Tuesday, March 11, 2008

CRAP: Offense

So we've made some projections for the outfield, infield, and catchers. If we've got performance, and we can take a swing at playing time, we can estimate runs.

I'm going to use the runs created cheat that I've used a few times. It's definitely not the best. It's not the most accurate. But, as usual, I'm looking for a range, not a number with three significant digits. When I did this little exercise last year, I came up with 697 runs scored. They scored 673. Off by 24... not great, but close enough.

The basic estimate formula I use is AB * on-base percentage * slugging average. As a spot check, last season, the Nats had 5520 ABS, a .325 OBP, and a .390 SLG. Multiply it out, and you're at 699, right about where my initial guess was. Close enough!

The key is estimating playing time. To do that, I looked at the average # of ABs each Nats position had last year, as well as what the league average # of ABs was, and assigned an AB total. I then divvied up the AB totals between the various contenders for playing time to get an AB estimate for each player at each position.

At catcher, for example, LoDuca gets 60% of the PT, Estrada 35% and Flores 5%. Multiply the OBP and SLG projections from the last few days' posts by those AB and sum up the position, and we can expect about 68 runs from our catchers this year. Last year, the Nats got about 64 runs, so it's a slight increase.

Rather than walking through the entire sheet position, by position, check it out yourself.

Quick note on a few of the PT assumptions:
-- NJ gets 65% of the time at 1B and Young gets about 200 ABs.
-- I've scaled back Lopez' playing time, giving him 25% at 2B and 30% at SS.
-- Dukes serves as the 4th outfielder, splitting time at all 3 positions, getting roughly 400 ABs all over the place.
-- For PHers and P, I just used the league average numbers. Because their AB totals are so low, even big sweeps here aren't going to change the overall numbers much.
-- I totaled up the team ABs, and they're within 50 of league average, within 1% of the total.

  • What's the bottom line?

    We're due for about 795 runs, an improvement of about 120 over last year. I'd guess that ~half that is due to the move from the park. The rest is due to position-by-position improvement.

    795 runs would've put the team 6th in the league, up from dead last. Since we're not looking for precision, it's probably better to look at the range: there are a lot of teams bunched up together just below Philly and COL. +/- 15 runs, and the Nats are likely to finish in the 3-7 range, well above league average.

  • Where does the improvement come from?

    If you look at the full chart, next to the positional box, I list the 2007 total for that position, using the same ab*obp*slg methodology. The two biggest improvements are in left and center, where the Nats are using Dukes to fill in the hundreds of deadweight ABs given to Logan, Langerhans, Fick and Casto. Replacing their cumulative line with Dukes' potential (if he hits it!) is a pretty big net plus for the team.

  • What if they trade NJ?
    Here's where it gets fun. We can make a few fun assumptions to get a rough estimate of various decisions.

    Trade NJ and let DY play full time? 779 runs, a drop of about 20.

    Cut Lopez? 793 runs, basically the same.

    Lo Duca gets suspended and Flores plays every day? 799 runs, a slight uptick.

    Zimmerman hits .300/40/120 and the Ladson Triple Crown? 840 runs

    Zimmerman breaks his hand early in the season and Boone fills in? 760 runs.

  • So a lot can really happen. And it ain't perfect. But there's a pretty good chance that this is going to be a good offensive club.

    Now about that pitching...


    • Was the SS combo really 15 runs better (in 30 more ABs) than Ryan Church in LF? Color me shocked

      By Anonymous Anonymous, at 3/11/2008 2:57 PM  

    • The splits by position are on BBREF.

      Church did all his damage early in the season as a CFer. He gave up hitting pretty much the day he shifted to LF: .268/ .333/ .426.

      Similar thing with Lopez. We know that Guzman hit well, but when he played SS, Lopez hit .260 .331 .356.

      LFers and SS had the same slugging, but the SS out-OBP'd the LFers by about 40 points.

      By Blogger Chris Needham, at 3/11/2008 3:00 PM  

    • Fun with pseudo-numbers! Interesting post, Chris - I might quibble that Dukes probably gets ~50-70 less ABs and Pena and Milledge get slightly more, but I don't think it would change things all that much.

      About that pitching - presuming the pitching is as bad as it was last year (783 runs allowed) yields a Pyth projection of 82 wins. Both of those are just WAGs, but I think a presumption that any improvement in the starters may be offset by the park isn't unreasonable. The overall result has that certain truthiness - I sort of feel like this team ought to be in that .500 range - if the starters turn out to be a lot better, I think we might surprise some people.

