Community Projection: Kenji Johjima
The results of the first community projection are in, as we had 142 people contribute their estimates for Johjima’s performance. Here are the results, and those of you on the email list can expect to receive the spreadsheet for Richie Sexson tomorrow.
Overall Projection: .291/.344/.468, 492 AB, 26 2B, 1 3B, 20 HR, 31 BB, 9 HBP, 51 K
High Projection: Mark S: .347/.429/.565
Low Projection: Natebracy: .268/.306/.403
Dave Projection: .276/.320/.434
PECOTA: .294/.351/.453
Besides Mark S, the projections were all fairly neutral, with no one projecting an OBP over .400 or an OBP under .300. The overall projection is extremely close to the PECOTA projection and maybe a tad bit optimistic on the power end of the spectrum.
Overall, though, I think it’s a good representation of what we can expect from Kenji this year. This kind of performance should again make him one of the best catchers in the American League.
Comments
30 Responses to “Community Projection: Kenji Johjima”
Leave a Reply
You must be logged in to post a comment.
What is to be expected of Kenji’s defensive performance? Is there any reason to expect a significant improvement or decline? And has any consensus been reached on the value/detriment of his defense?
And in a related note, what is everyone’s opinion of the different defensive metrics with regards to catchers? Is there a clear-cut best?
Who’s next? Richie?
Yeah, Richie. We’re snaking our way around the field.
Why do you have him breaking down so bad, Dave?
Did we get the “view only” situation resolved?
Why do you have him breaking down so bad, Dave?
I don’t.
The community projection gives him an isolated slugging percentage of .177. PECOTA gives him an ISO of .159. I give him an ISO of .168. In terms of power, we’re all in the same range.
I projected him to draw 26 walks, while the community had him drawing 31. For all intents and purposes, that’s a wash.
The difference between me and the community? 12 singles. You guys think he’ll get 12 more singles than I do over the course of 495 at-bats. There’s no way to construe that difference as a “break down”. It’s a bounce here or there, a bad call by an ump on a close play at first, and the difference in judgment by an official scorer on whether a ball was a hit or an error.
There’s very little difference in talent between what I’m projecting and what all you guys are projecting. I just don’t think he’ll get quite as lucky on balls in play as you all do.
Dave’s numbers seem closer to reality than the rest of the numbers to me…hard to imagine a 31+ year old catcher who has caught 90%+ of the games his team has played the last 5 years not having some slow down. Then on top of that, now there is a “book” on how to pitch Joh, and I doubt Joh can keep up his “Baylor-esque” HBP pace…but I thought Raul would colapse last year too.
Then on top of that, now there is a “book†on how to pitch Joh…
If “the book” makes that much of a difference on pitching to Johjima, it’s impossible to tell from his results last season. By August and September all of the teams in the league had lots and lots of film on him, and in September there were lots of games against divisional opponents that had seen him quite often already, and he hit roughly .300/.330/.440 in those last two months of the season while almost certainly battling fatigue. That line’s not meaningfully different from what Dave is projecting, which (as Dave correctly notes) isn’t meaningfully different from what the community projection is, just like it isn’t meaningfully different from his overall season total.
You’d be hard pressed to convince me that after the first four months none of the scouts could figure him out and then after the last two months, with the same results at the plate, they suddenly found a big weakness.
he is one of the very few good aquirements the Ms have made these past couple of years…
Wow, I didn’t look at the percentages. I just wrote down some raw numbers be (overly) optimistic.
I’ll be sure to look closer on the next spreadsheet.
I still think it’s interesting that as a group, the only real differences between this year and last are:
HBP (seemingly a crapshoot)
5 more strikeouts (goes along with a tad more power numbers)
but a 55% increase in walks.
I’m in there too. somebody tell me why we think he’s going to see such a marked imporvement in plate discipline?
if “the book” is out on how to pitch him, won’t it be harder, not easier to draw a lot of walks?
Two reasons:
1. Last year was, by far, his career low in walks. Since 2001, he drew 31, 30, 53, 49, and 33 walks. His 2006 total was 50% lower than his established career pattern in Japan.
2. Good hitters, just by the virtue of being good hitters, walk occassionally. Every once in a while, you get a ridiculous outliar year like 2005 Pudge Rodriguez, but for the most part, it’s hard to be a major league quality regular and draw less than 30 walks a year.
I hope you’re right about walks (and the community is right about HR); I wondered the same thing. What worries me about his walk rate is that he only walked 5 times after the all-star break (15 before).
My heart thinks Kenji will have a big year now that he’s adjusted/acclimated. But my head came up with a fairly low projection mainly because he seemed “unwilling” to take a walk.
20 walks is pretty much the floor for a hitter the quality of what we’re projecting Johjima to be. If you take the top 100 hitters by slugging percentage in MLB in 2006, Johjima was 99th in walks out of that group. Only Robinson Cano drew less walks (18), and Cano had 34 less plate appearances.
Even the biggest hacker in MLB among everyday players, Jeff Francouer, drew 23 walks last year. If you have any power at all, pitchers are going to be more careful with what they throw you, and you’ll draw some walks by default.
There’s almost no way Johjima draws less than 20 walks unless he just falls apart as a hitter.
