Thought Experiment : True* On-Base Percentage

Discuss all things Cardinals Baseball
User avatar
vinsanity
Chili dog truther
Posts: 8772
Joined: July 3 06, 2:19 pm
Location: Indianapolis

Thought Experiment : True* On-Base Percentage

Post by vinsanity »

After a semi-pedantic discussion with someone around here, I was actually musing - would a true "on-base" percentage statistic have any value?

Preface - This is not a critique of the current stat of "on-base percentage". I think it has a worthwhile measure but I do think there's some outliers that are arguably a skill.

If you hit a homerun - you're not getting the pitcher in to the stretch. Home runs are called rally killers and you're not technically ever "on-base".

I'd be willing to bet someone like Ichiro reached first more often on an error than Adam Dunn because of rushed or bad throws that may or may not be called IF singles.

What about guys who reach 1B on a FC because they leg out a GIDP? Or guys who hit absolute rockets the SS can't handle?

Is just "getting to 1B" currently unmeasured? Is it a skill? Does measuring have any value?

User avatar
Big Amoco Sign
Master of Hyperbole
Posts: 14402
Joined: December 1 17, 11:05 am

Re: Thought Experiment : True* On-Base Percentage

Post by Big Amoco Sign »

I'll bite. Does someone ahead of them get out? If so, swapping the runner for an out isn't a net positive on any level. Even if that is a speedy player being swapped, an out still hurts the team. Is there a fielder choice? Sacrifices are already not ABs that count against BA.

I don't get why you need an true 1B stat when you can use speed to get guys who leg out GIDP. O'Neill and Nootbaar types do it a lot. They shouldn't get any credit other than than their Statcast sprint speed. Why should a batter get credit for a lousy fielder? The biggest purpose of this knowledge is lineup placement. Put them in spots where they won't GIDP.

Fangraphs wrote about "reached on error" before:

https://blogs.fangraphs.com/the-error-o ... -on-error/

Why not use xwOBA and xwOBACON and xBA/xSLG to handle these issues and take fielders out entirely (well at least their non-normalized fielding play)? It even takes in speed of baserunner into account with the batted ball. That solves your problem. They're even normalized league average/park wide.

You can already see stats of how a pitcher does with runners on. Not sure you need baserunner's "true" on-base to get any more out of it than that.

User avatar
vinsanity
Chili dog truther
Posts: 8772
Joined: July 3 06, 2:19 pm
Location: Indianapolis

Re: Thought Experiment : True* On-Base Percentage

Post by vinsanity »

Big Amoco Sign wrote:
June 23 22, 4:23 pm
I'll bite.
A little passive aggressive so I will, too.
vinsanity wrote:
June 23 22, 4:06 pm
Is just "getting to 1B" currently unmeasured? Is it a skill?
Big Amoco Sign wrote:
June 23 22, 4:23 pm
Why not use xwOBA and xwOBACON and xBA/xSLG to handle these issues and take fielders out entirely (well at least their non-normalized fielding play)?
All you had to say...I was just trying to muse; not a gotcha or anything. Thought it'd be a fun thought experiment.
Big Amoco Sign wrote:
June 23 22, 4:23 pm
Does someone ahead of them get out? If so, swapping the runner for an out isn't a net positive on any level. Even if that is a speedy player being swapped, an out still hurts the team. Is there a fielder choice? Sacrifices already already are already not ABs that count against BA.
But a FC isn't a sacrifice and counts as an AB...so if you're saving an out by getting on 1B because of FC, that has value, no? Hell, I believe hitting behind a runner to advance them can still be ruled a FC and not a SAC by the scorer...but more than ever that's a skill. If you force a throwing error because you're fast, that has value, no? There's lot's of gaps with current stats and exceptions and none tell an entire story.
Big Amoco Sign wrote:
June 23 22, 4:23 pm
Fangraphs wrote about "reached on error" before:

https://blogs.fangraphs.com/the-error-o ... -on-error/
And that's what I was looking for.

So if Ichiro reaches first on a throwing error more than Adam Dunn - isn't that a skill? Why does a home-run count as on-base? Cause "on-base percentage" is just a misnomer for a "truly positive at-bat outcome"? I'm not sure it happens enough to be worth worrying about based everything else....but thirty years ago they said the same with adding walks to batting average.

Again, was just something I thought of and wanted to discuss. Nothing serious.

User avatar
Big Amoco Sign
Master of Hyperbole
Posts: 14402
Joined: December 1 17, 11:05 am

Re: Thought Experiment : True* On-Base Percentage

Post by Big Amoco Sign »

Not sure if serious on passive aggressive comment. It wasn't. This is a good post. More just mean I'll kick it off with a thought. I was just typing as I was thinking on most of that reply. Edited some of it. Wasn't too serious either.

My point with expected-stats was that you could get a lot of that true data you seek from using that + sprint speed.

ROE on the infield, should, at the very least, not count against the BA. That's the best point here I think. I really hate that that one is counted against the hitter when it's a scoring judgement. But, expected data cuts out that stuff of course.

User avatar
CardsofSTL
All Hail the GDT Master
Posts: 47817
Joined: April 26 11, 6:06 am
Location: Columbus, OH

Re: Thought Experiment : True* On-Base Percentage

Post by CardsofSTL »

Interesting subject. I'm not enough of a STAT head to delve into numbers but the speed factor comes into play. Runners like Bader and O'Neill will beat out a ball that's a GIDP by other hitters, so there's value there that's maybe not quantified.

Also big fan of bacon that I think was mentioned earlier.