      By Anonymous Anonymous, at 3/11/2008 3:01 PM  

    • The good thing about projecting the outfielders is that they're all basically of similar quality.

      some get on base more. some hit for more power.

      in this shorthand formula, little shifts in playing time of players of similar quality don't really much matter.

      By Blogger Chris Needham, at 3/11/2008 3:17 PM  

    • chris - what's your gut telling you re flop? obviously, it's telling you something.

      was he the recipient of the dreaded "get your ass in gear" speech from jimbo? you know, the speech in which he gets to take credit if said player does, in fact, get said ass in gear (i.e., bergmann, jason)

      By Blogger DCPowerGator, at 3/11/2008 4:10 PM  

    • I'm definitely more on FLop's side than half the fans, but...

      In my initial set of projections, I had given him a greater share of the infield ABs. I scaled him back because he's been dead so far in spring, and it's going to take him some time, if he does succeed, to wrest more PT away, barring trade or injury.

      By Blogger Chris Needham, at 3/11/2008 4:12 PM  

    • It seems you have had a slow but steady progression towards reality with Lopez. If we had cut him we would have lost 0 runs and get $5 Million to spend on pitching. Interesting.... now if we had only signed Adam Everett from Houston as a defensive replacement for Guz and maybe a pitcher like Jennings....oh well, Players' attitudes do matter in baseball and Lopez issues are so much more than one bad year. We can agree to disagree but at somepoint you may have to face the fact that Lopez is no longer good for what his past numbers say he should be.

      By Anonymous Anonymous, at 3/11/2008 4:25 PM  

    • Well, when he sucks all this year, too, I'll go along with it. Til then, we just don't know.

      I'm with you on Everett. But the team didn't want to do that.

      And saving money for pitching by cutting Lopez doesn't make sense. The team didn't NOT sign a FA pitcher because they didn't have the money. If they cut Lopez, they'd be saving money for the purposes of saving money. Yawn.

      By Blogger Chris Needham, at 3/11/2008 4:26 PM  

    • So do they think they have the pitching already is that it Chris?

      I really don't see why they are so adverse to getting some vet pitching in here to hold them over 2-3 years until the young guys are ready.

      On Lopez....we shall see...I know you do not give any weight to what his eyes told us last year but it seems more of the same this year (from the very little I can tell on a MLB.com feed.......)

      By Anonymous Anonymous, at 3/11/2008 4:56 PM  

    • Chris, did you buy Hanrahan some 'roids for Christmas?

      He just struck out 6 Braves today....

      By Blogger Michael Taylor, at 3/11/2008 8:57 PM  

    • It's only 5. He's a bum!

      By Blogger Chris Needham, at 3/11/2008 8:59 PM  

    • Never mind. That's 8 in 3 innings now...

      By Blogger Michael Taylor, at 3/11/2008 9:00 PM  

    • Get gameday audio--mlb.com's box scores are always slow.

      I just got to hear three innings of Hanrahan striking out the side!!!!!!!

      I'm not really sure how I feel now, but I'm kind of light-headed.

      By Blogger Michael Taylor, at 3/11/2008 9:02 PM  

    • ah, my gameday thing froze. Guess the data enterer took a nap.

      Hanrahan is out of of options, I think. I'm not sure they'd be able to sneak him down.

      I wonder what this means with some of the other guys in the pen...

      By Blogger Chris Needham, at 3/11/2008 9:04 PM  

    • Well, from a quick peek at baseball reference, Hanrahan never actually had a problem striking people out. Last year's short stint had him at around 7 K/9 and he had around 8.5 K/9 in his last year at Columbus. Looks like his real problems have been with control, even in the minors. Also, giving up on him totally might not be a good idea, since his 50 innings last year were just barely enough to get him acclimated while his minor league ERAs the past two years were in the mid 3s. He's not great, but he's also not hopeless. Maybe we'll find the roster space to give him a shot, maybe not. He and Lannan both seem to want it bad this spring though, that's for sure.

      By Blogger Michael Taylor, at 3/11/2008 9:20 PM  

    • Chris- I checked out your method over the past 3 years of National League baseball and it pretty consistently (46 times out of 48) overpredicts runs scored. It's not much but if you use a multiplier like say... .965 you'd get 767 runs scored not 795.

      Not that it's not an great jump but it puts the Nats near league average rather than clearly in the upper half.

      By Blogger Harper, at 3/11/2008 9:31 PM  

    • Thanks for checking that. It's not supposed to be an exact science -- there are better tools for that.

      But I appreciate you checking it out!

      770-something still has them in that same rough range, though they're at the bottom of that upper tier.

      By Blogger Chris Needham, at 3/11/2008 9:38 PM  

    • By Blogger Unknown, at 10/25/2009 8:18 PM  

    Post a Comment

    << Home