The forecasts for Bill James, ZIPS, Chone, and Marcel are all here:
http://www.fangraphs.com/statss.aspx?playerid=5811
Including PECOTA, the 5 forecasters ranged:
AVG: .281 – .294, with USSM readers at .291
OBP: .332 – .351, with USSM readers at .344
SLG: .435 – .459, with USSM readers at .468
If I do a modified version of OPS as: (1.72 * OBP + SLG) / 3, which scales along OBP lines, but is an “overall” measure, this is what we get:
0.353 USSM Readers
0.352 PECOTA
0.347 Bill James
0.342 CHONE
0.338 Marcel
0.338 ZiPS
There are two things to look at:
1 – Let’s see how the USSM Readers do with all the other players, to see if they are biased. If USSM Readers all overforecast their players, then we know they likely have a hometown bias
2 – If the forecast for the established players is in-line with the pro forecasters, then we may be able to conclude that USSM Readers used “tools” knowledge of Johjima, that the other forecasters simply didn’t have.
That is, when you have a large dataset, and the player is not injured, we don’t need to have toolsy knowledge, since 2000 PA trumps whatever it is that you can see with your eyes. But, with a player with limited PA in MLB, the fan can observe things beyond what the numbers will tell him.
I am thrilled that so many USSM Readers are passionate enough to participate. You are actually contributing beyond just this experiment.
I thought I was the only one who got the “view only” version-I never got to make my projection. Hopefully it works for Richie.
I gave him pretty average numbers and a low at bat total. I have a fear that Grover overworked him badly last year. Jojima came from a league with fewer games, less travel to a new league and a team that struggled from the outset. I think Grover will do it again and am terrified that by August he will breakdown and be on the DL. Not my wish but I think a real possibility.
Tango,
even if the USSM/LL readers don’t consistently overrate the players, this doesn’t necessarily mean that they used “tools” knowledge. I think Johjima is a well-liked player around these parts and this may tip the scale in his favor.
I think we’ll have to wait for projections for all hitters and then look for trends. I could imagine we may end up with groups of well-liked players with optimistic projections (Johjima, Ichiro, Lopez, maybe Beltre) disliked players with pessimistic projections (e.g. Blooquist, Vidro, Rivera) and neutral players with neutral projections (e.g. Guillen, Sexson). Or would such group bias fall into what you consider “tools” knowledge?
BTW, it’s great that so many participated. I did not b/c I didn’t feel up to the challenge of guessing precise numbers of hits, walks etc.
I’m not surprised that the projections on Johjima are all close. Despite a limited number of MLB PA, there’s a good history of what he can do and it seems hard for even the most optimistic or pessimistic to project him outside a narrow range.
I was looking at Sexson, and he’s another story. There’s a lot of MLB history, though only two years since a major injury and his 2006 season was wild. In April and May, he wasn’t even replacement level, June, July, and August are pretty good. Then his September/October statline looks like Pujols or Bonds. Looking at his monthly splits (and I realize you get into small sample size), it’s hard to believe they were all put up by the same player. Who knows which Sexson will show up? As a Mariners fan I’d like to believe April, May is a complete fluke and he’ll put up a season’s worth of July and August. The realist says April and May count.
Mark S’ disclosure that he owns and wears a Johjima jersey (in the thread above this one) could be seen as evidence of my point re likes and dislikes influencing projections.
I like the idea of adding a parameter of “well-liked, disliked, ambivalent” to the study. Maybe Dave can add that in at the end, asking the USSM readers: “Do you like this guy?”, and get a reading that way.
I’m really leary of that: it’s adding a qualitative judgment to a statistical query, and adding that question may cause people to reconsider the numbers they’re putting in.
Moreover, the question’s going to have the same issues as the numbers do, so I don’t see that we learn anything.
For those who were having problems editing the text –
I had this issue and just figured out that the problem was my Google account email address (personal) being different from the one I used to send the ‘registration’ email (work). I just changed my google account settings and its fixed. So that’s one thing to look at for those having the problem.
Actually, the question would be added at the end of the survey, after all the votes are in. So, there’s no reconsideration issue.
As for learning, we’re trying to see if one guy’s numbers may be more biased than another. Is it possible that a guy that is well-liked will come out with a bigger gap between his community forecast and the pro forecasters than a guy who is disliked?
There is no downside to asking the question.
Mark S’s projection is really just high on batting average. And his ISO is a bit high, but mostly it’s just the batting average.
I guess I don’t see that taking a perception survey after you’ve done the projections is going to be worth the effort.
Or take Bloomquist: I think he’s a good guy and all, but I think he’s one of the least talented players in baseball and I hate that he gets 200+ ABs. Do I rate him “like” or “detest”? Do we measure “opinion of him as a player” and if so, what does that gain us over a pure stat line?
Your opinion of him that would make you think he might be, or you want him to be, better than he is. He’s got hustle, he’s a character guy, he cures babies.
His stat line will show how much you detest him, but if alot of people like him, his stat line might have been bumped a notch more than it should.
See, and there’s the problem. We don’t know that opinions influence projections, so that detest/like bumps up or down projections. Asking people to rate their personal opinions may well end up being a tacit endorsement of bad judgments – go ahead and rate bad/good without trying to be rational, as long as later you tell us what that rating was. It’s a whole different survey than “try and make your best guess at what his actual performance will be”
Which is why we should let everyone know that we won’t ask their opinions afterwards. And then, ask everyone their opinions afterwards.