I don’t know why, but this is probably one of my favorite additions to the site in a while. These graphs show how players ranked in each statistical category. These stats are taken from each player’s 7 best seasons and their percentiles are averaged to give us the overall ranking.
Note that these stats are compared to contemporaries, so that is why Ty Cobb is in the 95th percentile in homeruns. Obviously, Cobb didn’t hit 40 HR’s a year, but he was better than 95% of his contemporaries in that category.
Take a look at Ichiro’s Career Batting Graph:
And for a pitching example, here is Carlos Marmol’s graph:
Dan –
A couple of questions on the new feature. When you say, “his contemporaries,” does that mean all players that were active during his career? Does it mean only players active during all seven selected seasons? Does it mean those seven seasons are taken as a group, and that subset is that to which the player is being compared?
Second, would it make more sense to have some of the pitching graphs inverted? For example, the HR% graph… ideally, a good player would want a low HR%. Shouldn’t he show up at the top of the graph? Likewise for H%, BB%, BR/9, ERA, etc. I’m just a bit confused, because on the hitter graph, ideally every bar would be high. On the pitcher side, that doesn’t appear to be the case.
Finally, when figuring SB/ToB, are HRs subtracted from ToB? I ask because it may make a difference for some of the extreme power/speed guys (Mays, Bonds, Ruth, etc.) in terms of their percentage if the large number of HRs are subtracted from hits.
Hope that’s not too many questions. And thanks again for the new feature! in spite of my above comments, I really like it!
1. Thanks for the reply. Here’s an example about what I mean by contemporaries (I’ll use Ichiro)…
I use batting Win Shares to determine the 7 best seasons. I compare the batters to 8 times the number of teams each season. So in 2001, there were 30 teams, and he is compared to the 240 (30×8) players with the most PA’s. Of those 240 players, Ichiro was 36th in Runs/27. In other words, he was better than 85% (204 players). I do this for all 7 years and find the average (weighted on PA’s)
2. The “inverted” graph for pitchers is an issue I went back and forth about. But I decided that the graph shows how good they are in each category. So a pitcher’s HR% is how good they are at avoiding HR’s and a batter’s K% is how good they are at avoiding K’s. For all categories, the more blue the better. (Except for Balls in Play%. The higher percentiles show more balls in play. Not necessarily better)
3. Times on Base is figured by Walks + Singles + Doubles + HBP. Triples and HR’s are not included. It’s not a perfect calculation because that doesn’t guarantee that the base ahead of them is unoccupied or that it is a steal situation. It also doesn’t include reached on error or fielders choices. Like I said, not perfect, but close enough
Thanks again for the reply!
Pingback: Career Batting and Pitching Graphs » Stathead » Blog Archive
Dan – thanks for the rapid reply. One more thing: I see that you’re using a DIPS-type ERA. I’m wondering what your method is. I don’t see it in the glossary, so I thought I’d ask. Is this the case where you’re ONLY using BB, HR, and Ks, and creating an ERA, (like Fangraphs WAR), or are you calculating, assuming a certain number of hits will be given up (as Sean Smith’s WAR calculates it)? I’m curious.
Dr. Doom,
I calculate the statistics using the DIPS 2.0 formula found in this link…
http://www.baseballthinkfactory.org/mccracken/dipsexpl.html
My method is much more similar to Fangraphs Pitching WAR
Dan