Speed, Power and Contribution to the Offense

[NorthSideSox72]

The thought is good but its already been done, far better than anyone here could hope w/o putting months and maybe years into their formulas. Its called EqA and VORP is good too, WARP is even better because it has defense.

VORP isn't really much like this stat. Its closest to EqA or Total Average. You are cetainly correct that people who put months or years into this kind of work will come up with much more intricate and detailed stats than I ever could, and I am sure more reliable for their purposes. But you are incorrect to say its been done, for the simple fact that it hasn't (at least not in wide use). Its not the same as some of the other ones, so its got slightly different results.

 

BTW, I did add in adjustors for SF & SH, the value of a hit over a walk and a hit over an SB. I'll break out the new numbers today or tomorrow for people to look at. It will also have a projected extra-base-on-hit adjustment.

 

[NorthSideSox72]

OK, I promised I'd make some adjustments based on some of the feedback. So I have. Below is the result of someone between assignments at work, bored in front of the computer. Final stats at the end...

 

OK, I am adding 4 adjustors to the measure.

 

1. As Mr. Civic points out, in some situations, a hit is better than a walk, double better than a walk and SB, etc. So let's do this: Situations where there are runners in scoring position, a walk or HBP is only half as valuable as a hit. Since runners in scoring position seems to be the case about 27% of the time, use that half-adjust on 27% of walks and HBP, and keep the others whole. That should adjust for the non-hit.

 

2. It is also true that SB's aren't as valuable as hits, with runners on (scoring position or not). But, since we are using TB and not H in the numerator, we are really looking at the difference between IsoD and SB-CS. Those two numbers should not be values the same - a 2B's value over a 1B is more than a SB's value is to a 1B, for example, with runners on. How much more valuable is arguable, but the base consideration is that an extra base via the hit helps the hitters in front AND behind, whereas the SB only helps the hitter behind. So let's use 50%. Runners on % among at bats is about 45%. So for 45% of SB's, multiply by .5. This also functions to make CS have a higher absolute value than SB, addressing some other folks’ concerns.

 

3. I have added SF’s and SH’s to the numerator to add value for those outs that create a base movement of a player on base. These do not need an adjustment as they only can be recorded with men on base anyway.

 

4. Finally, what to do about speed and taking an extra base on hits. This raw number is simply not available, but it certainly is part of the game. In order to project the number, you would need a formula like this:

 

XBOH = TPA*OBP*LBA*SUCCESS*SPEEDADJUST

 

Total plate appearances, OBP (how often actually on base), League Batting Average (how often a hit occurs), Success (how often, league-wide, taking the extra base occurs) and finally a speed adjustment factor. All of those except the speed factor are available as hard stats, and I used them. The question is, what is the difference between an Owens/Ozuna type, and a Konerko/AJP type, in terms of how often they can take the extra base. To be safe, I’d say that at most, over the same X number of chances, that the fastest players might take the extra base 3 times as often as the slowest. The league average sits in the middle. So, we establish a speed adjustment range of 0.5 to 1.5. Paul Konerko I rate a 0.5, so he is 50% as likely to take that extra base as a league average runner. Jerry Owens is 50% more likely than a league average runner to take that extra base. Here are the speed factors I assigned:

 

Owens: 1.4

Ozuna: 1.3

Pods: 1.3

Richar: 1.2

Terrero: 1.1

Erstad: 1.1

Fields: 1.0

Cintron: 1.0

Gonzalez: 0.9

Dye: 0.8

Uribe: 0.7

Hall: 0.6

Crede: 0.6

AJP: 0.6

Thome: 0.6

Konerko: 0.5

 

Given those factors, this is the number of extra bases each player has probably taken over their season (remember that this is greatly dictated by the number of times on base, because speed is useless if you are in the dugout):

 

Owens: 15.1

Pods: 7.5

Richar: 5.8

Terrero: 3.9

Erstad: 10.3

Fields: 10.5

Cintron: 5.0

Gonzalez: 4.9

Dye: 12.3

Uribe: 9.6

Hall: 1.2

AJP: 8.3

Thome: 11.3

Konerko: 9.7

 

That gives us the new formula:

 

TB + [bB – .5(.27 * BB)] + [HBP - .5(.27 * HBP)] + [sB – CS – .5(.45 * SB)] + SF + SH + XBOH – GIDP / Total PA

 

So for everyone, here are the final numbers, with data through yesterday’s action:

 

Jim Thome: .611

Paul Konerko: .518

Jermaine Dye: .514

Josh Fields: .507

Joe Crede: .505 (***05+06 numbers)

Luis Terrero: .475

Danny Richar: .471

Pablo Ozuna: .446 (***05+06 numbers)

