THE BOOK--Playing The Percentages In Baseball

I’m more suprised by the great showing by Rolen.  Didn’t follow his career very much, but wow!  Is he not a HOF-er?  If not, why not?

Tango,

As you know, Belle is exactly the kind of player that WAR helps us understand better.  Sure, he was a fearsome hitter, but he didn’t do anything else to help his team win: poor range, poor arm, GDP’d more than his share, did not ROE, wrong end of the defensive spectrum, even a tick below average in base running. 

The guy could mash, but WAR tells us how all that other stuff cancels out the offense.

John, right agreed.  Basically, everything ELSE he did was bad enough.

He’s like the anti-Ichiro or anti-Kenny Lofton.

Really puts in perspective how great Albert Belle was at the dish in 1994 (strike shortened season) when he was 6 WAR and -15 in total zone. Just 475 PA’s.

I’d add Eric Davis to this group.

Another thing that works against Belle is his best years came when the AL was averaging over 5 runs per game.

From the title I had just assumed this would be a Nomar inspired effort.  Where’s Nomah!

Too short?

I’ve always been a beleiever in Rolen as an under the radar HoFer.  I think the injury and quick offensive decline - though not really defensively except games played - are going to keep him out though.

I quickly put this together for Nomar:

I know these graphs using trend lines for best-to-worst WAR seasons have caught on.  And I see how they capture interesting information.  But I think using a trend line for seasons that are not sequential in time is pretty confusing, and I wonder how well they are understood by more casual saber-inclined fans. 

I think a better presentation for some purposes might be a cumulative WAR line starting with players’ first season (or you could use age), compared to the line for an average HOFer.  We would see whether players were keeping up (or exceeding) the cumulative total for a HOF player through various seasons/ages, and in this case could see when they began to fall short. At the end we could see how far short they fell on a career basis. And it would have the benefit of following an actual timeline, which is much more intuitive.

For those interested in seeing a player’s peak, the current graph clearly gives more valuable info, so I’m not arguing for one approach to the exclusion of the other.  But for many purposes—including this almost-HOF analysis—I think the cumulative graph might tell the story better.

ooooohhh… I like the cumulative approach… but not necessarily timeline driven (though that’s good too, for injury players because it shows a clear break).

I like it for cumulative best-to-worst.  This way, if you want to look at someone’s one-best or three-best or seven-best years, and how it compares, then you can see that.

For example, at what point does Scott Rolen go from being clear-HOF player to average/below-HOF player.  If you only count his 10 best years, then you would say, yup, he was on the path because he did what other HOF players did in their best 10 years. 

What sinks Rolen is that he didn’t perform “average enough” in his 11th - 15th best years.

Something like that…

And then, you see the possible silliness of that proposition: if only he were more average in his 11th-15th best year, he’d be a HOFer.  I consider that silly.  Others however think that’s very legitimate, that someone NEEDS to do that, unless he was a Bobby Orr type of player.

I’d like to see this chart actually for Sandy Koufax.