2015 Offensive Projections — Interactive
Using the FanGraph’s Depth Charts tool, I created a data visualization for batters’ offensive output using on-base plus slugging percentage (OPS) as the primary statistic. Pitchers have been excluded. The Depth Chart tool takes projections from both ZIPS and Steamer and accounts for playing time creating projections for counting stats (PA, WAR, etc.) and team performance. The Positional Power Rankings being released this week use the Depth Charts as well.
The histogram below creates an overview of the entire league showing the OPS distribution, which is normal with a skew to the right reflecting the infrequency of elite offensive players.
The data visualization includes both the individual player OPS (green circles) and the team OPS (yellow bars). The team OPS isn’t the simple average of the player OPS; it is calculated by weighing each player’s OPS by plate appearances, which are represented by the size of the circles. Finally, to give you an idea of the distribution for each team, the gray shading represents the 25th-75th percentiles.
The data visualization is INTERACTIVE with tooltips so please spend feel free to explore it.
The Rockies top the list due to park factors, which is why most of FanGraph’s preseason coverage uses park-adjusted stats like WAR to compare players. The elite offensive players really stand out from the pack being far beyond the 75th percentile and receiving well over 500 PA, and unsurprisingly Mike Trout has the highest OPS projection. You can spot defensive specialists and catchers by looking at the low end of the OPS projection and finding players with a high number of plate appearances.
The projections are current as of 3/23/2015.
I build things here.
I find these projections offensive!
Woah this is awesome.
2 things:
1. I was about to ask who that dot way to the left on Seattle’s bar was, but then I realized you can actually hover over the dots and it gives you information about them! That’s so awesome! (It’s John Baker, by the way)
2. Any particular reason why OPS was used for this as opposed to, say, wOBA?
Awesome as always, Sean!
I think OPS is a little more easily understood than wOBA. They both do a pretty good job summarizing offensive output. OPS has a larger scale, and I like comparing .800 OPS vs .550 OPS (elite player vs scrub) than .360 wOBA vs .250 wOBA especially in graphics and general data presentation. For more intense/rigorous analysis the nuances of wOBA might be important. But they are pretty much interchangeable for THIS purpose and extremely collinear.
Why OPS instead of wOBA or wRC+ ?
Reminder to self: after reading the article, and before posting a comment, always hit refresh.
It’s cool. I thought about wRC+, but it’s not on our Depth Charts, and I wanted to make something more tactical since all of our PPRs use WAR.
Maybe I misunderstood or there is a display error. The gray area is the 25th to 75th percentile team OPS projections. The yellow bar is the team OPS projection (e.g. 50th percentile). Yet somehow the yellow bars are showing consistently on the high side of the range and even above the range in some cases.
Team OPS is not the 50th percentile. The gray bar reflects the composition of the roster, while the yellow line reflects the team performance. Players are given different plate appearances, so I weighed the OPS by PA to estimate the Team OPS.
What you are seeing with the yellow line beyond the 25th-75th PCTL is the presence of a few players who are very good relative to the rest of the team which pull the team’s OPS upward.
Nice job Sean. I love the interactivity in here – so nice to be able to actually see who the dot dragging the Yankees down is…thanks Brendan Ryan, maybe we will miss a certain former SS a little bit after all.
OAK, STL, BAL, HOU and KC look to have the least variation…
I reckon that might explain quite a bit about how BAL has “outperformed” the projections.
And, it is one more reason I am long on HOU.
Terrance Gore had 2 PAs for the Royals last year and will be back in the low minors this year until they need him to steal bags when the big league rosters expand in September. I have practically as much chance of getting 56 PAs this year as he does.