User avatar
heyzeus
Everday Unicorn
Posts: 41342
Joined: April 21 06, 10:14 am
Location: Austin, TX
Contact:

Re: Thought Experiment : True* On-Base Percentage

Post by heyzeus »

I think i conceptually disagree with excluding home runs from whatever type of measure of utility this is, though, because a homerun is the thing of most utility. I'm of the opinion that it never "kills" a rally; it in fact contributes the most important thing for a rally - instant run(s).

User avatar
MrCrowesGarden
'Burb Boy
Posts: 23631
Joined: July 9 06, 11:33 am
Location: Out of the Loop

Re: Thought Experiment : True* On-Base Percentage

Post by MrCrowesGarden »

vinsanity wrote:
June 23 22, 4:06 pm
After a semi-pedantic discussion with someone around here, I was actually musing - would a true "on-base" percentage statistic have any value?

Preface - This is not a critique of the current stat of "on-base percentage". I think it has a worthwhile measure but I do think there's some outliers that are arguably a skill.

If you hit a homerun - you're not getting the pitcher in to the stretch. Home runs are called rally killers and you're not technically ever "on-base".

I'd be willing to bet someone like Ichiro reached first more often on an error than Adam Dunn because of rushed or bad throws that may or may not be called IF singles.

What about guys who reach 1B on a FC because they leg out a GIDP? Or guys who hit absolute rockets the SS can't handle?

Is just "getting to 1B" currently unmeasured? Is it a skill? Does measuring have any value?

I really don't have a thought about your experiment, but this line got me to thinking about other situations. How often do pitchers work from the stretch when there's a runner on 3rd, and only a runner on 3rd? The chance of a steal of home is astronomically low. Not quite as low, but still extremely low, is the likelihood of getting a good jump off 3rd making the difference in whether that run scores on a ball in play. And again, slightly higher-- but still low in the grand scheme of things-- is the possibility of a wild pitch or passed ball. Does the windup vs. stretch approach change if there are two outs?

User avatar
vinsanity
Chili dog truther
Posts: 8772
Joined: July 3 06, 2:19 pm
Location: Indianapolis

Re: Thought Experiment : True* On-Base Percentage

Post by vinsanity »

MrCrowesGarden wrote:
June 24 22, 10:13 am
How often do pitchers work from the stretch when there's a runner on 3rd, and only a runner on 3rd? Does the windup vs. stretch approach change if there are two outs?
I think this is an interesting outlier, too. I'm curious how often someone ends up with a runner on only third.

User avatar
vinsanity
Chili dog truther
Posts: 8772
Joined: July 3 06, 2:19 pm
Location: Indianapolis

Re: Thought Experiment : True* On-Base Percentage

Post by vinsanity »

Big Amoco Sign wrote:
June 23 22, 6:17 pm
Not sure if serious on passive aggressive comment. It wasn't. This is a good post. More just mean I'll kick it off with a thought.
My bad, the problem with text can be tone is tough to read. I read it wrong and that is wholly my mistake.
Big Amoco Sign wrote:
June 23 22, 6:17 pm
My point with expected-stats was that you could get a lot of that true data you seek from using that + sprint speed.

ROE on the infield, should, at the very least, not count against the BA. That's the best point here I think. I really hate that that one is counted against the hitter when it's a scoring judgement. But, expected data cuts out that stuff of course.
Yea, I think the thing is that it doesn't happen enough to not be captured elsewhere, at least secondarily. Things even out. So that's why I wanted to ruminate on it; see if anyone else had thought about it. I imagine there's some value to the skill but I can't imagine it adds up to even 10 whole runs a season for a player. Maybe a difference of 5-10 runs from best to worst over a season?

User avatar
Kincaid
Veteran Player
Posts: 654
Joined: June 15 09, 11:03 am

Re: Thought Experiment : True* On-Base Percentage

Post by Kincaid »

I also don't see the utility of excluding home runs since OBP is more useful as a measure of not making outs than of making the pitcher throw from the stretch. ROE and avoiding double plays are definitely skills related to not making outs, though, and it would make sense to measure them.

Run expectancy (RE24 and REW on FanGraphs) might be the closest to what you are talking about because it just looks at the base-out state before and after the play without regard for how it happened. If a batter ends up on first, the change in run expectancy is the same whether it was via hit, walk, or error (at least assuming there are no other runners). If he hits into a FC but beats out the double play attempt, the resulting run expectancy will be better than if he had hit into a double play. Run expectancy measures more than just on-base ability, though, so it's not really a replacement for OBP.

ROE not counting for a hitter's stats because it is a fielding mistake rather than a batting success is similar logic to why AVG (excluding walks) was initially chosen as the primary batting measure over OBP (it was assumed that walks were a pitching mistake rather than a batting success). Of course the difference is that for most hitters the impact of their ability to draw walks is far greater than the impact of their ability to induce fielding errors, and errors at least seem to be more random occurrences than walks.

Part of why GIDP and ROE are often ignored even by people who believe they matter is that the effect usually isn't that big, so it's not necessarily worth the extra effort of incorporating them into your source data when you could otherwise just use standard box score stats. And some of the skills relevant to GIDP and ROE are inversely related, so being better at one can be offset by being worse at the other (speed tends to improve both, but hitters who put the ball in play more often, hit more ground balls, and are right-handed tend to reach base on error more often but also hit into more double plays), which helps a little in letting you ignore them. That doesn't mean it is correct to ignore them, just that the convenience of ignoring them isn't usually a huge deal. For a small subset of players who are exceptionally good or bad at these skills, it will be important to consider, though.

Post Reply