress moving forward. there's obviously some guys that almost always outperform their FIP (ross mitchell, matt cain did it for years for the giants - though it caught up to him last year), and others that seem to always underperform their FIP (ricky nolasco is one that jumps out in MLB). so you do need to look at other contexts (i.e. the type of contact being made, whether their defenses emphasized defensive shifts - this is one reason the rays pitchers almost always outperform their FIPs), but it's generally a pretty solid rule of thumb for guesstimating future performance.

[dawgs]

I certainly wouldn't want to attempt to gather that myself. I'm auditing a class in sabermetrics in the summer, I can't decide if I should be excited or just worried how dull it could end up being. More sql and R work - hurray?

i think it would be pretty awesome, though i wonder how much ground they'd cover that you couldn't get from reading everything on fangraphs.

[KB21]

We just got beat in the national championship by a team with a really, really bad batting average. Not going to put much on stats. Nothing more than a talking point one way or another.

That's because batting average is a horrible indicator of actual hitting and offensive ability of a baseball team.

[lefty96]

i think it would be pretty awesome, though i wonder how much ground they'd cover that you couldn't get from reading everything on fangraphs.

https://www.edx.org/course/bux/bux-sabr101x-sabermetrics-101-1558

[dawgs]

That's because batting average is a horrible indicator of actual hitting and offensive ability of a baseball team.

that and if you are going to be as relatively weak on offense as they were, you have to have an elite pitching staff. they had an elite staff AND the CWS is played in a stadium that absolutely nullifies any hitting advantages, so pitchers can groove strikes all game and, if they have a good defensive team behind them, feel confident that they wouldn't get burned with extra base hits. accounting for the park, our hitting advantage was muted, and they had the better pitching staff.

[Esmerelda Villalobos]

You guys use whatever dumbass formula you want. Ill take the guy sitting at 8th in fip.

You guys take the 7 ahead of him. He would out pitch all 7, probably in consecutive days.

Greatest pitcher ever at msu is 8th best on team. Good to know. That tells you how worthless these stats are.

[dawgs]

Look who is missing the point!

We already said you can combine the eye test (type of contact) with metrics like FIP to get a true view of a guy. Also, pitching as a reliever only 1-2 IP at a time can skew things since guys tend to get more Ks pitching in shortened windows than over a full 5+ IP start (or long relief appearance).

[messageboardsuperhero]

I read this yesterday and thought about Mitchell, the piece about Burke Badenhop. . .
http://www.fangraphs.com/blogs/sunday-notes-cubs-as-sabr-analytics-red-sox-rays/

I will add on Mitchell that he has the lowest babip listed and has a defense behind him that is just above average compared to the SEC - that gives good context for the traditional view of babip.

Huh? We lead the SEC in fielding %. Before the Sunday game against Vandy, there was a stat floating around that we were fielding about .995 in our last 12-15 games. I don't know how much the Sunday game affected that, but it's still a very impressive number. Also, all of our fielders (infield and outfield) have good range, and I'd go as far to say that some of them have outstanding range.

Bottom line- we've easily been the best defense in the conference since making Heck our SS. We've also led the entire country in double plays turned since 2012. Don't sell us short defensively.

ETA: And that defense is a pretty big reason why pitchers like Brown and Mitchell are so successful here. I'm sure they'd still be good pitchers on other teams, but our defense makes them that much better.

[messageboardsuperhero]

We have a nice defense no doubt, but in all honestly if our official scorer knew the difference between a hit and an error we'd have 8-10 more errors this year.

Very true, but other team's scorers are just like ours in that regard.

[lefty96]

Huh? We lead the SEC in fielding %. Before the Sunday game against Vandy, there was a stat floating around that we were fielding about .995 in our last 12-15 games. I don't know how much the Sunday game affected that, but it's still a very impressive number. Also, all of our fielders (infield and outfield) have good range, and I'd go as far to say that some of them have outstanding range.

Bottom line- we've easily been the best defense in the conference since making Heck our SS. We've also led the entire country in double plays turned since 2012. Don't sell us short defensively.

ETA: And that defense is a pretty big reason why pitchers like Brown and Mitchell are so successful here. I'm sure they'd still be good pitchers on other teams, but our defense makes them that much better.

I would think that a .984 fielding % v/s the league avg of .974 isn't a huge gap from first to middle of the pack. If that's an incorrect conclusion, then I'm just wrong. The rest, I agree w/ you.

[messageboardsuperhero]

I would think that a .984 fielding % v/s the league avg of .974 isn't a huge gap from first to middle of the pack. If that's an incorrect conclusion, then I'm just wrong. The rest, I agree w/ you.

