Hi all,

I am planning to create my own cards for the 2006 baseball season for my Statis Pro baseball league.  According to the rules of the game, pitchers are to be ranked for PB rankings by ERA alone.  I don’t believe that ERA alone provides a good picture of pitching ability.

My question is this:  what would be the best way to evaluate pitchers (starters vs starters; relievers vs. relievers) statistically?  What statistics/sabermetrics should I look to?

Thank you,

sean

AndieMacfan - 29 October 2006 12:43 PM

Hi all,

I am planning to create my own cards for the 2006 baseball season for my Statis Pro baseball league.  According to the rules of the game, pitchers are to be ranked for PB rankings by ERA alone.  I don’t believe that ERA alone provides a good picture of pitching ability.

My question is this:  what would be the best way to evaluate pitchers (starters vs starters; relievers vs. relievers) statistically?  What statistics/sabermetrics should I look to?

Thank you,

sean

How about BB/IP, K/IP, HR/IP, BB:K and RA+ (adjusted run average, or its more available cousin, ERA+).  Plus inherited runners scored for relievers.  For the starters, you may wish to modify all of the above by including an innings pitched factor.

I’ve done some analysis of the ERA, which I plan to post soon (Sorry, life’s been crazy).

Using the Baseball Reference starts data (1994-2000), I found that the ERA is actually a surprisingly good measure of pitching ability. It’s positively correlated with a pitcher’s probability of giving up a walk and his probability of giving up a home run. It’s negatively correlated with a pitcher’s probability of striking out a batter. Interestingly, pitchers with high variance around their probability of giving up a walk and high variance around their probability of giving up a home run also have a higher ERA.

Those finding suggest that—by controlling walks, strike outs and home runs—a pitcher has some control over the runs he allows.

Of course, such a finding only holds when averaged over a whole season (so that the good luck cancels out with the bad luck). In a game-by-game analysis, I found that (on average) only 20 percent of the runs scored in a single game can be attributed to the pitcher. The other 80 percent is luck and fielding.

Like I said. I’ll post my findings soon, so stay tuned, but in the meantime ... “In the ERA we can trust.”

“I found that the ERA is actually a surprisingly good measure of pitching ability. It’s positively correlated with a pitcher’s probability of giving up a walk and his probability of giving up a home run.”

I don’t know why that would be surprising.  I mean, pitchers who allow the fewest runs are the best pitchers?  Who would have guessed?

ERA (context adjusted) is fine for ranking pitchers, but I would hate to use it as the measure of pitcher ability for a game though.

It tells you overall ability, but its not descriptive in the least.  Nolan Ryan and Jamie Moyer both are (were) good pitchers, but they go about it in completely different ways.  A game needs to at least consider a pitcher’s ability for walks, strikeouts, and homeruns.  That should be a minimum.  Next on the list would be how they control balls in play or groundball/flyball tendencies.

With Statis Pro, the ERA is THE tool to rank pitchers for the PB ranking, which is designed to replicate control (or so I’ve convinced myself of such).  It’s not that ERA wouldn’t suffice as a good device to rank the pitching; it’s just that I’m not convinced that it is a good measure of control.  Take two pitchers with very similar ERAs.  Let’s say the first pitcher has given up many more singles than pitcher number 2, who earned his ERA by giving up a lot of HRs.  Both pitchers gave up a similar number of walks and hit batters.  Which pitcher would you consider to be more of a control pitcher?  While pitcher number 2 gave up the homers on a consistent basis, pitcher number 1 gave up a lot of base hits but evidently pitched out of some jams . . .

Ah geez, I’m not sure that even makes sense.

I guess what I really want to know is this - is ERA the BEST way to rank pitchers?

Thanks for your posts.  I appreciate it.

I don’t understand the game.  Are you saying its using ERA to measure control instead of walk rate?

Maybe you should switch to APBA or Diamond mind.

No, I’d never switch from Statis Pro.  In my opinion, it’s the greatest baseball board game.

With Statis Pro, pitchers are given a PB ranking.  The ranks are 2-5 (really poor pitchers) 2-6, 2-7, 2-8, and 2-9 (really great pitchers).  Basically, Avalon Hills states that these rankings are awarded as follows:  Rank all pitchers by ERA; the top 5% are 2-9, next 10% are 2-8, next 40% are 2-7, next 30% are 2-6, and bottom 15% are 2-5.  I’ve tweaked those numbers a bit with my fantasy fiction league. 

With SP, the higher the PB ranking, the more the pitcher card controls the outcome of the play.

Anyway, a problem that I’ve had is that relievers and starters are lumped together in the ERA ranking, and, as a result, relievers make up a large percent of the higher-end pitchers.

What I’m really looking for is an alternative way of ranking pitchers, maybe even separating relievers and starters, that looks beyond just the ERA.

I know that there is a guy who creates seasonal cards and sells via ebay who ranks his pitchers by extra base hits given up.  That doesn’t exactly thrill me.  Several of his customers have complained that many pitchers have unrealistic performances.