Consistent players – suggestions requested
Posted by Andy on October 14, 2007
Readers, I am working on the development of a new stat that measures the consistency of a player over his entire career. This is something I hope to develop to a useful point such that Sean would consider incorporating it onto each player's main page.
You can help me by suggesting players who you consider to be very consistent over their careers. Look for players with consistent year-to-year batting averages, slugging percentage, and rates for runs, RBI, hits, 2B, 3B, HR, BB, and K (or some subset of those.)
Please post your suggestions in the comments. Thanks!
As a point of interest, I've tested it on some of the players we've discussed recently. Keith Hernandez and Tony Perez get quite consistent scores. Barry Bonds gets a fairly inconsistent score, since his power numbers and SLG went off the charts in the latter part of his career. Babe Ruth is incredibly inconsistent due to having some years with a very low HR rate (very early and very late in his career) and other years with a historically high HR rate.
October 14th, 2007 at 7:12 pm
Unless, you do something like age-neutralize stats or deal only with blocks of seasons and not whole careers, I'd imagine you'll find that players like Bonds, who had long careers will have a greater variety in their seasonal totals than those who started later and finished earlier.
Also someone who hit like this
.200
.400
.200
.200
.400
.400
.200
Should probably be regarded differently than someone who hit like this
.200
.200
.200
.400
.400
.400
.200
I would think that you'd consider the second player less consistent.
(PS I would love to see a measure like this on a game basis. What was the variation per game of Alex Rodriguez's value in 2007? Or something like that.)
October 14th, 2007 at 7:21 pm
I've always been in awe of the first 5 years of Albert Pujols' carrer and how close all his numbers were for those seasons...
October 14th, 2007 at 7:27 pm
I've got an idea about how to measure game-to-game consistency as well
October 14th, 2007 at 8:37 pm
Tries plugging these guys into your formula: Lou Gehrig, Hal Trosky, Jackie Robinson, Hank Sauer, Dom DiMaggio, Jackie Jensen, Earle Combs, Ralph Kiner, Ron Hunt, Kevin McReynolds, Larry Doby, Hank Bauer, Freddie Patek, and Carl Furillo.
October 14th, 2007 at 9:22 pm
Hank Aaron, Willie Randolph
October 14th, 2007 at 11:48 pm
I agree that there has to be some correlation with age with the beginning of careers and decline phases. I don't know exactly how to do that since every player will be different, but it doesn't have to make a large impact. Also does consistent also include playing time. If not,maybe use something like 15 year peak/span to take out the young and old, or subtract a total of 3 years of any combination from the beginning and end.
Manny Ramirez and AROD are guys who I think are very consistent, despite this year. Also will injuries count against them? Would you possibly divide the numbers by games played then add them up. So even though the rate stats might be the same, and RBI/HR per game played, etc... if the player plays 50 games. If you add it as a percentage as a whole it will count less than a full year, but not subtract from total value if there's a major injury.
Or maybe use a standard deviation from the previous season... this might account from age/decline or w/e. Like say the career as a whole averages 30 HR. The standard deviations in 1 season might shift from 2-6% so it's a 4% difference in that year. Well a consistent 3% increase and decrease isn't that much, so you can take the standard deviation of that also and come up with the final difference compared to the career averages. This can include percantages from injured seasons. This kind of makes my first thoughts irrelevent but i'll keep them. This is my first post and I've been reading this for a long time. I've always enjoyed stats and someday I'd like to create my own stats.
October 15th, 2007 at 9:33 am
I'd suggest Dale Murphy, Mike Schmidt, Cal Ripken, George Brett, Al Oliver, Willie Stargell, and Willie Wilson.
October 15th, 2007 at 9:46 pm
One thing to keep in mind is that statistics are discovered, not developed. Secondly, consider that measures of consistency aren't all that valuable unless you can do some prediction (ie it doesn't need 10 years of data). As a proof by counter example: Brady Anderson?
October 15th, 2007 at 11:45 pm
"One thing to keep in mind is that statistics are discovered, not developed."
This doesn't make sense. A statistic is just a numerical value. No one found batting average under a rock somewhere.
"Secondly, consider that measures of consistency aren’t all that valuable unless you can do some prediction (ie it doesn’t need 10 years of data)."
True, and I'm ambivalent on the value of a consistency stat (see my post on the other thread). But it doesn't have to be valuable to be interesting. We count balks and BA on Wednesdays and games finished. None of these necessarily matter or have any predictive value, but we keep track. A stat tells you what it tells you. It's up to us to decide how important it is. People always talk about players being consistent or inconsistent; why not have a number to actually try measuring it?
October 16th, 2007 at 4:54 am
Brady Anderson isn't an example of anything but a below-average ballplayer who probably cheated his way to one great year. Other than his speed, Anderson didn't have any well-above-average traits.
October 16th, 2007 at 7:35 am
The most consistent is Fred McGriff.
October 16th, 2007 at 1:04 pm
I'll second the suggestion of Fred McGriff.