That's a pretty big difference, especially in the college game- where defense is at a premium and routine groundballs aren't automatic outs like they are in MLB. And considering how much we pitch to contact, that just puts more pressure on our defense to make plays.

I just thought it was a little ridiculous to call what is probably the best defense in the conference "just above league average." Also remember (as someone else pointed out) that a .010 difference in fielding % is a much bigger difference than a .010 difference in other stats.

[lefty96]

I just thought it was a little ridiculous to call what is probably the best defense in the conference "just above league average." Also remember (as someone else pointed out) that a .010 difference in fielding % is a much bigger difference than a .010 difference in other stats.

Okay, looking at our numbers, we've had 1015 chances and recorded 693 put outs, 306 assists and 16 errors for a fielding % of .984. If we work backwards, if our fielding % was .974 that would be an additional 11 errors over 26 games, far worse sure, I get your point and should have done the math before I typed. I'm still not prepared to say that Mitchell is successful only b/c of our defense or that he is lucky, but I understand the correlation between his numbers and our fielding %.

[curmudgeon]

Awesome job. Why don't you do these for our softball team?

[messageboardsuperhero]

Okay, looking at our numbers, we've had 1015 chances and recorded 693 put outs, 306 assists and 16 errors for a fielding % of .984. If we work backwards, if our fielding % was .974 that would be an additional 11 errors over 26 games, far worse sure, I get your point and should have done the math before I typed. I'm still not prepared to say that Mitchell is successful only b/c of our defense or that he is lucky, but I understand the correlation between his numbers and our fielding %.

When did I claim that Ross Mitchell was only successful because of our defense or luck? I said he'd be a good pitcher on any other team, but when the best defense in the league is playing behind a contact pitcher, the defense in going to make the pitcher just that much better.


And that defense is a pretty big reason why pitchers like Brown and Mitchell are so successful here. I'm sure they'd still be good pitchers on other teams, but our defense makes them that much better.

Hell, he and Preston Brown be the first ones to tell you that our defense helps them out a lot.

[Will James]

Hard to quantify how we relate to other teams defensively with limited samples based on luck but fielding percentage is NOT it. One way to do it is look at teams pitching/defensive BABIP. The lower, the more balls were gotten to by the defense. But again it takes a lot of samples. There's really no way to know who has the best overall team D in the conference.

[The Croom Diaries]

Awesome job. Why don't you do these for our softball team?

Because Im as interested in the softball team as I am going shopping with my wife.

[The Croom Diaries]

Hard to quantify how we relate to other teams defensively with limited samples based on luck but fielding percentage is NOT it. One way to do it is look at teams pitching/defensive BABIP. The lower, the more balls were gotten to by the defense. But again it takes a lot of samples. There's really no way to know who has the best overall team D in the conference.

We'll have a pretty good idea next year when we can watch every game on the SEC Network.

[dawgs]

We'll have a pretty good idea next year when we can watch every game on the SEC Network.

You'll need to start charting ranges and how many balls different guys get to and all that stuff that goes into defensive metrics to get a true idea. Hope you're up to the challenge.

[The Croom Diaries]

You'll need to start charting ranges and how many balls different guys get to and all that stuff that goes into defensive metrics to get a true idea. Hope you're up to the challenge.

Or you can just watch the game and make your own determination like we do in football and basketball. Not everything has to be determined by a stat.

[Eric Nies Grind Time]

Something to chew on this week...

http://maroonandwhitenation.com/2014/03/24/baseball-advanced-stats-week-6/

Did you calculate the FIP constant or did you use the MLB constant 3.2?

[The Croom Diaries]

Did you calculate the FIP constant or did you use the MLB constant 3.2?

I used 3.2 because the SEC constant right now is all over the place with non-conference schedules not being created equal. Maybe half-way through SEC play I'll use the SEC only to determine the new number.

[Will James]

Interesting observations

1. Armstrong has a higher K% than Rea.

2. Bradford and Pirtle have combined for 6 walks in 190 plate appearances.

3. Henderson has a .452 BABIP

[dawgs]

Or you can just watch the game and make your own determination like we do in football and basketball. Not everything has to be determined by a stat.

actually recording data ensures you don't let bias or misperceptions slip in.

[The Croom Diaries]

actually recording data ensures you don't let bias or misperceptions slip in.

If y'all want to send me about $500 a week during baseball season next year to watch every SEC game and track all that stuff I'd be glad to.

[dawgs]

If y'all want to send me about $500 a week during baseball season next year to watch every SEC game and track all that stuff I'd be glad to.

ha, i'm just saying that watching a bunch of games on tv isn't the best way to make a determination on the best fielding team based on ranges, positioning, etc.