clock menu more-arrow no yes

Filed under:

BCB Projections: Introduction and Catchers

New, 26 comments

If you have read the Frosty Mug and the announcement post below, you are probably aware that Beyond the Boxscore is running a large-scale community project to attempt to accurately project the 2009 season. Based on new stats and ideas, we can determine how many wins above replacement level individual players are, and then use these WAR figures to estimate each team’s talent level so we can accurately compare teams and predict the 2009 season.

This gives us an exciting opportunity to build extremely comprehensive projections for the entire league. Standings estimates like this one, based on Marcel projections, are fine; however, they do not allow for individual editing and fine-tuning like this project will. We have an opportunity to tap into our fanbase's knowledge to come up with a better projection than a system alone can provide. Over the next few weeks, we’ll be projecting not only expected performance but playing time for position players, innings pitched for starters and relievers, and leverage for relievers. This will allow for a much more accurate and precise overall projection than a system that doesn’t know Seth McClung might bounce between the pen and rotation, or that Yovani Gallardo is a safe bet to throw a lot more innings than last year. 

Sky Kalkman has asked for a volunteer from each team to decide how all the players will be projected, and has asked repeatedly for that volunteer to get the team‘s online community involved in the projections for performance and playing time. jhmoore of the site Right Field Bleachers signed up for the Brewers, and to properly gauge the Brewer online community, we agreed to try to get as much input in this process as possible, and I'll be gathering the projections here.    

So with that being said, welcome to the first in the series of BCB Community Projections for the 2009 season. Our task this year is not just to pass time and guess what will happen next season, but to determine how the Brewers will perform and be utilized on the field as accurately as possible. 

The most important issue I want to stress in regard to making projections is that we want to keep extremes out of our projections. We’re going to be all about compromise. Say you have a hunch that Kendall will wear down and Salome will be called up and hit well in the second half. We certainly want to allow for this possibility in our projection. In a normal projection, you might just put down Salome for 300 plate appearances on that hunch. But we are looking to be as accurate as possible here. So in this situation, if you think there’s a 30% chance Salome gets 300 plate appearances, give him about 100 to account for both possibilities.

The principle is probably easier to explain with starting pitchers, who we’ll cover later. As an example, let’s say that in situation 1, every participant thinks there’s a 60% chance Manny Parra throws 190 innings and a 40% chance he throws 150 innings. If everyone projects 190 innings, we end up with an inaccurate projection. We’re also inaccurate if we all project 150. If everyone projects closer to 170 innings, we can increase the accuracy quite a bit and lower the potential for error.

Now I want to quickly disclaim the previous two paragraphs. We really want everyone’s individual opinions, ideas, analysis, and even hunches to be a big part of this process. These are just some basic guidelines attempting to get these projections to be as accurate as possible. 

We are going to start this series off with the catchers. Here we are veering off of the conventional model for community projections-- we’re not doing this one player at a time, and we’re asking for your projection of all the players who will get plate appearances as a catcher in 2009, their slash line, and their percentage of total available playing time at the catcher position. We’ll cover the projections of lineup position later in the series; obviously the catchers will likely hit eighth and receive fewer overall appearances than the second basemen or left fielders. For reference, Kendall had 587/643 plate appearances last year, for about 91% of the playing time available at catcher. 

Kendall and Rivera are not the only catchers available, either. I will not list the entire Chone projection for Angel Salome here, but feel free to project him for some playing time if you feel he will be in the majors at some point. The same goes for any other potential catcher, like Vinny Rottino. 

The format for the players capsules will be the same for all the projection posts. The first line is the player’s career line, the second line is their 2008 season line, and the five subsequent lines are the player’s 30th, 40th, 50th, 60th, and 70th percentile projections from the CHONE projection system. If your projection is below 30% or above 70%, that’s fine, we are just trying to provide some type of baseline to compare your projection to. The official CHONE projection is the 50th percentile, or the 3rd line in the block of five.

Jason Kendall

Career: .293/.371/.388
2008: .246/.327/.324

0.242 0.314 0.303
0.247 0.321 0.309
0.251 0.329 0.321
0.258 0.337 0.331
0.264 0.346 0.345

Mike Rivera

Career: .249/.299/.394
2008: .306/.377/.435

0.218 0.269 0.367
0.231 0.286 0.396
0.241 0.298 0.429
0.251 0.310 0.451
0.263 0.326 0.483

Angel Salome

Vinny Rottino

We're using Chone as the baseline because its simple, easily available, free, and takes minor league statistics and other trends into account more than a simpler system like Marcel.

Finally, the format. First, I’d prefer the projections to be in the bottom portion of the comment and not in the subject/title section. This is just because the subjects act as links to collapse the rest of the comment, and it will be easier to copy and paste the projections into a spreadsheet this way. You can just skip the subject line if you would like. For our purposes, we’d like to see them formatted like this:

.avg/.obp/.slg (PA)  Kendall
.avg/.obp/.slg (PA)  Rivera

Add as many players as needed, as long as the total plate appearances adds up to 100%. As a note, I’d prefer two spaces between the parentheses and the player name because we don’t actually need the player to be pasted into the spreadsheet, but this isn’t a big deal at all. We also don’t need a % sign after the number of PA, just a full number will be fine. Here’s an example:

.300/.350/.400 (75)  Starter last name
.270/.330/.400 (25)  Backup last name

Don’t feel as if you can’t discuss in these threads, either. Though I ask everyone to please participate and submit a projection, comments on this project in general are welcome and nested replies to an individual’s projection are absolutely fine too. If you want to point out something others might have missed, go right ahead.

We really want to encourage discussion and input on this project, and hopefully we can make the Brewers tab of the spreadsheet the best and most accurate projection of all the teams. Feel free to contact me or jhmoore in the threads or via email if you have any suggestions or comments.

The schedule is pretty flexible on the whole project right now, because the whole league will be making plenty of roster changes before opening day. As of now, the plan is to cycle through the positions, one every weekday, over the next two weeks. Then we’ll move into starting pitchers and relievers. We’ll also have to wrap up the whole thing with lineup projections and a few miscellaneous, informal discussions about base running modifiers and defense. You can expect the projection posts around 3:15 each afternoon, though I'm posting this one a little bit earlier.

Thanks in advance for your participation, and hopefully we can come up with a quality, accurate projection for the Beyond the Boxscore Community WAR spreadsheet. We'll leave this open through the weekend, and I'll post first basemen on Monday afternoon.