Forecasting OXS/OPS plus examination of Brady Clark

[1992casey]

In the "Who should we aim to sign?" thread on the Major League forum, Al wrote:

---

OXS is 98% accurate in forecasting runs scored. If it's not your god, you should change your place of worship. ...

---

I certainly accept the accuracy of OXS and OPS. But at the risk of splitting hairs, this quote has me wondering if it wouldn't be more accurate to say that OXS is 98% accurate in correlating to runs scored? Don't other stats do a better job at predicting OXS/OPS than OXS/OPS themselves?

[dadofandrew]

Forecasting OXS isn't difficult to get close, just by using career numbers and how old they are. Many folks don't get it though, as the number of folks who think Clark is having a "career" year seems to prove.

[1992casey]

Previous OXS/OPS and age?including a cursory examination of comparable players?is generally what I would look at. Don't the more complex systems rely a lot on other stats, though, such as K/BB ratio?

 

As far as Brady goes, he's a perfect example of why we rely on stats. People rely an awful lot on memory, and what sticks out in their minds is his career month in May.

[bjamesw22]

Many folks don't get it though, as the number of folks who think Clark is having a "career" year seems to prove.

 

Yeah, well you're saying this now after his descension after the All-Star break.

 

His value was at an all-time high at the end of June-mid July and that's when we should have traded him. He was having a career year up to that point and we should have traded him at his peek. We lost value on him now. The chances of him performing like he did for half a season are slim to none at the age of 33 next year.

[pogokat]

actually, Brady's OPS isn't that much higher than it was last year it's just that this season he has chosen to hit singles instead of taking walks, which pretty much evens out his seemingly better play this year..

[bjamesw22]

I know it isn't now....but I was advocating trading him at the break while he was at his peak value. He's lost value since then.

 

On June 23rd he had an .850 OPS and even during July he was consistently above .800 and had between an .809 and .818 OPS the whole last week before the trade deadline. He should have been traded.

[pogokat]

using that logic, we should have traded carlos lee and bill hall as well, and geoff jenkins asap...

 

if they'd traded clark in june or july,magruder would have started an additional 20 games or so since then, causing mass suicide amongst the faithful...

 

clark should still have plenty of value after the year...

[bjamesw22]

using that logic, we should have traded carlos lee and bill hall as well, and geoff jenkins asap...

 

Yeah, except Clark will be 33 next year and less likely than all of those players to repeat these stats.

 

Plus, Hart was just as ready at the end of July as he is now. I would have had no problem having Hart start if we could have packaged Clark with some other players(Ohka? Overbay? JDLR? Prospect?) to net us an upper tier pitcher like Pineiro, Sabbathia, et.

 

I think their is a fair chance that one of Jenkins or Lee will be traded this offseason also. We'll see what happens.

 

I don't think Clark will have nearly the value he did in late June and late July. Then again, I guess it all depends on how he performs at the end of the year.

[1992casey]

Clark's month by month OPS:

 

April: .806

May: .932

June: .696

July: .799

August: .530

 

Because of sample size, August can't really count. May is high, but lots of players have months like that. Nothing else really deviates very far from his .759 career number.

 

In a case like Brady, who can be signed year to year, I wouldn't be all that afraid to figure he'd stay close to his career numbers next year, even at 34. If we were talking about having to sign him for 3 years, I'd think a lot differently.

[1992casey]

---

it's just that this season he has chosen to hit singles instead of taking walks

---

Of course, singles are more valuable in the scheme of things.

 

But relating to the original question of predictability, wouldn't the drop in walks be a bad sign? Might some of the singles be attributable to luck?

[pogokat]

But relating to the original question of predictability, wouldn't the drop in walks be a bad sign? Might some of the singles be attributable to luck?

 

absolutely...if brady were slightly less lucky next year, he might slip to around .700 in ops...which would make him replacement level...

 

i think corey hart should start in cf next year...

[bjamesw22]

i think corey hart should start in cf next year...

 

Good call.

[1992casey]

---

if brady were slightly less lucky next year, he might slip to around .700 in ops...which would make him replacement level...

---

pogokat,

 

That addresses the kind of stuff I had in mind when I originally started this thread.

 

And it would seem to fit in with Al's comment when he says that "forecasting OXS isn't difficult to get close. But it would also seem, then, that unless we're either getting into a more sophisticated model or at least eyeballing things like walk rates, a 50 OPS point fluctuation either way due to luck shouldn't be an unexpected surprise at all.

 

I don't look at batting average all that often, but I wonder if this wouldn't be an appropriate time to consider Brady's 2005 BA vs. career (or recent) numbers, comparing the difference between it and his OBP.

 

I'm tending to agree with you; it's probably a very good time to consider Corey Hart for CF if he demonstrates a comfort level there over a larger sample of games.

 

Still, as I mentioned above, I don't think Brady is any huge contract risk for next year, even as a relatively pricey 4th outfielder. I wouldn't rush to trade him, but I think it's very possible that another GM might happily overvalue him, too.

[brettac1]

His walk rate was the exact reason why I liked him, now it's becoming a reason why I don't like him. I don't know his IsoOB's off-hand, but they definitely should be posted.