Juan Uribe: .438

Jerry Owens: .437

Darin Erstad: .423

Scott Podsednik: .413

AJ Pierzynski: .410

Alex Cintron: .358

Andy Gonalez: .310

Toby Hall: .238

 

Comments:

--The changes moved a few players around, but not that much

--AJ really isn't looking good

--Fields is very impressive for a rookie

--Terrero, Uribe and Crede are higher than I would have thought

--AJ is lower than I would have thought

 

Part of the reason for this was to see how much speed came into play. Owens, even with speed factored in, isn't great. For a rookie he isn't bad, but he is still nothing spectacular. These numbers are his full season, by the way, not just since his July recall - his number since his July recall is .466, around the same as Richar. Combine all that, and I think he is marginal as a starter right now, but he is also a rookie. I'd say he is best off as a 4th OF on a competitive team for now, but could start and probably get better with a lower bracket team. He might actually fetch something in trade, if the Sox pick up a better CF. Something to think about. If not, and of significant upgrades are done at SS and maybe other positions, I think you do could a lot worse than Owens in CF.

 

[Dogfood22]

NorthSideSox:

 

This is the best post I've ever seen here at SoxTalk. Great job!

 

(I meant the first post in the thread)

Edited September 24, 2007 by Dogfood22

[Dogfood22]

I think a good stat to come up with would be similar to what you first said. Just TB+HBP+BB / PA. Find out how many bases a player takes per PA and leave it at that. SB's are a huge help, especially for for leadoff men, but it is hard to make up any stat regarding the usefulness of a SB.

 

For example, let's say the leadoff man gets on base with no outs, steals 2B on a pitch out, and then the #2 hitter gets a shot at advancing him or scoring him with a 1-0 advantage. Contrast that to the same situation except the leadoff guy steals 2B but only after leaving the hitter at the plate with a two strike count. This immediately lowers the chance of the hitter getting a base hit and it also takes away an opportunity to bunt in order to advance the runner to 3B.

 

Also, you'd have to factor in situations where a SB occurs during a failed hit and run which results in a K of the hitter. While the SB is still productive, especially with less than 2 outs because it eliminates a GIDP possibility, it doesn't mean as much because an out was recorded in the process.

 

And then you have those situations where a guy will steal say a base with two outs in the inning. Again, the SB is helpful, but it still is reliant upon a basehit and thus can not be truly quantified at all. The only way you can judge the value of a SB w/ 2 outs would be if you gave the SB the same value as the hitter at the plate's BA with RSIP, because that average would most closely represent the chance of that baserunner scoring. Plus there are other factors that would have to be considered for a truer stat, like the arm strength and accuracy of the opposing defense, what base the baserunner is on (2B or 3B), and size of the ballpark.

 

You're never going to come up with a stat to truly explain the value of speed + power + contribution to an offense, so the best you can hope to do is find a relatively simple, sensical formula. You were much closer to that in the opening post than you are in the last one.

Edited September 24, 2007 by Dogfood22

[kapkomet]

Are you sure Konerko isn't a .01 on the speed factor?

 

Nice adjustment! But it doesn't really alter the #'s all that much - which I'm sort of surprised at.

 

[NorthSideSox72]

Are you sure Konerko isn't a .01 on the speed factor?

 

Nice adjustment! But it doesn't really alter the #'s all that much - which I'm sort of surprised at.

Well, as I mentioned, I made some adjustments to the speed issue - some would help speedier players (XBOH), others would hurt (value of SB slightly less than hit). I think both those adjustments are valid, though they happen to offset pretty well for many players.

 

I'm pretty happy with it either way, but I think the newer one is more accurate.

 

Toby Hall is not good.

 

[Dogfood22]

HBP is a big factor for a leadoff guy like Reed Johnson, who might be available this offseason--probably for a reasonable price.

 

http://torontosun.com/Sports/Baseball/2007...517163-sun.html

 

The sox have few internal options for a #1 or a #2 hitter-both huge sox needs. Free agency looks weak. It's going to take a few trades to get the necessary peices. I know he'll be a bit of risk following his surgery--much like Crede. The sox could acquire Johnson, saying they'll play him in CF, and have Fields in LF and Joe in 3b during spring. Joe proves he's healthy to other teams, then move him. I give Fields credit for moving to LF, and not having a drop off at the plate. Yet his long term future is probably at 3b for the sox as soon as 2009.

This whole situation is exactly why I want the Sox to just move forward without Crede. I love Joe, but the uncertainty of his situation is too much when the future is already very uncertain for the Sox. If the Sox get whatever they can for Joe, start Owens in LF and have him leadoff, start Fields at 3B, and either start Anderson/Sweeney or someone else in CF, we'll have a much better idea of our situation in 2009. Especially if we can find a way to deal both Contreras and Garland so that we can work in two of Floyd/Gio/Egbert/Broadway in the rotation. That way when 2009 comes around we'll know what we do or do not have in LF, CF, 3B, 2B, and 3 rotation spots including Danks. Plus we'll have better ideas in the bullpen and if we can pick up a good looking young SS we'll have a better idea there too.

