Returning Adjusted Pitching Runs Above Average

One of the things I wanted to look at with the Adjusted Runs Above Average pitching totals (aRAA) was what teams lost the most talent.

What I did was take the sums from my last post and then subtracted out all of the pitchers that were seniors, red shirt seniors, or juniors that signed with a Major League team in 2010. This left me with a returning aRAA total for the pitching staffs.

Some caveats. Obviously, these being my adjusted totals, they adjust for batting average on balls in play (BABIP) and thus, assume that pitchers have little control over their BABIP’s in college baseball. This is a truism for pro baseball, but whether this actually occurs in college ball, I am not sure. It’s just as likely that it does not extend to college baseball as it does.

Also, these aren’t projections. While I’ve done my best to adjust for ‘luck’ by assigning a league-average BABIP to all of the pitchers, this does not account for growth. In college basketball, the typical player improves the most from his freshman to his sophomore season. This is something that I think probably extends to baseball, too, but I’m not positive. Until I can study how pitchers “age” so-to-speak these are not ‘projections.’

That means these are just what the title implies: the total aRAA accumulated by each team’s pitchers in 2010 once you subtract graduates and early-draftees.

Team 2010 aRAA Returning aRAA
MSU 21 21
NW 21 21
MINN 38 19
PUR 45 18
OSU 16 -9
MICH 4 -10
IOWA -12 -19
ILL -31 -29
PSU -44 -53
IND -53 -56

I estimated that Michigan State to have the third most aRAA in 2010 and even though they lose AJ Achter to pro baseball, they still come out essentially the same. The reason for this is because I estimated Achter as being worth 7 aRAA last year while the two senior hurlers on the 2010 squad were Kyle Corcoran and Kurtis Frymier who combined to be worth an estimated -7 aRAA. So the loss of Achter’s positive contributions are balanced out by the loss of Corcoran/Frymier’s sub-average contributions.

The rest of the teams in bullet point fashion:

  • Northwestern is a team that my pitching run estimates like a lot more than reality did. I estimated them at +21 aRAA last year and their only loss is Joe Mursaki. However, his 4 runs is canceled out by the loss of David Jensen’s -3 runs.
  • Minnesota lost Seth Rosin a year early, but Scott Matyas’ returning for his senior season is an enormous help. If Matyas had signed professionally, Minnesota’s returning aRAA would’ve been just +6.
  • Purdue lost arguably the most dominant pitcher in the Big Ten last year, Matt Bischoff. His 25 aRAA is a substantial loss, but the other four seniors — Kevin Cahill, Matt Jansen, Connor Sestak and Drew Wurdack — were worth a cumulative +3 runs. The Boilermakers have a host of sophomores that all contributed a few runs each — and will-be-senior Matt Morgan put up 11 runs last year on the mound which puts Purdue in a good spot pitching-wise.
  • File this one under “things I didn’t expect.” Ohio State loses National Pitcher of the Year and still comes out looking good? Well, Wimmers posted +22 runs last year and they also lose  Eric Best’s +3. Still, losing only a net total of 9 runs would be better than I had assumed. Edit: When I wrote this paragraph, I was having a brain cramp since it was close to 2 in the morning. Ohio State is losing 25 runs — not good.
  • The Wolverines lost some talent in senior Alan Oakes, who posted a +7 aRAA total last year. That, however, was off set by Jeff DeCarlo’s -3 Eric Katzman’s -6. The key is losing Tyler Burgoon’s +16 from last years squad.
  • Iowa didn’t have a good pitching staff last year and barring major growth from an underclassman, they could struggle again. They lost Zach Robertson’s +11 and Michael Jacob’s -3. Luckily Jarred Hippen’s +10 is back.
  • Illinois basically returns anyone that helped the team on the mound last year. Their only losses are the league-average Mike Sterk and Kevin Manson’s -2. The problem is that no one was really productive on the mound aside from Bryan Roberts +5.
  • Penn State actually does have potential to field a worse pitching staff than last year. Senior David Lutz will be taking his +11 total from 2010 with him. Mike Wanamaker’s -1 total is also gone.
  • Finally, Indiana’s abysmal pitching staff that couldn’t strike anyone out or keep anyone off the bases — it was really like they had to face their own juggernaut offense all last year — might stay essentially the same. Their only loss is Chris Squires slightly-above-average +3 aRAA total. That mean’s Drew Leininger’s +12 is back, but so is Matt Igel’s -12, among others.

Again, these are no projections, so don’t take these as gospel. Just a fun look at what teams lost a lot in cumulative fashion.

Advertisements

16 Comments on “Returning Adjusted Pitching Runs Above Average”

  1. This is wonderful stuff, Mike. Great job.

    I was wondering what you thought about Hippen? Can Iowa reasonably expect him to improve on his +10 season or not? Is he a legitimate MLB prospect? Thoughts?

    • Mike says:

      Thanks.

