A Closer Look At The Rays Batting Order
BJ Upton, Dioner Navarro, Kelly Shoppach, Pat Burrell, Reid Brignac, Sean Rodriguez April 7th. 2010, 9:30amThere has been some discussion recently about the Rays batting lineup so we decided to look at the issue a little closer.
We used the projected stats for the Rays hitters using the CHONE projection system found at BaseballProjection.com. We then submitted the Rays projected totals to the lineup analysis tool at Baseball Musings.
First we used Joe Maddon’s opening day lineup and compared his batting order to the top four batting orders generated using the same players…
First of all, there are some weaknesses with the lineup analysis tool. The lineups generated using only a player’s OBP and SLG. So some very important factors are not considered such as speed and starting pitcher.
That being said, two things immediately stand out to us:
- BJ Upton should be leading off, despite his poor ’09 season and Jason Bartlett’s strong ’09 season.
- Pat Burrell should probably be batting 9th.
We then substituted Dioner Navarro and Reid Brignac in various combinations and just compared the top lineup of each combination…
Another factor that is not considered is defense. However, if we ignore defense, the best offensive lineup includes Sean Rodriguez and Kelly Shoppach and the worst lineup uses Reid Brignac and Dioner Navarro.
As a side note, 5.4 runs per game would translate to 877 runs scored during the season, which would be a 74 run increase over last season.
Related posts:
That’s probably why Maddon was so slow to pull the trigger on getting Upton out of the leadoff spot last year. Of course, this also shows the problem of depending too much on stats and not enough on instinct. As we saw last year, what *should be* on paper, doesn’t always translate to the field.
Professor: Please retract the Upton at lead off statement! It was a total killer last year. Unless and until he shows his pre-injury form, he cannot lead off. And even then, I’d rather have Bartlett up there because of his approach and discipline.
Great win last night. Crawford on a lefty is pretty impressive. Shoppach’s pinch ab was also very nice to see. Feeling a lot like 2008.
What about the concept that Burrell should not be in the line-up at all? Didn’t Stu and the fellas learn about the concept of a “sunk cost” on Wall Street. There are three guys on the bench and maybe a couple in Durham who deserve the DH at bats over Burrell at this point.
Like I said in another comment I think last year the Rays kept hoping Upton’s shoulder would get better and it never did. If he is doing well in June ad his shoulder looks strong I think Upton will be leading off again. He is an OBP machine when healthy and has speed to spare.
As for Burrell, I think they will have a short leash with Burrell but they are going to give him a month or two. If he still sucks then, others will start working in to the DH roll.
That being said, I still expect others to get occassional starts at DH early on including Rodriguez and Aybar.
I think last year they thought Upton’s shoulder would get better as the season went along and I’m not sure it ever did. I think if you see Upton hitting .300 with a .380 OBP in June, he’ll be moved back to leadoff no matter how well Bartlett is doing.
Am I missing something? Your “analysis” shows that Upton should bat leadoff, but you only show results with Upton at the top of the order. What does it look like with Bartlett at the top in various combination? Also, you know the difference in run production is not significant, right? I don’t understand how it can “immediately stand out” that Upton should be leadoff? This post makes no sense.
Why is analysis in quotes? I never claimed to do any analysis. I simply showed the numbers from somebody else’s lineup analysis tool.
As for a Bartlett-led lineup, I don’t know. The tool does not allow for forced lineups. I just know it would average less runs than an Upton-led lineup.
As far as significance, I never claimed any statistical significance because I don’t know the math behind the tool or how many simulations were run to generate the numbers. So I don’t know if they are significant just as you don’t know they are not. But just because differences are small does not mean they are not significant.
It’s not an analysis? You’re showing runs/game from a statistical modeling program and it’s not an analysis? What is it then? I put “analysis” in quotes to stress that you used a “lineup analysis tool” – not to be a jerk. Just after showing the chart you state the weaknesses in the “lineup analysis tool.” And then you state that two things stand out to you. I thought you meant the statistical analysis of runs/game based on differing lineups made those 2 things stand out to you. Obviously,this was just hunches you had, so I completely misinterpreted. But why would you show a numerical analysis tool and then be surprised I read it that way?
And how do you “know” an Upton-led lineup would yield more runs than a Bartlett-led lineup (OK – now I’m being a bit of a jerk, but you deserve it for that crap). This is an incredibly weak response from someone who usually does a pretty thorough job. I was genuinely interested in your opinion but this should have been couched as a baseless hunch and not jammed in between numerical charts that suggest otherwise.
Also, the biggest difference in runs/game you show is 0.047 runs/game between those lineups. I’m sure you could have done more combinations but that’s an extra run every 20 games – or 8 runs a season. Not much, but since we don’t have league-wide averages, you are correct – we don’t know whether it’s significant. Why not just do a post to talk about lineups and leave the numbers out since they are pointless here? Not trying to be mean but the post was confusing and your response seems kind of mean-spirited. Sorry if mine was testy.
If my comment was a little too defensive, then I apologize.
This post was never meant to draw any grand conclusions. More than anything I was just throwing it out there as something to think about knowing full well that there are some weaknesses. I included the numbers for express purpose of showing that there is not a lot of difference between each lineup. And I would have lived to have seen what the numbers for a Bartlett-led lineup would be, but their tool won’t allow me to do that.
But while there are weaknesses and no strong conclusions to be drawn, I take exception to the idea that it is meaningless.
Maybe my wording on Upton was too strong. Maybe I should have said “from a purely mathematical point of view MAYBE Upton should be leading off.” fact is, I don’t think Upton should be leading off right now. This tool is using projections that are really just educated guesses based on similar players in the past. I want to see Upton hitting and getting on base first. If that happens, then maybe he should be leading off.
This post is getting repped on 620WDAE Big Dog today FYI.
One thing that struck me about this year’s order last night was the absolute “black hole” created at 5-6-7 w/ Pena/Upton/Burrell when it comes to K’s. However, I do like splitting Pena/Burrell up w/ Upton because it allows Pena to do like he did last night and drop a bunt to beat the shift. I don’t want him doing that with PtB behind him.