 

I think the worst thing the Sox can do in 2008 is exactly what they are planning to do. If they don't commit to at least one year of relative rebuilding they are going to f*** themselves in 2009, which will f*** them over in 2010, and who knows how long that can carry on. We don't need to trade everyone on the team; let's just please find out what we have first and go from there.

[witesoxfan]

This whole situation is exactly why I want the Sox to just move forward without Crede. I love Joe, but the uncertainty of his situation is too much when the future is already very uncertain for the Sox.

 

I really don't think you would have seen Fields move to LF if the Sox felt there was something that could have gone wrong. They have to like where Joe is at in his rehab. That, or they just felt Fields was never going to cut it at 3B, so the move to LF happens while the team is struggling so he can get adjusted to it.

 

If the Sox get whatever they can for Joe

 

That's part of it too, they'll get absolutely jacks*** for him. He doesn't have negative value, but his salary next year is going to be atleast $4 mill, he's coming off of back surgery, his offensive production depends entirely upon his average because he doesn't walk and he's not a strong enough power hitter - unlike say even Josh Fields - to be of much value otherwise, and there's no ability for long-term investment because he'll be a free agent after next season...his value has diminished incredibly.

 

If the Sox are serious about competing next year, I'm not sure you'll find a better option at 3B for the price; it's a huge risk, and without any insurance at all at 3B - which the organization really doesn't, unless you count Fields, and then if you're counting Fields you have to depend on any number of AAAA players (or Ryan Sweeney, and some will probably argue he is a AAAA player) to play LF.

[Balta1701]

Just out of curiosity, what happens to Owens's numbers when you only include his numbers since his most recent callup (and thus exclude his terrible numbers during his first callup?)

[beautox]

The thought is good but its already been done, far better than anyone here could hope w/o putting months and maybe years into their formulas. Its called EqA and VORP is good too, WARP is even better because it has defense.

 

whats the formula for warp3 ?

[Vance Law]

4. Finally, what to do about speed and taking an extra base on hits. This raw number is simply not available, but it certainly is part of the game.

 

Great stuff, Northside. I mentioned this before, and I think it's what you're looking for. In the Bill James Handbook, he lists the raw numbers for taking extra bases on hits.

 

 

League Average #s for 2006:

A) 1st to 3rd on a single- 28%

B) Scoring from 2nd on a single- 60%

C) Scoring from 1st on a double- 42%

 

ex.

Pods:

A) 6/21 29%

B) 11/22 50%

C) 4/7 57%

 

Konerko

1/26 4%

9/29 31%

1/6 17%

[NorthSideSox72]

Great stuff, Northside. I mentioned this before, and I think it's what you're looking for. In the Bill James Handbook, he lists the raw numbers for taking extra bases on hits.

League Average #s for 2006:

A) 1st to 3rd on a single- 28%

B) Scoring from 2nd on a single- 60%

C) Scoring from 1st on a double- 42%

 

ex.

Pods:

A) 6/21 29%

B) 11/22 50%

C) 4/7 57%

 

Konerko

1/26 4%

9/29 31%

1/6 17%

Yeah, I used numbers from a site that were very close to those, to determine the SUCCESS part of that equation.

 

[GoSox05]

Here's a number - 188. That's how many at bats Andy Gonzalez has this year. That is not a good recipe for winning.

[NorthSideSox72]

Just out of curiosity, what happens to Owens's numbers when you only include his numbers since his most recent callup (and thus exclude his terrible numbers during his first callup?)

As I noted in the last graf of my very, very long post (sorry), his number since his July callup is .466. That compares most closely with Richar (.471), and of course its 40+ points better than Erstad and 50+ better than Pods (for OF comparison's sake).

 

I personally think that we're seeing pretty darn good seasons from 3 rookies this year - Fields, Richar and Owens. In a lousy year, that is something to smile at.

 

Just for fun, I ran Chris Young's number for this year, his first full season (he played 30 games in 2006). Assuming the same speed factor as Owens (1.4), he is at .568, which is far higher than everyone on the Sox other than Thome. Very, very impressive for a rookie.

 

[witesoxfan]

Here's a number - 188. That's how many at bats Andy Gonzalez has this year. That is not a good recipe for winning.

 

It was a pretty piss poor team before Andy Gonzalez had 1 at bat. He was merely a stopgap to guarantee the team lost and secured a very good pick in next year's draft.