      I’d say that Hippen can probably improve on his +10 from last year. He reduced his walks by 4% from his freshman to his sophomore year while maintaining his K-rate and facing about 140 more hitters in 2010 versus 2009. So, I wouldn’t be surprised to see his K% pick up a bit but that is 100% speculation.

      Like I said, my next project with Big Ten numbers is to finish putting in the classes of each player in each year and then finding how each of the components change over the course of 4 years. That may take a while because I think I’d like to add as many years as possible. I have individual data for 2008-2010 and team data from 2007-2010, so it’d be all manual input for any years prior to 2008 when it comes to individual data.

      As far as his pro prospects, I don’t know much about his arsenal. Obviously his fastball velo will have a lot to do with it, but eyeballing some other kids that were drafted, he should probably be getting his name called in the future. Probably not a high round pick, I’d guess, though.

    • Mike says:

      Forgot that he was a LHP. So he’s most certainly going to get drafted so long as he can get it up over 85 MPH with regularity.

  2. Yeah, being a LHP is going to get him taken in the draft. I’m anxious to see him pitch this year.

    Basically for Iowa, they have to get a little lucky to contend and that’s probably true about every year. I’m really surprised about OSU as you mentioned in the post. I’d think losing Wimmers would have hurt them more than it did.

    • Mike says:

      Yeah. I’m still learning about the Big Ten and college baseball in general — reason for starting this blog — but the Hawkeye’s pitching staff puts them behind the eight ball.

      I did finish up labeling the class for every season I have from 2008-2010 for pitchers. So, in the near future I can look at the comparative strikeout rates, walk rates, etc etc put up by freshman, sophomores, juniors and seniors. Pretty excited for that.

  3. I’m looking forward to that.

    I was thinking, couldn’t we use strikeout rate as a proxy for velocity? I know it wouldn’t be 100% accurate, but it probably would be relatively accurate. I think THT wrote a piece a year or two ago about how much a 1 mph increase in velocity affects strikeout rate. When I get time later today I’ll try and find that article.

    There would be outliers of course, but I wonder if that would work.

    • Mike says:

      Oooh good thought. I don’t have the data in front of me but Hippen was around 16% in his first two seasons.

      But I like that thought on that. I don’t know how I’d test it without accurate scouting reports on all pitchers (or most). For instance, I don’t have one for Hippen at all and any underclassmen is difficult. I would guess a PG Crosscheckers subscription would help a lot. Not sure I want to fork out the cash for that, though it has crossed my mind in the past.

  4. That’s the problem since we don’t have accurate scouting reports. The only way to test it is to test the group that we do have that information with and then make adjustments based on research. For example, it’s been shown that a pitcher’s FB velocity will pick up by about 1 mph until he’s 25 years old. This isn’t true for all pitchers of course. A guy with 1000 innings by the time he’s 25 isn’t at all likely to see the same benefit as a guy making his big league debut, but since these are college pitchers we’re interested with, we don’t really have to worry about that.

    I looked around only briefly for the article I mentioned and didn’t have luck. I’ll look around more when I have more time, but just looking over the fangraphs leaderboards, it appears the top 15 or so in velocity strikeout about 22% of the batters they face and then the next 15 are at about 20%. I imagine if we include multiple years we’d find a linear relationship. I’d be shocked if we didn’t. There’s been research that has shown that babip skill for pitchers is loosely tied with velocity. Batters have less time to think and make more defensive swings resulting in weaker contact. It only seems logical that those same defensive swings and that lack of confidence in what you’re swinging at will result in more swings and misses.

    So based on what we know and knowing college players are between the ages of 18 and 22 (mostly), we could take the average velocity at the MLB level and work backwards. Only thing though, we’d have to consider the simply fact that the overwhelming majority of college pitchers won’t pitch in the big leagues. It’s very, very likely their velocity is significantly lower than that of the MLB pitchers. In fact, based on the several years of reading Baseball America’s scouting reports for the top 500 draft prospects, it seems to me there is a considerable difference in terms of average velocity. We’d never be able to get a perfect estimate, but our end result isn’t necessarily intended to be perfect, is it? It’s an estimated velocity we’re trying to come up for the most part so it naturally comes with large variance.

    I’m just making numbers up here, but let’s say the average FB velocity at the MLB level is 88 mph. That consists of some players whose velocity is peaking and a lot more whose velocity peaked years ago. Would it then be reasonable to guess that the velocity at the college level is 85 mph?

    Surely there are ways to work with what data we do have and work backwards. Thoughts?

    • Mike says:

      David, your final paragraph where you made up numbers were nearly identical to what was popping in my head while reading your reply. I was thinking 88 and 86-ish. I could probably just email Harry Pavlidis and ask what the average FB velo from 07-2010 is in the majors (or I’m sure it’s been published or I could get it with one of the PITCHf/x online tools like Joe Lefkowitz’s). But yes, I agree with you on all points above.