 

In seasons with over 100 AB:

2001: .109

2003: .057

2004: .105

2005: .061

Career: .076

Minor Leagues: .092

[pogokat]

yeah, what to do with those numbers???? Maybe last year was the fluke...was he batting 8th alot last year???could he have inflated walk totals for some other reason...is he hacking earlier in counts this season???have his k's gone up???

 

Whatever the dif. is, all signs point to massive decline next year...

 

maybe we can dig up devon white and play him in cf...

[1992casey]

Wow. It appears that Brady's allegedly increased production is essentially entirely due to singles. Home run and doubles rates are almost exactly the same. (Two triples in three years make them irrelevant.)

 

Home runs, 2004: 7 homers in 420 PS = 1 per 60.0 PA

Home runs, 2005: 9 homers in 528 PA = 1 per 56.8 PA

 

Doubles, 2004: 18 doubles in 420 PA = 1 per 23.3 PA

Doubles, 2005: 24 doubles in 528 PA = 1 per 22 PA

 

Would his BABIP (batting average on balls in play) vs. the team be something to look at?

[brettac1]

Would his BABIP (batting average on balls in play) vs. the team be something to look at?

 

Team BABIP: .297

Clark's 2005 BABIP: .329

Clark's 2004 BABIP: .309

Clark's career BABIP entering 2005: .291

 

EDIT: Added more BABIP numbers

[1992casey]

Can we assume that the fact that Brady's BABIP vs. the team is another warning sign, or should we be looking at something more complex, such as groundball, flyball, and line drive rates?

[rluzinski]

Dan Fox breaks down different types of BIPs (balls in play) and how often they are hits HERE. Here is the most important table:

 BIP Pct Outs Pct Ground 60234 42% 45433 75% Line 25654 18% 6712 26% Fly 47693 33% 37186 78% Pop 11010 8% 10784 98% Single Double Triple Homerun Ground 45% 14% 42% 0% Line 46% 50% 28% 14% Fly 8% 35% 30% 86% Pop 1% 0% 0% 0% 

 

Obviously, the higher the LD% you have the higher BA you should have, as well as more extra base hits. I don't have time to get into this like I could but Lyle Overbay led the league last year in LD% with 28%. Clark has a 27.1% LD%, so either the gentlemen in Milwaukee who designates the balls is extremely generous to the Crew or Clark has earned his good stats. It doesn't seem like a fluke, since 24% of his hits were linedrives last year too.

 

Read the article above and HERE and do a search at hardballtimes.com if you want to learn more. Hardballtimes actually used an equation to eastimate what a person's BA,OBP and SLG should be. Bill Hall was one of the leaders in "luck."

[clancyphile]

Until he got hurt earlier this month, Brady Clark was a solid CF. He was hitting more singles as opposed to walks, and a single not only will score someone from third, it can also score someone from second.

 

What would I look at to predict? I'd look at the OPS as well as the components - taking into account not just the trends, but also where the OPS comes from. In 2003, his .733 OPS had a slight lean towards SLG, while in 2004, the .782 OPS was pretty well balanced between OBP and SLG.

this year, his .811 OPS seems to be leaning slightly toward SLG again. All in all, I think he's going to be a solid 4th OF for any contending club, or a starter for a team looking for someone to fill a hole for a year or two.

[1992casey]

---

Read the article above and HERE and do a search at hardballtimes.com if you want to learn more. Hardballtimes actually used an equation to eastimate what a person's BA,OBP and SLG should be. Bill Hall was one of the leaders in "luck."

---

rluzinski,

 

I'm still doing that homework assignment! http://forum.brewerfan.net/images/smilies/smile.gif

 

 

---

He was hitting more singles as opposed to walks, and a single not only will score someone from third, it can also score someone from second.

---

clancyphile,

 

I think it's a given that singles are more valuable than walks when related to the outcome of a game.

 

I think what a lot of us are suggesting is that there's more luck (statistical variance) involved in hitting singles than walks, meaning it's probably very reasonable to expect a swing in OPS.

 

I did some playing around, subtracting singles from Brady's numbers while leaving everything else the same. This work is a couple days old, but for the sake of illustration, it ought to be fine.

 

As of a couple days ago, Brady had a .304 batting average. It appears that its not unusual for a .281 career hitter to hit .304 for a season. Likewise, .265 wouldn't be out of the question, either.

 

When I gathered these numbers, Brady had an OBP/SLG/OPS of .370/.419/.789. Those numbers, of course, are above average for a CF.

 

Let's say Brady hits at his career batting average (.281), keeping doubles, homers, walks, and HBP exactly the same. His OBP/SLG/OPS turns into .349/.396/.746. That's OK, but not quite so impressive.

 

Now let's say Brady hits .265, keeping everything else the same. That turns his OBP/SLG/OPS into .334/.380/.714?not really good at all for a CF.

 

I'm probably being generous here. Brady's been hit by 16 pitches this season. Although on one hand, I think there's a degree of skill involved in being hit, I wouldn't expect him to be hit 16 times again.

 

One of us is going to have to take a good look at rluzinski's article and see what conclusions we can draw. What seemed to trip me up is that various sources split up the types of balls in play differently.