Nationals Baseball: How much better? - Part 2

Thursday, January 28, 2010

How much better? - Part 2

AsI blogged yesterday, the offense is not going to turn this ship around next year. Right now it will probably only be very very slightly better than last year. That isn't bad, because last year the Nats weren't bad offensively, but it's also not going to help turn the Nats into a respectable team (re: under 90 losses) by itself. What the Nats need is a massive turnaround in the pitching. They've made a fair number of changes, but is it enough?

First off, we need to figure out how many innings the Nats starters will pitch and how many runs they will give up doing it. The better the starting pitching is - the less the Nats will have to rely on relief pitching, which as we saw last year, is very... uhh... variable. Last year Nats starters threw 901.1 innings with an ERA of just under 5.00. However, 206.1 innings/3.89 ERA of that was by John Lannan (iBrow?). The rest of the rag tag bunch of misfits threw only 695 innings at a 5.30 ERA pace. This helps us get a sense of what the "average" Nats starter will do. Divide by 4 and you got your base Nats starter - 173.2 IP with a 5.30 ERA. Are the Nats going to have several pitchers throw nearly 30 games with those stats? No. But the "everyone else" pot will likely combine into this menagerie of below average slop. (If Zimmermann were around, or I could be promised Strasburg, things would be different. But he's isn't and I haven't been, so things aren't.)

The Nats have two sure things on the staff - Jason Marquis and John Lannan. How will they do next year? Well pitching is hard to grasp. What the projections say is probably about 190+ IP for Marquis with an ERA around 4.40 and high 180s IP at a 4.25 ERA for Lannan. Do I buy that? Sure. Seems reasonable and conservative. But I also think the Nats will want to push these guys a little more - get that 200IP to help the bullpen rest. So how about we make it easy - 200IP for both at a 4.50 ERA. Any one want to vociferously argue that? Follow this up with three Nats "average starters" and what does your rotation look like?

920 IP, 507 runs scored and a 4.96 ERA. The ERA is only ever so slightly better. It's only 20 more IP too. Will that make a big difference?

The average team throws around 1440 innings total (the worse you are generally the fewer innings your throw. All those bottom of the ninths away are lost- but right now lets stick with the averages. ) So the Nats relievers are going to need to throw 520 innings. That's roughly about as much as before but the pitchers throwing those innings are much better. Let's look at the projections for the guys the Nats are counting on:

Capps about 57 IP and a 3.70 ERA
Bruney 45 IP, 4.20 ERA
Bergmann 68 IP, 4.35 ERA
Clippard 75 IP, 4.00 ERA
Burnett 55 IP, 4.50 ERA

That's 300 IP and around 138 runs given up. There is still 220 IP left for the dregs. Now the average Nats reliever last year (including some of the guys above) would be a 5.09 ERA reliever. Can I use that? I think so. That number does include some of the good pitchers up there (so it would be higher). But it also includes good deal of quality innings from McDougal and Beimel. If the Nats could have another "find" (Hello, Drew Storen) they could easily hit that number. Let's just say the rest of the pen is 5.00 ERA worth. 5.00 ERA in 220 innings... that's 122 runs.

So let's add this all up....
122 runs given up by bad relievers, 138 by good ones, 307 by bad starters, 200 by good ones...
That's 767 runs given up. Last year the Nats gave up 874 runs. 874! I think will see some difference.

Last year the Nats scored 710 runs and gave up 874.
Next year the Nats project to score about 722 runs and will give up 767 runs. Based on the pythagoream theorem that would push the Nats to... 76-77 wins. Hey, how about that! Now do I believe that?

I think I do. Let's say the Nats were a 66 win team last year, based on runs scored. This is because they had an average offense coupled with an incredibly bad pitching staff, specifically in relief. So bad in fact, that probably just sitting on the guys they had that were ok (Clippard, Bergmann, and Burnett) and moving on from there would save them 30 runs. They aren't going to be that bad again. They aren't going to give up 50 more runs than the next worse NL team. It just won't happen. That "correction" in itself would happen to bring the Nats close to 70 wins. Add in the actual improvement made in both the rotation and relief corps... and well 76-77 seems a bit high but not out of the question.

My guess is that they'll lose a couple more thanks to the competetiveness of the NL East. so 74-75.

But, BUT there is an exciting part to be had. What if they do add Orlando Hudson? And what if they do add another 200 IP 4.50 ERA type pitcher (like Doug Davis in a bad year). I said yesterday - Orlando probably brings the run created total up by 30 rather than 12. Another solid starter... adds up to about 10 runs more of savings I think (a few runs off that rotation spot, and a few more lost off the back of the bad bullpen). That gets you to 742-757 and that my friends is getting dangerously near .500. Not quite, but in the realm of possibility if things break right. It ALSO gets the Nats so far away from <70 wins that that becomes a worst-case scenario rather than a strong chance. That in itself, making it no chance of another non-competitive year, might be worth spending that money.

2 comments:

Unknown said...

I have 2 concerns with your projections: 1) I think you're over-rating the "good relievers" the Nats have. I don't see Capps pitching 3.7 ERA, I see him touching 4 or above because of his peripheral stats. Bruney I think will be worse than that.

2) I think the Nats will give up fewer un-earned runs defensively than they did last year. I believe our error totals were well above what they can be expected to be, so regression-to-the-mean on returning players will work in the Nats favor. On top of that, we will not be playings Dukes/Kearns/Millege in CF. That's fewer runs given up independent of the pitching. We won't be playing Dunn in LF (and I think he's a better 1B defensively than LF- not a good 1B, but better than he is as a LF). We won't be putting Kearns or Willingham or Desmond in RF. In the infield, we won't have Anderson Hernandez taking grounders off the nuts. We won't have Josh Bard who can't stand up from a crouch behind the plate. Either Alex Gonzalez returns to the solid defensive player he was supposed to be when we acquired him, or he doesn't play (I hope). Maybe we acquire Hudson or other defensive 2B player who can prevent more runs than the garbage we threw out there last year.

Those are all changes that will prevent runs defensively not accounted for in your estimates. I don't know enough about statistics to put a number on these changes, but it could have a more-positive effect than you've already spelled out! Hell, it's not even February, we hope still springs eternal...

Harper said...

1) Completely possible - I was just eyeballing an average from the combination of Bill James, CHONE and Marcel projections. I thought they are a little on the brightside too. Then I took a look at the team stats and the total combined ERA for the relievers I thought would be GOOD on tha Nats was 4.14 which is below average compared to other teams entire staffs. So are those stats expecting too much? I had to say no. I guess for every Bruney or Capps that might pitch a run worse - there's going to be someone (Storen? Walker? Batista) to fill in that gap with adequate 4.00+ ERA pitching.

2) I agree. But I don't see the difference being as huge as people think. Dunn out of LF is GREAT, but it means he'll be at first where he's not good. Willingham in LF will be bad, Dukes in RF isn't good (easily worse than Kearns), and Morgan in Center is great but the Nats had about 60% of last year covered by him or the talented Willie Harris. (milledge and Kearns barely played there)

As for the IF and C. Catcher yes, infield we'll see. Desmond + Guzman is not guaranteed to be THAT much better defensively than Guzman + ? was last year.