      Is this the article you were talking about? http://www.hardballtimes.com/main/article/lose-a-tick-gain-a-tick/

      It was the one that sprung to mind when you were describing what you read in an earlier comment and I remember something similar; THT has put out so many PITCHf/x based stuff that it sometimes gets muddled (and I love it all) in my mind as to what I’ve read and what I haven’t.

      That article says the average velo in the Majors for a starter from 02-09 was around 90.5 MPH. So, I’d say something like 85-86 seems right for the Big Ten for a couple reasons:

      1) If a guy can throw consistently harder than 90, he’s likely drafted high enough to sign pro ball.

      2) The talent base in the Big Ten isn’t as good as the SEC or Pac-10.

      Also, I think it’s important to look at the average MLB starter’s velo as we’ll be interested in starters velo in the Big Ten because, let’s face it, if you’re a college reliever, you have to be ELITE to get drafted in the top 5 rounds, or show the flexibility to become a starter at some point and succeed. Someone like Chance Ruffin of Texas can get selected in the 2nd round by the Tigers because he’s shown that flexibility in the past.

      I really do like this idea. I’ll also send a shout out to patriot from Walk Saber at some point about his thoughts. I’m just not sure how to secure velocity information on pitchers in the Big Ten short of a PG Crosschecker subscription. Maybe there’s a chance at emailing some SID’s at the schools to see if we could get that info, though I don’t know how willing they’d be to dish out “hey can you ask the coach how hard each of his pitchers throw?” Probably wouldn’t hurt to feel it out and get a bunch of “uhh, no’s,” haha.

      So glad you’ve stumbled upon this place. Definitely nicer to bounce ideas off someone else like this.

  5. I’ll respond in more detail a bit later today, but I was given the opportunity to interview someone on The Hawkeye State about college recruits. I assumed it was basketball and football, but it’s worth checking into now to see if there would also be someone on the baseball side of things. We know Iowa isn’t among the best in the B10 so if we have an idea of the average velocity for Hawkeyes pitchers, we can probably add a tick to the conference. Let me look into this.

    If nothing else, I think I might go ahead and get a subscription to PG Crosschecker.

    Anyway, I’ll be back with some more thoughts later on.

    • Mike says:

      Awesome. Honestly, we should email PG Crosschecker as see if there’s a subscription that would work on two IP’s and I would go splits-ville on a subscription with you. PG is kind of expensive.

  6. Yes, excellent point about limiting it to the starters. I hadn’t considered that, but it wouldn’t make sense to include the relievers. I think that THT article is the one I was thinking about. I also found an article that Dave Cameron did on Fangraphs that showed velocity does correlate with strikeouts though the correlation is by no means that high. It’s just one factor of course.

    That got me to thinking that we know other factors too. Strikeouts being one of them and lower babip is another too. Downward movement is another, but there’s no way to get that info. If we included both babip and K%, we’d have two known factors that relate to velocity. The great thing is that those stats are available or can be easily calculated.

    I’m thinking this weekend I may take a look at MLB starters and apply some method to estimate their velocity based on their babip and K%. Then I can check that against their actual velocity to see how far off they are. I’m expecting many to be off, but many more to be fairly accurate.

    And really that’s what we want in my opinion. Short of lengthy discussions with numerous scouts or spending a lot of money obtaining the information, we’re left to work with what we have. I don’t even know how many people would find the information useful, but I sure would. If I could have a spreadsheet set up and update the data and have it spit out an estimated FB velocity, I’d consider that real knowledge. It comes with error and even the chance of being flat out wrong, but if it works for MLB players, I’d feel comfortable reporting those results.

    I’m definitely glad I found this place. I find myself more interested in college baseball than I am in MLB.

  7. I was just checking out PG and we should contact them. I’d think if we’re up front with them that they’d be willing to allow two IP’s.

    • Mike says:

      Absolutely agreed on the first part. I’ve found college baseball more interesting to me than the MLB for the last year or so but I rarely did much with it until I decided to limit my scope to the big ten because it’s easy to take the data from their site and because I have a rooting interest in one of their teams. As much as I loved watching Trevor Bauer pitch in the CWS I had no attachment to UCLA. I am an MSU fan so that helps drive this.

      The other thing that I love is just the research part of baseball. So with things to uncover in college baseball intrigues me a ton.

      I’d be interested in the results of trying to guesstimate velo from K% and BABIP and whatnot.

      One place to get movement, is probably just Boston College (maybe Notre Dame now that Aoki took that job this year). Great watch: http://www.youtube.com/watch?v=Efsg1YTvprw

      What I would give to have the money to give to a big ten school and say “let me play with around with this data please.”

      Yeah, i’m down to contact PG for sure.

  8. Dave says:

    This is excellent stuff, but how will it apply with the new bat requirements this spring?

    • Mike says:

      Dave, I honestly have no idea. Reports I’ve been reading is that the ball jumps off the bat far less than it used to. It’s just something I won’t know until it happens.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s