Will the Bucs be Better this Year? A final look at PECOTA Projections

With the Pirates having set their Opening Day roster, it is time to take a final look at projections for the 2013 team. (If you haven't already, be sure to check out my ZiPS/Steamer WAR calculator and ZiPS/Bill James Runs Scored-Runs Allowed, Pythagorean Record calculator.)

In the tables below I use the most recent (3/27) update of Baseball Prospectus' PECOTA projection system to:

1. Compare the 2012 Pirates to the projected 2013 Pirates.

2. Examine which National League teams are projected to be the most improved/suffer the largest decline in 2013.

3. Show how the recent roster decisions have changed the Pirates projected WARP-by-position.

PECOTA

The acronym PECOTA stands for Player Empirical Comparison and Optimization Test Algorithm and is Baseball Prospectus' proprietary projection system. It was first developed by Nate Silver in 2003, although Silver no longer runs it.

Since PECOTA is a proprietary system that requires a subscription to Baseball Prospectus to see the data, I have refrained from directly posting any of their forecasts regarding projected wins and losses for the 2013 season. To see that information and other content that any serious baseball fan would enjoy, I strongly recommend a subscription to Baseball Prospectus.

One of the neat things about working with the PECOTA system is that Baseball Prospectus constantly updates their team projections to reflect roster changes and evolving estimates of playing time. After the Pirates finalized their Opening Day roster, BP posted revised projections based on new estimates. So, all the data presented here reflect PECOTA's WARP (Wins Above Replacement Player) projections adjusted for individual player estimates of positional playing time.

Now to the results of my study:

Positional Comparison: PECOTA 2013 Pirates and 2012 Pirates

The table below displays the projected improvement/decline of the 2013 Pirates by position. I calculated these results by simply comparing the latest playing time estimates with the 2012 end-of-season totals. Overall, the Pirates are projected to improve by 4.83 WARP. (Interpretation example: Compared to 2012, the Pirates are projected to gain .4 WARP from the second base position in 2013.)

Observations:

C: Martin and McKenry should team up to improve the catcher's position, but PECOTA does not forecast as much improvement as I think many Pirates fans are hoping for.

SS: Barmes is projected to have something of a bounce-back year (or at least he should not be as awful), creating 1.2 additional wins.

LF: A full season of Marte in LF is projected to have the largest positional impact in terms of wins.

CF: McCutchen is still good, really good. But some drop-off is expected.

Pitching: Here is the problem. PECOTA is not optimistic about the Pirates pitching staff improving on 2012's performance.

Positional Comparison to National League Average: PECOTA 2013 Pirates and 2012 Pirates

The table below shows differences from National League average in terms of positional WARP. (Interpretation example: In 2012 the Pirates were 1 game below average WARP at the catcher's position.)

Same as above with Pirates positional averages excluded from calculation of league average.

Observations:

The 2013 Pirates are projected to be above league average in only three of the ten positions. However, in comparison to 2012, they should be less below average in many positions. Depressingly, the significantly-above-average production in CF is projected to be balanced out by significantly-below-average pitching.

Most Improved/Largest Decline Teams

The table below is derived by subtracting the projected total WARP for each National League team from their total 2012 WARP.

Observations: The five National League Central teams make up five of the six positions at either end of the table.

Most Improved/Largest Decline Teams Broken Down By Position and Pitcher Contributions

The table below takes the results from the table above and divides pitchers from position players.

Observations:

Pittsburgh: Pirates positional players show the second best improvement in the National League. Pitching declines.

Chicago: PECOTA sees the revamped and strengthened pitching staff as adding close to ten wins.

Cincinnati: PECOTA projects the Reds position players to add eight more wins this year, improving on what was a slightly below league average offense in terms of runs (slightly above in terms of OPS) in 2012.

Colorado: The Rockies improvement in position players is projected to be offset by pitching decline.

Milwaukee: As has been widely discussed, the Brewers pitching staff projects to be a worse this year - costing them six wins.

San Francisco: Very interesting results here. The pitching improves significantly; but position players decline. They mostly offset each other, and the team loses about 1.5 wins.

Pirates Roster Decisions

This table is reflects a comparison between the Pirates positional WARP estimates published on 3/27 and those published on 3/3.

Observation: There has been much hand-wringing recently about the final roster decisions made by the Pirates. As we can see, PECOTA forecasts the current roster as adding .1 WARP from what it produced on 3/3. In other words, recent developments have not changed the overall forecast for the team significantly.

Conclusion

Entering the 2013 season, the Pirates appear to have a stronger team than the 2012 edition, to the tune of 4.83 wins. However, since the 2012 team ended with a record slightly above what its underlying statistics would predict, we should not conclude that this team should be expected to win 83 games. Indeed, Baseball Prospectus forecasts a slightly lower number of wins.

This study confirms what many on this site have been discussing for days, i.e. pitching is a concern. There appeared to be opportunities to improve the staff in the off season, with rumors of Capuano and Porcello swirling about. Apparently, however, the asking price was too high for each. The freak injury to Franciso Liriano, plus the constant uncertainty surrounding Jeff Karsten's health hurts the staff in the short run. Add the expected slight regression of AJ Burnett coupled with the high-beta nature of James McDonald, and you have the formula for a rocky ride. It should be interesting.

Follow me on Twitter @DavidManel and Bucs Dugout @Bucsdugout