Saturday, January 11, 2014

Studying major league pitchers

I haven't analyzed MLB data a lot. I think there is a lot to learn about players at the big league level but not a hell of a lot unless I'm breaking down individual players or have access to more detailed information. I wanted to see if I was missing anything with the current mostly basic data I have. So I've been looking at catagorizing hitters and pitchers in different ways.

From 2009-2013 there were 3302 pitcher seasons in the majors. I catagorized them in semi-logical ways. I isolated starters, high end closers, other starters, eliminated short timers, then looked at the rest of the relievers to try to make sense of how roster spots and players fit beyond the impact players. 

Here are the 18 groups:
1  -  top 150 gs
2  -  next 150 gs
3  -  next 150 gs
4  -  next 150 gs
5  -  over 10 starts
cl  -  top 150 svs
small time closers  -  10 or more saves
most holds  -  most holds, top 150
the third #5  -  over 1/2 games are starts, then most starts (min. 5 Starts), next 150
part time starters  -  more than half their games are starts, avg over 1.5 IP/G (a couple exceptions)
cup o coffee  -  less than 10 IP and less than 10 games
specialist  -  less than .8 ip/g,  top 150
swing man/ Long relief  -  less than half their games are starts, avg over 1.4 IP/G, top 300
60+ IP relievers  - 
long relief 1  -  over 1 IP/G, over 30 IP season-290
long relief 2  -  over 1 IP/G, under 30 IP season-236
short relief 1  -  under 1 IP/G, over 30 IP season-258
short relief 2  -  under 1 IP/G, under 30 IP season-147

On average, teams used about 22 pitchers per season. The statitstical averages are also interesting. There is an obvious regression from the "1" starter on each team to the "4" starter on each team.

I've seen this in other data, but the "5" starter doesn't really exist. It's kind of its own monster and usually encompasses three different pitchers. It's usually two pitchers getting 15 or so starts each and a third guy getting 5-7 more. There just aren't enough starters to go around. Well, that and injury.

Closers are usually isolated to one per team, as are many other roles. The dominance of players with saves and holds is kind of impressive comparatively.

There are a few more types that are good pitchers and some areas that just are horrible. Essentially, the more innings you can get out of the best 7 pitchers on the roster, the better off you are. Nothing ground breaking here, just confirmation of the existing facts.

What did surprise me is the stuff that each possessed. I lined up what each player with their PithcFX data and there is very little noticable differnce in stuff from the #1 to the #5. Starters often have lesser velocity but it's only about a mile per hour. The top 3 starters are all at 91 mph. The 4,5 and spot starters are all in the 90's. Relievers are usually in the 92 range. Closers are the highest at 93.5. Their secondaries are similarly..similar. Some variation but no big steps down. Movement is a little better with the better production as well but nothing major.

I didn't think I'd get to this point but I want to analyze this more.


  1. This comment has been removed by a blog administrator.

  2. Sorry Dan, hit a button and I don't know how to restore it but I appreciate the kind words.